{
  "architecture": {
    "adaptive_apical_parameters": 2260992,
    "base_width": 16,
    "blocks_per_stage": 3,
    "bn_eps": 1e-05,
    "bn_momentum": 0.1,
    "depth": 20,
    "family": "CIFAR 6n+2 ResNet, option-A shortcuts",
    "fixed_traffic_coefficients": 188416,
    "forward_parameters": 269722,
    "hidden_shapes": [
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        16,
        32,
        32
      ],
      [
        32,
        16,
        16
      ],
      [
        32,
        16,
        16
      ],
      [
        32,
        16,
        16
      ],
      [
        32,
        16,
        16
      ],
      [
        32,
        16,
        16
      ],
      [
        32,
        16,
        16
      ],
      [
        64,
        8,
        8
      ],
      [
        64,
        8,
        8
      ],
      [
        64,
        8,
        8
      ],
      [
        64,
        8,
        8
      ],
      [
        64,
        8,
        8
      ],
      [
        64,
        8,
        8
      ]
    ],
    "normalization": "batchnorm",
    "predictor_parameters": 376832,
    "residual_scale": 1.0,
    "vectorizer_mode": "spatial_template",
    "vectorizer_parameters": 1884160
  },
  "args": {
    "a_scale": 1.0,
    "a_warmup_steps": 0,
    "alignment_probe": 32,
    "apical_seed": null,
    "augment_train": 1,
    "batch_size": 128,
    "bn_eps": 1e-05,
    "bn_momentum": 0.1,
    "data_dir": "/home/yurenh2/sdrn/data",
    "depth": 20,
    "device": "cuda:0",
    "epochs": 200,
    "eta_A": 0.01,
    "eta_P": 0.01,
    "eval_every": 20,
    "eval_split": "validation",
    "learn_P": 0,
    "loader_seed": 0,
    "lr": 0.03,
    "lr_gamma": 0.1,
    "lr_milestones": "100,150",
    "lr_schedule": "step",
    "max_steps": 0,
    "mode": "dfa",
    "momentum": 0.9,
    "normalization": "batchnorm",
    "nuisance_scale": 0.0,
    "out": "results/oral_a_dev/dfa_full_r20_s0.json",
    "output_lr": 0.1,
    "pert_directions": 1,
    "pert_every": 4,
    "pert_sigma": 0.01,
    "perturb_seed": 1000,
    "predictor_warmup_steps": 0,
    "residual_scale": null,
    "seed": 0,
    "split_seed": 2027,
    "train_limit": 0,
    "use_residual": 1,
    "val_examples": 5000,
    "vectorizer_mode": "spatial_template",
    "warmup_epochs": 0,
    "weight_decay": 0.0001,
    "weight_scale": 1.0,
    "width": 16
  },
  "counters": {
    "apical_warmup_examples": 0,
    "calibration_event_examples": 0,
    "causal_scalar_observations": 0,
    "logical_batch_loss_queries": 0,
    "ordinary_examples": 9000000,
    "per_example_loss_terms": 0,
    "perturbation_events": 0,
    "perturbation_forward_examples": 0,
    "predictor_warmup_examples": 0
  },
  "diagnostics": {
    "early_third_mean": 0.00047278958178746205,
    "innovation_negative_gradient_cosine": [
      0.0010253445943817496,
      0.0004478447372093797,
      -0.0006283384864218533,
      0.0007250487105920911,
      0.00023278198204934597,
      0.0010340559529140592,
      0.0008693559793755412,
      0.00824977271258831,
      0.00294140400364995,
      0.005062468349933624,
      0.004748804494738579,
      0.008493781089782715,
      0.005644433666020632,
      0.012445333413779736,
      0.018638355657458305,
      0.014637157320976257,
      0.01852312870323658,
      0.022409353405237198,
      0.032944779843091965
    ],
    "normalization_state": "training_batch_stats_without_running_update",
    "raw_negative_gradient_cosine": [
      0.0010253445943817496,
      0.0004478447372093797,
      -0.0006283384864218533,
      0.0007250487105920911,
      0.00023278198204934597,
      0.0010340559529140592,
      0.0008693559793755412,
      0.00824977271258831,
      0.00294140400364995,
      0.005062468349933624,
      0.004748804494738579,
      0.008493781089782715,
      0.005644433666020632,
      0.012445333413779736,
      0.018638355657458305,
      0.014637157320976257,
      0.01852312870323658,
      0.022409353405237198,
      0.032944779843091965
    ],
    "teaching_negative_gradient_cosine": [
      0.0010253445943817496,
      0.0004478447372093797,
      -0.0006283384864218533,
      0.0007250487105920911,
      0.00023278198204934597,
      0.0010340559529140592,
      0.0008693559793755412,
      0.00824977271258831,
      0.00294140400364995,
      0.005062468349933624,
      0.004748804494738579,
      0.008493781089782715,
      0.005644433666020632,
      0.012445333413779736,
      0.018638355657458305,
      0.014637157320976257,
      0.01852312870323658,
      0.022409353405237198,
      0.032944779843091965
    ],
    "wall_s": 0.14631915092468262
  },
  "epochs": [
    {
      "epoch": 1,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 352,
      "train_examples": 45000,
      "train_loss": 2.556540602959527
    },
    {
      "epoch": 2,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 704,
      "train_examples": 45000,
      "train_loss": 3.100372151099311
    },
    {
      "epoch": 3,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 1056,
      "train_examples": 45000,
      "train_loss": 4.778685990058051
    },
    {
      "epoch": 4,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 1408,
      "train_examples": 45000,
      "train_loss": 7.7772000684950084
    },
    {
      "epoch": 5,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 1760,
      "train_examples": 45000,
      "train_loss": 11.414061186557346
    },
    {
      "epoch": 6,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 2112,
      "train_examples": 45000,
      "train_loss": 14.56060789320204
    },
    {
      "epoch": 7,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 2464,
      "train_examples": 45000,
      "train_loss": 16.539933412000867
    },
    {
      "epoch": 8,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 2816,
      "train_examples": 45000,
      "train_loss": 21.91490387776693
    },
    {
      "epoch": 9,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 3168,
      "train_examples": 45000,
      "train_loss": 28.971424170600045
    },
    {
      "epoch": 10,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 3520,
      "train_examples": 45000,
      "train_loss": 30.374662603081596
    },
    {
      "epoch": 11,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 3872,
      "train_examples": 45000,
      "train_loss": 41.90342777777778
    },
    {
      "epoch": 12,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 4224,
      "train_examples": 45000,
      "train_loss": 51.25492545979818
    },
    {
      "epoch": 13,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 4576,
      "train_examples": 45000,
      "train_loss": 56.11107198079427
    },
    {
      "epoch": 14,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 4928,
      "train_examples": 45000,
      "train_loss": 68.55007108018663
    },
    {
      "epoch": 15,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 5280,
      "train_examples": 45000,
      "train_loss": 76.00376447211372
    },
    {
      "epoch": 16,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 5632,
      "train_examples": 45000,
      "train_loss": 85.58951192220051
    },
    {
      "epoch": 17,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 5984,
      "train_examples": 45000,
      "train_loss": 100.25683562282985
    },
    {
      "epoch": 18,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 6336,
      "train_examples": 45000,
      "train_loss": 125.66919279649522
    },
    {
      "epoch": 19,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 6688,
      "train_examples": 45000,
      "train_loss": 127.25568057793511
    },
    {
      "epoch": 20,
      "eval_accuracy": 0.2624,
      "eval_loss": 125.2142234375,
      "eval_split": "validation",
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 7040,
      "train_examples": 45000,
      "train_loss": 132.27207181125218
    },
    {
      "epoch": 21,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 7392,
      "train_examples": 45000,
      "train_loss": 151.6751389811198
    },
    {
      "epoch": 22,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 7744,
      "train_examples": 45000,
      "train_loss": 169.5348867702908
    },
    {
      "epoch": 23,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 8096,
      "train_examples": 45000,
      "train_loss": 191.51776351453992
    },
    {
      "epoch": 24,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 8448,
      "train_examples": 45000,
      "train_loss": 218.43964941134982
    },
    {
      "epoch": 25,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 8800,
      "train_examples": 45000,
      "train_loss": 223.38382630208332
    },
    {
      "epoch": 26,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 9152,
      "train_examples": 45000,
      "train_loss": 280.8093923258464
    },
    {
      "epoch": 27,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 9504,
      "train_examples": 45000,
      "train_loss": 297.94869933810764
    },
    {
      "epoch": 28,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 9856,
      "train_examples": 45000,
      "train_loss": 312.76720779079864
    },
    {
      "epoch": 29,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 10208,
      "train_examples": 45000,
      "train_loss": 312.5946175998264
    },
    {
      "epoch": 30,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 10560,
      "train_examples": 45000,
      "train_loss": 351.4095493733724
    },
    {
      "epoch": 31,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 10912,
      "train_examples": 45000,
      "train_loss": 375.9645059407552
    },
    {
      "epoch": 32,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 11264,
      "train_examples": 45000,
      "train_loss": 418.8832517578125
    },
    {
      "epoch": 33,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 11616,
      "train_examples": 45000,
      "train_loss": 401.78692121310763
    },
    {
      "epoch": 34,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 11968,
      "train_examples": 45000,
      "train_loss": 459.99783371310764
    },
    {
      "epoch": 35,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 12320,
      "train_examples": 45000,
      "train_loss": 470.8934236816406
    },
    {
      "epoch": 36,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 12672,
      "train_examples": 45000,
      "train_loss": 535.8501281141494
    },
    {
      "epoch": 37,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 13024,
      "train_examples": 45000,
      "train_loss": 568.164088547092
    },
    {
      "epoch": 38,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 13376,
      "train_examples": 45000,
      "train_loss": 578.0671964301215
    },
    {
      "epoch": 39,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 13728,
      "train_examples": 45000,
      "train_loss": 586.6293734266493
    },
    {
      "epoch": 40,
      "eval_accuracy": 0.2272,
      "eval_loss": 685.036525,
      "eval_split": "validation",
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 14080,
      "train_examples": 45000,
      "train_loss": 651.2344521484375
    },
    {
      "epoch": 41,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 14432,
      "train_examples": 45000,
      "train_loss": 697.9699587944879
    },
    {
      "epoch": 42,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 14784,
      "train_examples": 45000,
      "train_loss": 666.9690939561632
    },
    {
      "epoch": 43,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 15136,
      "train_examples": 45000,
      "train_loss": 813.9267259440104
    },
    {
      "epoch": 44,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 15488,
      "train_examples": 45000,
      "train_loss": 730.138217686632
    },
    {
      "epoch": 45,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 15840,
      "train_examples": 45000,
      "train_loss": 785.480466265191
    },
    {
      "epoch": 46,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 16192,
      "train_examples": 45000,
      "train_loss": 914.0447047743055
    },
    {
      "epoch": 47,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 16544,
      "train_examples": 45000,
      "train_loss": 886.4085672526041
    },
    {
      "epoch": 48,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 16896,
      "train_examples": 45000,
      "train_loss": 974.3970221354167
    },
    {
      "epoch": 49,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 17248,
      "train_examples": 45000,
      "train_loss": 938.4398298828125
    },
    {
      "epoch": 50,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 17600,
      "train_examples": 45000,
      "train_loss": 1086.7354953667534
    },
    {
      "epoch": 51,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 17952,
      "train_examples": 45000,
      "train_loss": 1172.6876183810764
    },
    {
      "epoch": 52,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 18304,
      "train_examples": 45000,
      "train_loss": 1124.182160329861
    },
    {
      "epoch": 53,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 18656,
      "train_examples": 45000,
      "train_loss": 1166.5601384331596
    },
    {
      "epoch": 54,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 19008,
      "train_examples": 45000,
      "train_loss": 1149.0074053819444
    },
    {
      "epoch": 55,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 19360,
      "train_examples": 45000,
      "train_loss": 1256.0298586588542
    },
    {
      "epoch": 56,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 19712,
      "train_examples": 45000,
      "train_loss": 1302.0698480794272
    },
    {
      "epoch": 57,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 20064,
      "train_examples": 45000,
      "train_loss": 1460.6422423828126
    },
    {
      "epoch": 58,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 20416,
      "train_examples": 45000,
      "train_loss": 1472.918828938802
    },
    {
      "epoch": 59,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 20768,
      "train_examples": 45000,
      "train_loss": 1589.164417860243
    },
    {
      "epoch": 60,
      "eval_accuracy": 0.329,
      "eval_loss": 1347.2895,
      "eval_split": "validation",
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 21120,
      "train_examples": 45000,
      "train_loss": 1661.3754509114583
    },
    {
      "epoch": 61,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 21472,
      "train_examples": 45000,
      "train_loss": 1479.4850687282985
    },
    {
      "epoch": 62,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 21824,
      "train_examples": 45000,
      "train_loss": 1608.6262785373265
    },
    {
      "epoch": 63,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 22176,
      "train_examples": 45000,
      "train_loss": 1673.533507139757
    },
    {
      "epoch": 64,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 22528,
      "train_examples": 45000,
      "train_loss": 1933.4802288194444
    },
    {
      "epoch": 65,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 22880,
      "train_examples": 45000,
      "train_loss": 1885.1568971788195
    },
    {
      "epoch": 66,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 23232,
      "train_examples": 45000,
      "train_loss": 2055.0665389756946
    },
    {
      "epoch": 67,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 23584,
      "train_examples": 45000,
      "train_loss": 1991.6385318576388
    },
    {
      "epoch": 68,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 23936,
      "train_examples": 45000,
      "train_loss": 1979.9061404513889
    },
    {
      "epoch": 69,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 24288,
      "train_examples": 45000,
      "train_loss": 2016.5147117838542
    },
    {
      "epoch": 70,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 24640,
      "train_examples": 45000,
      "train_loss": 2136.3028917751735
    },
    {
      "epoch": 71,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 24992,
      "train_examples": 45000,
      "train_loss": 2168.3169892144097
    },
    {
      "epoch": 72,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 25344,
      "train_examples": 45000,
      "train_loss": 2428.1194118055555
    },
    {
      "epoch": 73,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 25696,
      "train_examples": 45000,
      "train_loss": 2350.8536308159723
    },
    {
      "epoch": 74,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 26048,
      "train_examples": 45000,
      "train_loss": 2417.361424435764
    },
    {
      "epoch": 75,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 26400,
      "train_examples": 45000,
      "train_loss": 2694.048535503472
    },
    {
      "epoch": 76,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 26752,
      "train_examples": 45000,
      "train_loss": 2560.427761892361
    },
    {
      "epoch": 77,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 27104,
      "train_examples": 45000,
      "train_loss": 2634.9838090711805
    },
    {
      "epoch": 78,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 27456,
      "train_examples": 45000,
      "train_loss": 2847.5417817708335
    },
    {
      "epoch": 79,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 27808,
      "train_examples": 45000,
      "train_loss": 2805.50854765625
    },
    {
      "epoch": 80,
      "eval_accuracy": 0.1308,
      "eval_loss": 5932.2877,
      "eval_split": "validation",
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 28160,
      "train_examples": 45000,
      "train_loss": 2874.5113036458333
    },
    {
      "epoch": 81,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 28512,
      "train_examples": 45000,
      "train_loss": 3193.7074268229167
    },
    {
      "epoch": 82,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 28864,
      "train_examples": 45000,
      "train_loss": 3132.8926462239583
    },
    {
      "epoch": 83,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 29216,
      "train_examples": 45000,
      "train_loss": 3113.0640076605905
    },
    {
      "epoch": 84,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 29568,
      "train_examples": 45000,
      "train_loss": 3188.840806315104
    },
    {
      "epoch": 85,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 29920,
      "train_examples": 45000,
      "train_loss": 3214.120521875
    },
    {
      "epoch": 86,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 30272,
      "train_examples": 45000,
      "train_loss": 3177.258704513889
    },
    {
      "epoch": 87,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 30624,
      "train_examples": 45000,
      "train_loss": 3702.192488498264
    },
    {
      "epoch": 88,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 30976,
      "train_examples": 45000,
      "train_loss": 3671.4719335069444
    },
    {
      "epoch": 89,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 31328,
      "train_examples": 45000,
      "train_loss": 3359.0455736979166
    },
    {
      "epoch": 90,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 31680,
      "train_examples": 45000,
      "train_loss": 3451.624914453125
    },
    {
      "epoch": 91,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 32032,
      "train_examples": 45000,
      "train_loss": 3519.3614254774307
    },
    {
      "epoch": 92,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 32384,
      "train_examples": 45000,
      "train_loss": 3716.5331078125
    },
    {
      "epoch": 93,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 32736,
      "train_examples": 45000,
      "train_loss": 3814.598072829861
    },
    {
      "epoch": 94,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 33088,
      "train_examples": 45000,
      "train_loss": 4416.5923919704865
    },
    {
      "epoch": 95,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 33440,
      "train_examples": 45000,
      "train_loss": 3868.4341471788193
    },
    {
      "epoch": 96,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 33792,
      "train_examples": 45000,
      "train_loss": 4481.933284201389
    },
    {
      "epoch": 97,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 34144,
      "train_examples": 45000,
      "train_loss": 4258.458974001736
    },
    {
      "epoch": 98,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 34496,
      "train_examples": 45000,
      "train_loss": 4381.8194765625
    },
    {
      "epoch": 99,
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 34848,
      "train_examples": 45000,
      "train_loss": 4606.362935763889
    },
    {
      "epoch": 100,
      "eval_accuracy": 0.2856,
      "eval_loss": 4711.7743,
      "eval_split": "validation",
      "lr": 0.03,
      "output_lr": 0.1,
      "step": 35200,
      "train_examples": 45000,
      "train_loss": 4435.167835026042
    },
    {
      "epoch": 101,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 35552,
      "train_examples": 45000,
      "train_loss": 1033.1025149739583
    },
    {
      "epoch": 102,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 35904,
      "train_examples": 45000,
      "train_loss": 718.2667085069445
    },
    {
      "epoch": 103,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 36256,
      "train_examples": 45000,
      "train_loss": 618.5412251627604
    },
    {
      "epoch": 104,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 36608,
      "train_examples": 45000,
      "train_loss": 563.2143335286459
    },
    {
      "epoch": 105,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 36960,
      "train_examples": 45000,
      "train_loss": 513.7835886121961
    },
    {
      "epoch": 106,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 37312,
      "train_examples": 45000,
      "train_loss": 493.456696484375
    },
    {
      "epoch": 107,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 37664,
      "train_examples": 45000,
      "train_loss": 460.3670596137153
    },
    {
      "epoch": 108,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 38016,
      "train_examples": 45000,
      "train_loss": 431.1101678005642
    },
    {
      "epoch": 109,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 38368,
      "train_examples": 45000,
      "train_loss": 412.06136221788194
    },
    {
      "epoch": 110,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 38720,
      "train_examples": 45000,
      "train_loss": 384.5155812988281
    },
    {
      "epoch": 111,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 39072,
      "train_examples": 45000,
      "train_loss": 392.2792468098958
    },
    {
      "epoch": 112,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 39424,
      "train_examples": 45000,
      "train_loss": 359.34863812934026
    },
    {
      "epoch": 113,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 39776,
      "train_examples": 45000,
      "train_loss": 359.1387890733507
    },
    {
      "epoch": 114,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 40128,
      "train_examples": 45000,
      "train_loss": 365.8967673231337
    },
    {
      "epoch": 115,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 40480,
      "train_examples": 45000,
      "train_loss": 346.3132191623264
    },
    {
      "epoch": 116,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 40832,
      "train_examples": 45000,
      "train_loss": 341.6049950737847
    },
    {
      "epoch": 117,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 41184,
      "train_examples": 45000,
      "train_loss": 338.83940558810764
    },
    {
      "epoch": 118,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 41536,
      "train_examples": 45000,
      "train_loss": 347.8370302354601
    },
    {
      "epoch": 119,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 41888,
      "train_examples": 45000,
      "train_loss": 339.90422545030384
    },
    {
      "epoch": 120,
      "eval_accuracy": 0.3714,
      "eval_loss": 251.859665625,
      "eval_split": "validation",
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 42240,
      "train_examples": 45000,
      "train_loss": 349.63033754611547
    },
    {
      "epoch": 121,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 42592,
      "train_examples": 45000,
      "train_loss": 362.5899243923611
    },
    {
      "epoch": 122,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 42944,
      "train_examples": 45000,
      "train_loss": 363.9039738986545
    },
    {
      "epoch": 123,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 43296,
      "train_examples": 45000,
      "train_loss": 361.1454255425347
    },
    {
      "epoch": 124,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 43648,
      "train_examples": 45000,
      "train_loss": 350.4127308865017
    },
    {
      "epoch": 125,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 44000,
      "train_examples": 45000,
      "train_loss": 331.0137764702691
    },
    {
      "epoch": 126,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 44352,
      "train_examples": 45000,
      "train_loss": 342.88085292697485
    },
    {
      "epoch": 127,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 44704,
      "train_examples": 45000,
      "train_loss": 353.7506278266059
    },
    {
      "epoch": 128,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 45056,
      "train_examples": 45000,
      "train_loss": 355.71841215820314
    },
    {
      "epoch": 129,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 45408,
      "train_examples": 45000,
      "train_loss": 357.59253021375866
    },
    {
      "epoch": 130,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 45760,
      "train_examples": 45000,
      "train_loss": 331.9203523383247
    },
    {
      "epoch": 131,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 46112,
      "train_examples": 45000,
      "train_loss": 389.8919418511285
    },
    {
      "epoch": 132,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 46464,
      "train_examples": 45000,
      "train_loss": 374.8937126030816
    },
    {
      "epoch": 133,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 46816,
      "train_examples": 45000,
      "train_loss": 368.5728384548611
    },
    {
      "epoch": 134,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 47168,
      "train_examples": 45000,
      "train_loss": 348.7659294921875
    },
    {
      "epoch": 135,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 47520,
      "train_examples": 45000,
      "train_loss": 338.869308070204
    },
    {
      "epoch": 136,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 47872,
      "train_examples": 45000,
      "train_loss": 359.9369916693793
    },
    {
      "epoch": 137,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 48224,
      "train_examples": 45000,
      "train_loss": 343.4246781521267
    },
    {
      "epoch": 138,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 48576,
      "train_examples": 45000,
      "train_loss": 331.4392785264757
    },
    {
      "epoch": 139,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 48928,
      "train_examples": 45000,
      "train_loss": 351.22242747938367
    },
    {
      "epoch": 140,
      "eval_accuracy": 0.319,
      "eval_loss": 287.6015625,
      "eval_split": "validation",
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 49280,
      "train_examples": 45000,
      "train_loss": 359.108759242079
    },
    {
      "epoch": 141,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 49632,
      "train_examples": 45000,
      "train_loss": 332.1290716796875
    },
    {
      "epoch": 142,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 49984,
      "train_examples": 45000,
      "train_loss": 402.5178586751302
    },
    {
      "epoch": 143,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 50336,
      "train_examples": 45000,
      "train_loss": 343.6958641276042
    },
    {
      "epoch": 144,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 50688,
      "train_examples": 45000,
      "train_loss": 360.3065439290365
    },
    {
      "epoch": 145,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 51040,
      "train_examples": 45000,
      "train_loss": 360.4169939127604
    },
    {
      "epoch": 146,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 51392,
      "train_examples": 45000,
      "train_loss": 326.6301142361111
    },
    {
      "epoch": 147,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 51744,
      "train_examples": 45000,
      "train_loss": 402.5602715657552
    },
    {
      "epoch": 148,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 52096,
      "train_examples": 45000,
      "train_loss": 346.8896536512587
    },
    {
      "epoch": 149,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 52448,
      "train_examples": 45000,
      "train_loss": 383.8439869140625
    },
    {
      "epoch": 150,
      "lr": 0.003,
      "output_lr": 0.010000000000000002,
      "step": 52800,
      "train_examples": 45000,
      "train_loss": 382.86647420247397
    },
    {
      "epoch": 151,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 53152,
      "train_examples": 45000,
      "train_loss": 137.74708860405815
    },
    {
      "epoch": 152,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 53504,
      "train_examples": 45000,
      "train_loss": 116.78161181504991
    },
    {
      "epoch": 153,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 53856,
      "train_examples": 45000,
      "train_loss": 107.33565850016276
    },
    {
      "epoch": 154,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 54208,
      "train_examples": 45000,
      "train_loss": 99.61256242133247
    },
    {
      "epoch": 155,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 54560,
      "train_examples": 45000,
      "train_loss": 95.76746183268229
    },
    {
      "epoch": 156,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 54912,
      "train_examples": 45000,
      "train_loss": 90.10048072781034
    },
    {
      "epoch": 157,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 55264,
      "train_examples": 45000,
      "train_loss": 88.16128125542535
    },
    {
      "epoch": 158,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 55616,
      "train_examples": 45000,
      "train_loss": 83.83129876437717
    },
    {
      "epoch": 159,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 55968,
      "train_examples": 45000,
      "train_loss": 82.67561102498372
    },
    {
      "epoch": 160,
      "eval_accuracy": 0.4054,
      "eval_loss": 72.1459671875,
      "eval_split": "validation",
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 56320,
      "train_examples": 45000,
      "train_loss": 79.51378596869574
    },
    {
      "epoch": 161,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 56672,
      "train_examples": 45000,
      "train_loss": 76.7619601006402
    },
    {
      "epoch": 162,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 57024,
      "train_examples": 45000,
      "train_loss": 73.90739558580186
    },
    {
      "epoch": 163,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 57376,
      "train_examples": 45000,
      "train_loss": 71.83256978488498
    },
    {
      "epoch": 164,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 57728,
      "train_examples": 45000,
      "train_loss": 69.75941671820746
    },
    {
      "epoch": 165,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 58080,
      "train_examples": 45000,
      "train_loss": 68.5504772772895
    },
    {
      "epoch": 166,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 58432,
      "train_examples": 45000,
      "train_loss": 67.27026283365885
    },
    {
      "epoch": 167,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 58784,
      "train_examples": 45000,
      "train_loss": 65.27212860243056
    },
    {
      "epoch": 168,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 59136,
      "train_examples": 45000,
      "train_loss": 62.43467607014974
    },
    {
      "epoch": 169,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 59488,
      "train_examples": 45000,
      "train_loss": 62.754966387261284
    },
    {
      "epoch": 170,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 59840,
      "train_examples": 45000,
      "train_loss": 61.26306409505208
    },
    {
      "epoch": 171,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 60192,
      "train_examples": 45000,
      "train_loss": 60.49156853705512
    },
    {
      "epoch": 172,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 60544,
      "train_examples": 45000,
      "train_loss": 59.510354464382594
    },
    {
      "epoch": 173,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 60896,
      "train_examples": 45000,
      "train_loss": 57.63853626437717
    },
    {
      "epoch": 174,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 61248,
      "train_examples": 45000,
      "train_loss": 57.52931418457031
    },
    {
      "epoch": 175,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 61600,
      "train_examples": 45000,
      "train_loss": 56.75467317911784
    },
    {
      "epoch": 176,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 61952,
      "train_examples": 45000,
      "train_loss": 54.795722674560544
    },
    {
      "epoch": 177,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 62304,
      "train_examples": 45000,
      "train_loss": 56.41205893961588
    },
    {
      "epoch": 178,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 62656,
      "train_examples": 45000,
      "train_loss": 54.108699124484595
    },
    {
      "epoch": 179,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 63008,
      "train_examples": 45000,
      "train_loss": 53.167111499701605
    },
    {
      "epoch": 180,
      "eval_accuracy": 0.3482,
      "eval_loss": 48.12962734375,
      "eval_split": "validation",
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 63360,
      "train_examples": 45000,
      "train_loss": 54.02441080457899
    },
    {
      "epoch": 181,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 63712,
      "train_examples": 45000,
      "train_loss": 52.32608561062283
    },
    {
      "epoch": 182,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 64064,
      "train_examples": 45000,
      "train_loss": 52.743165169949
    },
    {
      "epoch": 183,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 64416,
      "train_examples": 45000,
      "train_loss": 52.80967100219726
    },
    {
      "epoch": 184,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 64768,
      "train_examples": 45000,
      "train_loss": 51.2523949883355
    },
    {
      "epoch": 185,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 65120,
      "train_examples": 45000,
      "train_loss": 51.99414753350152
    },
    {
      "epoch": 186,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 65472,
      "train_examples": 45000,
      "train_loss": 52.79423908148871
    },
    {
      "epoch": 187,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 65824,
      "train_examples": 45000,
      "train_loss": 50.42084914008247
    },
    {
      "epoch": 188,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 66176,
      "train_examples": 45000,
      "train_loss": 51.72486465657552
    },
    {
      "epoch": 189,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 66528,
      "train_examples": 45000,
      "train_loss": 53.04346135864258
    },
    {
      "epoch": 190,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 66880,
      "train_examples": 45000,
      "train_loss": 50.21678973490397
    },
    {
      "epoch": 191,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 67232,
      "train_examples": 45000,
      "train_loss": 52.006068816460505
    },
    {
      "epoch": 192,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 67584,
      "train_examples": 45000,
      "train_loss": 51.55635069715712
    },
    {
      "epoch": 193,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 67936,
      "train_examples": 45000,
      "train_loss": 50.32145675591363
    },
    {
      "epoch": 194,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 68288,
      "train_examples": 45000,
      "train_loss": 50.85280059678819
    },
    {
      "epoch": 195,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 68640,
      "train_examples": 45000,
      "train_loss": 51.28352748684353
    },
    {
      "epoch": 196,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 68992,
      "train_examples": 45000,
      "train_loss": 49.64204266628689
    },
    {
      "epoch": 197,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 69344,
      "train_examples": 45000,
      "train_loss": 49.435947619628905
    },
    {
      "epoch": 198,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 69696,
      "train_examples": 45000,
      "train_loss": 48.43252044949002
    },
    {
      "epoch": 199,
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 70048,
      "train_examples": 45000,
      "train_loss": 51.39490638292101
    },
    {
      "epoch": 200,
      "eval_accuracy": 0.3306,
      "eval_loss": 48.06651171875,
      "eval_split": "validation",
      "lr": 0.00030000000000000003,
      "output_lr": 0.0010000000000000002,
      "step": 70400,
      "train_examples": 45000,
      "train_loss": 48.690658324517145
    }
  ],
  "evaluation_protocol": {
    "test_evaluations": 0,
    "test_used_for_selection": false,
    "validation_evaluations": 11
  },
  "final": {
    "accuracy": 0.3306,
    "epoch": 200,
    "evaluation_split": "validation",
    "finite": true,
    "loss": 48.06651171875,
    "step": 70400
  },
  "hardware": {
    "cuda_device_name": "NVIDIA GeForce GTX 1080",
    "cuda_visible_devices": "5",
    "device": "cuda:0",
    "device_total_memory_bytes": 8507949056,
    "peak_memory_allocated_bytes": 2170440704,
    "peak_memory_reserved_bytes": 2812280832,
    "torch_version": "2.3.1+cu118"
  },
  "protocol_family": "oral_a_cifar_local_resnet_development",
  "provenance": {
    "git_commit": "a1d060e6d96d4ab973d16cd75149dd61899c7c95",
    "git_tracked_dirty": false
  },
  "schema_version": 1,
  "split": {
    "cifar_source_files": [
      {
        "bytes": 31035704,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_1",
        "sha256": "54636561a3ce25bd3e19253c6b0d8538147b0ae398331ac4a2d86c6d987368cd"
      },
      {
        "bytes": 31035320,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_2",
        "sha256": "766b2cef9fbc745cf056b3152224f7cf77163b330ea9a15f9392beb8b89bc5a8"
      },
      {
        "bytes": 31035999,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_3",
        "sha256": "0f00d98ebfb30b3ec0ad19f9756dc2630b89003e10525f5e148445e82aa6a1f9"
      },
      {
        "bytes": 31035696,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_4",
        "sha256": "3f7bb240661948b8f4d53e36ec720d8306f5668bd0071dcb4e6c947f78e9682b"
      },
      {
        "bytes": 31035623,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_5",
        "sha256": "d91802434d8376bbaeeadf58a737e3a1b12ac839077e931237e0dcd43adcb154"
      },
      {
        "bytes": 31035526,
        "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/test_batch",
        "sha256": "f53d8d457504f7cff4ea9e021afcf0e0ad8e24a91f3fc42091b8adef61157831"
      }
    ],
    "dataset": "cifar10",
    "evaluation_split": "validation",
    "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"
  },
  "timing": {
    "apical_warmup_wall_s": 0.0,
    "evaluation_wall_s": 3.8021719455718994,
    "predictor_warmup_wall_s": 0.0,
    "timing_excludes_data_loading_hashing_and_model_construction": true,
    "total_timed_wall_s": 3289.903744697571,
    "train_wall_s": 3286.076476097107
  },
  "work": {
    "apical_macs_per_example": 1884160,
    "causal_scalar_observations": 0,
    "components": {
      "apical_projection_macs": 16957440000000,
      "apical_regression_macs": 0,
      "bp_reverse_macs_estimate": 0,
      "local_weight_correlation_macs": 364959360000000,
      "ordinary_forward_macs": 364959360000000,
      "perturbation_forward_macs": 0,
      "warmup_clean_forward_macs": 0
    },
    "definition": "multiply-accumulates in conv/linear maps; one local weight correlation equals one forward-weight MAC count; BP reverse is estimated as one weight-gradient plus one activation-gradient convolution per forward convolution; elementwise nonlinearities and optimizer arithmetic excluded",
    "forward_macs_per_example": 40551040,
    "logical_batch_loss_queries": 0,
    "per_example_cross_entropy_terms": 0,
    "total_clean_forward_examples": 9000000,
    "total_forward_equivalent_examples": 9000000,
    "total_macs_estimate": 746876160000000
  }
}