summaryrefslogtreecommitdiff
path: root/results/kp_innovation_stability/closed_form_innovation.json
diff options
context:
space:
mode:
authorYurenHao0426 <Blackhao0426@gmail.com>2026-07-22 16:54:12 -0500
committerYurenHao0426 <Blackhao0426@gmail.com>2026-07-22 16:54:12 -0500
commit7954dc472683aad665d05b00ab3d9f32902e140c (patch)
treefb4dfe88b0985c6a48f365f78cecf69beafb4bd1 /results/kp_innovation_stability/closed_form_innovation.json
parent0ae690e7b3144da2c5b3384cd7169e86529fd26f (diff)
diagnostic: record pre-grid stability candidates
Diffstat (limited to 'results/kp_innovation_stability/closed_form_innovation.json')
-rw-r--r--results/kp_innovation_stability/closed_form_innovation.json32980
1 files changed, 32980 insertions, 0 deletions
diff --git a/results/kp_innovation_stability/closed_form_innovation.json b/results/kp_innovation_stability/closed_form_innovation.json
new file mode 100644
index 0000000..17c0c11
--- /dev/null
+++ b/results/kp_innovation_stability/closed_form_innovation.json
@@ -0,0 +1,32980 @@
+{
+ "first_any_nonfinite_groups": {},
+ "first_any_nonfinite_step": null,
+ "first_nonfinite_by_group": {},
+ "first_nonfinite_by_metric": {},
+ "first_nonfinite_metrics": [],
+ "first_training_failure_groups": {},
+ "first_training_failure_step": null,
+ "initial_parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4011031687259674,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3956821858882904,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1162850856781006,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1175870895385742e-08,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "max_steps": 352,
+ "predictor_mode": "closed_form",
+ "predictor_warmup": {
+ "closed_form_fit": {
+ "max_absolute_residual_soma_slope": 1.6851760875624677e-08,
+ "mse": 1.477951640859762e-18,
+ "observations": 64,
+ "residual_traffic_rms_ratio": 7.21125499270486e-08
+ },
+ "first_mse": 1.477951640859762e-18,
+ "last_mse": 1.477951640859762e-18,
+ "post_warmup_traffic_residual_rms_ratio": 7.21125499270486e-08,
+ "steps": 1,
+ "task_loader_state_restored": true
+ },
+ "protocol": "kp_mixed_traffic_nonfinite_diagnosis_v1",
+ "provenance": {
+ "git_commit": "72f6b758540e8c1d7a44fed63bafe71784f6dfa2",
+ "git_tracked_dirty": false
+ },
+ "rule": "innovation",
+ "scope": "training_only_no_validation_or_test_evaluation",
+ "split": {
+ "dataset": "cifar10",
+ "input_layout": "NCHW",
+ "input_shape": [
+ 3,
+ 32,
+ 32
+ ],
+ "loader_seed": 0,
+ "normalization_mean": [
+ 0.49140000343322754,
+ 0.4821999967098236,
+ 0.4465000033378601
+ ],
+ "normalization_std": [
+ 0.24699999392032623,
+ 0.2434999942779541,
+ 0.26159998774528503
+ ],
+ "split_from_training_only": true,
+ "split_seed": 2027,
+ "test_examples": 10000,
+ "train_examples": 45000,
+ "training_augmentation": "random_crop_32_padding_4_zero_then_horizontal_flip_p0.5",
+ "validation_class_counts": {
+ "0": 500,
+ "1": 500,
+ "2": 500,
+ "3": 500,
+ "4": 500,
+ "5": 500,
+ "6": 500,
+ "7": 500,
+ "8": 500,
+ "9": 500
+ },
+ "validation_examples": 5000,
+ "validation_index_sha256": "8328b206a97c420e49e54e3eca4abe3274c4756b084355784ea3fb8059e4515b"
+ },
+ "test_evaluations": 0,
+ "traffic_calibration": {
+ "instruction_rms": [
+ 0.008471464738249779,
+ 0.006617916747927666,
+ 0.003988116513937712,
+ 0.0043865954503417015,
+ 0.002713605063036084,
+ 0.003158062230795622,
+ 0.002004207344725728,
+ 0.003757704282179475,
+ 0.0027085731271654367,
+ 0.0030200635083019733,
+ 0.002056022873148322,
+ 0.0023899395018815994,
+ 0.0016842411132529378,
+ 0.0034010226372629404,
+ 0.002549725119024515,
+ 0.0030542907770723104,
+ 0.0020918569061905146,
+ 0.002605808200314641,
+ 0.0018916188273578882
+ ],
+ "realized_traffic_instruction_rms_ratio": [
+ 4.0,
+ 4.0,
+ 4.0,
+ 4.0,
+ 4.0,
+ 4.0,
+ 3.999999523162842,
+ 4.0,
+ 4.000000476837158,
+ 3.999999761581421,
+ 4.0,
+ 4.0,
+ 3.999999761581421,
+ 4.0,
+ 4.000000476837158,
+ 4.000000476837158,
+ 4.0,
+ 4.0,
+ 4.0
+ ],
+ "target_ratio": 4.0,
+ "traffic_gain": [
+ 0.0443243607878685,
+ 0.035817068070173264,
+ 0.013778294436633587,
+ 0.022762222215533257,
+ 0.007214079145342112,
+ 0.017218582332134247,
+ 0.004616298712790012,
+ 0.02042512409389019,
+ 0.007327491417527199,
+ 0.01611751690506935,
+ 0.004703814629465342,
+ 0.012562948279082775,
+ 0.003324519144371152,
+ 0.01805141195654869,
+ 0.006063441745936871,
+ 0.016296451911330223,
+ 0.0044285994954407215,
+ 0.013874167576432228,
+ 0.0036007820162922144
+ ],
+ "unscaled_traffic_rms": [
+ 0.7644973993301392,
+ 0.7390796542167664,
+ 1.157796859741211,
+ 0.7708553671836853,
+ 1.504616141319275,
+ 0.7336404919624329,
+ 1.736635684967041,
+ 0.7358984351158142,
+ 1.478581428527832,
+ 0.7495108246803284,
+ 1.7483876943588257,
+ 0.7609485983848572,
+ 2.0264477729797363,
+ 0.7536302804946899,
+ 1.6820316314697266,
+ 0.7496824264526367,
+ 1.8894071578979492,
+ 0.7512690424919128,
+ 2.10134220123291
+ ]
+ },
+ "trajectory": [
+ {
+ "batch_loss": 2.9134020805358887,
+ "innovation_rms": 0.0042155604681470464,
+ "instruction_rms": 0.004215397894925855,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.40285658836364746,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9875131845474243,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.053474169224500656,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4058851897716522,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6322960257530212,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.00811514537781477,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.08115145564079285,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0037773847579956,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.03945024311542511,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.05994534492492676,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3794170618057251,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.632293164730072,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1170727014541626,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 1.3374497281403603e-09,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.01731826254949561,
+ "step": 1,
+ "teaching_rms": 0.0042155604681470464,
+ "traffic_rms": 0.01680093257108933
+ },
+ {
+ "batch_loss": 2.4060206413269043,
+ "innovation_rms": 0.004091979791127017,
+ "instruction_rms": 0.004091855717102748,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7660105228424072,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.962428092956543,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.05502914637327194,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4140651524066925,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3981689214706421,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.014180497266352177,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.062249813228845596,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0088073015213013,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.050298769026994705,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.08588835597038269,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3788767158985138,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39816349744796753,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1178287267684937,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017319218640961538,
+ "step": 2,
+ "teaching_rms": 0.004091979791127017,
+ "traffic_rms": 0.016830575365087055
+ },
+ {
+ "batch_loss": 2.635755777359009,
+ "innovation_rms": 0.004280882571916788,
+ "instruction_rms": 0.0042807399556906425,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0908452272415161,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.803679943084717,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.059588588774204254,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4310036599636078,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3125655949115753,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.017003284767270088,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.05397982895374298,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0124577283859253,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.058454759418964386,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.12444570660591125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3962392508983612,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3126407861709595,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1186662912368774,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017367767665854697,
+ "step": 3,
+ "teaching_rms": 0.004280882571916788,
+ "traffic_rms": 0.016834309883729782
+ },
+ {
+ "batch_loss": 2.8387465476989746,
+ "innovation_rms": 0.004141767496510404,
+ "instruction_rms": 0.004141622534104046,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3725169897079468,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.7627177238464355,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.07006050646305084,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4422183930873871,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.23453189432621002,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.019318245351314545,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.07771710306406021,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0175281763076782,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.08348087966442108,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11777545511722565,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3977593779563904,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2345178723335266,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1191269159317017,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017322043898484612,
+ "step": 4,
+ "teaching_rms": 0.004141767496510404,
+ "traffic_rms": 0.016820494119753265
+ },
+ {
+ "batch_loss": 2.7703216075897217,
+ "innovation_rms": 0.004121630316859195,
+ "instruction_rms": 0.004121494756286382,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6269456148147583,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.646733283996582,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.07763772457838058,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4483798146247864,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4080396294593811,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0276226494461298,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.08518961817026138,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0239425897598267,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.09665486216545105,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.12530282139778137,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3851257264614105,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.408022940158844,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1194764375686646,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01735644607001741,
+ "step": 5,
+ "teaching_rms": 0.004121630316859195,
+ "traffic_rms": 0.016864356326022626
+ },
+ {
+ "batch_loss": 2.6939163208007812,
+ "innovation_rms": 0.00405820661368265,
+ "instruction_rms": 0.004057911339026368,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.849908709526062,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.50608491897583,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.07965654134750366,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4589293599128723,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4636334776878357,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.03593522682785988,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.08797040581703186,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0324904918670654,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0907355546951294,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.12392328679561615,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3915436863899231,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4636010527610779,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1197863817214966,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017346697090116353,
+ "step": 6,
+ "teaching_rms": 0.00405820661368265,
+ "traffic_rms": 0.016868334446286024
+ },
+ {
+ "batch_loss": 2.5562174320220947,
+ "innovation_rms": 0.004106378832790066,
+ "instruction_rms": 0.004105980114442889,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1297059059143066,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.180209159851074,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11324872076511383,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.47391656041145325,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6083535552024841,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.04298514872789383,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.09073657542467117,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0452340841293335,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.12743616104125977,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11740318685770035,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39905864000320435,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6082929372787476,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1196560859680176,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017376262411764796,
+ "step": 7,
+ "teaching_rms": 0.004106378832790066,
+ "traffic_rms": 0.01687783260866868
+ },
+ {
+ "batch_loss": 2.6328983306884766,
+ "innovation_rms": 0.0039552406116064726,
+ "instruction_rms": 0.003954820305362132,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.414527416229248,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.730833053588867,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.13951970636844635,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.49402716755867004,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5742970705032349,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.04839029908180237,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.09636467695236206,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0631275177001953,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1789344847202301,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.17809614539146423,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39275550842285156,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5742810964584351,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1190859079360962,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01740131227070936,
+ "step": 8,
+ "teaching_rms": 0.0039552406116064726,
+ "traffic_rms": 0.01694032170039137
+ },
+ {
+ "batch_loss": 2.8354854583740234,
+ "innovation_rms": 0.0038066679401576897,
+ "instruction_rms": 0.0038058435770175056,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.6981747150421143,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.311980247497559,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.18914268910884857,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5308004021644592,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6741237640380859,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.053289610892534256,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.12330227345228195,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0879368782043457,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.24809400737285614,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1894720494747162,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3938608467578888,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.674206018447876,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1189571619033813,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01743969784968324,
+ "step": 9,
+ "teaching_rms": 0.0038066679401576897,
+ "traffic_rms": 0.017016972010574432
+ },
+ {
+ "batch_loss": 2.5795114040374756,
+ "innovation_rms": 0.003411647747041196,
+ "instruction_rms": 0.0034104946524406697,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.9664218425750732,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.8352689743042,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22492200136184692,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5668005347251892,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5925586819648743,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.059577826410532,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.15927714109420776,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1152677536010742,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27613478899002075,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2744438052177429,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.398246169090271,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5926461219787598,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.118929147720337,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01746617310928646,
+ "step": 10,
+ "teaching_rms": 0.003411647747041196,
+ "traffic_rms": 0.017121798117182705
+ },
+ {
+ "batch_loss": 2.9323935508728027,
+ "innovation_rms": 0.0035422110084162603,
+ "instruction_rms": 0.003540679348208058,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.2369801998138428,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.870978355407715,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.24631331861019135,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5499615669250488,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6828886270523071,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.07304336130619049,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.182850643992424,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1464817523956299,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.32981863617897034,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3281099498271942,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4137158691883087,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.68284010887146,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1192632913589478,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01756752150632741,
+ "step": 11,
+ "teaching_rms": 0.0035422110084162603,
+ "traffic_rms": 0.01720946856364748
+ },
+ {
+ "batch_loss": 2.3742740154266357,
+ "innovation_rms": 0.003066461246923823,
+ "instruction_rms": 0.0030645078628753703,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.487191677093506,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.107603073120117,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2631068527698517,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.540130078792572,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5768554210662842,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.09118994325399399,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2347237467765808,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.175665259361267,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.37117981910705566,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3532576560974121,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.41690877079963684,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5768046975135803,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1196684837341309,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01757949632574908,
+ "step": 12,
+ "teaching_rms": 0.003066461246923823,
+ "traffic_rms": 0.017301854686027556
+ },
+ {
+ "batch_loss": 2.7443814277648926,
+ "innovation_rms": 0.003104139123506068,
+ "instruction_rms": 0.0031012259437220796,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.723893642425537,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.847227096557617,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2889328896999359,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5433750152587891,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39185935258865356,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11064150929450989,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27764618396759033,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2020100355148315,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.43435293436050415,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39944711327552795,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.41358500719070435,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39180660247802734,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1200388669967651,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017679801026698805,
+ "step": 13,
+ "teaching_rms": 0.003104139123506068,
+ "traffic_rms": 0.017408859298089045
+ },
+ {
+ "batch_loss": 2.760450839996338,
+ "innovation_rms": 0.003264266662167202,
+ "instruction_rms": 0.0032612553408262624,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.9565203189849854,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.100886344909668,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.312777578830719,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5593229532241821,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6146011352539062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.13215889036655426,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2358931303024292,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2288312911987305,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5316415429115295,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3160385191440582,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.40624040365219116,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6145534515380859,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.120679259300232,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017806782135157385,
+ "step": 14,
+ "teaching_rms": 0.003264266662167202,
+ "traffic_rms": 0.01750919466592939
+ },
+ {
+ "batch_loss": 2.390294313430786,
+ "innovation_rms": 0.0031973566588540782,
+ "instruction_rms": 0.0031932068062097683,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.1814751625061035,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.205536842346191,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.32848480343818665,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5915842056274414,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.48606380820274353,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.15089307725429535,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2408883422613144,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.253474473953247,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6546567678451538,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4822673797607422,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4053812623023987,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.48597395420074463,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1215593814849854,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.017882206367036687,
+ "step": 15,
+ "teaching_rms": 0.0031973566588540782,
+ "traffic_rms": 0.017594439150033114
+ },
+ {
+ "batch_loss": 2.499338150024414,
+ "innovation_rms": 0.0027314716652300196,
+ "instruction_rms": 0.002724101781537995,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.3955888748168945,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.519990921020508,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3223620355129242,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6112402677536011,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6568875908851624,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1713712513446808,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2901841998100281,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.334059238433838,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8058475852012634,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5154536366462708,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.436866819858551,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6569380164146423,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1225414276123047,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.0003153845318593085,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01807934583116801,
+ "step": 16,
+ "teaching_rms": 0.0027314716652300196,
+ "traffic_rms": 0.017879520457487353
+ },
+ {
+ "batch_loss": 2.372697114944458,
+ "innovation_rms": 0.002453181137562051,
+ "instruction_rms": 0.0024430717349943148,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.604691982269287,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.946907997131348,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.29982179403305054,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6155865788459778,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1037086248397827,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1929243952035904,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.34248465299606323,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.429958462715149,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9589927792549133,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.44344985485076904,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4194991886615753,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1039767265319824,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1234971284866333,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 4.9573829198187074e-08,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.018258765929505105,
+ "step": 17,
+ "teaching_rms": 0.002453181137562051,
+ "traffic_rms": 0.018097707486157088
+ },
+ {
+ "batch_loss": 2.966481924057007,
+ "innovation_rms": 0.002626741137803833,
+ "instruction_rms": 0.002617023871033276,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.781918525695801,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.133254051208496,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.29185765981674194,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6121950745582581,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.45661425590515137,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20417754352092743,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3939765393733978,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5443110466003418,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1435253620147705,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.37990516424179077,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4209553897380829,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4563407897949219,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1246222257614136,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01840762769110877,
+ "step": 18,
+ "teaching_rms": 0.002626741137803833,
+ "traffic_rms": 0.018238385274236046
+ },
+ {
+ "batch_loss": 2.539519786834717,
+ "innovation_rms": 0.0023978059063201555,
+ "instruction_rms": 0.0023860750280373065,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.933766841888428,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.204654693603516,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2663789391517639,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6084650158882141,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1335527896881104,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.24664592742919922,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4476110339164734,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6783114671707153,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3400042057037354,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4886690378189087,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4469318985939026,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1332745552062988,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1256836652755737,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.018471955371142847,
+ "step": 19,
+ "teaching_rms": 0.0023978059063201555,
+ "traffic_rms": 0.018320735134734888
+ },
+ {
+ "batch_loss": 2.5693371295928955,
+ "innovation_rms": 0.0022789799861239156,
+ "instruction_rms": 0.0022624175807791583,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.049882411956787,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.284788131713867,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2783513367176056,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6129173040390015,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4527481198310852,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2976106107234955,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5096468925476074,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8352769613265991,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5696547031402588,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.32260870933532715,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4439314901828766,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4524655342102051,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1267844438552856,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01850159085698167,
+ "step": 20,
+ "teaching_rms": 0.0022789799861239156,
+ "traffic_rms": 0.0183581768144954
+ },
+ {
+ "batch_loss": 2.813922643661499,
+ "innovation_rms": 0.0020388620260393415,
+ "instruction_rms": 0.0020134926272382153,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.144930362701416,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.06090545654297,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3052089214324951,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6319825649261475,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5588269233703613,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3575979173183441,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5998730063438416,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.022397518157959,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8712055683135986,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4660817086696625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.44456157088279724,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5588585138320923,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1274603605270386,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.018573525221096267,
+ "step": 21,
+ "teaching_rms": 0.0020388620260393415,
+ "traffic_rms": 0.018462791244348596
+ },
+ {
+ "batch_loss": 2.5997934341430664,
+ "innovation_rms": 0.001722749007804901,
+ "instruction_rms": 0.0016877974602493176,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.239352226257324,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.724449157714844,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3211499750614166,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6529640555381775,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5088138580322266,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4268565773963928,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6925865411758423,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.241236925125122,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1883931159973145,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3510974943637848,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4422110617160797,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5089699029922485,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1281388998031616,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.018657582314540743,
+ "step": 22,
+ "teaching_rms": 0.001722749007804901,
+ "traffic_rms": 0.018597868295724972
+ },
+ {
+ "batch_loss": 2.377486228942871,
+ "innovation_rms": 0.0015274287987295054,
+ "instruction_rms": 0.0014810663240569766,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.2756781578063965,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 50.454349517822266,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3958193063735962,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6642628908157349,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6023381948471069,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5072360038757324,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8037945032119751,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4970622062683105,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.558253288269043,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39585763216018677,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.43711206316947937,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6024078130722046,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1288468837738037,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01884103452723237,
+ "step": 23,
+ "teaching_rms": 0.0015274287987295054,
+ "traffic_rms": 0.01879422267083811
+ },
+ {
+ "batch_loss": 2.519068717956543,
+ "innovation_rms": 0.0015539535905935678,
+ "instruction_rms": 0.0014884740555743167,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.365382194519043,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.41975021362305,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5265324115753174,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6646087169647217,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.49619781970977783,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6015908122062683,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9435480833053589,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.8016581535339355,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.045958995819092,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5266553163528442,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.440369188785553,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.49602699279785156,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1295040845870972,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01878732357970989,
+ "step": 24,
+ "teaching_rms": 0.0015539535905935678,
+ "traffic_rms": 0.018733003818666633
+ },
+ {
+ "batch_loss": 2.4134912490844727,
+ "innovation_rms": 0.0015181679861462772,
+ "instruction_rms": 0.0014300949813765812,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.426975727081299,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 62.707462310791016,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5889586210250854,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6626064777374268,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5045179724693298,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7111784815788269,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0958764553070068,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.1586475372314453,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.569894790649414,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5890192985534668,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4511707127094269,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5042850375175476,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1301672458648682,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01887763674492342,
+ "step": 25,
+ "teaching_rms": 0.0015181679861462772,
+ "traffic_rms": 0.018821619324871187
+ },
+ {
+ "batch_loss": 2.1513545513153076,
+ "innovation_rms": 0.0014808927926199071,
+ "instruction_rms": 0.0013564777233801475,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.4548540115356445,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.04145050048828,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7377861738204956,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6633495092391968,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5040757656097412,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8395453095436096,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2836681604385376,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.5891213417053223,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.304738998413086,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7378472685813904,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.47198665142059326,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5038410425186157,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1307259798049927,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.018981568904481497,
+ "step": 26,
+ "teaching_rms": 0.0014808927926199071,
+ "traffic_rms": 0.01893045325357228
+ },
+ {
+ "batch_loss": 2.5521321296691895,
+ "innovation_rms": 0.001618313117661261,
+ "instruction_rms": 0.001460687109813608,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.505712509155273,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.54241180419922,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9004915952682495,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6681128740310669,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.47095921635627747,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9896760582923889,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5013076066970825,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.108846187591553,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.1972479820251465,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9005530476570129,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.49181830883026123,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4709010124206543,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1310385465621948,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01907346488702647,
+ "step": 27,
+ "teaching_rms": 0.001618313117661261,
+ "traffic_rms": 0.019019131427712613
+ },
+ {
+ "batch_loss": 2.32607102394104,
+ "innovation_rms": 0.001734717008075824,
+ "instruction_rms": 0.0015439323092846754,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.581539154052734,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.0711441040039,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5860507488250732,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6759477853775024,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3560085594654083,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1644923686981201,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7481633424758911,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.725391864776611,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.165456771850586,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5860090255737305,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.502753496170044,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35594815015792847,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.131016731262207,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.019085816668597693,
+ "step": 28,
+ "teaching_rms": 0.001734717008075824,
+ "traffic_rms": 0.01902718417092033
+ },
+ {
+ "batch_loss": 2.1079533100128174,
+ "innovation_rms": 0.001614208422847363,
+ "instruction_rms": 0.0013879725209924342,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.655301094055176,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.73936462402344,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9515906572341919,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6850367188453674,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3998200297355652,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8331910371780396,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.358466625213623,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9397430419921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.387650012969971,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.622581481933594,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.95154869556427,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5205430388450623,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3998992443084717,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1308608055114746,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01908608820952148,
+ "step": 29,
+ "teaching_rms": 0.001614208422847363,
+ "traffic_rms": 0.01903741706923829
+ },
+ {
+ "batch_loss": 2.4398152828216553,
+ "innovation_rms": 0.0018684299261216928,
+ "instruction_rms": 0.0015840354028558773,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.698953151702881,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79.24505615234375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4339640140533447,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7020197510719299,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.37158241868019104,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.9502741694450378,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5811188220977783,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.226522445678711,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.143819808959961,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.561697483062744,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.433885335922241,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5161225199699402,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.37157946825027466,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.130989909172058,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.019178149811696178,
+ "step": 30,
+ "teaching_rms": 0.0018684299261216928,
+ "traffic_rms": 0.019127195133454768
+ },
+ {
+ "batch_loss": 2.3253462314605713,
+ "innovation_rms": 0.001788298802936639,
+ "instruction_rms": 0.001449563260319728,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.742127895355225,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.76603698730469,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9212884902954102,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7136139869689941,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.36551153659820557,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.14240300655365,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.821723222732544,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4060440063476562,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.925755500793457,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.819357872009277,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9213452339172363,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5063400864601135,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.36544498801231384,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1310813426971436,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.019313134773148905,
+ "step": 31,
+ "teaching_rms": 0.001788298802936639,
+ "traffic_rms": 0.01926499254371355
+ },
+ {
+ "batch_loss": 2.380310535430908,
+ "innovation_rms": 0.001988100819788187,
+ "instruction_rms": 0.0015841581083066878,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.739208698272705,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.56456756591797,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9271314144134521,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7039700150489807,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.257015198469162,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3351161479949951,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0821731090545654,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.604498863220215,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.739916801452637,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.14161205291748,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9271882772445679,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5060418844223022,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2569483816623688,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2761509418487549,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.004851209465414286,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.01939135662203683,
+ "step": 32,
+ "teaching_rms": 0.001988100819788187,
+ "traffic_rms": 0.019330560063914322
+ },
+ {
+ "batch_loss": 2.204509973526001,
+ "innovation_rms": 0.0021966655013596096,
+ "instruction_rms": 0.00149694468195795,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.7255024909973145,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78.12850952148438,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.191646099090576,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6886186003684998,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3075675964355469,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4585866928100586,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.3925247192382812,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.1035170555114746,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.759930610656738,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.200140953063965,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1916518211364746,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5022019743919373,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.30755722522735596,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3996272087097168,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 2.585193887640085e-06,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.01947853235276175,
+ "step": 33,
+ "teaching_rms": 0.0021966655013596096,
+ "traffic_rms": 0.019420931377611958
+ },
+ {
+ "batch_loss": 2.3175172805786133,
+ "innovation_rms": 0.0023349538988556673,
+ "instruction_rms": 0.001808198949383306,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.724668025970459,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.736572265625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.728097915649414,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6793885827064514,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2493831366300583,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6114938259124756,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.7340359687805176,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.415111541748047,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.8848237991333,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.248933792114258,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.728424310684204,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5104270577430725,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.24931100010871887,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5525400638580322,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.0195202133290542,
+ "step": 34,
+ "teaching_rms": 0.0023349538988556673,
+ "traffic_rms": 0.01943597903624241
+ },
+ {
+ "batch_loss": 2.058091163635254,
+ "innovation_rms": 0.0023462935833058543,
+ "instruction_rms": 0.0015564900497075924,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.7286858558654785,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.56428527832031,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4223194122314453,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.676106333732605,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.25796088576316833,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7676796913146973,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.118908166885376,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.8487207889556885,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.173614501953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.887908935546875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.422416925430298,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5183413028717041,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.25795793533325195,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.708731770515442,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.019697766417200342,
+ "step": 35,
+ "teaching_rms": 0.0023462935833058543,
+ "traffic_rms": 0.019636765218064345
+ },
+ {
+ "batch_loss": 2.199193239212036,
+ "innovation_rms": 0.002600908246636117,
+ "instruction_rms": 0.0017726901728169148,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.201139450073242,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.58634948730469,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.334897994995117,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6777284741401672,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3074975609779358,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8926100730895996,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.5607967376708984,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.418884754180908,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.709510803222656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15.358966827392578,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.3349485397338867,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5314429998397827,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3074129819869995,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8336679935455322,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.019895631647437646,
+ "step": 36,
+ "teaching_rms": 0.002600908246636117,
+ "traffic_rms": 0.019820392096708397
+ },
+ {
+ "batch_loss": 2.040970802307129,
+ "innovation_rms": 0.0028156293536198865,
+ "instruction_rms": 0.0016947090309600535,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.88612699508667,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.34931182861328,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.122631072998047,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6787092685699463,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.26846808195114136,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.079721689224243,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.071847915649414,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.110510349273682,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.540473937988281,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.309627532958984,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.1225433349609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5433096885681152,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.26838332414627075,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.020785331726074,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.020115239283186384,
+ "step": 37,
+ "teaching_rms": 0.0028156293536198865,
+ "traffic_rms": 0.02004368357922975
+ },
+ {
+ "batch_loss": 2.1023881435394287,
+ "innovation_rms": 0.003334321862602586,
+ "instruction_rms": 0.001531414006765337,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.2105937004089355,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 84.8583984375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.9010181427001953,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6809340715408325,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.26986435055732727,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.149989128112793,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.66050910949707,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.886610984802246,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16.686119079589844,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.456451416015625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.901067018508911,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.549298882484436,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2697794437408447,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0910587310791016,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.02075299374874943,
+ "step": 38,
+ "teaching_rms": 0.003334321862602586,
+ "traffic_rms": 0.020699453381049435
+ },
+ {
+ "batch_loss": 2.0099165439605713,
+ "innovation_rms": 0.003537141055369514,
+ "instruction_rms": 0.001698314737619188,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.308614730834961,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91.82675170898438,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.015223741531372,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6801835894584656,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22674185037612915,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1379549503326416,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.356091022491455,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.955818176269531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.33684539794922,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.507261276245117,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.0152878761291504,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5522624850273132,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22690927982330322,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0790302753448486,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.021082425522408552,
+ "step": 39,
+ "teaching_rms": 0.003537141055369514,
+ "traffic_rms": 0.02101331331448059
+ },
+ {
+ "batch_loss": 2.0152196884155273,
+ "innovation_rms": 0.00409521288753361,
+ "instruction_rms": 0.0017857084471213092,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.331830024719238,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.36001586914062,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.653810977935791,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6761200428009033,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.23667769134044647,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1774535179138184,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.184297561645508,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.28206729888916,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.666475296020508,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.296302795410156,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.653931617736816,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5591883063316345,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.23678140342235565,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1122381687164307,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.021728567682548282,
+ "step": 40,
+ "teaching_rms": 0.00409521288753361,
+ "traffic_rms": 0.021652491132571644
+ },
+ {
+ "batch_loss": 2.0925281047821045,
+ "innovation_rms": 0.005129964994619575,
+ "instruction_rms": 0.001851708561969891,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.93542194366455,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 137.0222930908203,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.976522445678711,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6651917099952698,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3333132863044739,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.617431640625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.13846492767334,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.54167366027832,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.52328109741211,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 38.568050384521484,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.97627067565918,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5607180595397949,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.333417147397995,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.558518409729004,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.022800187908210083,
+ "step": 41,
+ "teaching_rms": 0.005129964994619575,
+ "traffic_rms": 0.022727280865286068
+ },
+ {
+ "batch_loss": 2.1183021068573,
+ "innovation_rms": 0.002951957092535396,
+ "instruction_rms": 0.0014796906295925982,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.810994148254395,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 409.62493896484375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.10142707824707,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6583411693572998,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27649450302124023,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.2480177879333496,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.054064750671387,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.155994415283203,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.067380905151367,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35.44099807739258,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.101175308227539,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5507796406745911,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2766783535480499,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.2708704471588135,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.025141185176954264,
+ "step": 42,
+ "teaching_rms": 0.002951957092535396,
+ "traffic_rms": 0.025098144686580656
+ },
+ {
+ "batch_loss": 2.0149037837982178,
+ "innovation_rms": 0.002735980525409247,
+ "instruction_rms": 0.001401138792865662,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.537160873413086,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1560.9248046875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.859293937683105,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6572230458259583,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3521784842014313,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.102819442749023,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.906126976013184,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.520625114440918,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.28163528442383,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.142539978027344,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.859041213989258,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5349798798561096,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3523625433444977,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.356774806976318,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.026775218639381555,
+ "step": 43,
+ "teaching_rms": 0.002735980525409247,
+ "traffic_rms": 0.026739451422194555
+ },
+ {
+ "batch_loss": 2.06754732131958,
+ "innovation_rms": 0.002624742829125229,
+ "instruction_rms": 0.0012474109555980965,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.771754264831543,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4846.61962890625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.83598518371582,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6618534326553345,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4002056121826172,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.086418151855469,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.696064949035645,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.899374961853027,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36.208534240722656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29.26897430419922,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.835733413696289,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5337828397750854,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.40009379386901855,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.340348243713379,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.02855988130603714,
+ "step": 44,
+ "teaching_rms": 0.002624742829125229,
+ "traffic_rms": 0.028531371603229192
+ },
+ {
+ "batch_loss": 1.9831970930099487,
+ "innovation_rms": 0.0027614787270841506,
+ "instruction_rms": 0.0011461203688579564,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.775663375854492,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9412.7900390625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.881804466247559,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6568214297294617,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.36849328875541687,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.974598407745361,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.431102752685547,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.350380897521973,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 38.881343841552734,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.728097915649414,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.881551742553711,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5441074371337891,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.36832907795906067,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.228503227233887,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.030177513522591366,
+ "step": 45,
+ "teaching_rms": 0.0027614787270841506,
+ "traffic_rms": 0.030156916575088056
+ },
+ {
+ "batch_loss": 2.220123529434204,
+ "innovation_rms": 0.003499385905151872,
+ "instruction_rms": 0.0014088857436442297,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31.53508758544922,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14838.7705078125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.055091857910156,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6429234743118286,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2765609920024872,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.780107498168945,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.120888710021973,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.454262733459473,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.316864013671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24.355188369750977,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.054839134216309,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5347049832344055,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27649977803230286,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.033987045288086,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.03171000101192745,
+ "step": 46,
+ "teaching_rms": 0.003499385905151872,
+ "traffic_rms": 0.031682429287722524
+ },
+ {
+ "batch_loss": 2.0825812816619873,
+ "innovation_rms": 0.003866424422555478,
+ "instruction_rms": 0.001320558508569626,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43.48346710205078,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19871.623046875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.3899383544921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6304598450660706,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.21657994389533997,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.519101142883301,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.779350280761719,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.771522045135498,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43.56111145019531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.442481994628906,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.389685153961182,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5257382988929749,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.21682511270046234,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.772955417633057,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.03326161655382081,
+ "step": 47,
+ "teaching_rms": 0.003866424422555478,
+ "traffic_rms": 0.03323720140323781
+ },
+ {
+ "batch_loss": 1.9411752223968506,
+ "innovation_rms": 0.0045724994552830975,
+ "instruction_rms": 0.0013657805311849656,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.04855728149414,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24655.91796875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.740541458129883,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.62469482421875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1743798404932022,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.193155288696289,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.420040130615234,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.33426570892334,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 45.66083526611328,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.021343231201172,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.740288257598877,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5254156589508057,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1744949221611023,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.44698429107666,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.034664515405893326,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.03495245347428118,
+ "step": 48,
+ "teaching_rms": 0.0045724994552830975,
+ "traffic_rms": 0.034925786134215424
+ },
+ {
+ "batch_loss": 2.184627056121826,
+ "innovation_rms": 0.005420115623403887,
+ "instruction_rms": 0.0019650060232990017,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.28936004638672,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29372.66015625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.231871128082275,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6241801977157593,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2581639289855957,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.812821388244629,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.032024383544922,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.844401359558105,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47.6019287109375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.27933120727539,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.23192834854126,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5299937129020691,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.25827908515930176,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.066625595092773,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 2.5517491571477293e-05,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.0371751262272479,
+ "step": 49,
+ "teaching_rms": 0.005420115623403887,
+ "traffic_rms": 0.037122664049294
+ },
+ {
+ "batch_loss": 2.167476177215576,
+ "innovation_rms": 0.004648903859758439,
+ "instruction_rms": 0.0016945470099312767,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 81.49446868896484,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32677.076171875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.902511119842529,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6281175017356873,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.37021586298942566,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.382397651672363,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.623291015625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.330876350402832,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.4387321472168,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.145946502685547,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.902568340301514,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.544355034828186,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3702772259712219,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.636176109313965,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.03899368342774569,
+ "step": 50,
+ "teaching_rms": 0.004648903859758439,
+ "traffic_rms": 0.03895624964529248
+ },
+ {
+ "batch_loss": 2.10048770904541,
+ "innovation_rms": 0.005490770513520117,
+ "instruction_rms": 0.0016777067148821632,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94.04544067382812,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35629.68359375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.700067520141602,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.625889778137207,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3031086027622223,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.908103942871094,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.202095985412598,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.100746154785156,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.17665481567383,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23.711755752563477,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.700113296508789,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5709730386734009,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.30317002534866333,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.161856651306152,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.0413203442372364,
+ "step": 51,
+ "teaching_rms": 0.005490770513520117,
+ "traffic_rms": 0.04128572501293771
+ },
+ {
+ "batch_loss": 2.0741970539093018,
+ "innovation_rms": 0.006841989136812824,
+ "instruction_rms": 0.0023904357430147745,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.22465515136719,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37867.7734375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.738127708435059,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.622815728187561,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2285844385623932,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.399454116821289,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.765153884887695,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.074268341064453,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.792640686035156,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.25682830810547,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.738235473632812,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5927773118019104,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22907042503356934,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.653181076049805,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.04480536464608587,
+ "step": 52,
+ "teaching_rms": 0.006841989136812824,
+ "traffic_rms": 0.04474038086496156
+ },
+ {
+ "batch_loss": 2.165985584259033,
+ "innovation_rms": 0.0068542671980913,
+ "instruction_rms": 0.0022540302828774035,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.30413055419922,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 48784.01171875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.90695858001709,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.619274377822876,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.15705493092536926,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.851248741149902,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15.830488204956055,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.636003494262695,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54.29010009765625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.18691062927246,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.907066822052002,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6063031554222107,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.15697839856147766,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.104950904846191,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.049027635905685145,
+ "step": 53,
+ "teaching_rms": 0.0068542671980913,
+ "traffic_rms": 0.04897908801142792
+ },
+ {
+ "batch_loss": 1.909328579902649,
+ "innovation_rms": 0.008773025838657541,
+ "instruction_rms": 0.0017726973371559535,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 133.2132568359375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61003.24609375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.328192710876465,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6176231503486633,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2035539299249649,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.265533447265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.3225154876709,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.920268058776855,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.6806526184082,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 38.067665100097656,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.328350067138672,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6082965135574341,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20355036854743958,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.519209861755371,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.053615947770543695,
+ "step": 54,
+ "teaching_rms": 0.008773025838657541,
+ "traffic_rms": 0.05358646151834602
+ },
+ {
+ "batch_loss": 1.955823302268982,
+ "innovation_rms": 0.008105473946378255,
+ "instruction_rms": 0.001323462681213481,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.8180694580078,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74070.375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.865716934204102,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6113687753677368,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3338838517665863,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.64564037322998,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.07297134399414,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.504566192626953,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.97142791748047,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46.90654373168945,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.86596965789795,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.610686719417572,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3337692320346832,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.8992919921875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.05672090889917317,
+ "step": 55,
+ "teaching_rms": 0.008105473946378255,
+ "traffic_rms": 0.05670490649960291
+ },
+ {
+ "batch_loss": 2.062808036804199,
+ "innovation_rms": 0.00903515924683132,
+ "instruction_rms": 0.0013554101918051187,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157.97601318359375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86031.03125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.591849327087402,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6060581207275391,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.28502556681632996,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.993281364440918,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.116300582885742,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.433290481567383,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60.94546890258789,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57.06481170654297,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14.591830253601074,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6090942025184631,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2849109470844269,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.246907234191895,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.060652591821530835,
+ "step": 56,
+ "teaching_rms": 0.00903515924683132,
+ "traffic_rms": 0.060636953713783685
+ },
+ {
+ "batch_loss": 2.1083431243896484,
+ "innovation_rms": 0.011159874124058014,
+ "instruction_rms": 0.0011888566922161922,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.91371154785156,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98787.7109375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.154611587524414,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6044140458106995,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22934368252754211,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.312887191772461,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23.55345344543457,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24.37152862548828,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 68.31675720214844,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73.71290588378906,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.15441131591797,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6075281500816345,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.22947824001312256,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.566487312316895,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.06560065851605605,
+ "step": 57,
+ "teaching_rms": 0.011159874124058014,
+ "traffic_rms": 0.06558895139992797
+ },
+ {
+ "batch_loss": 2.138840436935425,
+ "innovation_rms": 0.011212803983990306,
+ "instruction_rms": 0.001200115388915514,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 181.47586059570312,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110812.34375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.381765365600586,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6033912897109985,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.26858121156692505,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.613241195678711,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.33758544921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27.8413143157959,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.9240951538086,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.07341003417969,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19.381933212280273,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6075416207313538,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2687157988548279,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.866815567016602,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.07031084944813094,
+ "step": 58,
+ "teaching_rms": 0.011212803983990306,
+ "traffic_rms": 0.07029984979347467
+ },
+ {
+ "batch_loss": 1.9688163995742798,
+ "innovation_rms": 0.016004864480133205,
+ "instruction_rms": 0.0011652585452782706,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 192.6074676513672,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 119876.0703125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.403688430786133,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6002151966094971,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.40579813718795776,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.896987915039062,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29.718576431274414,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.80991744995117,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88.53753662109375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 116.13439178466797,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.403648376464844,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.60763019323349,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4060295522212982,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.15053653717041,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.07563021243988294,
+ "step": 59,
+ "teaching_rms": 0.016004864480133205,
+ "traffic_rms": 0.07562074301106612
+ },
+ {
+ "batch_loss": 1.8338898420333862,
+ "innovation_rms": 0.018945460871524173,
+ "instruction_rms": 0.001289803491752858,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 203.00831604003906,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127343.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23.516315460205078,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5975149273872375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2735362648963928,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.161669731140137,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.72859573364258,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.1002082824707,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 103.1207275390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.83193969726562,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23.516178131103516,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6053431034088135,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27376770973205566,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.415192604064941,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.08133528239982217,
+ "step": 60,
+ "teaching_rms": 0.018945460871524173,
+ "traffic_rms": 0.08132480337780036
+ },
+ {
+ "batch_loss": 2.310033082962036,
+ "innovation_rms": 0.023051502820084783,
+ "instruction_rms": 0.0014982873546807139,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 212.2045135498047,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 135014.78125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.25456237792969,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5957058072090149,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2876763343811035,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.950289726257324,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 38.539859771728516,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 48.11262893676758,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122.10171508789062,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 189.8098602294922,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.25456237792969,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.595278263092041,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2876116931438446,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.866286277770996,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.08828280656267228,
+ "step": 61,
+ "teaching_rms": 0.023051502820084783,
+ "traffic_rms": 0.08827067666711817
+ },
+ {
+ "batch_loss": 1.934715986251831,
+ "innovation_rms": 0.015426241684492262,
+ "instruction_rms": 0.0012456050235322552,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 221.028564453125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140009.421875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.25745391845703,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5917008519172668,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3218923807144165,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16.622678756713867,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43.4822998046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.42438507080078,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140.51246643066406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 184.10748291015625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.25745391845703,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5976938605308533,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.32182776927948,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16.369205474853516,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.08744499878636605,
+ "step": 62,
+ "teaching_rms": 0.015426241684492262,
+ "traffic_rms": 0.08743669872472609
+ },
+ {
+ "batch_loss": 2.130141258239746,
+ "innovation_rms": 0.01644073814809705,
+ "instruction_rms": 0.0016062666268935762,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 227.8116912841797,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 141200.71875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.6978874206543,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.587117075920105,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2967767119407654,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.90494155883789,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 48.49421691894531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 50.11915588378906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 158.10057067871094,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175.88104248046875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.69788360595703,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6044679284095764,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.29671210050582886,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.6514949798584,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.0935400847637206,
+ "step": 63,
+ "teaching_rms": 0.01644073814809705,
+ "traffic_rms": 0.09352673588965851
+ },
+ {
+ "batch_loss": 2.0911762714385986,
+ "innovation_rms": 0.021432630715508078,
+ "instruction_rms": 0.0018273850066030752,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233.4797821044922,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 141017.734375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.57431411743164,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5787665843963623,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20179972052574158,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24.693313598632812,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.62386703491211,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.29651641845703,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175.07020568847656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.6962890625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.574310302734375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6129800081253052,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20173512399196625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24.439891815185547,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.1464596837759018,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.10153289271767962,
+ "step": 64,
+ "teaching_rms": 0.021432630715508078,
+ "traffic_rms": 0.10151643278849748
+ },
+ {
+ "batch_loss": 2.4405720233917236,
+ "innovation_rms": 0.024026229770380388,
+ "instruction_rms": 0.0017331503693605102,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 276.6404113769531,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 141970.65625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.606407165527344,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5754231214523315,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2690553069114685,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28.105710983276367,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.979129791259766,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.552616119384766,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191.85145568847656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167.81256103515625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.60627365112305,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6194607019424438,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2690582871437073,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27.852313995361328,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 0.0005742584990645012,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.109565195804465,
+ "step": 65,
+ "teaching_rms": 0.024026229770380388,
+ "traffic_rms": 0.109551578043064
+ },
+ {
+ "batch_loss": 2.1966357231140137,
+ "innovation_rms": 0.020314257907502523,
+ "instruction_rms": 0.0017367750069105735,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 351.6284484863281,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 181466.15625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.94084167480469,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5768557190895081,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27829739451408386,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31.090808868408203,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64.38126373291016,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54.02133560180664,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 208.00498962402344,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 161.5353546142578,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.940757751464844,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6226761937141418,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27830037474632263,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.83743667602539,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.1162157813527165,
+ "step": 66,
+ "teaching_rms": 0.020314257907502523,
+ "traffic_rms": 0.11620296567512042
+ },
+ {
+ "batch_loss": 2.2899105548858643,
+ "innovation_rms": 0.02367071903984054,
+ "instruction_rms": 0.0018553008472907404,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 434.93096923828125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 328448.84375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.21030044555664,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5847180485725403,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3573927879333496,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.70689010620117,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.77857208251953,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.97309494018555,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 223.2581787109375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.0856170654297,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.210296630859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6222277879714966,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35699647665023804,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33.45354461669922,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.1237314109129431,
+ "step": 67,
+ "teaching_rms": 0.02367071903984054,
+ "traffic_rms": 0.12371785895317305
+ },
+ {
+ "batch_loss": 2.1973819732666016,
+ "innovation_rms": 0.02362025072164488,
+ "instruction_rms": 0.0014989647304078488,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 516.864013671875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 488706.0625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.509033203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5877630710601807,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3358531892299652,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35.96508026123047,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.19926452636719,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54.20695495605469,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 237.73971557617188,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 178.15213012695312,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.508949279785156,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6228611469268799,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.33545687794685364,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35.711761474609375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.13238737348753743,
+ "step": 68,
+ "teaching_rms": 0.02362025072164488,
+ "traffic_rms": 0.13237877082179988
+ },
+ {
+ "batch_loss": 2.144948720932007,
+ "innovation_rms": 0.035168302448189456,
+ "instruction_rms": 0.0015326875375211483,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 597.361083984375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 619578.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.3314208984375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5848925709724426,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2915516197681427,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.15971374511719,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 80.75312805175781,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.538612365722656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 251.80055236816406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 232.09939575195312,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.33121490478516,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6264827251434326,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2915479838848114,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.244571685791016,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.14422385958112982,
+ "step": 69,
+ "teaching_rms": 0.035168302448189456,
+ "traffic_rms": 0.14421560770973438
+ },
+ {
+ "batch_loss": 2.2181198596954346,
+ "innovation_rms": 0.03055994738719929,
+ "instruction_rms": 0.0015397416090407968,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 671.7391357421875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 708965.75,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.2371826171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5817073583602905,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2550879716873169,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46.1738166809082,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.56404113769531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.10914611816406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 265.8994140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228.40960693359375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.23696899414062,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6327337026596069,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2550843358039856,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46.258663177490234,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.14786247228388458,
+ "step": 70,
+ "teaching_rms": 0.03055994738719929,
+ "traffic_rms": 0.14785362991012302
+ },
+ {
+ "batch_loss": 2.295954465866089,
+ "innovation_rms": 0.030817208948726003,
+ "instruction_rms": 0.001382648425383141,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 747.18359375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 878752.9375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82.12838745117188,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5779213309288025,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20458054542541504,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.681583404541016,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.5521011352539,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.88063430786133,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 279.744873046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233.04156494140625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82.128173828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6351819634437561,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.20471301674842834,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.766422271728516,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.15505052957518348,
+ "step": 71,
+ "teaching_rms": 0.030817208948726003,
+ "traffic_rms": 0.15504447105152466
+ },
+ {
+ "batch_loss": 2.1647696495056152,
+ "innovation_rms": 0.031191667966970715,
+ "instruction_rms": 0.0011596139466605237,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 817.3475952148438,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1030900.125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.05909729003906,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5768850445747375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3909012973308563,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.72922897338867,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98.78349304199219,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 62.31390380859375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 293.57696533203125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 249.7273406982422,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.05887603759766,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6409686803817749,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.39069652557373047,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.814056396484375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.1688425551747201,
+ "step": 72,
+ "teaching_rms": 0.031191667966970715,
+ "traffic_rms": 0.16883843848834307
+ },
+ {
+ "batch_loss": 2.222933769226074,
+ "innovation_rms": 0.04298763799764033,
+ "instruction_rms": 0.0011495505933842105,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 870.3493041992188,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1037879.75,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91.10125732421875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.582819402217865,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.2730547785758972,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64.3931884765625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105.55812072753906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.74629211425781,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308.6510925292969,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 324.028564453125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91.10124206542969,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6467307806015015,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.27285003662109375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64.4780044555664,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.18708129244180396,
+ "step": 73,
+ "teaching_rms": 0.04298763799764033,
+ "traffic_rms": 0.18707769853901693
+ },
+ {
+ "batch_loss": 2.164555549621582,
+ "innovation_rms": 0.0362830609762782,
+ "instruction_rms": 0.0010849826778216718,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 929.0654296875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1228772.125,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66.46115112304688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.586223840713501,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35348862409591675,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.61058807373047,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112.52466583251953,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.6654281616211,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 323.4552307128906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 312.9791259765625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66.46112060546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6416953206062317,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35344940423965454,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.69539642333984,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.1841427742224879,
+ "step": 74,
+ "teaching_rms": 0.0362830609762782,
+ "traffic_rms": 0.18413961711041882
+ },
+ {
+ "batch_loss": 2.2325124740600586,
+ "innovation_rms": 0.04072368955716243,
+ "instruction_rms": 0.0010395588662053242,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 975.428955078125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1340142.375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.87126922607422,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5940329432487488,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35207340121269226,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.39981079101562,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 119.80200958251953,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 72.77342987060547,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 338.4229736328125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 302.13031005859375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59.871238708496094,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6316984295845032,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3522058427333832,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.48461151123047,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.1968657176494399,
+ "step": 75,
+ "teaching_rms": 0.04072368955716243,
+ "traffic_rms": 0.19686273808450533
+ },
+ {
+ "batch_loss": 2.112586498260498,
+ "innovation_rms": 0.0468078412935187,
+ "instruction_rms": 0.0009561888307314549,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1010.5224609375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1505050.625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.7944564819336,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.601161003112793,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35902512073516846,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78.81360626220703,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127.54849243164062,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.46479797363281,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 354.04766845703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 304.8291015625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.79454040527344,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6214930415153503,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.35904932022094727,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78.89839935302734,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.21496656916637863,
+ "step": 76,
+ "teaching_rms": 0.0468078412935187,
+ "traffic_rms": 0.2149643411805441
+ },
+ {
+ "batch_loss": 2.143122911453247,
+ "innovation_rms": 0.05155829026566517,
+ "instruction_rms": 0.0009432132610974052,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1033.042724609375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1884185.5,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 144.04995727539062,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6028051972389221,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3587197959423065,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82.89802551269531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 135.95970153808594,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 84.11204528808594,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 371.1506652832031,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 344.5229187011719,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 144.050048828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6151719689369202,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.3586321175098419,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82.9828109741211,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.2376305682721329,
+ "step": 77,
+ "teaching_rms": 0.05155829026566517,
+ "traffic_rms": 0.23762866895511864
+ },
+ {
+ "batch_loss": 2.019437313079834,
+ "innovation_rms": 0.06750500327339162,
+ "instruction_rms": 0.0010765712307637938,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1036.66650390625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2089120.375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100.63165283203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5985782742500305,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5197781324386597,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.7525405883789,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.5342254638672,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.74519348144531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 392.7328796386719,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 409.9132080078125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100.63169860839844,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6160281896591187,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5198174715042114,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.83731842041016,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.2791535864835858,
+ "step": 78,
+ "teaching_rms": 0.06750500327339162,
+ "traffic_rms": 0.2791513954772867
+ },
+ {
+ "batch_loss": 2.1966910362243652,
+ "innovation_rms": 0.10033700863457708,
+ "instruction_rms": 0.0013117574563697794,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1013.1668701171875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2122544.25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 199.98887634277344,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5933515429496765,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5691739916801453,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 90.55086517333984,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157.32923889160156,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117.95014190673828,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 431.02557373046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 573.3977661132812,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 199.9889373779297,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6170535683631897,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.569213330745697,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 90.63563537597656,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.3278705755504002,
+ "step": 79,
+ "teaching_rms": 0.10033700863457708,
+ "traffic_rms": 0.3278677105604888
+ },
+ {
+ "batch_loss": 2.18965220451355,
+ "innovation_rms": 0.12258525422210353,
+ "instruction_rms": 0.0011827614084501687,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 961.724609375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2082555.625,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197.6600341796875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5953396558761597,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.48927706480026245,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94.57557678222656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 171.83724975585938,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.08009338378906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 506.1573791503906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 751.3179931640625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197.65985107421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6098588109016418,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.4893164038658142,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94.66033935546875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8856685161590576,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.3615311962555354,
+ "step": 80,
+ "teaching_rms": 0.12258525422210353,
+ "traffic_rms": 0.3615291490051247
+ },
+ {
+ "batch_loss": 2.217668056488037,
+ "innovation_rms": 0.15360248371435553,
+ "instruction_rms": 0.0013545042322846382,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 912.9450073242188,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2041051.75,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258.4854736328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.607008695602417,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5756626129150391,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 114.95745086669922,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 190.345458984375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 185.08213806152344,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 597.2694091796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 911.1200561523438,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258.48565673828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5909577012062073,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5757495164871216,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 114.76780700683594,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 0.023591889351786778,
+ "predictor_updated": true,
+ "raw_apical_rms": 0.4215401336817219,
+ "step": 81,
+ "teaching_rms": 0.15360248371435553,
+ "traffic_rms": 0.4215379055130173
+ },
+ {
+ "batch_loss": 2.1351571083068848,
+ "innovation_rms": 0.15006985624878177,
+ "instruction_rms": 0.0012976697625070215,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 940.8123779296875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2144135.75,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 372.8627014160156,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6261988878250122,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5582478046417236,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 142.4016876220703,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 212.0081787109375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 216.6272430419922,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 706.622802734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1093.533935546875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 372.8627624511719,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5958163738250732,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5582340359687805,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 142.21206665039062,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.48048031319157225,
+ "step": 82,
+ "teaching_rms": 0.15006985624878177,
+ "traffic_rms": 0.48047847999250953
+ },
+ {
+ "batch_loss": 2.238063097000122,
+ "innovation_rms": 0.0991057749981345,
+ "instruction_rms": 0.0011528285570985475,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 977.873046875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2071537.875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 305.49932861328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6422209739685059,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5078710913658142,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167.86207580566406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233.9481658935547,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219.39991760253906,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 810.8685913085938,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1042.4580078125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 305.4993896484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6044920682907104,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5080034732818604,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167.67247009277344,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.5740881986185401,
+ "step": 83,
+ "teaching_rms": 0.0991057749981345,
+ "traffic_rms": 0.5740870278184781
+ },
+ {
+ "batch_loss": 2.176492691040039,
+ "innovation_rms": 0.16056316384767053,
+ "instruction_rms": 0.0011096235997764175,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 988.9700317382812,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1952339.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258.6731872558594,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6590030789375305,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7902647256851196,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 193.72940063476562,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 257.90557861328125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 239.573974609375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 917.2728271484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1120.292724609375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258.67340087890625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6104620695114136,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7902255058288574,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 193.53981018066406,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.6551347021609294,
+ "step": 84,
+ "teaching_rms": 0.16056316384767053,
+ "traffic_rms": 0.6551337714869452
+ },
+ {
+ "batch_loss": 2.416860818862915,
+ "innovation_rms": 0.2032459858655871,
+ "instruction_rms": 0.0013231373552311714,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 967.7470092773438,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1827308.875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 380.9134521484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6684628129005432,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6335961818695068,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 220.56666564941406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 285.466552734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 275.6095886230469,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1025.2755126953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1360.3079833984375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 380.91326904296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6080571413040161,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6336545944213867,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 220.37710571289062,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.750846207651044,
+ "step": 85,
+ "teaching_rms": 0.2032459858655871,
+ "traffic_rms": 0.7508450917377039
+ },
+ {
+ "batch_loss": 2.328747272491455,
+ "innovation_rms": 0.2537151850633394,
+ "instruction_rms": 0.001194178612744988,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 925.1856079101562,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1695066.5,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 781.7706909179688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6717215776443481,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7590739727020264,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 247.63160705566406,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 318.47198486328125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 330.0542907714844,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1180.3896484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1723.962158203125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 781.7706298828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.599278450012207,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7589863538742065,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 247.4420623779297,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 0.855260589870617,
+ "step": 86,
+ "teaching_rms": 0.2537151850633394,
+ "traffic_rms": 0.8552597838429511
+ },
+ {
+ "batch_loss": 2.1462583541870117,
+ "innovation_rms": 0.32244929469096356,
+ "instruction_rms": 0.0012199727960697994,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 824.8753662109375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1607032.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.368408203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6844914555549622,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1322641372680664,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 275.62274169921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 359.6727600097656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 412.00775146484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1418.942138671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2385.525146484375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.368408203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5971534252166748,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1321766376495361,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 275.4332275390625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1.0034681525587874,
+ "step": 87,
+ "teaching_rms": 0.32244929469096356,
+ "traffic_rms": 1.0034674408493789
+ },
+ {
+ "batch_loss": 2.3443851470947266,
+ "innovation_rms": 0.3783060823712267,
+ "instruction_rms": 0.0011524885283502714,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 764.659912109375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1588797.875,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 720.7196044921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7074704170227051,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5794150233268738,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 302.0154113769531,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 409.411376953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 497.3861999511719,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1722.2010498046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3032.58935546875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 720.719482421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6085814237594604,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5795018672943115,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 301.8258972167969,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1.1974156668435803,
+ "step": 88,
+ "teaching_rms": 0.3783060823712267,
+ "traffic_rms": 1.197415059865353
+ },
+ {
+ "batch_loss": 2.3994855880737305,
+ "innovation_rms": 0.42632160974128513,
+ "instruction_rms": 0.0012468787679541214,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 895.7922973632812,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1745265.375,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.727294921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7324374914169312,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.060755729675293,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 339.9569396972656,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 468.8896484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 594.7828369140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2085.711669921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3635.10498046875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.72705078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6238681674003601,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0608432292938232,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 340.08587646484375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1.3314423821876749,
+ "step": 89,
+ "teaching_rms": 0.42632160974128513,
+ "traffic_rms": 1.3314417588576086
+ },
+ {
+ "batch_loss": 2.290465831756592,
+ "innovation_rms": 0.3037356704575101,
+ "instruction_rms": 0.001346935422543042,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745.35009765625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2667787.75,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 951.1387939453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.77045738697052,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3626724481582642,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 414.6609802246094,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527.939697265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 590.500244140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2428.69677734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3429.850830078125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 951.1385498046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6444383859634399,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3627599477767944,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 414.7898864746094,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1.514360221673792,
+ "step": 90,
+ "teaching_rms": 0.3037356704575101,
+ "traffic_rms": 1.5143594836045442
+ },
+ {
+ "batch_loss": 2.1979851722717285,
+ "innovation_rms": 0.340028992653145,
+ "instruction_rms": 0.0012900615788502862,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 865.9905395507812,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5467646.5,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 796.6319580078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8220745325088501,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2668670415878296,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 483.9232177734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 586.3785400390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 584.3886108398438,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2750.030517578125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3213.337646484375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 796.6317138671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6899359822273254,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2669546604156494,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 484.0521240234375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1.7680210134521388,
+ "step": 91,
+ "teaching_rms": 0.340028992653145,
+ "traffic_rms": 1.7680204282739935
+ },
+ {
+ "batch_loss": 2.349238395690918,
+ "innovation_rms": 0.44528443569283616,
+ "instruction_rms": 0.0012438194500479175,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 997.0281372070312,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10248871.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.6923217773438,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.823173999786377,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7876071929931641,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 550.243896484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 646.8541870117188,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 604.7561645507812,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3063.8193359375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3137.886962890625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.692138671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6955245733261108,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7876946926116943,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 550.372802734375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2.0418802068018684,
+ "step": 92,
+ "teaching_rms": 0.44528443569283616,
+ "traffic_rms": 2.0418799158837744
+ },
+ {
+ "batch_loss": 2.4867260456085205,
+ "innovation_rms": 0.5356692206792382,
+ "instruction_rms": 0.001113609853911327,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1440.202880859375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16611360.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 695.8977661132812,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.799161434173584,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0353561639785767,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 619.0330200195312,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 711.6090087890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 647.548095703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3373.465087890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3715.53076171875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 695.8975219726562,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7104668021202087,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.035191297531128,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 619.1619262695312,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2.229240675768782,
+ "step": 93,
+ "teaching_rms": 0.5356692206792382,
+ "traffic_rms": 2.229240457671836
+ },
+ {
+ "batch_loss": 2.4813480377197266,
+ "innovation_rms": 0.6508469171252628,
+ "instruction_rms": 0.0011023811688492641,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2043.39599609375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22446440.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1097.430908203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7925569415092468,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5441315174102783,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.3995971679688,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 783.3268432617188,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 717.1784057617188,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3693.419677734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4491.61767578125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1097.4310302734375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.709028422832489,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.5441899299621582,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 691.5285034179688,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2.510457767687137,
+ "step": 94,
+ "teaching_rms": 0.6508469171252628,
+ "traffic_rms": 2.5104575507841727
+ },
+ {
+ "batch_loss": 2.7437257766723633,
+ "innovation_rms": 0.857481096154837,
+ "instruction_rms": 0.0015527622495360132,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2318.647216796875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26189258.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1698.5147705078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8006237745285034,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4708662033081055,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 762.0848999023438,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 868.4046630859375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 850.7783813476562,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4067.5986328125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5462.5732421875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1698.5146484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6738852858543396,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4707787036895752,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 762.2138061523438,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2.856063057870281,
+ "step": 95,
+ "teaching_rms": 0.857481096154837,
+ "traffic_rms": 2.8560626961161772
+ },
+ {
+ "batch_loss": 2.4597184658050537,
+ "innovation_rms": 1.110497796906918,
+ "instruction_rms": 0.0012712606993432235,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2291.801025390625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39214728.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1823.5810546875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8431665301322937,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0053348541259766,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 831.8360595703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 971.1906127929688,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1027.8594970703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4534.63134765625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6699.669921875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1823.5810546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.711094081401825,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.005247116088867,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 831.9649658203125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.859949588775635,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3.1818432150734037,
+ "step": 96,
+ "teaching_rms": 1.110497796906918,
+ "traffic_rms": 3.1818429366604946
+ },
+ {
+ "batch_loss": 2.1501476764678955,
+ "innovation_rms": 1.4216190184242123,
+ "instruction_rms": 0.0012270292341346307,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1919.141845703125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 70576960.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3495.75244140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.875916600227356,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5944384336471558,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 899.4194946289062,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1103.485107421875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1322.944580078125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5206.1796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8261.69921875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3495.75244140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7438573837280273,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5943505764007568,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 899.5484008789062,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 2.0209991321013336,
+ "predictor_updated": true,
+ "raw_apical_rms": 3.656212536034317,
+ "step": 97,
+ "teaching_rms": 1.4216190184242123,
+ "traffic_rms": 3.656212305601598
+ },
+ {
+ "batch_loss": 2.620966672897339,
+ "innovation_rms": 1.2464244780429912,
+ "instruction_rms": 0.0015138446407044978,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1944.3289794921875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 108237152.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2299.4697265625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8574458360671997,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6258525848388672,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1024.720947265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1256.475341796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1529.90234375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5983.9501953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9606.068359375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2299.469482421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7253998517990112,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.6258285641670227,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1024.8497314453125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4.487926814214475,
+ "step": 98,
+ "teaching_rms": 1.2464244780429912,
+ "traffic_rms": 4.487926661521476
+ },
+ {
+ "batch_loss": 2.8113272190093994,
+ "innovation_rms": 1.5516075505939397,
+ "instruction_rms": 0.0014020250496828734,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2176.20654296875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197473856.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2202.65576171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8109462857246399,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.758334159851074,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1244.6353759765625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1433.910400390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1774.35107421875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6907.23095703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11319.91796875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2202.655517578125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7106890082359314,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.7584221363067627,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1244.76416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5.378576423103973,
+ "step": 99,
+ "teaching_rms": 1.5516075505939397,
+ "traffic_rms": 5.3785762455035
+ },
+ {
+ "batch_loss": 2.0472519397735596,
+ "innovation_rms": 1.604016085942292,
+ "instruction_rms": 0.0015382625811004957,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2413.76171875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 315425824.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2760.589599609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7834523916244507,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.2177839279174805,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1456.0850830078125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1632.145751953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1982.35302734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8148.3447265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13052.85546875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2760.58935546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7491268515586853,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.217871904373169,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1456.2138671875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5.953939475232596,
+ "step": 100,
+ "teaching_rms": 1.604016085942292,
+ "traffic_rms": 5.9539392744251565
+ },
+ {
+ "batch_loss": 3.0180697441101074,
+ "innovation_rms": 1.7873479220421948,
+ "instruction_rms": 0.002107559983585464,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3344.232666015625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 401582912.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1971.9208984375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7985683679580688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9382182359695435,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1653.2772216796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1847.4134521484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2206.486328125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9669.5791015625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15212.3427734375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1971.9207763671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7132946848869324,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9382421970367432,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1653.406005859375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6.836476206182247,
+ "step": 101,
+ "teaching_rms": 1.7873479220421948,
+ "traffic_rms": 6.836475913606616
+ },
+ {
+ "batch_loss": 2.8296351432800293,
+ "innovation_rms": 2.228832937360745,
+ "instruction_rms": 0.002084619185754151,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4937.97314453125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 624461824.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2372.458251953125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8184917569160461,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0599780082702637,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1823.173095703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2084.038330078125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2612.863525390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11477.2548828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18076.76171875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2372.458251953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7022076845169067,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0601427555084229,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1823.3018798828125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8.127145497764827,
+ "step": 102,
+ "teaching_rms": 2.228832937360745,
+ "traffic_rms": 8.127145259564115
+ },
+ {
+ "batch_loss": 2.6632587909698486,
+ "innovation_rms": 2.9288757340111204,
+ "instruction_rms": 0.0018562231454161034,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6341.95849609375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1122117504.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4395.87451171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8339239954948425,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8607509136199951,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1924.682373046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2361.8349609375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3092.669677734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13656.5341796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21792.796875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4395.87451171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7141236066818237,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8609155416488647,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1924.8111572265625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 9.283686991602789,
+ "step": 103,
+ "teaching_rms": 2.9288757340111204,
+ "traffic_rms": 9.28368681855184
+ },
+ {
+ "batch_loss": 2.9444141387939453,
+ "innovation_rms": 3.6895724078524528,
+ "instruction_rms": 0.0020187410979078894,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7688.21533203125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1885203072.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6007.0830078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8476161360740662,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4281933307647705,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1994.7490234375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2706.369140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3682.7099609375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16290.0546875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26335.20703125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6007.08349609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7500485777854919,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.428105592727661,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1994.8778076171875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 10.862873879801272,
+ "step": 104,
+ "teaching_rms": 3.6895724078524528,
+ "traffic_rms": 10.862873720416747
+ },
+ {
+ "batch_loss": 2.6947290897369385,
+ "innovation_rms": 4.017002979585962,
+ "instruction_rms": 0.0016833769227208395,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9667.9287109375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2887171328.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6501.16064453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8489416241645813,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9427778720855713,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2392.143310546875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3103.4794921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4373.66796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19435.376953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31453.23046875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6501.16015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7657685279846191,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9426900148391724,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2392.27197265625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 12.44155855767709,
+ "step": 105,
+ "teaching_rms": 4.017002979585962,
+ "traffic_rms": 12.441558445829804
+ },
+ {
+ "batch_loss": 4.7782979011535645,
+ "innovation_rms": 4.569659009582034,
+ "instruction_rms": 0.001917948572891992,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12364.7607421875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3792890624.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10303.90234375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.8287279605865479,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.686251640319824,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2983.379150390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3554.064208984375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5220.478515625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23204.505859375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37691.296875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10303.90234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.7978639006614685,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.686339378356934,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2983.5078125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 14.541453983243583,
+ "step": 106,
+ "teaching_rms": 4.569659009582034,
+ "traffic_rms": 14.541453883823007
+ },
+ {
+ "batch_loss": 5.574229717254639,
+ "innovation_rms": 6.088966590401986,
+ "instruction_rms": 0.0023222323066953636,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14677.7412109375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6803484672.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10361.041015625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4171088933944702,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.740379333496094,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3624.7431640625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4110.76708984375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6213.34033203125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27729.251953125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 45247.453125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10361.041015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3297100067138672,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.740466117858887,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3624.690185546875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 17.575071285100996,
+ "step": 107,
+ "teaching_rms": 6.088966590401986,
+ "traffic_rms": 17.575071163050985
+ },
+ {
+ "batch_loss": 7.9123640060424805,
+ "innovation_rms": 7.556105460056522,
+ "instruction_rms": 0.0047106100209058175,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17183.123046875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11251217408.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10888.9443359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8329648971557617,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.158559799194336,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4690.12646484375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4803.78076171875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7437.5283203125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33191.46875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54622.16796875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10888.9443359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7455745935440063,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.158646106719971,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4690.0732421875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 21.233427210837753,
+ "step": 108,
+ "teaching_rms": 7.556105460056522,
+ "traffic_rms": 21.23342671110632
+ },
+ {
+ "batch_loss": 9.464959144592285,
+ "innovation_rms": 9.095928299888127,
+ "instruction_rms": 0.005073874217734828,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20558.04296875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19150190592.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11796.45703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.053941011428833,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.552899360656738,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5828.2578125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5647.40966796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8828.4267578125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39682.64453125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64911.7578125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11796.45703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9665592908859253,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.552695274353027,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5828.20458984375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 24.382135899526773,
+ "step": 109,
+ "teaching_rms": 9.095928299888127,
+ "traffic_rms": 24.382135218196996
+ },
+ {
+ "batch_loss": 11.438346862792969,
+ "innovation_rms": 10.182342892401987,
+ "instruction_rms": 0.005265823483569773,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26400.583984375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32740337664.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12529.5634765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0784902572631836,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.831203460693359,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7081.21435546875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6634.7421875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10705.6591796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47583.52734375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79008.8125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12529.5634765625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9911171197891235,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.831037998199463,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7081.1611328125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 28.686567479290012,
+ "step": 110,
+ "teaching_rms": 10.182342892401987,
+ "traffic_rms": 28.686567038393978
+ },
+ {
+ "batch_loss": 14.117121696472168,
+ "innovation_rms": 13.64039972172638,
+ "instruction_rms": 0.007977974541339344,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32246.779296875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59591122944.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29882.5078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9050095081329346,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.6026811599731445,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8268.9248046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7914.14794921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12881.919921875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57209.44140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 96259.125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29882.5078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8176449537277222,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.602516174316406,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8268.87109375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 36.18939272436381,
+ "step": 111,
+ "teaching_rms": 13.64039972172638,
+ "traffic_rms": 36.189391865486606
+ },
+ {
+ "batch_loss": 13.894920349121094,
+ "innovation_rms": 14.986234741336904,
+ "instruction_rms": 0.005618148314464034,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40153.9921875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102494535680.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34239.015625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.609477162361145,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.448822975158691,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9424.3330078125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9457.4228515625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15432.748046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 68865.15625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 116557.1640625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34239.015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5221213102340698,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.448846817016602,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9424.279296875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.01618957519531,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 44.75747567060569,
+ "step": 112,
+ "teaching_rms": 14.986234741336904,
+ "traffic_rms": 44.757475335274414
+ },
+ {
+ "batch_loss": 13.045845031738281,
+ "innovation_rms": 16.934652964614543,
+ "instruction_rms": 0.006476501276143489,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51772.06640625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 166565265408.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23839.1171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1447389125823975,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.043609619140625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10485.2822265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11325.423828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18680.013671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 83016.03125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 141508.71875,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23839.1171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2590136528015137,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.043647766113281,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10485.228515625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 286.78242963785647,
+ "predictor_updated": true,
+ "raw_apical_rms": 51.098181824768396,
+ "step": 113,
+ "teaching_rms": 16.934652964614543,
+ "traffic_rms": 51.09818143398691
+ },
+ {
+ "batch_loss": 15.704752922058105,
+ "innovation_rms": 16.07396846773465,
+ "instruction_rms": 0.0062020648129239506,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67161.640625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 252891430912.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26316.462890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1693341732025146,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.584365844726562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12492.7451171875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13461.5966796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21361.7265625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 99239.8203125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 162237.890625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26316.462890625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2038772106170654,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.584354400634766,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12492.62109375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 56.89500852177463,
+ "step": 114,
+ "teaching_rms": 16.07396846773465,
+ "traffic_rms": 56.89500820102468
+ },
+ {
+ "batch_loss": 21.14615249633789,
+ "innovation_rms": 21.201593280692787,
+ "instruction_rms": 0.008766515602190623,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85178.671875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 392958345216.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32226.83984375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.421084403991699,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.738245964050293,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15715.4287109375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15953.2548828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24916.5859375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 118251.046875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 190112.234375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32226.83984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.434847116470337,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.738234519958496,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15715.3046875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 69.49266475603234,
+ "step": 115,
+ "teaching_rms": 21.201593280692787,
+ "traffic_rms": 69.49266423159888
+ },
+ {
+ "batch_loss": 29.51326560974121,
+ "innovation_rms": 27.065078421715224,
+ "instruction_rms": 0.011974316434148919,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 108355.8671875,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 662993960960.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47205.9453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.4880921840667725,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.650062561035156,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20436.0234375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18871.453125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29181.98828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140662.3125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 224112.625,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47205.9453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.5018537044525146,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13.649859428405762,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20435.8984375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 85.11740213252877,
+ "step": 116,
+ "teaching_rms": 27.065078421715224,
+ "traffic_rms": 85.11740133474682
+ },
+ {
+ "batch_loss": 23.224014282226562,
+ "innovation_rms": 33.321999930249994,
+ "instruction_rms": 0.01853994690192504,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 131519.703125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1094833405952.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75031.1640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.315685749053955,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.747611999511719,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27939.140625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22328.923828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34574.70703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167175.34375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 265130.34375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75031.1640625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.329445838928223,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.747525215148926,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27939.015625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 98.71648529963448,
+ "step": 117,
+ "teaching_rms": 33.321999930249994,
+ "traffic_rms": 98.71648356889297
+ },
+ {
+ "batch_loss": 43.88111114501953,
+ "innovation_rms": 37.71147579789755,
+ "instruction_rms": 0.017126742241418077,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 161186.03125,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1726278533120.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176760.078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.889939785003662,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.834362983703613,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36413.671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26367.462890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40385.390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197831.796875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 306564.53125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176760.078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.903698444366455,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10.834451675415039,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36413.546875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 110.19258618798648,
+ "step": 118,
+ "teaching_rms": 37.71147579789755,
+ "traffic_rms": 110.19258490134293
+ },
+ {
+ "batch_loss": 60.61956787109375,
+ "innovation_rms": 41.033298538930744,
+ "instruction_rms": 0.03576730640683992,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 193656.734375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2815033868288.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 212209.703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.178188323974609,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.815339088439941,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46166.25390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31056.826171875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46893.62890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 232721.46875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 348896.75,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 212209.703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.191945552825928,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11.815349578857422,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46166.203125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 128.25323868754015,
+ "step": 119,
+ "teaching_rms": 41.033298538930744,
+ "traffic_rms": 128.2532337382272
+ },
+ {
+ "batch_loss": 97.18505096435547,
+ "innovation_rms": 45.27826640649538,
+ "instruction_rms": 0.02098441253704733,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 365797.59375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7845247451136.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196074.515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.070223331451416,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.960845947265625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65773.703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36530.29296875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54734.6640625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 273315.59375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 405941.25,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196074.515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.083979606628418,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.960735321044922,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65773.65625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 153.20396986405206,
+ "step": 120,
+ "teaching_rms": 45.27826640649538,
+ "traffic_rms": 153.20396847713303
+ },
+ {
+ "batch_loss": 120.7970962524414,
+ "innovation_rms": 58.500066144425716,
+ "instruction_rms": 0.056124863130299095,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 495113.34375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19137868857344.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196713.296875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.665016174316406,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.203554153442383,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85445.03125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43062.0546875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65317.62890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 322040.96875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 487253.8125,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196713.296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.678771018981934,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.20344352722168,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85444.984375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 183.19229150663418,
+ "step": 121,
+ "teaching_rms": 58.500066144425716,
+ "traffic_rms": 183.19228296617158
+ },
+ {
+ "batch_loss": 139.6154327392578,
+ "innovation_rms": 70.14583879830953,
+ "instruction_rms": 0.09367637872756546,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 648482.5625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44462826848256.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207222.875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.9525184631347656,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.36603355407715,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106167.3203125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 50837.8671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77758.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 380204.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 581631.4375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207222.875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.06156063079834,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.365924835205078,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106167.2734375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 210.5981507165663,
+ "step": 122,
+ "teaching_rms": 70.14583879830953,
+ "traffic_rms": 210.5981299560893
+ },
+ {
+ "batch_loss": 104.40152740478516,
+ "innovation_rms": 88.28801070248723,
+ "instruction_rms": 0.047074407503202416,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 856646.5625,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88912886235136.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246404.515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.179629325866699,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16.482711791992188,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125945.0390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60024.1171875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92283.34375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 450231.875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 720627.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246404.515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.288660526275635,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16.482769012451172,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125944.9921875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 253.7691883035316,
+ "step": 123,
+ "teaching_rms": 88.28801070248723,
+ "traffic_rms": 253.76918397002507
+ },
+ {
+ "batch_loss": 128.8788299560547,
+ "innovation_rms": 96.34822456674128,
+ "instruction_rms": 0.0550170119583707,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1178511.375,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153524713291776.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 263326.90625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.994475841522217,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.67235565185547,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 143718.859375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 70865.9609375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112263.4765625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 534133.0625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 872778.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 263326.9375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.103496074676514,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.672393798828125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 143718.8125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 265.0044291633943,
+ "step": 124,
+ "teaching_rms": 96.34822456674128,
+ "traffic_rms": 265.0044234840111
+ },
+ {
+ "batch_loss": 164.70367431640625,
+ "innovation_rms": 80.98133039306805,
+ "instruction_rms": 0.06346142869712318,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1901469.25,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219682158149632.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 349181.90625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.726107597351074,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.92202758789062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157924.703125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82540.5625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 123608.890625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 621112.8125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 908831.5,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 349181.90625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.778613090515137,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.92206573486328,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157924.65625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 298.74193104751964,
+ "step": 125,
+ "teaching_rms": 80.98133039306805,
+ "traffic_rms": 298.74192434571825
+ },
+ {
+ "batch_loss": 145.857666015625,
+ "innovation_rms": 123.52754965839237,
+ "instruction_rms": 0.1195997381551323,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2187684.75,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 367808516980736.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 413747.9375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.120316505432129,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63.94208526611328,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 96243.828125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 143876.46875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 726908.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1067375.5,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 413747.875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12.081297874450684,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63.94211959838867,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175431.953125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 358.2426501252812,
+ "step": 126,
+ "teaching_rms": 123.52754965839237,
+ "traffic_rms": 358.2426301935958
+ },
+ {
+ "batch_loss": 217.91761779785156,
+ "innovation_rms": 153.9838779350764,
+ "instruction_rms": 0.24827904748135535,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2475714.25,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 535158763552768.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 382494.625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.305587768554688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.852718353271484,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 200087.296875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112122.53125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 165888.671875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 853062.9375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1336967.375,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 382494.5625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.266572952270508,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.852752685546875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 200087.25,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 418.6845063817715,
+ "step": 127,
+ "teaching_rms": 153.9838779350764,
+ "traffic_rms": 418.68443279644185
+ },
+ {
+ "batch_loss": 257.57696533203125,
+ "innovation_rms": 191.6368729696335,
+ "instruction_rms": 0.49683766382409994,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3042425.5,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 784843868733440.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 418211.5625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.593942642211914,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 42.883541107177734,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 225603.53125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 130890.390625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 206446.359375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1004821.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1703217.75,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 418211.5625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21.554929733276367,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 42.883575439453125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 225603.484375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493.0826416015625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 505.8744684914152,
+ "step": 128,
+ "teaching_rms": 191.6368729696335,
+ "traffic_rms": 505.8742245395045
+ },
+ {
+ "batch_loss": 260.2485656738281,
+ "innovation_rms": 234.55136173925388,
+ "instruction_rms": 0.30768395080365446,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3740766.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1219179751407616.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 407522.8125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25.224340438842773,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36.30398178100586,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 250092.640625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 152841.28125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 251978.46875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1189396.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2155672.75,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 407522.8125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25.185331344604492,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36.304019927978516,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 250092.59375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 55014.24662497769,
+ "predictor_updated": true,
+ "raw_apical_rms": 593.3747396293944,
+ "step": 129,
+ "teaching_rms": 234.55136173925388,
+ "traffic_rms": 593.3746598932111
+ },
+ {
+ "batch_loss": 317.91943359375,
+ "innovation_rms": 242.07435317646485,
+ "instruction_rms": 0.1355439048154094,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4440511.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1854236031713280.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 386730.03125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28.27865982055664,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.543190002441406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 277430.78125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179359.296875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 282391.71875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1442610.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2595894.5,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 386730.03125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28.239654541015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.543228149414062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 277430.71875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 710.6708921384561,
+ "step": 130,
+ "teaching_rms": 242.07435317646485,
+ "traffic_rms": 710.6708792614301
+ },
+ {
+ "batch_loss": 403.0797119140625,
+ "innovation_rms": 279.42695926799985,
+ "instruction_rms": 0.2574608766701381,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5374299.5,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2781748312145920.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1523628.5,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.314956665039062,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.59923553466797,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347887.375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 212558.421875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 331991.3125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1745452.875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3028423.75,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1523628.25,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.275955200195312,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.59931945800781,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347887.3125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 816.9719083450698,
+ "step": 131,
+ "teaching_rms": 279.42695926799985,
+ "traffic_rms": 816.9718678500976
+ },
+ {
+ "batch_loss": 400.1134338378906,
+ "innovation_rms": 359.9929377855082,
+ "instruction_rms": 0.2167463459426942,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6898254.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5231343180972032.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1327504.125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31.665632247924805,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44.05791473388672,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 422705.40625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 252312.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 397543.3125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2119910.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3744578.75,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1327503.875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31.62663459777832,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44.05800247192383,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 422705.34375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1009.1117036955022,
+ "step": 132,
+ "teaching_rms": 359.9929377855082,
+ "traffic_rms": 1009.1116804886268
+ },
+ {
+ "batch_loss": 414.2424621582031,
+ "innovation_rms": 442.6781401201619,
+ "instruction_rms": 0.4771650086068726,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9488191.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9996633018204160.0,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1192802.25,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.2691650390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37.33203887939453,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 490998.875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296299.21875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 439864.53125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2585246.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4653360.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1192802.125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.23017120361328,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37.332130432128906,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 490998.8125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1202.0108468705748,
+ "step": 133,
+ "teaching_rms": 442.6781401201619,
+ "traffic_rms": 1202.0107521939437
+ },
+ {
+ "batch_loss": 295.88763427734375,
+ "innovation_rms": 535.7778343242588,
+ "instruction_rms": 0.7432970784468234,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12655387.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7845894057558016e+16,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1073489.125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.618953704833984,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.343177795410156,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 558292.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 341238.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 449392.84375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3159423.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5741763.5,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1073489.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.57996368408203,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26.343265533447266,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 558292.1875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1439.455100199018,
+ "step": 134,
+ "teaching_rms": 535.7778343242588,
+ "traffic_rms": 1439.4549082811307
+ },
+ {
+ "batch_loss": 234.62496948242188,
+ "innovation_rms": 645.0304917084283,
+ "instruction_rms": 0.272898247883326,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16732724.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.0775588159488e+16,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 966526.75,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.87628173828125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.663898468017578,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 632882.9375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 388953.5625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 477150.46875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3863060.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7036369.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 966526.625,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.83729553222656,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22.664098739624023,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 632882.875,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1726.3071115434927,
+ "step": 135,
+ "teaching_rms": 645.0304917084283,
+ "traffic_rms": 1726.3070899937052
+ },
+ {
+ "batch_loss": 188.0745849609375,
+ "innovation_rms": 795.9170110597166,
+ "instruction_rms": 0.5495187998017678,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20338944.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.360803803194982e+16,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 872951.375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.84515380859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.112346649169922,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 704360.8125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 438982.6875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 500291.15625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4733841.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8707816.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 872951.25,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.80617141723633,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.112430572509766,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 704360.75,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2095.993781446279,
+ "step": 136,
+ "teaching_rms": 795.9170110597166,
+ "traffic_rms": 2095.993709455994
+ },
+ {
+ "batch_loss": 134.19541931152344,
+ "innovation_rms": 962.1109800002436,
+ "instruction_rms": 0.38969601764572903,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24463868.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.146694638554317e+16,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 826826.6875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.723609924316406,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.360971450805664,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 787043.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 491725.09375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527424.1875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5804733.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10708916.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 826826.6875,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.68463134765625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17.36105728149414,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 787043.4375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2524.5752019129063,
+ "step": 137,
+ "teaching_rms": 962.1109800002436,
+ "traffic_rms": 2524.5751718561605
+ },
+ {
+ "batch_loss": 53.268577575683594,
+ "innovation_rms": 1173.3253195464886,
+ "instruction_rms": 0.28212996085698827,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27908640.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.533313976190894e+17,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 841853.8125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.55638885498047,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.57724952697754,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 871228.875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 545734.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 540096.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7124451.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13197178.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 841853.8125,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.51741409301758,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20.577049255371094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 871228.8125,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3071.4407478757757,
+ "step": 138,
+ "teaching_rms": 1173.3253195464886,
+ "traffic_rms": 3071.440734927448
+ },
+ {
+ "batch_loss": 62.499786376953125,
+ "innovation_rms": 1427.6975581360664,
+ "instruction_rms": 0.41637522485950185,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31529926.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.5522611228246016e+17,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 740050.375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.37872314453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.072351455688477,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 944327.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 601898.4375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 561637.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8749009.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16245580.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 740050.375,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.339752197265625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.07215118408203,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 944327.0625,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3724.412966477346,
+ "step": 139,
+ "teaching_rms": 1427.6975581360664,
+ "traffic_rms": 3724.4129432077816
+ },
+ {
+ "batch_loss": 485.0879821777344,
+ "innovation_rms": 1318.148560290764,
+ "instruction_rms": 2.3413324256652115,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37863084.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.6184515353247744e+17,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7394703.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.052978515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.45564651489258,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1123824.375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 659710.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 578118.375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10502476.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17534666.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7394702.5,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.01401138305664,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.4556999206543,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1123824.375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4130.022110129677,
+ "step": 140,
+ "teaching_rms": 1318.148560290764,
+ "traffic_rms": 4130.021446503258
+ },
+ {
+ "batch_loss": 714.5146484375,
+ "innovation_rms": 742.5364494546625,
+ "instruction_rms": 0.8104153534490492,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46318016.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.368454545725194e+17,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9063570.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31.004793167114258,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73.7625961303711,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1543651.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 716744.9375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 570346.625,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12103981.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16015053.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9063570.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30.965829849243164,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73.76251220703125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1543651.375,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2134.3451707138606,
+ "step": 141,
+ "teaching_rms": 742.5364494546625,
+ "traffic_rms": 2134.3450169030825
+ },
+ {
+ "batch_loss": 405.4831237792969,
+ "innovation_rms": 2370.489763904843,
+ "instruction_rms": 0.5561079447780907,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49173168.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.35715665737941e+17,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9154501.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29.626996994018555,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64.23834991455078,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2459101.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 797425.8125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 806808.6875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14278308.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21743274.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9154501.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29.588037490844727,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64.23826599121094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2459101.5,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6168.133575664372,
+ "step": 142,
+ "teaching_rms": 2370.489763904843,
+ "traffic_rms": 6168.133550615724
+ },
+ {
+ "batch_loss": 387.27392578125,
+ "innovation_rms": 2733.012487542847,
+ "instruction_rms": 0.26165269865662916,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55706380.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.310218325271249e+18,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8515086.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28.300413131713867,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 48.749263763427734,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3310610.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 901659.9375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1042341.1875,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17058582.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27802736.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8515086.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28.261457443237305,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 48.749176025390625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3310610.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 7194.976832582406,
+ "step": 143,
+ "teaching_rms": 2733.012487542847,
+ "traffic_rms": 7194.976827831729
+ },
+ {
+ "batch_loss": 1629.581787109375,
+ "innovation_rms": 3272.3871382340667,
+ "instruction_rms": 0.5205027195407613,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 83650680.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1747019209066414e+18,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8330770.5,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25.825908660888672,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.34416198730469,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4143687.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1084271.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1826112.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20561130.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35025480.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8330770.5,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25.786956787109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.34405517578125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4143687.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7005.14453125,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8587.015230498655,
+ "step": 144,
+ "teaching_rms": 3272.3871382340667,
+ "traffic_rms": 8587.015214756373
+ },
+ {
+ "batch_loss": 5285.1689453125,
+ "innovation_rms": 2004.8261524862369,
+ "instruction_rms": 1.449208737869915,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 171811392.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.4237480921328517e+18,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9174470.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.517005920410156,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 254.7462921142578,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4959992.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1273453.375,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1891821.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23931928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33707980.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9174470.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18.478057861328125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 254.74615478515625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4959992.5,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 4019325.627430734,
+ "predictor_updated": true,
+ "raw_apical_rms": 7270.428132588029,
+ "step": 145,
+ "teaching_rms": 2004.8261524862369,
+ "traffic_rms": 7270.427988294935
+ },
+ {
+ "batch_loss": 2762.796630859375,
+ "innovation_rms": 3707.917259464779,
+ "instruction_rms": 0.3439411358508167,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228241856.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.353920466754732e+18,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13398584.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.59889221191406,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 218.06948852539062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5704091.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1538441.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2649881.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27990816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40588884.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13398585.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32.76275634765625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 218.06935119628906,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5704091.5,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 11495.720573630502,
+ "step": 146,
+ "teaching_rms": 3707.917259464779,
+ "traffic_rms": 11495.720568510378
+ },
+ {
+ "batch_loss": 2436.4892578125,
+ "innovation_rms": 3312.4026816833953,
+ "instruction_rms": 0.602516249876021,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 245606768.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1024821783407624e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19952526.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.20235824584961,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.03465270996094,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6399774.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1816245.125,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2778036.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32218986.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 42281704.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19952526.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51.366207122802734,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.03451538085938,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6399774.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 10112.607724347743,
+ "step": 147,
+ "teaching_rms": 3312.4026816833953,
+ "traffic_rms": 10112.607706441515
+ },
+ {
+ "batch_loss": 1597.145751953125,
+ "innovation_rms": 5528.796313829351,
+ "instruction_rms": 0.3275608181361852,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 251809952.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.791561728533096e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17205054.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.09564971923828,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 158.93287658691406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7291637.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2142704.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3264594.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37732424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55134384.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17205056.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.25948333740234,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 158.9327392578125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7291637.5,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 16024.299380065291,
+ "step": 148,
+ "teaching_rms": 5528.796313829351,
+ "traffic_rms": 16024.299376727473
+ },
+ {
+ "batch_loss": 3266.970703125,
+ "innovation_rms": 6023.352615461024,
+ "instruction_rms": 1.3204567698844993,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 261489616.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.6754565972254786e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15945334.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79.64862060546875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.52967834472656,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8080251.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2508223.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3655186.75,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44407164.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66747380.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15945336.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79.81243896484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.529541015625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8080252.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 18154.268819014855,
+ "step": 149,
+ "teaching_rms": 6023.352615461024,
+ "traffic_rms": 18154.26877103229
+ },
+ {
+ "batch_loss": 3344.693603515625,
+ "innovation_rms": 7258.739586540186,
+ "instruction_rms": 1.9363816087227759,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 278150496.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.842052068239999e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12884779.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.20817565917969,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98.8020248413086,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9368729.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2922357.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4141337.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52524660.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 81174960.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12884780.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.37197875976562,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98.80196380615234,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9368730.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 21356.18912487728,
+ "step": 150,
+ "teaching_rms": 7258.739586540186,
+ "traffic_rms": 21356.189037110194
+ },
+ {
+ "batch_loss": 2288.597412109375,
+ "innovation_rms": 8997.173011851104,
+ "instruction_rms": 0.3463343357959186,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 315357664.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.6307842035990135e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13421648.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.01397705078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78.05801391601562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10710894.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3398389.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4760322.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 62634384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 101097232.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13421649.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.17776489257812,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78.05787658691406,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10710895.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 26051.1175563177,
+ "step": 151,
+ "teaching_rms": 8997.173011851104,
+ "traffic_rms": 26051.117554020482
+ },
+ {
+ "batch_loss": 4809.966796875,
+ "innovation_rms": 10348.739833146146,
+ "instruction_rms": 1.8857177512152004,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 451640000.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.796641350606127e+19,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34908848.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.57881164550781,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 161.79037475585938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12457105.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4016891.25,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6185020.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74850352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122159696.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34908848.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.74258422851562,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 161.79043579101562,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12457106.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 30548.76414863054,
+ "step": 152,
+ "teaching_rms": 10348.739833146146,
+ "traffic_rms": 30548.76409051202
+ },
+ {
+ "batch_loss": 3684.880126953125,
+ "innovation_rms": 12280.580394013557,
+ "instruction_rms": 0.25318450503118134,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 522067296.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.174921236942831e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25018436.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.2930679321289,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127.00410461425781,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14274278.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4736955.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7200641.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89541880.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146915264.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25018436.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.45682525634766,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127.00416564941406,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14274280.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 36076.21898050063,
+ "step": 153,
+ "teaching_rms": 12280.580394013557,
+ "traffic_rms": 36076.218979625395
+ },
+ {
+ "batch_loss": 5480.837890625,
+ "innovation_rms": 14531.876441035974,
+ "instruction_rms": 0.3121558674568225,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 648868992.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8979651851022814e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19480398.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.69390869140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.54198455810547,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16222318.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5591215.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8542602.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107104816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175629376.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19480398.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.85765075683594,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.54204559326172,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16222320.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 42453.26705711168,
+ "step": 154,
+ "teaching_rms": 14531.876441035974,
+ "traffic_rms": 42453.267055984295
+ },
+ {
+ "batch_loss": 5207.16650390625,
+ "innovation_rms": 17165.619874204524,
+ "instruction_rms": 0.5520526552764263,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 813755840.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.053209034188852e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40257320.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112.45194244384766,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 178.2096405029297,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20248050.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6635309.5,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10440939.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 128017440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 209126272.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40257320.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112.61566925048828,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 178.20965576171875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20248052.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 50219.682855484614,
+ "step": 155,
+ "teaching_rms": 17165.619874204524,
+ "traffic_rms": 50219.68285247151
+ },
+ {
+ "batch_loss": 8661.58984375,
+ "innovation_rms": 18747.316964964346,
+ "instruction_rms": 0.6606389977975461,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1065454080.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.44909692394257e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75743712.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111.49309539794922,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 149.71279907226562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20554954.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7820906.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11855967.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151896256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 238788080.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75743712.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111.65680694580078,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 149.7128143310547,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20554956.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 59172.73965478753,
+ "step": 156,
+ "teaching_rms": 18747.316964964346,
+ "traffic_rms": 59172.73965112593
+ },
+ {
+ "batch_loss": 5937.43603515625,
+ "innovation_rms": 17230.18966449186,
+ "instruction_rms": 0.4855176441768551,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1182876672.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.415492350201404e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 166764416.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.82281494140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.9533920288086,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37231396.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9086052.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12651457.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176475984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 245797280.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 166764416.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 107.98651123046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.95337677001953,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37231396.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 50423.873886776935,
+ "step": 157,
+ "teaching_rms": 17230.18966449186,
+ "traffic_rms": 50423.87388445377
+ },
+ {
+ "batch_loss": 7968.1337890625,
+ "innovation_rms": 26688.54721961249,
+ "instruction_rms": 0.4273337386086636,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1360546048.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.249625207932579e+20,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 142279744.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100.5020523071289,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.87489318847656,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51459372.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10546342.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14602900.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 206140480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296644992.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 142279744.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100.66573333740234,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.87509155273438,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 51459372.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 82570.75313264415,
+ "step": 158,
+ "teaching_rms": 26688.54721961249,
+ "traffic_rms": 82570.75313154815
+ },
+ {
+ "batch_loss": 8076.53515625,
+ "innovation_rms": 31746.58949387972,
+ "instruction_rms": 0.606160901672398,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1615305728.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.50754690195251e+21,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 119082656.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.82271575927734,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106.79336547851562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63367636.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12191214.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16448724.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241995920.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 358554336.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 119082656.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.98638153076172,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106.79353332519531,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63367636.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 97461.64832029103,
+ "step": 159,
+ "teaching_rms": 31746.58949387972,
+ "traffic_rms": 97461.64831841475
+ },
+ {
+ "batch_loss": 12940.822265625,
+ "innovation_rms": 34061.31015024355,
+ "instruction_rms": 0.897437449383954,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1962319104.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.337105871926993e+21,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196581472.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.1024169921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 230.06942749023438,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67560168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14049323.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18581090.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 283164608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 411686752.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 196581472.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.26606750488281,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 230.0692138671875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67560168.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31536.119140625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 109949.19532939345,
+ "step": 160,
+ "teaching_rms": 34061.31015024355,
+ "traffic_rms": 109949.19532574217
+ },
+ {
+ "batch_loss": 9805.611328125,
+ "innovation_rms": 27304.88706092028,
+ "instruction_rms": 0.5449548264554801,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2420591872.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.55790694708283e+21,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 156793536.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.47748947143555,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 199.91966247558594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 83239520.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15999099.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19497764.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 323955264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 407906464.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 156793536.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.6411247253418,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 199.91944885253906,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 83239520.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 745556901.6344929,
+ "predictor_updated": true,
+ "raw_apical_rms": 80053.4853447783,
+ "step": 161,
+ "teaching_rms": 27304.88706092028,
+ "traffic_rms": 80053.48534293506
+ },
+ {
+ "batch_loss": 7779.52880859375,
+ "innovation_rms": 38421.424434758344,
+ "instruction_rms": 0.7337379968603895,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3019728128.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.7868058662532777e+21,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 286714880.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43.42588424682617,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 190.11602783203125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111911008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18135992.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21368924.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 369660864.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 457056064.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 286714880.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43.44983673095703,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 190.11618041992188,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111911008.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 124594.03309453768,
+ "step": 162,
+ "teaching_rms": 38421.424434758344,
+ "traffic_rms": 124594.03309238119
+ },
+ {
+ "batch_loss": 8256.4169921875,
+ "innovation_rms": 47664.96250090274,
+ "instruction_rms": 10.774513710634011,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3957487872.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.720222387316306e+21,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 275706592.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.39825439453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 195.41183471679688,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139481664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20333910.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21979182.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 423649312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 539884544.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 275706592.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.60108184814453,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 195.4119873046875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139481664.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 163575.19479360443,
+ "step": 163,
+ "teaching_rms": 47664.96250090274,
+ "traffic_rms": 163575.19443878988
+ },
+ {
+ "batch_loss": 9157.5263671875,
+ "innovation_rms": 62127.250838670065,
+ "instruction_rms": 2.7281928370643294,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4894334464.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.141950510734591e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 248196064.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.01014709472656,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233.47463989257812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 164301264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 22955272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26213610.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 491524288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 678749888.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 248196064.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.212955474853516,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233.4748077392578,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 164301264.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 203283.70840800522,
+ "step": 164,
+ "teaching_rms": 62127.250838670065,
+ "traffic_rms": 203283.70838970164
+ },
+ {
+ "batch_loss": 13881.8505859375,
+ "innovation_rms": 42490.02942855806,
+ "instruction_rms": 1.0877123505979645,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6336157184.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8131883912961198e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228686176.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.3397102355957,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 377.0568542480469,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 187169888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25529806.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 25745336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 558795456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 672711744.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228686176.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55.54249954223633,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 377.0567626953125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 187169888.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 209799.24264457254,
+ "step": 165,
+ "teaching_rms": 42490.02942855806,
+ "traffic_rms": 209799.24264176158
+ },
+ {
+ "batch_loss": 10619.5263671875,
+ "innovation_rms": 25007.706360465665,
+ "instruction_rms": 0.8519612180651772,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8102618624.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.7034899145722415e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207515136.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.422142028808594,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 293.7848815917969,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207921408.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27952312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24225050.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 620164160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 613687104.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207515136.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.62491226196289,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 293.7847900390625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207921408.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 220820.6804697935,
+ "step": 166,
+ "teaching_rms": 25007.706360465665,
+ "traffic_rms": 220820.6804681587
+ },
+ {
+ "batch_loss": 13707.16015625,
+ "innovation_rms": 30582.04301936531,
+ "instruction_rms": 0.5128456052498355,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9251202048.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.042800771057164e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186742832.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.67054748535156,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 221.1544189453125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 226595696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 30132330.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21800180.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 675396032.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 552318528.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186742832.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 56.779109954833984,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 221.1544189453125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 226595696.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 242901.11196924964,
+ "step": 167,
+ "teaching_rms": 30582.04301936531,
+ "traffic_rms": 242901.11196871978
+ },
+ {
+ "batch_loss": 13760.8125,
+ "innovation_rms": 36129.631955115874,
+ "instruction_rms": 0.6829641946902115,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10944724992.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.268398191750408e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 168722944.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73.16261291503906,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167.2322235107422,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 243467984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32134468.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20021372.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 725477632.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 509703968.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 168722944.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73.27117156982422,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167.2322998046875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 243467984.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 272831.6694959134,
+ "step": 168,
+ "teaching_rms": 36129.631955115874,
+ "traffic_rms": 272831.6694950718
+ },
+ {
+ "batch_loss": 14037.734375,
+ "innovation_rms": 40304.6642573313,
+ "instruction_rms": 0.8226597012706106,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12893383680.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.724544312747596e+22,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151788928.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.2812271118164,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.12777709960938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258646880.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 33941816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19162662.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 770554176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 591044608.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151788928.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 86.3897705078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.12771606445312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 258646880.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 295247.98089912115,
+ "step": 169,
+ "teaching_rms": 40304.6642573313,
+ "traffic_rms": 295247.9808979899
+ },
+ {
+ "batch_loss": 14106.294921875,
+ "innovation_rms": 55177.41759175107,
+ "instruction_rms": 0.959362496145216,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15506385920.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4823562144692969e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 137914000.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 96.95873260498047,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 443.89990234375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 272438272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35782988.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21406610.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 813910400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 685474688.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 137914000.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.06726837158203,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 443.89984130859375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 272438272.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 318726.39169147296,
+ "step": 170,
+ "teaching_rms": 55177.41759175107,
+ "traffic_rms": 318726.39169003954
+ },
+ {
+ "batch_loss": 21901.537109375,
+ "innovation_rms": 47091.44407568113,
+ "instruction_rms": 1.759542324188182,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18162956288.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1967980720121125e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 124087784.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 101.356689453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347.10858154296875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 284847040.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 37417584.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 23139386.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 852932224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 755657792.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 124087784.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 101.4652099609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347.1085205078125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 284847040.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 341377.8665849882,
+ "step": 171,
+ "teaching_rms": 47091.44407568113,
+ "traffic_rms": 341377.86658046814
+ },
+ {
+ "batch_loss": 11273.9404296875,
+ "innovation_rms": 61355.63055679654,
+ "instruction_rms": 1.1614301157454776,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 21005772800.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.129033286958625e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111647832.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 104.54678344726562,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296.860107421875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296011808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 38869340.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24495694.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 888052544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 887256768.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111647832.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 104.6552963256836,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296.86016845703125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 296011808.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 361941.6632016744,
+ "step": 172,
+ "teaching_rms": 61355.63055679654,
+ "traffic_rms": 361941.663199818
+ },
+ {
+ "batch_loss": 15025.1318359375,
+ "innovation_rms": 71707.84435549322,
+ "instruction_rms": 5.465969618709177,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 24226232320.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.303377153984978e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100436592.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105.77552032470703,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 267.86920166015625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 306055456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40158384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 26551778.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 919662336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1046347328.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100436592.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105.88401794433594,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 267.869384765625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 306055456.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 381993.8094658023,
+ "step": 173,
+ "teaching_rms": 71707.84435549322,
+ "traffic_rms": 381993.80942669994
+ },
+ {
+ "batch_loss": 5542.59033203125,
+ "innovation_rms": 81629.8392110438,
+ "instruction_rms": 2.0370757743881627,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 27466153984.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.744349611334382e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106399312.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.85159301757812,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 243.21243286132812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 315091456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41302172.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29434488.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 948111616.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1221807104.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106399312.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.9376220703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 243.21261596679688,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 315091456.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 407130.0495035423,
+ "step": 174,
+ "teaching_rms": 81629.8392110438,
+ "traffic_rms": 407130.0494984581
+ },
+ {
+ "batch_loss": 11897.9775390625,
+ "innovation_rms": 93483.5296129134,
+ "instruction_rms": 4.004047763404685,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31420309504.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.417259250933218e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 133792240.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.09680938720703,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 238.9092559814453,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 323220448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 42313332.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31375436.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1082801152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1419330944.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 133792240.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.18283081054688,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 238.909423828125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 323220448.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 415792.3486894973,
+ "step": 175,
+ "teaching_rms": 93483.5296129134,
+ "traffic_rms": 415792.3486702496
+ },
+ {
+ "batch_loss": 22650.484375,
+ "innovation_rms": 55920.561664746376,
+ "instruction_rms": 16.2981084106909,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 35972087808.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.399939082981553e+23,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117954712.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 138.1602020263672,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 192.77801513671875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 330532512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 43204980.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31510020.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1220475008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1376739072.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117954712.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 138.2462158203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 192.7779998779297,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 330532512.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308940.40625,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 360287.27219388576,
+ "step": 176,
+ "teaching_rms": 55920.561664746376,
+ "traffic_rms": 360287.27182526403
+ },
+ {
+ "batch_loss": 119132.453125,
+ "innovation_rms": 36435.77776052217,
+ "instruction_rms": 50.7402852609786,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66821107712.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0596942541317043e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153416960.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 133.1339874267578,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1666.7921142578125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 337110048.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44007460.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 31854776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1348114816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1276397440.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153416960.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 133.22000122070312,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1666.792236328125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 337110048.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 1327563326.437902,
+ "predictor_updated": true,
+ "raw_apical_rms": 434051.9217723354,
+ "step": 177,
+ "teaching_rms": 36435.77776052217,
+ "traffic_rms": 434051.91880670155
+ },
+ {
+ "batch_loss": 99260.5625,
+ "innovation_rms": 5187.266516968814,
+ "instruction_rms": 2.593849145411237,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78128963584.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3048706501913183e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 144391712.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246.20570373535156,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1460.916748046875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 343026464.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 44729692.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29023966.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1463065600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149507712.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 144391712.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246.118896484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1460.916748046875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 343026464.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 500230.1478612452,
+ "step": 178,
+ "teaching_rms": 5187.266516968814,
+ "traffic_rms": 500230.1478545432
+ },
+ {
+ "batch_loss": 218361.28125,
+ "innovation_rms": 13201.490909943346,
+ "instruction_rms": 3.112417551910885,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88792465408.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.567012142076034e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 369457408.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 358.658447265625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1124.5274658203125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 348347808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46950132.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29042700.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1566976000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1039104000.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 369457408.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 358.5716552734375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1124.5274658203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 348347808.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 446906.9329274823,
+ "step": 179,
+ "teaching_rms": 13201.490909943346,
+ "traffic_rms": 446906.9329166826
+ },
+ {
+ "batch_loss": 22795.705078125,
+ "innovation_rms": 51433.819192916664,
+ "instruction_rms": 13.315208873390059,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88650129408.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8621818086487863e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1171670784.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 459.8055419921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1011.4710693359375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 353133536.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49881396.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 29312638.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1666957824.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 999817792.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1171670784.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 459.71875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1011.4710693359375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 353133536.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 487462.2549674079,
+ "step": 180,
+ "teaching_rms": 51433.819192916664,
+ "traffic_rms": 487462.2547855575
+ },
+ {
+ "batch_loss": 322596.84375,
+ "innovation_rms": 12791.086332833544,
+ "instruction_rms": 1168.8114854181654,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105283674112.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.194746916569026e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1070580288.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 529.3699340820312,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1743.3177490234375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 357437152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52718024.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28366282.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1757313792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 903559104.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1070580288.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 529.2831420898438,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1743.3177490234375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 357437152.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 576390.4980420044,
+ "step": 181,
+ "teaching_rms": 12791.086332833544,
+ "traffic_rms": 576389.3129758787
+ },
+ {
+ "batch_loss": 574736.5,
+ "innovation_rms": 25870.078658834696,
+ "instruction_rms": 18.204752872640363,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 297402499072.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.4328330171873516e+24,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1011583424.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 565.2599487304688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2486.8525390625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 370027264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55582844.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28648200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1840068736.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 827548928.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1011583424.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 565.1731567382812,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2486.852294921875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 370027264.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 583512.791376227,
+ "step": 182,
+ "teaching_rms": 25870.078658834696,
+ "traffic_rms": 583512.7910924614
+ },
+ {
+ "batch_loss": 476236.46875,
+ "innovation_rms": 21056.740900055574,
+ "instruction_rms": 4.646448552420547,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493398949888.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1443683258406202e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 931802816.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 578.806396484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2011.471435546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 463207552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58479892.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 28970468.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1915469952.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 754011840.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 931802816.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 578.7196044921875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2011.47119140625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 463207552.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 608614.3043103317,
+ "step": 183,
+ "teaching_rms": 21056.740900055574,
+ "traffic_rms": 608614.3042927695
+ },
+ {
+ "batch_loss": 798995.9375,
+ "innovation_rms": 40430.79759272372,
+ "instruction_rms": 5.529113564499056,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 761677676544.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.330258658501054e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 909089152.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 567.3839721679688,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2318.01904296875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 554116480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61758928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 32790360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1986477440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 710074304.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 909089152.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 567.2971801757812,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2318.01904296875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 554116480.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 634818.5605955822,
+ "step": 184,
+ "teaching_rms": 40430.79759272372,
+ "traffic_rms": 634818.5605717914
+ },
+ {
+ "batch_loss": 1015305.875,
+ "innovation_rms": 41997.66783835554,
+ "instruction_rms": 9.629725632026954,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1125617893376.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.898150739092442e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 871136768.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 542.0552368164062,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2296.635009765625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 641230144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65446960.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 36880328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2053191168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 667137600.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 871136768.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 541.9684448242188,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2296.634765625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 641230144.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 638171.5701128984,
+ "step": 185,
+ "teaching_rms": 41997.66783835554,
+ "traffic_rms": 638171.5700409964
+ },
+ {
+ "batch_loss": 1888426.25,
+ "innovation_rms": 45974.294717420365,
+ "instruction_rms": 6.755896944334688,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1609688416256.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.4032542533743385e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 821961536.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 525.101318359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3055.83935546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723426304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69695472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 42485152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2115904256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 627130240.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 821961536.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 525.3037109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3055.83935546875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723426304.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 602027.3567709264,
+ "step": 186,
+ "teaching_rms": 45974.294717420365,
+ "traffic_rms": 602027.3567344954
+ },
+ {
+ "batch_loss": 2689511.5,
+ "innovation_rms": 45952.63685090035,
+ "instruction_rms": 4.705042174125604,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2012389310464.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.553859610172933e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 775370112.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 633.7174682617188,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2575.93359375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 800963328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74220320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 45248472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2175067136.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 591629504.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 775370112.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 633.7561645507812,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2575.933837890625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 800963328.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 639001.2467680144,
+ "step": 187,
+ "teaching_rms": 45952.63685090035,
+ "traffic_rms": 639001.2467526093
+ },
+ {
+ "batch_loss": 2357202.5,
+ "innovation_rms": 49414.80343883996,
+ "instruction_rms": 4.07219731226606,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2370223472640.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0132273554494167e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745501248.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 820.3973999023438,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2025.212646484375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 875513472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79015152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47948320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2232198656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 571316352.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745501248.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 820.3837280273438,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2025.212890625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 875513472.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 659645.646965888,
+ "step": 188,
+ "teaching_rms": 49414.80343883996,
+ "traffic_rms": 659645.646955383
+ },
+ {
+ "batch_loss": 2162250.0,
+ "innovation_rms": 50096.55166652376,
+ "instruction_rms": 4.6030391630889,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2665272311808.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2958038967762568e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 714429056.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 775.847412109375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1835.8062744140625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 946956352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 84070792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 50556420.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2287761408.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 555627136.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 714429056.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 775.833740234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1835.8062744140625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 946956352.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 671261.7812943315,
+ "step": 189,
+ "teaching_rms": 50096.55166652376,
+ "traffic_rms": 671261.7812795321
+ },
+ {
+ "batch_loss": 2115707.75,
+ "innovation_rms": 47877.14169105058,
+ "instruction_rms": 2.6058019966723793,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2872775802880.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6292964954593075e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 680005632.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 535.1529541015625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2697.5927734375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1014956928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89304744.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52339508.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2341569792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 538084288.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 680005632.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 535.1659545898438,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2697.5927734375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1014956928.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 726784.7958133165,
+ "step": 190,
+ "teaching_rms": 47877.14169105058,
+ "traffic_rms": 726784.7958093765
+ },
+ {
+ "batch_loss": 1119620.75,
+ "innovation_rms": 49060.79161488805,
+ "instruction_rms": 2.3796676570429733,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3029047705600.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9505690485307293e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 644039104.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 576.1715087890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2266.7158203125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1079360896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94568368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52636232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2393899520.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 523296512.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 644039104.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 576.334716796875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2266.7158203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1079360896.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 729052.2337418521,
+ "step": 191,
+ "teaching_rms": 49060.79161488805,
+ "traffic_rms": 729052.2337385041
+ },
+ {
+ "batch_loss": 970403.625,
+ "innovation_rms": 59944.23522482849,
+ "instruction_rms": 1.9524613951308272,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3108905943040.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.2810579638762578e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 623155712.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 610.164306640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1668.5589599609375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141676416.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 100066592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54982256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2446698496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527990464.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 623155712.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 610.3275146484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1668.5589599609375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141676416.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 347188.09375,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 757384.6627311262,
+ "step": 192,
+ "teaching_rms": 59944.23522482849,
+ "traffic_rms": 757384.6627289273
+ },
+ {
+ "batch_loss": 945437.875,
+ "innovation_rms": 58041.47588039863,
+ "instruction_rms": 3.7727876209136935,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3186762448896.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.54661212647304e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 594748608.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 625.5211181640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1625.8726806640625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1201151232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105782512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57159176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2499446784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527482432.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 594748608.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 625.684326171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1625.87255859375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1201151232.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 3368812551.9665337,
+ "predictor_updated": true,
+ "raw_apical_rms": 782163.931170862,
+ "step": 193,
+ "teaching_rms": 58041.47588039863,
+ "traffic_rms": 782163.9311617881
+ },
+ {
+ "batch_loss": 761272.5625,
+ "innovation_rms": 71170.80295941772,
+ "instruction_rms": 7.127305939064361,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3197213605888.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.761256595840317e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583875904.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 706.4270629882812,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1585.9703369140625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1259538816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111763408.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59808928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2555565312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 561185792.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583875904.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 706.5137939453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1585.97021484375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1259538816.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 800788.3471439905,
+ "step": 194,
+ "teaching_rms": 71170.80295941772,
+ "traffic_rms": 800788.3471123357
+ },
+ {
+ "batch_loss": 1056214.125,
+ "innovation_rms": 50505.82025964118,
+ "instruction_rms": 6.548520899437225,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3161276022784.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.921709511739935e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 546016064.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723.45556640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1322.7811279296875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1314140416.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117707984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 59445788.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2610716160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 551507200.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 546016064.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723.5422973632812,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1322.7813720703125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1314140416.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 793511.2957640969,
+ "step": 195,
+ "teaching_rms": 50505.82025964118,
+ "traffic_rms": 793511.2957371813
+ },
+ {
+ "batch_loss": 844307.9375,
+ "innovation_rms": 76215.74335411403,
+ "instruction_rms": 3.1450791251529364,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3070520721408.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.9870504617931846e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527173216.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 726.9530639648438,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1182.6129150390625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1366857728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 123893472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61854888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2669828864.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 591127680.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 527173216.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 726.85107421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1182.6131591796875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1366857728.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 828687.4127282148,
+ "step": 196,
+ "teaching_rms": 76215.74335411403,
+ "traffic_rms": 828687.4127225141
+ },
+ {
+ "batch_loss": 617291.875,
+ "innovation_rms": 67223.07356175063,
+ "instruction_rms": 2.01350969498224,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2902203564032.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.971328301819162e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 491710336.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 776.0635986328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1387.35693359375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1416028800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 130009224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61157480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2730232320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 604034816.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 491710336.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 775.9616088867188,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1387.357177734375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1416028800.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 827412.6797246151,
+ "step": 197,
+ "teaching_rms": 67223.07356175063,
+ "traffic_rms": 827412.6797222713
+ },
+ {
+ "batch_loss": 316148.6875,
+ "innovation_rms": 30799.36978695539,
+ "instruction_rms": 2.5462750700724146,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2720670679040.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.852508396021821e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 444567200.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 820.2553100585938,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1127.1728515625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1460485504.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 135754192.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57449628.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2786079744.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 558473664.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 444567200.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 820.1533203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1127.173095703125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1460485504.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 823451.4535489413,
+ "step": 198,
+ "teaching_rms": 30799.36978695539,
+ "traffic_rms": 823451.4535450707
+ },
+ {
+ "batch_loss": 320162.0,
+ "innovation_rms": 52535.585300321014,
+ "instruction_rms": 5.4528305453963695,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2654606458880.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.6734329385774708e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 402943488.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 860.0196533203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 968.5377197265625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1500779904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 141226624.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 54724276.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2840526848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 544470592.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 402943488.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 859.9176635742188,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 968.5379028320312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1500779904.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 885592.0388782317,
+ "step": 199,
+ "teaching_rms": 52535.585300321014,
+ "traffic_rms": 885592.0388615166
+ },
+ {
+ "batch_loss": 273565.4375,
+ "innovation_rms": 113319.09671027833,
+ "instruction_rms": 1.2771917914865192,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2472206925824.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.4659181357946826e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 401349984.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 895.7935180664062,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 864.6165771484375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1535953792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 147228976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60023460.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2908765696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 682388032.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 401349984.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 895.6915283203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 864.6167602539062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1535953792.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 860118.2866447137,
+ "step": 200,
+ "teaching_rms": 113319.09671027833,
+ "traffic_rms": 860118.2866438447
+ },
+ {
+ "batch_loss": 253270.375,
+ "innovation_rms": 98626.04063531834,
+ "instruction_rms": 1.769662641334844,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2328309792768.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.251516056644534e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 465266016.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 927.9805908203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 709.3147583007812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1566060928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153636704.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64077272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2984410112.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 756443136.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 465266016.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 927.8786010742188,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 709.31494140625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1566060928.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 906340.7132979949,
+ "step": 201,
+ "teaching_rms": 98626.04063531834,
+ "traffic_rms": 906340.7132963178
+ },
+ {
+ "batch_loss": 180857.265625,
+ "innovation_rms": 158095.47295069208,
+ "instruction_rms": 0.9401387586666713,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2135601577984.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.0343237229857964e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 392837248.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 956.9410400390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 627.1572875976562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1599979520.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 160958656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73219472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3087970304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1035600960.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 392837248.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 956.8390502929688,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 627.1575317382812,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1599979520.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 826338.9355085812,
+ "step": 202,
+ "teaching_rms": 158095.47295069208,
+ "traffic_rms": 826338.9355080903
+ },
+ {
+ "batch_loss": 134478.15625,
+ "innovation_rms": 119071.5056488582,
+ "instruction_rms": 1.2432869747554978,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1941382496256.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.837044649554628e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 404990336.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 982.9837646484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 557.4359130859375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1631860864.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 168430528.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74718768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3200701440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1127310848.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 404990336.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 982.8817749023438,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 557.4361572265625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1631860864.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 960363.5878168899,
+ "step": 203,
+ "teaching_rms": 119071.5056488582,
+ "traffic_rms": 960363.5878161356
+ },
+ {
+ "batch_loss": 144423.375,
+ "innovation_rms": 115185.20846254077,
+ "instruction_rms": 0.6346487227908811,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1812235419648.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.655548425885485e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 733190656.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1006.4149169921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 501.6938781738281,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1658815232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176234880.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78043496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3319698944.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1189975040.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 733190656.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1006.3129272460938,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 501.694091796875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1658815232.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 785617.8266883984,
+ "step": 204,
+ "teaching_rms": 115185.20846254077,
+ "traffic_rms": 785617.8266881788
+ },
+ {
+ "batch_loss": 116380.0078125,
+ "innovation_rms": 97096.96311028075,
+ "instruction_rms": 1.0378341191888705,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1686520332288.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4922792455681336e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 751663296.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1027.492919921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 451.2641296386719,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1678741248.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 184205120.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79702416.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3438870528.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1191716224.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 751663296.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1027.390869140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 451.26434326171875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1678741248.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 936488.3826378679,
+ "step": 205,
+ "teaching_rms": 97096.96311028075,
+ "traffic_rms": 936488.3826373154
+ },
+ {
+ "batch_loss": 71357.0078125,
+ "innovation_rms": 121702.94669827726,
+ "instruction_rms": 0.6044368854360402,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1542696730624.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.346594239571605e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 660776960.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1046.4676513671875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 406.0023498535156,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1695223168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 192722304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85171760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3564442624.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1255722112.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 660776960.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1046.3656005859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 406.0025634765625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1695223168.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1015439.9726485325,
+ "step": 206,
+ "teaching_rms": 121702.94669827726,
+ "traffic_rms": 1015439.9726483745
+ },
+ {
+ "batch_loss": 80374.46875,
+ "innovation_rms": 94904.79751503971,
+ "instruction_rms": 0.5410298618837721,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1411318022144.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2181263176888326e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 574531584.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1063.5494384765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 365.39007568359375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1707095808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 201534368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88120648.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3688197888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1237553152.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 574531584.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1063.4473876953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 365.3902587890625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1707095808.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1066119.5957371686,
+ "step": 207,
+ "teaching_rms": 94904.79751503971,
+ "traffic_rms": 1066119.595737056
+ },
+ {
+ "batch_loss": 69256.5390625,
+ "innovation_rms": 89760.60868614505,
+ "instruction_rms": 0.9089898737616051,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1282984116224.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.10595046793824e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 633481728.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1078.912353515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 328.6080627441406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1717871360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 210311360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 87769896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3808861952.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1206641664.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 633481728.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1078.810302734375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 328.6082458496094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1717871360.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 906052.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1068446.291088194,
+ "step": 208,
+ "teaching_rms": 89760.60868614505,
+ "traffic_rms": 1068446.2910878316
+ },
+ {
+ "batch_loss": 121090.984375,
+ "innovation_rms": 160241.19600416833,
+ "instruction_rms": 0.8323561546900885,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165078560768.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0061611456645415e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745975168.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1092.934326171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 295.2411804199219,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1728889600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219969376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 96580080.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3946093312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1372313984.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745975168.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1092.832275390625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 295.2413330078125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1728889600.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 25677240895.97954,
+ "predictor_updated": true,
+ "raw_apical_rms": 1012488.0048602055,
+ "step": 209,
+ "teaching_rms": 160241.19600416833,
+ "traffic_rms": 1012488.0048598892
+ },
+ {
+ "batch_loss": 66127.015625,
+ "innovation_rms": 141670.30858780738,
+ "instruction_rms": 1.01908670201803,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1072786833408.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.108153673370386e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 616551040.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1105.543212890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 265.66741943359375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1737222656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 230411488.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 104421184.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4093651456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1475581952.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 616551040.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1105.441162109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 265.6675720214844,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1737222656.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 936881.1222890378,
+ "step": 210,
+ "teaching_rms": 141670.30858780738,
+ "traffic_rms": 936881.1222885374
+ },
+ {
+ "batch_loss": 93691.578125,
+ "innovation_rms": 172186.5805262067,
+ "instruction_rms": 0.7454939392954986,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 996454367232.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.270351441414293e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 711822336.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1117.0333251953125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 331.86199951171875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1744883456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241721184.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 113097016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4260822528.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1671711744.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 711822336.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1116.931396484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 331.86212158203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1744883456.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 954562.5730809878,
+ "step": 211,
+ "teaching_rms": 172186.5805262067,
+ "traffic_rms": 954562.5730807529
+ },
+ {
+ "batch_loss": 44659.90234375,
+ "innovation_rms": 105404.58425665613,
+ "instruction_rms": 0.9467041185893881,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 943043117056.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.479550591020809e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 572092416.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1127.400634765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 271.5803527832031,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1751303040.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 254814016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 130928320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4423828992.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1630062720.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 572092416.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1127.2987060546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 271.5804748535156,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1751303040.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1120040.6538203482,
+ "step": 212,
+ "teaching_rms": 105404.58425665613,
+ "traffic_rms": 1120040.6538199275
+ },
+ {
+ "batch_loss": 104215.03125,
+ "innovation_rms": 156196.0559562529,
+ "instruction_rms": 10.248760816330076,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 904435728384.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.782852577339141e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583710656.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.34521484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 244.229736328125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1737453056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 270448608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 156346048.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4596907008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1730779904.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583710656.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.2432861328125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 244.2298583984375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1737453056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1129079.5804153958,
+ "step": 213,
+ "teaching_rms": 156196.0559562529,
+ "traffic_rms": 1129079.5803673442
+ },
+ {
+ "batch_loss": 47800.28515625,
+ "innovation_rms": 189116.9150474485,
+ "instruction_rms": 0.512059467854485,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 835613949952.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.225519715134349e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 710348480.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.421630859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219.70220947265625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1722571008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 286723232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 162746256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4789602304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1926950912.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 710348480.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.3197021484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219.70233154296875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1722571008.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1244989.91749145,
+ "step": 214,
+ "teaching_rms": 189116.9150474485,
+ "traffic_rms": 1244989.9174913655
+ },
+ {
+ "batch_loss": 41312.75,
+ "innovation_rms": 217482.56095029146,
+ "instruction_rms": 0.48585357990726313,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 765520248832.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.709349378824234e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1071348480.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.9996337890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197.66616821289062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1705305472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 303594656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 168714352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5009817088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2202146816.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1071348480.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.897705078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197.66629028320312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1705305472.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1270681.3497499863,
+ "step": 215,
+ "teaching_rms": 217482.56095029146,
+ "traffic_rms": 1270681.3497499065
+ },
+ {
+ "batch_loss": 38430.75,
+ "innovation_rms": 225696.4219415755,
+ "instruction_rms": 0.39203296375766394,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 696345755648.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.208438189211892e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 610349696.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.8126220703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.96109008789062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1685439232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 321724096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 181294368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5256270336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2464531712.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 610349696.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.710693359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.9613494873047,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1685439232.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1180849.1941875713,
+ "step": 216,
+ "teaching_rms": 225696.4219415755,
+ "traffic_rms": 1180849.1941875166
+ },
+ {
+ "batch_loss": 51027.265625,
+ "innovation_rms": 214266.44365856273,
+ "instruction_rms": 0.47520982189753425,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 685810384896.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.805463069729883e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1081762304.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.66259765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 222.45643615722656,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1665258368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 340688512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 189644144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5519436288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2631657472.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1081762304.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.5606689453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 222.45635986328125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1665258368.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1189327.914961958,
+ "step": 217,
+ "teaching_rms": 214266.44365856273,
+ "traffic_rms": 1189327.914961876
+ },
+ {
+ "batch_loss": 45476.49609375,
+ "innovation_rms": 254888.6274166417,
+ "instruction_rms": 0.38733621718085903,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 719931113472.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.41180171133069e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583531648.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1171.004638671875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 216.5256805419922,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1639724800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 361405792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207172704.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5812416512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2929800192.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583531648.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1170.9027099609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 216.52561950683594,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1639724800.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1407624.5413050342,
+ "step": 218,
+ "teaching_rms": 254888.6274166417,
+ "traffic_rms": 1407624.5413049886
+ },
+ {
+ "batch_loss": 27503.505859375,
+ "innovation_rms": 263251.5037146663,
+ "instruction_rms": 0.36472135625658947,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 737265844224.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.155884496772895e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 492630400.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1175.0572509765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.89772033691406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1613091712.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 384736928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233311200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6132399104.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3199827712.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 492630400.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1174.955322265625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.8976593017578,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1613091712.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1508540.8392483024,
+ "step": 219,
+ "teaching_rms": 263251.5037146663,
+ "traffic_rms": 1508540.8392482705
+ },
+ {
+ "batch_loss": 55581.140625,
+ "innovation_rms": 296684.52888961794,
+ "instruction_rms": 0.50872477047691,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 750899167232.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.929785060504446e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 514094496.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1178.294189453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 177.72689819335938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1582716160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 409297760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 245608176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6488207360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3558083072.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 514094496.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1178.1922607421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 177.7269744873047,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1582716160.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1518657.2626282887,
+ "step": 220,
+ "teaching_rms": 296684.52888961794,
+ "traffic_rms": 1518657.2626282168
+ },
+ {
+ "batch_loss": 27574.515625,
+ "innovation_rms": 212497.7518498037,
+ "instruction_rms": 0.4095952099489695,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 771949133824.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.7263177039557307e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 458313504.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1179.979736328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.73226928710938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1555374976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 435822976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 265252000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6841206272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3529986560.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 458313504.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1179.8779296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.73220825195312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1555374976.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1572965.3033546133,
+ "step": 221,
+ "teaching_rms": 212497.7518498037,
+ "traffic_rms": 1572965.3033545741
+ },
+ {
+ "batch_loss": 39644.875,
+ "innovation_rms": 315205.9326966031,
+ "instruction_rms": 0.4582393706943739,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 784292708352.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.5639102632877826e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 397365376.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1180.4080810546875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 131.86959838867188,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1528096512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 463938816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 281158496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7227470848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3862648064.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 397365376.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1180.3062744140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 131.86952209472656,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1528096512.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1708664.252457604,
+ "step": 222,
+ "teaching_rms": 315205.9326966031,
+ "traffic_rms": 1708664.2524575526
+ },
+ {
+ "batch_loss": 95400.796875,
+ "innovation_rms": 280337.230130306,
+ "instruction_rms": 0.7959594758851971,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 799532974080.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.6083461639183397e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 384347840.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1179.150390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146.60006713867188,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1504327936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 495892288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 319534592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7626571776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3991009792.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 384347840.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1179.048583984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146.60025024414062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1504327936.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1789481.9948931236,
+ "step": 223,
+ "teaching_rms": 280337.230130306,
+ "traffic_rms": 1789481.9948929762
+ },
+ {
+ "batch_loss": 61372.3828125,
+ "innovation_rms": 227155.88074555885,
+ "instruction_rms": 0.5486655909039636,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 826916601856.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.6978767445271862e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 445440128.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1177.876220703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 260.46142578125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1485831936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 529995072.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 341027968.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8017356800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3907849472.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 445440128.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1177.7744140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 260.46136474609375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1485831936.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2517333.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2032787.0858247776,
+ "step": 224,
+ "teaching_rms": 227155.88074555885,
+ "traffic_rms": 2032787.0858247136
+ },
+ {
+ "batch_loss": 47320.8046875,
+ "innovation_rms": 292524.7436785483,
+ "instruction_rms": 0.5569194746397228,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 838377078784.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.807186154716165e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 485643072.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1176.646728515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 172.31448364257812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1470932224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 566479808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 364847328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8419267584.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4019106304.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 485643072.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1176.544921875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 172.31442260742188,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1470932224.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 85570725663.89018,
+ "predictor_updated": true,
+ "raw_apical_rms": 2002933.8762728567,
+ "step": 225,
+ "teaching_rms": 292524.7436785483,
+ "traffic_rms": 2002933.8762727836
+ },
+ {
+ "batch_loss": 63551.8515625,
+ "innovation_rms": 377103.2010462588,
+ "instruction_rms": 0.43205523181391575,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 844909445120.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.936466932376546e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 852530560.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1172.9996337890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 203.85333251953125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1453840640.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 606206848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 397270432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8869918720.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4506509824.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 852530560.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1172.8978271484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 203.85345458984375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1453840640.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2033106.8579130948,
+ "step": 226,
+ "teaching_rms": 377103.2010462588,
+ "traffic_rms": 2033106.857913067
+ },
+ {
+ "batch_loss": 53065.78515625,
+ "innovation_rms": 353942.19966151274,
+ "instruction_rms": 0.3719480515461305,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 860967665664.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.019279439378048e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 788934528.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1170.087646484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 226.94032287597656,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1442941440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 649858432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 436515584.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9349801984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4798836736.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 788934528.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1169.98583984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 226.94046020507812,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1442941440.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 1859683.4415927771,
+ "step": 227,
+ "teaching_rms": 353942.19966151274,
+ "traffic_rms": 1859683.441592739
+ },
+ {
+ "batch_loss": 40476.15625,
+ "innovation_rms": 363146.48905380035,
+ "instruction_rms": 0.3844038494175006,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 879610036224.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.108421946608444e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 714277760.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1167.455078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 320.03533935546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1433134208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 695821312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 459628928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9855869952.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5060679680.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 714277760.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1167.3533935546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 320.03515625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1433134208.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2283538.09111623,
+ "step": 228,
+ "teaching_rms": 363146.48905380035,
+ "traffic_rms": 2283538.091116197
+ },
+ {
+ "batch_loss": 42382.60546875,
+ "innovation_rms": 420990.89532110206,
+ "instruction_rms": 0.46736560768069707,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 883795951616.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.244668214831057e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1195297920.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.552001953125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 341.15771484375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1421157760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 745730368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 499090560.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10405702656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5498322432.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1195297920.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.4503173828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 341.15753173828125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1421157760.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2412778.0152376425,
+ "step": 229,
+ "teaching_rms": 420990.89532110206,
+ "traffic_rms": 2412778.015237597
+ },
+ {
+ "batch_loss": 35443.48828125,
+ "innovation_rms": 452342.9219581274,
+ "instruction_rms": 0.39857901221686254,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 912677273600.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.284546847338204e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1553581184.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.647216796875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 260.7667541503906,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1417332224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 800424832.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 546944768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11003756544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5980542464.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1553581184.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.5455322265625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 260.7666015625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1417332224.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2285493.9935735306,
+ "step": 230,
+ "teaching_rms": 452342.9219581274,
+ "traffic_rms": 2285493.9935735017
+ },
+ {
+ "batch_loss": 29823.654296875,
+ "innovation_rms": 360409.6094603252,
+ "instruction_rms": 0.4003598762410513,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 923340374016.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.356150651471919e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1474475264.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.7213134765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228.92355346679688,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1413980288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 856431040.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 560062016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11603954688.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6001984000.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1474475264.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.61962890625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 228.9233856201172,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1413980288.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2740768.499246286,
+ "step": 231,
+ "teaching_rms": 360409.6094603252,
+ "traffic_rms": 2740768.49924626
+ },
+ {
+ "batch_loss": 25020.474609375,
+ "innovation_rms": 429043.8884785636,
+ "instruction_rms": 0.3899761055049495,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 933111857152.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.450767230006188e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1323940352.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.7764892578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 211.84378051757812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1416701056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 914769984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 583389696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12226720768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6227665408.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1323940352.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.6748046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 211.84365844726562,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1416701056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2723547.3775852034,
+ "step": 232,
+ "teaching_rms": 429043.8884785636,
+ "traffic_rms": 2723547.377585176
+ },
+ {
+ "batch_loss": 26773.701171875,
+ "innovation_rms": 552767.5592564093,
+ "instruction_rms": 0.39341679643041794,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 933835309056.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.573941674209569e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6974867968.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.814453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.651123046875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1429583872.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 975490624.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 607206656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12918496256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6917754880.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6974867968.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.7127685546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.6511993408203,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1429583872.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 2640288.2293402706,
+ "step": 233,
+ "teaching_rms": 552767.5592564093,
+ "traffic_rms": 2640288.229340253
+ },
+ {
+ "batch_loss": 45497.70703125,
+ "innovation_rms": 116656.69562317015,
+ "instruction_rms": 0.33613290212667246,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 941200441344.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.634773502141441e+25,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6128275968.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.8369140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 115.86117553710938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1442705920.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1031835200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 563445440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 13546566656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6280705536.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6128275968.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.7352294921875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 115.86106872558594,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1442705920.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3326859.489486546,
+ "step": 234,
+ "teaching_rms": 116656.69562317015,
+ "traffic_rms": 3326859.4894865337
+ },
+ {
+ "batch_loss": 38880.703125,
+ "innovation_rms": 60644.96976045326,
+ "instruction_rms": 0.4712603075575942,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 949783887872.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2758963550393502e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5493879296.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.8468017578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.51997375488281,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1831979904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1083456512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 516213344.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14113206272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5666393600.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5493879296.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.7451171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.52040100097656,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1831979904.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3477409.3715051236,
+ "step": 235,
+ "teaching_rms": 60644.96976045326,
+ "traffic_rms": 3477409.371505096
+ },
+ {
+ "batch_loss": 36402.28515625,
+ "innovation_rms": 48838.18042056937,
+ "instruction_rms": 0.6234455676231996,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 979473530880.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.10900794891238e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4936522240.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.666015625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.8526611328125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2325632256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1131032960.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 475764288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 14623985664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5107791360.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4936522240.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.5643310546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.852783203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2325632256.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3747452.6000049957,
+ "step": 236,
+ "teaching_rms": 48838.18042056937,
+ "traffic_rms": 3747452.60000495
+ },
+ {
+ "batch_loss": 12604.3515625,
+ "innovation_rms": 37221.4593177079,
+ "instruction_rms": 0.45975671638222576,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 993278427136.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.0377812216051747e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4438140928.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.49169921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139.8726348876953,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2769446400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1174214272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 431813664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15084129280.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4601432064.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4438140928.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.3900146484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139.8727569580078,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2769446400.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3756804.2488317257,
+ "step": 237,
+ "teaching_rms": 37221.4593177079,
+ "traffic_rms": 3756804.2488316996
+ },
+ {
+ "batch_loss": 35365.375,
+ "innovation_rms": 36475.079716712666,
+ "instruction_rms": 0.4504275625935183,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1044392312832.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.435364500783445e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3991708672.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.2877197265625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146.42507934570312,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3168617216.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1213375232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 391609088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15498660864.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4145314304.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3991708672.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.18603515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146.4251708984375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3168617216.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3908299.983493553,
+ "step": 238,
+ "teaching_rms": 36475.079716712666,
+ "traffic_rms": 3908299.9834935307
+ },
+ {
+ "batch_loss": 33107.6484375,
+ "innovation_rms": 33255.90165077961,
+ "instruction_rms": 0.3325512386839938,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1259090477056.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.509601100364288e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3589673728.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.17724609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136.35504150390625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3527584512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1248928896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 355537024.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 15872082944.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3734225664.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3589673728.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.0755615234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136.3551483154297,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3527584512.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 3953457.0148460725,
+ "step": 239,
+ "teaching_rms": 33255.90165077961,
+ "traffic_rms": 3953457.014846061
+ },
+ {
+ "batch_loss": 22428.5703125,
+ "innovation_rms": 43196.33435037459,
+ "instruction_rms": 0.35494974529631135,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1498018480128.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2027300010021279e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3227069952.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1166.06884765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 180.533935546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3850291456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1281127168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 321982816.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16208794624.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3367119104.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3227069952.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.9671630859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 180.53384399414062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3850291456.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5191884.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4094302.808851666,
+ "step": 240,
+ "teaching_rms": 43196.33435037459,
+ "traffic_rms": 4094302.808851653
+ },
+ {
+ "batch_loss": 16168.6474609375,
+ "innovation_rms": 30654.40133822433,
+ "instruction_rms": 0.3285195341541306,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1938812174336.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.003961009429769e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2902155776.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.9818115234375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175.31924438476562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4140507136.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1309996544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 288693952.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16512154624.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3033596416.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2902155776.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.8802490234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175.31903076171875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4140507136.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 939692493.8131386,
+ "predictor_updated": true,
+ "raw_apical_rms": 4129797.2747581596,
+ "step": 241,
+ "teaching_rms": 30654.40133822433,
+ "traffic_rms": 4129797.2747581494
+ },
+ {
+ "batch_loss": 16667.37890625,
+ "innovation_rms": 44857.23548647746,
+ "instruction_rms": 0.3244161272255543,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2131074744320.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.1449776610100757e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2608418048.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.891845703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151.8714141845703,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4401349120.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1335974784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 259782096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 16786050048.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2738954752.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2608418048.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.790283203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151.87120056152344,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4401349120.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4277761.013813225,
+ "step": 242,
+ "teaching_rms": 44857.23548647746,
+ "traffic_rms": 4277761.013813213
+ },
+ {
+ "batch_loss": 15875.578125,
+ "innovation_rms": 25695.28793292796,
+ "instruction_rms": 0.30108026654901493,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2577480024064.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.2211961280189125e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2346007808.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.84130859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.72335815429688,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4635950080.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1359277440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 233026288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17032767488.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2467171072.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2346007808.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.73974609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.7232666015625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4635950080.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4321116.781488087,
+ "step": 243,
+ "teaching_rms": 25695.28793292796,
+ "traffic_rms": 4321116.781488077
+ },
+ {
+ "batch_loss": 18893.787109375,
+ "innovation_rms": 60110.24710292211,
+ "instruction_rms": 0.46538129286147295,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3012261576704.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.925974498134396e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2106777088.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.8255615234375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 163.81040954589844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4846627840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1380423680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 211462992.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17255905280.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2231380224.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2106777088.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.7239990234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 163.8102264404297,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4846627840.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4440015.530286347,
+ "step": 244,
+ "teaching_rms": 60110.24710292211,
+ "traffic_rms": 4440015.530286325
+ },
+ {
+ "batch_loss": 25534.232421875,
+ "innovation_rms": 29072.03549010647,
+ "instruction_rms": 0.3317594136184242,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3448670519296.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.187399161086992e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1894207360.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.7996826171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.6917266845703,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5036048384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1399570688.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191469936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17456996352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2010904320.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1894207360.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.6982421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.69190979003906,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5036048384.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4459752.291894875,
+ "step": 245,
+ "teaching_rms": 29072.03549010647,
+ "traffic_rms": 4459752.291894864
+ },
+ {
+ "batch_loss": 23782.787109375,
+ "innovation_rms": 32827.6327414407,
+ "instruction_rms": 0.48139222446137014,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3826294980608.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.850413512323787e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1703640576.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.780517578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.29315185546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5206412288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1416858752.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 172880672.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17638291456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1812951424.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1703640576.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.6790771484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.29335021972656,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5206412288.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4391830.220164255,
+ "step": 246,
+ "teaching_rms": 32827.6327414407,
+ "traffic_rms": 4391830.220164233
+ },
+ {
+ "batch_loss": 18786.4921875,
+ "innovation_rms": 73578.17574637274,
+ "instruction_rms": 0.3039491095660204,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4118648717312.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.428370369358564e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1528039040.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.7515869140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 206.51309204101562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5359216128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1432833920.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 159751296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17803059200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1647680896.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1528039040.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.650146484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 206.5130157470703,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5359216128.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4632013.564697299,
+ "step": 247,
+ "teaching_rms": 73578.17574637274,
+ "traffic_rms": 4632013.564697292
+ },
+ {
+ "batch_loss": 20138.71484375,
+ "innovation_rms": 40712.07793254022,
+ "instruction_rms": 0.3609414143890433,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4305598283776.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.61496110412552e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1372129408.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.5511474609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 201.7037353515625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5496429056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1447299712.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 144658480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 17951907840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1488488960.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1372129408.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.44970703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 201.70391845703125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5496429056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4671178.833625905,
+ "step": 248,
+ "teaching_rms": 40712.07793254022,
+ "traffic_rms": 4671178.833625892
+ },
+ {
+ "batch_loss": 16898.19921875,
+ "innovation_rms": 26097.84145426883,
+ "instruction_rms": 0.3262146225004968,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4706391293952.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.066212208329553e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1233363968.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.2633056640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 287.4990234375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5619765248.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1460336384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 130366976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18086084608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1341766016.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1233363968.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1165.1619873046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 287.49920654296875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5619765248.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4673597.820826593,
+ "step": 249,
+ "teaching_rms": 26097.84145426883,
+ "traffic_rms": 4673597.820826583
+ },
+ {
+ "batch_loss": 27242.8984375,
+ "innovation_rms": 18797.758002963572,
+ "instruction_rms": 0.30434390477175133,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4813822623744.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.755883137770526e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1109054080.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.99267578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112.81078338623047,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5730670592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1472160512.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 118240984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18206926848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1208420864.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1109054080.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.891357421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 112.81069946289062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5730670592.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4632596.223839875,
+ "step": 250,
+ "teaching_rms": 18797.758002963572,
+ "traffic_rms": 4632596.223839867
+ },
+ {
+ "batch_loss": 31675.142578125,
+ "innovation_rms": 22569.09957971517,
+ "instruction_rms": 0.3541924432614959,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5119111331840.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.9485421757391e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 996957248.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.7374267578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.73797607421875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5830366208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1482853888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106933320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18315816960.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1088907776.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 996957248.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.6361083984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 169.73806762695312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5830366208.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4778356.600810407,
+ "step": 251,
+ "teaching_rms": 22569.09957971517,
+ "traffic_rms": 4778356.6008103965
+ },
+ {
+ "batch_loss": 28190.23828125,
+ "innovation_rms": 100457.38563895186,
+ "instruction_rms": 0.3702324803714586,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5372838412288.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0616066386967036e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 893730752.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.49609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 259.3755798339844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5919739392.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1493115904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102619608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18417039360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1012220736.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 893730752.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.394775390625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 259.37542724609375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5919739392.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4807803.888919082,
+ "step": 252,
+ "teaching_rms": 100457.38563895186,
+ "traffic_rms": 4807803.888919069
+ },
+ {
+ "batch_loss": 21967.595703125,
+ "innovation_rms": 102722.29924708209,
+ "instruction_rms": 0.3433439093017331,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5447508557824.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0439534883104403e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 798210624.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.295654296875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191.9983673095703,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5999560704.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1502931840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98159992.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18511267840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 942285248.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 798210624.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.1943359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191.99819946289062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5999560704.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4762533.925711685,
+ "step": 253,
+ "teaching_rms": 102722.29924708209,
+ "traffic_rms": 4762533.925711675
+ },
+ {
+ "batch_loss": 20291.03515625,
+ "innovation_rms": 87535.80305688016,
+ "instruction_rms": 0.3304990730468482,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5772906856448.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.119870724125701e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723818048.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.1036376953125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 194.83494567871094,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6070644736.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1512131456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91996400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18598393856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 871267520.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 723818048.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1164.0023193359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 194.83486938476562,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6070644736.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4817340.9989024345,
+ "step": 254,
+ "teaching_rms": 87535.80305688016,
+ "traffic_rms": 4817340.998902425
+ },
+ {
+ "batch_loss": 37330.21875,
+ "innovation_rms": 60160.92276924165,
+ "instruction_rms": 0.3367557003252622,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5841287643136.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1777011782524089e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 655960320.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.919189453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 170.2043914794922,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6134323712.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1520859648.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 87281552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18677811200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 794166336.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 655960320.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.81787109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 170.20431518554688,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6134323712.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4861252.8157174885,
+ "step": 255,
+ "teaching_rms": 60160.92276924165,
+ "traffic_rms": 4861252.815717479
+ },
+ {
+ "batch_loss": 32670.896484375,
+ "innovation_rms": 31434.40579808684,
+ "instruction_rms": 0.3964382509095306,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5927480590336.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1516900294872766e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 589485824.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.741455078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 174.3753204345703,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6191443968.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1528669056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78093712.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18749560832.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 717499328.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 589485824.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.64013671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 174.3751983642578,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6191443968.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5011309.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5060788.900529707,
+ "step": 256,
+ "teaching_rms": 31434.40579808684,
+ "traffic_rms": 5060788.900529694
+ },
+ {
+ "batch_loss": 14684.40625,
+ "innovation_rms": 153625.7911759871,
+ "instruction_rms": 0.3457337040068099,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6077520805888.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2068758403244155e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 564233600.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.5699462890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 135.3002471923828,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6241851904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1536434560.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77655088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18820653056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 710927744.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 564233600.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.4686279296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 135.30010986328125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6241851904.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 23600875176.58963,
+ "predictor_updated": true,
+ "raw_apical_rms": 4905830.843107249,
+ "step": 257,
+ "teaching_rms": 153625.7911759871,
+ "traffic_rms": 4905830.843107238
+ },
+ {
+ "batch_loss": 22307.0,
+ "innovation_rms": 24713.61551200399,
+ "instruction_rms": 0.3266619247626138,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6066600935424.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2148403475160993e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 506526944.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.4039306640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 217.77610778808594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6287077376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1543411840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69773200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18884812800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 641604736.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 506526944.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.3026123046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 217.7760009765625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6287077376.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4808724.546894984,
+ "step": 258,
+ "teaching_rms": 24713.61551200399,
+ "traffic_rms": 4808724.546894975
+ },
+ {
+ "batch_loss": 8998.3876953125,
+ "innovation_rms": 20766.37295287977,
+ "instruction_rms": 0.30810872697100095,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6387038945280.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3972751636598056e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 455192096.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.242919921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 420.52093505859375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6327670272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1549589248.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61773976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18942672896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 578597120.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 455192096.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.1416015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 420.52081298828125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6327670272.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4772981.176546472,
+ "step": 259,
+ "teaching_rms": 20766.37295287977,
+ "traffic_rms": 4772981.176546462
+ },
+ {
+ "batch_loss": 29020.8984375,
+ "innovation_rms": 38590.366103194385,
+ "instruction_rms": 0.299116611473922,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6765274988544.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5462168316420831e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 410512096.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1163.0863037109375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 311.21722412109375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6364014592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1555039360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 55979600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 18995218432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 525459648.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 410512096.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.9849853515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 311.21710205078125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6364014592.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4964871.287782483,
+ "step": 260,
+ "teaching_rms": 38590.366103194385,
+ "traffic_rms": 4964871.287782474
+ },
+ {
+ "batch_loss": 10212.9365234375,
+ "innovation_rms": 24690.16668799856,
+ "instruction_rms": 0.3160813242332674,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6708008058880.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.569414984752532e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 368284928.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.9337158203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 279.9462890625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6396589568.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1559987840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58686724.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19042668544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 474490944.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 368284928.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.8323974609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 279.9461669921875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6396589568.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5102460.410022555,
+ "step": 261,
+ "teaching_rms": 24690.16668799856,
+ "traffic_rms": 5102460.410022546
+ },
+ {
+ "batch_loss": 21458.712890625,
+ "innovation_rms": 31340.457511443958,
+ "instruction_rms": 0.299063643652926,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6983143915520.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6832275587645526e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 330935136.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.7847900390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 250.29600524902344,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6425793024.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1564297344.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 62447184.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19085623296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 429547200.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 330935136.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.6834716796875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 250.29588317871094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6425793024.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5023820.246208812,
+ "step": 262,
+ "teaching_rms": 31340.457511443958,
+ "traffic_rms": 5023820.246208804
+ },
+ {
+ "batch_loss": 31694.39453125,
+ "innovation_rms": 48051.11434982608,
+ "instruction_rms": 0.31536368750773885,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7164634595328.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6984828095099768e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 299906496.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.61767578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191.05303955078125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6451870208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1567987968.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65999348.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19124865024.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 392422176.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 299906496.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.516357421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 191.05282592773438,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6451870208.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4893281.780255636,
+ "step": 263,
+ "teaching_rms": 48051.11434982608,
+ "traffic_rms": 4893281.780255628
+ },
+ {
+ "batch_loss": 30249.234375,
+ "innovation_rms": 37570.19868389409,
+ "instruction_rms": 0.4791700592327943,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7583737839616.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.736908705581087e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 268367376.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.4549560546875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 202.26058959960938,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6475123200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1571040128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65892128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19160547328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 356825408.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 268367376.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.3536376953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 202.26068115234375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6475123200.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5119380.710832737,
+ "step": 264,
+ "teaching_rms": 37570.19868389409,
+ "traffic_rms": 5119380.710832718
+ },
+ {
+ "batch_loss": 45324.26171875,
+ "innovation_rms": 38572.09670829838,
+ "instruction_rms": 0.5052851824223913,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7959664918528.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7604682377866375e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241898688.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.341064453125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 301.904541015625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6495878144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1573501056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67701536.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19193012224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 324657664.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241898688.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.23974609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 301.9045104980469,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6495878144.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4970777.840102316,
+ "step": 265,
+ "teaching_rms": 38572.09670829838,
+ "traffic_rms": 4970777.8401022935
+ },
+ {
+ "batch_loss": 44527.3359375,
+ "innovation_rms": 41395.37181599149,
+ "instruction_rms": 0.363737336857288,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8102706413568.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8114863240843197e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 218683456.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.2281494140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 257.0477294921875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6514288128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1575614720.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 73624800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19222767616.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 297559840.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 218683456.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.1268310546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 257.0476989746094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6514288128.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4961609.849836937,
+ "step": 266,
+ "teaching_rms": 41395.37181599149,
+ "traffic_rms": 4961609.849836926
+ },
+ {
+ "batch_loss": 51913.6484375,
+ "innovation_rms": 30168.22507330105,
+ "instruction_rms": 0.3393226108418625,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8198374817792.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.887037576497158e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 195653632.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.1148681640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 231.3328399658203,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6530732032.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1577368576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 80388296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19249696768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 269285376.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 195653632.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.0135498046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 231.3328094482422,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6530732032.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4947700.205039102,
+ "step": 267,
+ "teaching_rms": 30168.22507330105,
+ "traffic_rms": 4947700.205039091
+ },
+ {
+ "batch_loss": 11321.0537109375,
+ "innovation_rms": 52089.44205622624,
+ "instruction_rms": 0.31188105751742545,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8131292168192.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8887404618508808e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176011824.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1162.0048828125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 194.19468688964844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6545355264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1579013632.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85123232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19274573824.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 277454912.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176011824.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.9036865234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 194.1946563720703,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6545355264.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5254237.699227288,
+ "step": 268,
+ "teaching_rms": 52089.44205622624,
+ "traffic_rms": 5254237.69922728
+ },
+ {
+ "batch_loss": 25115.96484375,
+ "innovation_rms": 34757.40799760692,
+ "instruction_rms": 0.3522857684442455,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8182441181184.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9221673788223874e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157678432.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.8399658203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 165.8302001953125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6558416384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1580259456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91316416.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19297069056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 328311712.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 157678432.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.73876953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 165.83016967773438,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6558416384.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5066378.166577411,
+ "step": 269,
+ "teaching_rms": 34757.40799760692,
+ "traffic_rms": 5066378.166577401
+ },
+ {
+ "batch_loss": 28549.693359375,
+ "innovation_rms": 39609.22178430456,
+ "instruction_rms": 0.3285028856507734,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8384745046016.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8905704969221549e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 171164368.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.6610107421875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 147.0679931640625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6570011136.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1580909696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 98505832.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19317573632.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 364831040.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 171164368.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.559814453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 147.06817626953125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6570011136.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5083061.977052491,
+ "step": 270,
+ "teaching_rms": 39609.22178430456,
+ "traffic_rms": 5083061.977052482
+ },
+ {
+ "batch_loss": 28431.220703125,
+ "innovation_rms": 47477.59880472841,
+ "instruction_rms": 0.30712453120475514,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8273550376960.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8907547872741489e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 253815712.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.48828125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.2524642944336,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6580203008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1581543680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 103461000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19336423424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 414159488.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 253815712.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.3870849609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126.25234985351562,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6580203008.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5143440.888784564,
+ "step": 271,
+ "teaching_rms": 47477.59880472841,
+ "traffic_rms": 5143440.888784556
+ },
+ {
+ "batch_loss": 50933.0390625,
+ "innovation_rms": 42683.13508510996,
+ "instruction_rms": 0.33889529050954076,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8605498605568.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9527549729460686e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136462336.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.3692626953125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 259.04937744140625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6589198848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1581754368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111302016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19353751552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 421065216.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136462336.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.26806640625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 259.0491943359375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6589198848.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4750556.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5096238.338105236,
+ "step": 272,
+ "teaching_rms": 42683.13508510996,
+ "traffic_rms": 5096238.338105226
+ },
+ {
+ "batch_loss": 15174.265625,
+ "innovation_rms": 42423.91314163517,
+ "instruction_rms": 0.32424469830473596,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8188621488128.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8272389580795521e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 342951552.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.25048828125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207.44473266601562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6597139456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1581863808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 121670576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19369543680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 460421952.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 342951552.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.1492919921875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 207.44454956054688,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6597139456.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 1799788227.8843603,
+ "predictor_updated": true,
+ "raw_apical_rms": 5246583.461879788,
+ "step": 273,
+ "teaching_rms": 42423.91314163517,
+ "traffic_rms": 5246583.46187978
+ },
+ {
+ "batch_loss": 34124.01171875,
+ "innovation_rms": 31093.000181020114,
+ "instruction_rms": 0.39580261541356937,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7958014459904.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7444249421343845e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 274925856.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.1063232421875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 154.07412719726562,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6604130304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1581779328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 115964600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19383986176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 439484416.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 274925856.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1161.005126953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 154.07394409179688,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6604130304.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5093130.509491906,
+ "step": 274,
+ "teaching_rms": 31093.000181020114,
+ "traffic_rms": 5093130.5094918925
+ },
+ {
+ "batch_loss": 14405.111328125,
+ "innovation_rms": 28494.855657507836,
+ "instruction_rms": 0.3617020718648119,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7897891733504.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7551626590431335e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246070640.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.90283203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176.76861572265625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6610262016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1581545088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 108805112.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19397208064.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 408137696.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 246070640.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.8016357421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176.76852416992188,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6610262016.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5095394.268922249,
+ "step": 275,
+ "teaching_rms": 28494.855657507836,
+ "traffic_rms": 5095394.268922238
+ },
+ {
+ "batch_loss": 33702.01171875,
+ "innovation_rms": 29891.09376666936,
+ "instruction_rms": 0.39261232764858944,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8023545741312.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.7319491582416153e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 223413552.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.6744384765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140.29153442382812,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6615607296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1580862208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 101738352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19409389568.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 376639072.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 223413552.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.5732421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140.2914581298828,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6615607296.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5062260.540829828,
+ "step": 276,
+ "teaching_rms": 29891.09376666936,
+ "traffic_rms": 5062260.540829815
+ },
+ {
+ "batch_loss": 13361.306640625,
+ "innovation_rms": 26368.51642970773,
+ "instruction_rms": 0.33827903327145287,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7714069544960.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6083640009707822e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 202467568.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.425537109375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.982177734375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6620262912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1579806976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94744728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19420573696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 345837536.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 202467568.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.3243408203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 186.9822540283203,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6620262912.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5058527.714485493,
+ "step": 277,
+ "teaching_rms": 26368.51642970773,
+ "traffic_rms": 5058527.714485483
+ },
+ {
+ "batch_loss": 9362.634765625,
+ "innovation_rms": 27919.727588981543,
+ "instruction_rms": 0.3234287051692243,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7655262781440.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.576539028769427e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 184398720.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.12451171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 134.3793487548828,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6624286208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1578528256.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 88407608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19430885376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 317972480.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 184398720.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1160.0233154296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 134.37942504882812,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6624286208.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5082041.175827217,
+ "step": 278,
+ "teaching_rms": 27919.727588981543,
+ "traffic_rms": 5082041.175827209
+ },
+ {
+ "batch_loss": 15459.287109375,
+ "innovation_rms": 25021.42970220968,
+ "instruction_rms": 0.28926554361228485,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7813413732352.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5809896230612076e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167488448.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.8243408203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 118.1325912475586,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6627772928.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1576902016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 82565424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19440367616.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 291572576.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 167488448.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.72314453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 118.13247680664062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6627772928.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5227868.257235712,
+ "step": 279,
+ "teaching_rms": 25021.42970220968,
+ "traffic_rms": 5227868.257235705
+ },
+ {
+ "batch_loss": 16656.490234375,
+ "innovation_rms": 21499.610173379526,
+ "instruction_rms": 0.3122682504136604,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7608719638528.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5140212717309848e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153327424.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.5341796875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.13111877441406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6630807552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1575061760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 78744160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19449008128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 271614304.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 153327424.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.4329833984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 145.13113403320312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6630807552.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 4993222.707797375,
+ "step": 280,
+ "teaching_rms": 21499.610173379526,
+ "traffic_rms": 4993222.707797365
+ },
+ {
+ "batch_loss": 14447.544921875,
+ "innovation_rms": 59386.85366706049,
+ "instruction_rms": 0.30685598114387536,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7561497018368.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4718734425164008e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140232528.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.208740234375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136.9683074951172,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6633190400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1573194496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74705384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19457910784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 252138608.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 140232528.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1159.1075439453125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136.96873474121094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6633190400.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5205791.769878242,
+ "step": 281,
+ "teaching_rms": 59386.85366706049,
+ "traffic_rms": 5205791.769878234
+ },
+ {
+ "batch_loss": 8560.4677734375,
+ "innovation_rms": 32195.78210967037,
+ "instruction_rms": 0.420227164199938,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7060313866240.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3395990128579255e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 128586280.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.8897705078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.59022521972656,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6635127296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1571162112.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 72082856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19466240000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 236193168.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 128586280.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.78857421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.59065246582031,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6635127296.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5412868.95132445,
+ "step": 282,
+ "teaching_rms": 32195.78210967037,
+ "traffic_rms": 5412868.951324437
+ },
+ {
+ "batch_loss": 13977.88671875,
+ "innovation_rms": 36769.517518258195,
+ "instruction_rms": 0.300714819310176,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6842064306176.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2731755045606638e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120432808.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.5343017578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.82494354248047,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6636615680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1569194112.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69194096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19474176000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 221167232.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120432808.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.43310546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.82537078857422,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6636615680.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5365109.910211456,
+ "step": 283,
+ "teaching_rms": 36769.517518258195,
+ "traffic_rms": 5365109.910211449
+ },
+ {
+ "batch_loss": 18154.0078125,
+ "innovation_rms": 30432.665066783353,
+ "instruction_rms": 0.31875092751799194,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6690678243328.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2337821136563756e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 128072512.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.1414794921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79.97711181640625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637774336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1567215232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67233960.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19481585664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 208289488.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 128072512.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1158.040283203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 79.9771728515625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637774336.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5560327.410949521,
+ "step": 284,
+ "teaching_rms": 30432.665066783353,
+ "traffic_rms": 5560327.410949512
+ },
+ {
+ "batch_loss": 15619.244140625,
+ "innovation_rms": 26295.4336410921,
+ "instruction_rms": 0.29645368145298207,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6568736194560.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2225832576605744e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122004904.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1157.760009765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.33749389648438,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6638681088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1565289856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65474116.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19488423936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197524336.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122004904.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1157.6588134765625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 120.33748626708984,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6638681088.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5293406.288020419,
+ "step": 285,
+ "teaching_rms": 26295.4336410921,
+ "traffic_rms": 5293406.2880204115
+ },
+ {
+ "batch_loss": 9000.8876953125,
+ "innovation_rms": 34819.79695137443,
+ "instruction_rms": 0.311719506050369,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6239458164736.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1274108086297369e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 123973576.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1157.3392333984375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.31179809570312,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639290368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1563340160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 64359588.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19494924288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 188989120.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 123973576.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1157.238037109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 148.31179809570312,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639290368.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5463911.678745798,
+ "step": 286,
+ "teaching_rms": 34819.79695137443,
+ "traffic_rms": 5463911.678745791
+ },
+ {
+ "batch_loss": 7649.99658203125,
+ "innovation_rms": 37446.26350589938,
+ "instruction_rms": 0.298455359387703,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6035666370560.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.068757718261713e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136784832.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.948974609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139.8455810546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639588352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1561426944.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63566508.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19501191168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 181759376.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 136784832.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.8477783203125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 139.8455810546875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639588352.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5466623.489168017,
+ "step": 287,
+ "teaching_rms": 37446.26350589938,
+ "traffic_rms": 5466623.48916801
+ },
+ {
+ "batch_loss": 5813.4541015625,
+ "innovation_rms": 28013.43886867528,
+ "instruction_rms": 0.2752939263061882,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6132032602112.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.055582669951993e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 131249208.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.5709228515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.69163513183594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639733760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1559199360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63472356.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19507005440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 178181824.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 131249208.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.4698486328125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 125.69163513183594,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639733760.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4667962.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5305036.729367915,
+ "step": 288,
+ "teaching_rms": 28013.43886867528,
+ "traffic_rms": 5305036.729367909
+ },
+ {
+ "batch_loss": 9953.83203125,
+ "innovation_rms": 26116.170171876416,
+ "instruction_rms": 0.28979563492200255,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6339365437440.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0904462488667506e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126797264.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.140380859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106.27611541748047,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639759360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1556680448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 63845140.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19512365056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 175723856.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 126797264.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156.039306640625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 106.27618408203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639759360.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 682054366.09748,
+ "predictor_updated": true,
+ "raw_apical_rms": 5240161.078208261,
+ "step": 289,
+ "teaching_rms": 26116.170171876416,
+ "traffic_rms": 5240161.078208253
+ },
+ {
+ "batch_loss": 9837.3173828125,
+ "innovation_rms": 29476.416583391954,
+ "instruction_rms": 0.29854587293559265,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6350212431872.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1110570173812351e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146041904.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1155.6639404296875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.70054626464844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639643648.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1554116480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65027776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19517372416.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 178096208.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 146041904.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1155.5628662109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.70060729980469,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639643648.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5316569.0586697925,
+ "step": 290,
+ "teaching_rms": 29476.416583391954,
+ "traffic_rms": 5316569.058669785
+ },
+ {
+ "batch_loss": 9293.9921875,
+ "innovation_rms": 28089.05257153578,
+ "instruction_rms": 0.2744948507989897,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6444149112832.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.110153274495576e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 160581888.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1155.2236328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.45584869384766,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639412224.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1551320576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65216008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19522056192.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 177621872.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 160581888.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1155.12255859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.45594787597656,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639412224.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5375610.81719164,
+ "step": 291,
+ "teaching_rms": 28089.05257153578,
+ "traffic_rms": 5375610.817191633
+ },
+ {
+ "batch_loss": 5329.16259765625,
+ "innovation_rms": 28884.685048583815,
+ "instruction_rms": 0.3339362163438091,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6187855642624.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0341306118485852e+28,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176426704.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1154.809326171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.44818115234375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639069696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1548445696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65757868.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19526461440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 177834704.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 176426704.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1154.708251953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 102.44828796386719,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6639069696.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5322461.111822759,
+ "step": 292,
+ "teaching_rms": 28884.685048583815,
+ "traffic_rms": 5322461.1118227495
+ },
+ {
+ "batch_loss": 6299.64404296875,
+ "innovation_rms": 41940.65562013781,
+ "instruction_rms": 0.2594832563267098,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6083904012288.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.862729102899825e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197317952.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1154.387451171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.07569885253906,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6638455296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1545624320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66496336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19531003904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 181741728.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197317952.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1154.286376953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.07579803466797,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6638455296.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5421744.323425832,
+ "step": 293,
+ "teaching_rms": 41940.65562013781,
+ "traffic_rms": 5421744.323425828
+ },
+ {
+ "batch_loss": 7099.45263671875,
+ "innovation_rms": 31879.52310978517,
+ "instruction_rms": 0.28805380193760394,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5775357378560.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.120404377470665e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241040896.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.9962158203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 81.86488342285156,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637795328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1542658944.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 71258528.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19535204352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197196992.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 241040896.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.8951416015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 81.86498260498047,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637795328.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5533294.510877066,
+ "step": 294,
+ "teaching_rms": 31879.52310978517,
+ "traffic_rms": 5533294.51087706
+ },
+ {
+ "batch_loss": 9338.900390625,
+ "innovation_rms": 36907.51489520035,
+ "instruction_rms": 0.2718689574382092,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5643854938112.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.730125580085524e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 298292352.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.632568359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 71.8516616821289,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637080576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1539610752.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75951376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19539161088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 219611456.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 298292352.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.531494140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 71.85150909423828,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6637080576.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5419196.281501278,
+ "step": 295,
+ "teaching_rms": 36907.51489520035,
+ "traffic_rms": 5419196.281501273
+ },
+ {
+ "batch_loss": 8134.41259765625,
+ "innovation_rms": 40068.56072146317,
+ "instruction_rms": 0.2862345001891908,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5635933470720.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.657858025206549e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 371793152.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.293701171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 109.01856994628906,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6636306432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1536510336.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 81545184.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19542892544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 249666256.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 371793152.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1153.192626953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 109.01847076416016,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6636306432.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5408101.245042165,
+ "step": 296,
+ "teaching_rms": 40068.56072146317,
+ "traffic_rms": 5408101.245042158
+ },
+ {
+ "batch_loss": 3481.25634765625,
+ "innovation_rms": 50652.509749472934,
+ "instruction_rms": 0.27735963071673,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5242772520960.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.872229789425286e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 474888256.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.9771728515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.86774444580078,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6635437056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1533546496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91100264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19546507264.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 307916224.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 474888256.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.876220703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.86764526367188,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6635437056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5411371.2271033125,
+ "step": 297,
+ "teaching_rms": 50652.509749472934,
+ "traffic_rms": 5411371.227103306
+ },
+ {
+ "batch_loss": 6335.203125,
+ "innovation_rms": 62353.98957859104,
+ "instruction_rms": 0.280089351498284,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5073064165376.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.366561737918656e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 388259392.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.6807861328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 72.13352966308594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6634490368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1530488576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 111159448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19549997056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 418272512.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 388259392.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.579833984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 72.1333999633789,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6634490368.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5331566.664173202,
+ "step": 298,
+ "teaching_rms": 62353.98957859104,
+ "traffic_rms": 5331566.664173193
+ },
+ {
+ "batch_loss": 7658.0107421875,
+ "innovation_rms": 57189.72789488267,
+ "instruction_rms": 0.2568957183650676,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4885770665984.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.973898145459666e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 240265200.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.402587890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.04751586914062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6633465344.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1527473408.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 132328728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19553359872.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 488183872.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 240265200.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.3016357421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 110.04740905761719,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6633465344.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5270486.746913842,
+ "step": 299,
+ "teaching_rms": 57189.72789488267,
+ "traffic_rms": 5270486.746913835
+ },
+ {
+ "batch_loss": 4398.3466796875,
+ "innovation_rms": 60054.672255843696,
+ "instruction_rms": 0.28325664883746116,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4683582144512.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6.494408532550824e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308258560.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.05081939697266,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6632323584.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1524469760.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 151535792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19556698112.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 551675392.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 308258560.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1152.0396728515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 77.05071258544922,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6632323584.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5358184.90570292,
+ "step": 300,
+ "teaching_rms": 60054.672255843696,
+ "traffic_rms": 5358184.905702912
+ },
+ {
+ "batch_loss": 5005.2197265625,
+ "innovation_rms": 64712.34226226986,
+ "instruction_rms": 0.27414106303361613,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4409515835392.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.95282629364616e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 432408864.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.893310546875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91.14080810546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6631007232.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1521511040.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 168645088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19560114176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 617390336.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 432408864.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.7923583984375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 91.140625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6631007232.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5394376.468717898,
+ "step": 301,
+ "teaching_rms": 64712.34226226986,
+ "traffic_rms": 5394376.468717892
+ },
+ {
+ "batch_loss": 7563.0888671875,
+ "innovation_rms": 78994.98656954884,
+ "instruction_rms": 0.26385935960559426,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4265152610304.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.561944213942832e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 412805024.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.6593017578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117.11109924316406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6629587968.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1518429824.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 188414432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19563495424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 754348352.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 412805024.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.558349609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 117.11091613769531,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6629587968.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5748133.025704507,
+ "step": 302,
+ "teaching_rms": 78994.98656954884,
+ "traffic_rms": 5748133.025704501
+ },
+ {
+ "batch_loss": 10906.65625,
+ "innovation_rms": 66530.64975667647,
+ "instruction_rms": 0.2469861308876273,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4120665915392.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.224416021058776e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 236176160.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.4371337890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105.35260009765625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6628209664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1515223168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 197046160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19566624768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 823479616.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 236176160.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.336181640625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 105.35240936279297,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6628209664.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5704594.340303589,
+ "step": 303,
+ "teaching_rms": 66530.64975667647,
+ "traffic_rms": 5704594.340303584
+ },
+ {
+ "batch_loss": 9151.359375,
+ "innovation_rms": 71562.21115122947,
+ "instruction_rms": 0.27958138805504196,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4165537628160.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.489798850296602e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 356933312.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.2257080078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94.77873992919922,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6626825728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1512056064.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 205649312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19569616896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 893915456.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 356933312.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1151.124755859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 94.77854919433594,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6626825728.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4418653.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5771790.827099878,
+ "step": 304,
+ "teaching_rms": 71562.21115122947,
+ "traffic_rms": 5771790.827099872
+ },
+ {
+ "batch_loss": 20161.162109375,
+ "innovation_rms": 80050.41046003632,
+ "instruction_rms": 0.29558318701858693,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4357043781632.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.629067931130341e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 433448064.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1150.6553955078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.1385269165039,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6625265152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1508830592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 217510304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19572811776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 965729344.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 433448064.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1150.554443359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.13835144042969,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6625265152.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 6408068211.973412,
+ "predictor_updated": true,
+ "raw_apical_rms": 5927413.220073428,
+ "step": 305,
+ "teaching_rms": 80050.41046003632,
+ "traffic_rms": 5927413.220073424
+ },
+ {
+ "batch_loss": 4621.6923828125,
+ "innovation_rms": 78761.45231740986,
+ "instruction_rms": 0.28444147171330375,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4233039446016.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.372174145941284e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 229293792.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1150.130615234375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.96017456054688,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6623462912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1505719680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 231746432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19576242176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1037973760.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 229293792.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1150.0296630859375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 74.96025848388672,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6623462912.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5900977.750260071,
+ "step": 306,
+ "teaching_rms": 78761.45231740986,
+ "traffic_rms": 5900977.750260064
+ },
+ {
+ "batch_loss": 4077.45654296875,
+ "innovation_rms": 81228.45037684891,
+ "instruction_rms": 0.313071949521369,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4165488869376.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.412890389756587e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 401610112.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.650634765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.42526245117188,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6621440000.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1502760448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 247501696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19579865088.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1107401728.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 401610112.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.5496826171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 92.42518615722656,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6621440000.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 5993930.579147112,
+ "step": 307,
+ "teaching_rms": 81228.45037684891,
+ "traffic_rms": 5993930.579147104
+ },
+ {
+ "batch_loss": 5606.4423828125,
+ "innovation_rms": 92340.4040224468,
+ "instruction_rms": 0.3160377391465887,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3981865123840.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5.008759710885568e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 415402816.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.2071533203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 80.81185150146484,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6619018240.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1499816320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 262656896.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19583862784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1205041536.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 415402816.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1149.106201171875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 80.81177520751953,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6619018240.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6321191.109026512,
+ "step": 308,
+ "teaching_rms": 92340.4040224468,
+ "traffic_rms": 6321191.109026505
+ },
+ {
+ "batch_loss": 8863.2236328125,
+ "innovation_rms": 92025.44710261338,
+ "instruction_rms": 0.27764163086148724,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4060801138688.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.978292182929714e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 225636672.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1148.7965087890625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 103.3206558227539,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6616504320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1496742912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 270788320.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19587954688.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1294303232.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 225636672.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1148.695556640625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 103.3205795288086,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6616504320.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6274378.302993644,
+ "step": 309,
+ "teaching_rms": 92025.44710261338,
+ "traffic_rms": 6274378.3029936375
+ },
+ {
+ "batch_loss": 8888.9140625,
+ "innovation_rms": 97673.73355067195,
+ "instruction_rms": 0.2775079122299352,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4147207733248.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.9048998193841004e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 429587904.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1148.3660888671875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 236.83810424804688,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6613828096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1493539328.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 282133440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19592318976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1379768448.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 429587904.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1148.26513671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 236.83804321289062,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6613828096.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6494338.522630928,
+ "step": 310,
+ "teaching_rms": 97673.73355067195,
+ "traffic_rms": 6494338.5226309225
+ },
+ {
+ "batch_loss": 2556.673095703125,
+ "innovation_rms": 117211.8772635478,
+ "instruction_rms": 0.2878273666207913,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3999552765952.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.5418847194440917e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 316048992.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.9671630859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 205.9840545654297,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6610343936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1490808448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 304635168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19597998080.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1492270848.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 316048992.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.8662109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 205.9840087890625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6610343936.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6700239.071988012,
+ "step": 311,
+ "teaching_rms": 117211.8772635478,
+ "traffic_rms": 6700239.071988006
+ },
+ {
+ "batch_loss": 2934.625732421875,
+ "innovation_rms": 110351.63361390613,
+ "instruction_rms": 0.25850803674469863,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3995281129472.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.2589205197905425e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 495667552.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.59765625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.9338836669922,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6606688768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1488864384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 321941152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19603873792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1590397184.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 495667552.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.4967041015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 179.933837890625,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6606688768.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6602720.175034146,
+ "step": 312,
+ "teaching_rms": 110351.63361390613,
+ "traffic_rms": 6602720.17503414
+ },
+ {
+ "batch_loss": 1917.767333984375,
+ "innovation_rms": 127754.36929574661,
+ "instruction_rms": 0.2545822638625968,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3931637022720.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.03780131704037e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 761056128.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.2535400390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 159.04437255859375,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6602727424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1486744576.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 341095744.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19610322944.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1738302976.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 761056128.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1147.152587890625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 159.0443115234375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6602727424.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6902988.0983317075,
+ "step": 313,
+ "teaching_rms": 127754.36929574661,
+ "traffic_rms": 6902988.098331702
+ },
+ {
+ "batch_loss": 5690.9970703125,
+ "innovation_rms": 126310.62616347874,
+ "instruction_rms": 0.23799042881531138,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3874291449856.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.87868590076656e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 609125696.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.9337158203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122.01190185546875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6598829056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1484433024.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 366727936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19616620544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1877747712.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 609125696.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.832763671875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 122.0118408203125,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6598829056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 6896731.87136378,
+ "step": 314,
+ "teaching_rms": 126310.62616347874,
+ "traffic_rms": 6896731.871363777
+ },
+ {
+ "batch_loss": 5334.87060546875,
+ "innovation_rms": 144043.68175707106,
+ "instruction_rms": 0.2769439242245611,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3952291086336.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.0615403582018503e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 962906432.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.6085205078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.7251968383789,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6594925056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1482004352.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 391673728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19622762496.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2078571392.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 962906432.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.507568359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 89.72513580322266,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6594925056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 7205609.366521382,
+ "step": 315,
+ "teaching_rms": 144043.68175707106,
+ "traffic_rms": 7205609.366521377
+ },
+ {
+ "batch_loss": 2183.252685546875,
+ "innovation_rms": 141798.0735336687,
+ "instruction_rms": 0.22909643686344847,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4219186446336.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.4604292102768835e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1288730624.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.304443359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.25318908691406,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6590637056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1479517312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 428863488.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19629262848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2201477120.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1288730624.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1146.2034912109375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.25300598144531,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6590637056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 7385233.090317353,
+ "step": 316,
+ "teaching_rms": 141798.0735336687,
+ "traffic_rms": 7385233.090317349
+ },
+ {
+ "batch_loss": 4893.96533203125,
+ "innovation_rms": 136413.80970558288,
+ "instruction_rms": 0.3016890927434854,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4203224236032.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.671807941860327e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1363938432.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.9664306640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127.32282257080078,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6586398720.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1477070592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 440523072.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19635531776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2285201664.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1363938432.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.865478515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 127.32289123535156,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6586398720.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 7845514.67685723,
+ "step": 317,
+ "teaching_rms": 136413.80970558288,
+ "traffic_rms": 7845514.676857224
+ },
+ {
+ "batch_loss": 2262.17138671875,
+ "innovation_rms": 185375.17498016686,
+ "instruction_rms": 0.2708078738612783,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4013754155008.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.3612606947282417e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1610516736.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.6507568359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 115.07486724853516,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6580821504.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1474917248.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 493689152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19643129856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2546712064.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1610516736.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.5498046875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 115.07493591308594,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6580821504.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8015591.2681412175,
+ "step": 318,
+ "teaching_rms": 185375.17498016686,
+ "traffic_rms": 8015591.268141212
+ },
+ {
+ "batch_loss": 8800.71875,
+ "innovation_rms": 114075.60691919156,
+ "instruction_rms": 0.2893312339653874,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4050914115584.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.326770891120603e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 754543168.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.355224609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.38102722167969,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6574553600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1472756096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 482281728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19651430400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2438429440.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 754543168.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1145.2542724609375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.3814468383789,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6574553600.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8471667.195154864,
+ "step": 319,
+ "teaching_rms": 114075.60691919156,
+ "traffic_rms": 8471667.195154859
+ },
+ {
+ "batch_loss": 5892.416015625,
+ "innovation_rms": 115869.43164122428,
+ "instruction_rms": 0.3353744535273775,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3904941326336.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4.019320040661754e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 659738816.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.7867431640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.10179138183594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6566270464.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1471011840.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 464540672.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19661725696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2293558272.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 659738816.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.685791015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 95.10186004638672,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6566270464.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4378609.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8764071.179773435,
+ "step": 320,
+ "teaching_rms": 115869.43164122428,
+ "traffic_rms": 8764071.17977343
+ },
+ {
+ "batch_loss": 6720.646484375,
+ "innovation_rms": 117462.21184623124,
+ "instruction_rms": 0.26009194607905745,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3916351668224.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.898284016818374e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1305555968.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.22265625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 84.4930191040039,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6557949952.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1469639808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 459266752.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19671906304.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2218995456.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1305555968.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1144.1217041015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 84.49308013916016,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6557949952.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 13797371078.17604,
+ "predictor_updated": true,
+ "raw_apical_rms": 9064243.880802449,
+ "step": 321,
+ "teaching_rms": 117462.21184623124,
+ "traffic_rms": 9064243.880802447
+ },
+ {
+ "batch_loss": 4735.1669921875,
+ "innovation_rms": 151747.84284266224,
+ "instruction_rms": 0.3390365004254848,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3812799021056.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.622448229370717e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2409520384.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1143.629150390625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.95000457763672,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6548483072.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1469209856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 466686688.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19683227648.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2266340352.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2409520384.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1143.5281982421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 69.95006561279297,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6548483072.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 9418338.352269592,
+ "step": 322,
+ "teaching_rms": 151747.84284266224,
+ "traffic_rms": 9418338.352269588
+ },
+ {
+ "batch_loss": 7697.0498046875,
+ "innovation_rms": 153465.16663743427,
+ "instruction_rms": 0.25551488044856696,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3827808075776.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.7800557349969653e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1393462016.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1143.0836181640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.64370727539062,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6538205696.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1468822144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 511198912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19695634432.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2305539584.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1393462016.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1142.982666015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 75.64353942871094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6538205696.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 8990228.264897212,
+ "step": 323,
+ "teaching_rms": 153465.16663743427,
+ "traffic_rms": 8990228.26489721
+ },
+ {
+ "batch_loss": 5840.205078125,
+ "innovation_rms": 126083.15963869357,
+ "instruction_rms": 0.24317196373988975,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3763250397184.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.768164225897289e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1042311552.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1142.582763671875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66.80610656738281,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6527789056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1468492544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 542503552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19708237824.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2247376128.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1042311552.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1142.4818115234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 66.80601501464844,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6527789056.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 9163179.76171719,
+ "step": 324,
+ "teaching_rms": 126083.15963869357,
+ "traffic_rms": 9163179.761717185
+ },
+ {
+ "batch_loss": 1365.5123291015625,
+ "innovation_rms": 158318.4824499235,
+ "instruction_rms": 0.24769751206525567,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3616612286464.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.503321517325942e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1681976576.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1142.1204833984375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60.41381072998047,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6515974144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1468375552.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 584147968.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19722412032.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2269224448.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1681976576.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1142.01953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60.413719177246094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6515974144.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 9583760.571297992,
+ "step": 325,
+ "teaching_rms": 158318.4824499235,
+ "traffic_rms": 9583760.571297988
+ },
+ {
+ "batch_loss": 2260.481689453125,
+ "innovation_rms": 176203.9342368827,
+ "instruction_rms": 0.27455205445959163,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3781198348288.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.3671532603840164e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2263634688.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141.6929931640625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.31819152832031,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6504069632.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1468115456.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 628508608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19736709120.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2380380160.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2263634688.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141.592041015625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.31801223754883,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6504069632.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 10143614.263768166,
+ "step": 326,
+ "teaching_rms": 176203.9342368827,
+ "traffic_rms": 10143614.263768163
+ },
+ {
+ "batch_loss": 5584.48291015625,
+ "innovation_rms": 208274.28232011234,
+ "instruction_rms": 0.26392558400977206,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3909373919232.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.2179081803552145e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1329780736.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141.296875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.82945251464844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6491579904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1478708736.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 644187776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19751718912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2597039360.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1329780736.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1141.1959228515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.82936096191406,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6491579904.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 11208618.293386389,
+ "step": 327,
+ "teaching_rms": 208274.28232011234,
+ "traffic_rms": 11208618.293386387
+ },
+ {
+ "batch_loss": 2681.362548828125,
+ "innovation_rms": 210968.9450997936,
+ "instruction_rms": 0.2574936545883225,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4077609287680.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3.0249372333219967e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1647895808.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.928955078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57.94127655029297,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6481975808.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1544538752.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 658299712.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19767089152.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2785917440.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1647895808.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.8280029296875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57.94133758544922,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6481975808.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 11640992.062336402,
+ "step": 328,
+ "teaching_rms": 210968.9450997936,
+ "traffic_rms": 11640992.062336398
+ },
+ {
+ "batch_loss": 2604.351318359375,
+ "innovation_rms": 254543.78582485966,
+ "instruction_rms": 0.2653329534609749,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4161159823360.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.7887153122637965e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1878806400.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.58642578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61.949676513671875,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6480182784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1614937472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 703987072.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19783823360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3128725504.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1878806400.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.4854736328125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 61.94973373413086,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6480182784.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 11875628.145596776,
+ "step": 329,
+ "teaching_rms": 254543.78582485966,
+ "traffic_rms": 11875628.145596772
+ },
+ {
+ "batch_loss": 2441.088623046875,
+ "innovation_rms": 237636.69609603152,
+ "instruction_rms": 0.4222082864985507,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4335684812800.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.5419158805797786e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1218033152.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.2667236328125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60.6102409362793,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6478503936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1692126976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 771894656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19799689216.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3365217792.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1218033152.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1140.165771484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 60.61030197143555,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6478503936.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 11803227.710564885,
+ "step": 330,
+ "teaching_rms": 237636.69609603152,
+ "traffic_rms": 11803227.710564878
+ },
+ {
+ "batch_loss": 6252.21923828125,
+ "innovation_rms": 277030.8147674175,
+ "instruction_rms": 1.210810272230304,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4488481996800.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.304302632296563e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2201141760.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.9676513671875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57.114044189453125,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6476926976.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1776208768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 840818176.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19817220096.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3695826432.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2201141760.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.86669921875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 57.114105224609375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6476926976.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 12380388.841647934,
+ "step": 331,
+ "teaching_rms": 277030.8147674175,
+ "traffic_rms": 12380388.841647875
+ },
+ {
+ "batch_loss": 1431.85107421875,
+ "innovation_rms": 292715.62464314915,
+ "instruction_rms": 0.2944302433834218,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4657330520064.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2.104295064055964e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 996177088.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.68701171875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.426124572753906,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6475442688.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1863686784.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 874780544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19835709440.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4050569472.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 996177088.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.5860595703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 52.426185607910156,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6475442688.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 13696913.312854562,
+ "step": 332,
+ "teaching_rms": 292715.62464314915,
+ "traffic_rms": 13696913.312854558
+ },
+ {
+ "batch_loss": 4672.2978515625,
+ "innovation_rms": 335373.263140621,
+ "instruction_rms": 0.3085876012123727,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4800086802432.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.9796220801510115e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2094073856.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.423095703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65.08384704589844,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6473929728.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1957345792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 936589888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19856543744.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4535314944.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2094073856.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.3221435546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 65.08390808105469,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6473929728.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 14332555.672845269,
+ "step": 333,
+ "teaching_rms": 335373.263140621,
+ "traffic_rms": 14332555.672845265
+ },
+ {
+ "batch_loss": 933.2939453125,
+ "innovation_rms": 304027.5629265624,
+ "instruction_rms": 0.3662433123987456,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5044625211392.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.8010323623716102e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1399998848.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.1741943359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.55024719238281,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6472454656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2057145472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 997996288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19883888640.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4688652800.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1399998848.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1139.0732421875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 58.55030822753906,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6472454656.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 14136259.540260281,
+ "step": 334,
+ "teaching_rms": 304027.5629265624,
+ "traffic_rms": 14136259.540260276
+ },
+ {
+ "batch_loss": 1390.27197265625,
+ "innovation_rms": 339223.1197906009,
+ "instruction_rms": 0.27283918551459774,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5257952755712.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6859844468024614e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3160869376.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.938720703125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.39664459228516,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6470971904.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2165099776.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1079542528.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19912542208.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5041449472.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3160869376.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.8377685546875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 67.39706420898438,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6470971904.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 15019332.584809005,
+ "step": 335,
+ "teaching_rms": 339223.1197906009,
+ "traffic_rms": 15019332.584809002
+ },
+ {
+ "batch_loss": 2000.961181640625,
+ "innovation_rms": 335080.1615810615,
+ "instruction_rms": 0.2940024437203431,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5587304185856.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.6121645615161957e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1794429696.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.7154541015625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.21929931640625,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6469417472.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2275633664.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1105337600.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19941140480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5305181696.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1794429696.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.614501953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.21971130371094,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6469417472.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4322443.5,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 16472182.087483803,
+ "step": 336,
+ "teaching_rms": 335080.1615810615,
+ "traffic_rms": 16472182.087483801
+ },
+ {
+ "batch_loss": 1946.3153076171875,
+ "innovation_rms": 396334.50866232725,
+ "instruction_rms": 0.28163140988074803,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5830209961984.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.575758952856038e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4257774592.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.5030517578125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85.34903717041016,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6467500032.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2391278592.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1156448768.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 19971375104.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5781923840.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4257774592.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.4022216796875,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 85.34944915771484,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6467500032.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": 157081019783.38776,
+ "predictor_updated": true,
+ "raw_apical_rms": 17297679.447868995,
+ "step": 337,
+ "teaching_rms": 396334.50866232725,
+ "traffic_rms": 17297679.44786899
+ },
+ {
+ "batch_loss": 1888.3758544921875,
+ "innovation_rms": 259167.33299852573,
+ "instruction_rms": 0.3034358217573253,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6256502767616.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.5141363498992143e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2693653760.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.300537109375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.01705169677734,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6465478656.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2511376384.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1200977280.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20002093056.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5658652672.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2693653760.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.19970703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 76.0174560546875,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6465478656.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 16301237.643009989,
+ "step": 338,
+ "teaching_rms": 259167.33299852573,
+ "traffic_rms": 16301237.643009987
+ },
+ {
+ "batch_loss": 2419.53955078125,
+ "innovation_rms": 216939.19764861357,
+ "instruction_rms": 0.26332365551848824,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6743163142144.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.4771168335962836e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1919302784.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.10693359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.975120544433594,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6463244288.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2633502464.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1221261312.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20033992704.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5372797952.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1919302784.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1138.006103515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.97552490234375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6463244288.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 16724059.92672756,
+ "step": 339,
+ "teaching_rms": 216939.19764861357,
+ "traffic_rms": 16724059.926727558
+ },
+ {
+ "batch_loss": 1186.956298828125,
+ "innovation_rms": 225898.8273510374,
+ "instruction_rms": 0.2629253696071341,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7239323615232.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3781670226453796e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1349988608.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.9212646484375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.13748550415039,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6460996608.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2757927680.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1244252544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20068075520.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5109138944.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1349988608.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.8204345703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 49.13789367675781,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6460996608.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 17338420.591024045,
+ "step": 340,
+ "teaching_rms": 225898.8273510374,
+ "traffic_rms": 17338420.591024045
+ },
+ {
+ "batch_loss": 1597.16845703125,
+ "innovation_rms": 248893.48062603368,
+ "instruction_rms": 0.29816489797127127,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7730983075840.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3720403424296666e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2123345536.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.7427978515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.75210952758789,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6458553856.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2886160128.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1282325248.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20103956480.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4937037824.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2123345536.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.6419677734375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 53.75190353393555,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6458553856.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 18152839.43187297,
+ "step": 341,
+ "teaching_rms": 248893.48062603368,
+ "traffic_rms": 18152839.431872968
+ },
+ {
+ "batch_loss": 1943.8702392578125,
+ "innovation_rms": 291588.7142196957,
+ "instruction_rms": 0.2731420590653395,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8200723628032.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.3091893338914864e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3284448256.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.57080078125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47.92499542236328,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6456054272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3019828992.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1336689792.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20141989888.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4894154752.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3284448256.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.469970703125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47.9247932434082,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6456054272.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 19009020.274486333,
+ "step": 342,
+ "teaching_rms": 291588.7142196957,
+ "traffic_rms": 19009020.27448633
+ },
+ {
+ "batch_loss": 1925.236083984375,
+ "innovation_rms": 401801.80162921746,
+ "instruction_rms": 0.28499054883669306,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8513663270912.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.2453315758487394e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4958513152.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.4046630859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.24825668334961,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6453498880.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3161306368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1414774400.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20183087104.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5279034368.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4958513152.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.3038330078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 40.248050689697266,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6453498880.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 21053756.51087505,
+ "step": 343,
+ "teaching_rms": 401801.80162921746,
+ "traffic_rms": 21053756.51087505
+ },
+ {
+ "batch_loss": 2957.44580078125,
+ "innovation_rms": 362975.05460025836,
+ "instruction_rms": 0.2909139107042191,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8970216407040.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.1440504653818e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2891158272.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.2437744140625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34.426029205322266,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6451127296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3300252672.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1389463168.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20227895296.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5395477504.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2891158272.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.1429443359375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 34.42582321166992,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6451127296.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 23263085.869731642,
+ "step": 344,
+ "teaching_rms": 362975.05460025836,
+ "traffic_rms": 23263085.86973164
+ },
+ {
+ "batch_loss": 2467.40966796875,
+ "innovation_rms": 318676.79230389115,
+ "instruction_rms": 0.3182401850138222,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9482362945536.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1.0633653660340863e+27,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2011907200.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1137.037353515625,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47.34452819824219,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6448918016.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3436239360.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1359865984.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20272521216.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5342146560.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2011907200.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.9365234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 47.34434509277344,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6448918016.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 23854770.051083423,
+ "step": 345,
+ "teaching_rms": 318676.79230389115,
+ "traffic_rms": 23854770.05108342
+ },
+ {
+ "batch_loss": 5813.15185546875,
+ "innovation_rms": 317991.29094273434,
+ "instruction_rms": 0.3895949925658205,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9990943277056.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.952410236610429e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3270264576.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.6024169921875,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46.01380920410156,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6446849536.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3568584192.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1323449344.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20318695424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5255916544.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3270264576.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.5015869140625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 46.01362991333008,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6446849536.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 24877937.127732377,
+ "step": 346,
+ "teaching_rms": 317991.29094273434,
+ "traffic_rms": 24877937.127732374
+ },
+ {
+ "batch_loss": 2637.25146484375,
+ "innovation_rms": 454822.8609710206,
+ "instruction_rms": 0.32232423051339343,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10419377799168.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.232847760350559e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5555488768.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.1995849609375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.89439010620117,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6444920832.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3703830272.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1352460160.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20377806848.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5578864640.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5555488768.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1136.0987548828125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.89421081542969,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6444920832.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 26056030.356604625,
+ "step": 347,
+ "teaching_rms": 454822.8609710206,
+ "traffic_rms": 26056030.356604625
+ },
+ {
+ "batch_loss": 3686.61376953125,
+ "innovation_rms": 527431.0686480213,
+ "instruction_rms": 0.8526723179965882,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10583922442240.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.511169073610553e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4292083968.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.82568359375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.187076568603516,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6443116544.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3850319104.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1464887424.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20442232832.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6237528576.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4292083968.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.724853515625,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 39.18689727783203,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6443116544.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 26294987.641613,
+ "step": 348,
+ "teaching_rms": 527431.0686480213,
+ "traffic_rms": 26294987.641612988
+ },
+ {
+ "batch_loss": 2900.048095703125,
+ "innovation_rms": 497401.0790649294,
+ "instruction_rms": 0.2686649193541375,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11010857500672.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.177454191661414e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2433182720.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.477783203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 45.13093948364258,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6441402368.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4011436800.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1611176448.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20509065216.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6679847424.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2433182720.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.376953125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 45.13105392456055,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6441402368.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 25795998.26022303,
+ "step": 349,
+ "teaching_rms": 497401.0790649294,
+ "traffic_rms": 25795998.26022303
+ },
+ {
+ "batch_loss": 1366.9373779296875,
+ "innovation_rms": 575012.3657529651,
+ "instruction_rms": 0.2757519895049111,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11381300527104.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7.61451490276402e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5365909504.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.1533203125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.59339141845703,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6439795200.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4188709376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1772725376.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20579151872.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 7369723904.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 5365909504.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1135.052490234375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 41.593505859375,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6439795200.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 27315500.33674384,
+ "step": 350,
+ "teaching_rms": 575012.3657529651,
+ "traffic_rms": 27315500.33674384
+ },
+ {
+ "batch_loss": 5513.25341796875,
+ "innovation_rms": 635037.0445547941,
+ "instruction_rms": 0.2611506105488838,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 11652020830208.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8.699711769048413e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2676862208.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1134.8499755859375,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.51261901855469,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6438173184.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4374182912.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1854734464.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20650987520.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8284314624.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 2676862208.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1134.7491455078125,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 97.51254272460938,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6438173184.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 30438775.56197408,
+ "step": 351,
+ "teaching_rms": 635037.0445547941,
+ "traffic_rms": 30438775.56197408
+ },
+ {
+ "batch_loss": 10160.5595703125,
+ "innovation_rms": 602561.1308105083,
+ "instruction_rms": 0.28054503801115005,
+ "parameter_state": {
+ "bn_running_mean": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 12079995027456.0,
+ "nonfinite_indices": []
+ },
+ "bn_running_var": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 9.975884825248869e+26,
+ "nonfinite_indices": []
+ },
+ "feedback_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3755489792.0,
+ "nonfinite_indices": []
+ },
+ "feedback_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1134.565673828125,
+ "nonfinite_indices": []
+ },
+ "feedback_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 109.74665832519531,
+ "nonfinite_indices": []
+ },
+ "feedback_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6436647936.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 4564779008.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_bias_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1905963392.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 20723206144.0,
+ "nonfinite_indices": []
+ },
+ "forward_bn_scale_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 8868926464.0,
+ "nonfinite_indices": []
+ },
+ "forward_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 3755489792.0,
+ "nonfinite_indices": []
+ },
+ "forward_readout": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 1134.46484375,
+ "nonfinite_indices": []
+ },
+ "forward_readout_momentum": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 109.74657440185547,
+ "nonfinite_indices": []
+ },
+ "forward_weight": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 6436647936.0,
+ "nonfinite_indices": []
+ },
+ "predictor_bias": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 10107099.0,
+ "nonfinite_indices": []
+ },
+ "predictor_slope": {
+ "all_finite": true,
+ "max_abs_over_finite_tensors": 0.11251720041036606,
+ "nonfinite_indices": []
+ }
+ },
+ "predictor_mse": null,
+ "predictor_updated": false,
+ "raw_apical_rms": 31843694.140763804,
+ "step": 352,
+ "teaching_rms": 602561.1308105083,
+ "traffic_rms": 31843694.140763804
+ }
+ ],
+ "trajectory_metrics_finite": true,
+ "validation_evaluations": 0
+}
generated by cgit v1.2.3 (git 2.25.1) at 2026-08-01 03:38:44 +0000