diff options
Diffstat (limited to 'results/transformer_crossover/development_depth100.json')
| -rw-r--r-- | results/transformer_crossover/development_depth100.json | 147 |
1 files changed, 147 insertions, 0 deletions
diff --git a/results/transformer_crossover/development_depth100.json b/results/transformer_crossover/development_depth100.json new file mode 100644 index 0000000..9b062ea --- /dev/null +++ b/results/transformer_crossover/development_depth100.json @@ -0,0 +1,147 @@ +{ + "status": "development_only", + "claim_policy": "This short same-rate pilot checks mechanics and direction only. It is excluded from T1 selection, the primary crossover, figures, and reviewer-score changes.", + "source_commit": "08554657946e54d8d908a2866e82504e9b194617", + "configuration": { + "task": "tiny Shakespeare character language modeling", + "train_sha256": "6ec305602a99ac2802745a134e1f5e33e2231b4855525b00b9aebb730ac2626f", + "validation_sha256": "d37d30cc0c8327c270d493299c3dca54135f6d5f1c9ef60cda78076e311204b1", + "depths": [ + 4, + 8, + 12 + ], + "methods": [ + "bp", + "fa", + "dfa", + "clean_kp", + "sdil" + ], + "width": 128, + "heads": 4, + "context_length": 64, + "batch_size": 8, + "train_steps": 100, + "learning_rate": 0.001, + "schedule": "constant", + "validation_batches": 10, + "validation_tokens": 5120, + "test_evaluations": 0 + }, + "records": [ + { + "depth": 4, + "method": "bp", + "validation_nll": 2.5736704111099242, + "wall_seconds": 3.6037685871124268, + "first_nonfinite_step": null + }, + { + "depth": 4, + "method": "fa", + "validation_nll": 2.7691116333007812, + "wall_seconds": 4.258021593093872, + "first_nonfinite_step": null + }, + { + "depth": 4, + "method": "dfa", + "validation_nll": 3.252720332145691, + "wall_seconds": 4.3479225635528564, + "first_nonfinite_step": null + }, + { + "depth": 4, + "method": "clean_kp", + "validation_nll": 2.642165517807007, + "wall_seconds": 5.114134073257446, + "first_nonfinite_step": null + }, + { + "depth": 4, + "method": "sdil", + "validation_nll": 2.6421636819839476, + "wall_seconds": 5.935818672180176, + "first_nonfinite_step": null + }, + { + "depth": 8, + "method": "bp", + "validation_nll": 2.8022268295288084, + "wall_seconds": 5.785389184951782, + "first_nonfinite_step": null + }, + { + "depth": 8, + "method": "fa", + "validation_nll": 2.8849047660827636, + "wall_seconds": 7.843722105026245, + "first_nonfinite_step": null + }, + { + "depth": 8, + "method": "dfa", + "validation_nll": 3.3771399736404417, + "wall_seconds": 7.509201526641846, + "first_nonfinite_step": null + }, + { + "depth": 8, + "method": "clean_kp", + "validation_nll": 2.691094470024109, + "wall_seconds": 9.399603366851807, + "first_nonfinite_step": null + }, + { + "depth": 8, + "method": "sdil", + "validation_nll": 2.691092610359192, + "wall_seconds": 10.85186505317688, + "first_nonfinite_step": null + }, + { + "depth": 12, + "method": "bp", + "validation_nll": 2.771989607810974, + "wall_seconds": 9.70148754119873, + "first_nonfinite_step": null + }, + { + "depth": 12, + "method": "fa", + "validation_nll": 2.8445181608200074, + "wall_seconds": 11.526866912841797, + "first_nonfinite_step": null + }, + { + "depth": 12, + "method": "dfa", + "validation_nll": 3.3635837554931642, + "wall_seconds": 12.529596090316772, + "first_nonfinite_step": null + }, + { + "depth": 12, + "method": "clean_kp", + "validation_nll": 2.6933349132537843, + "wall_seconds": 13.416383028030396, + "first_nonfinite_step": null + }, + { + "depth": 12, + "method": "sdil", + "validation_nll": 2.6933418989181517, + "wall_seconds": 15.886815071105957, + "first_nonfinite_step": null + } + ], + "depth12_minus_depth4_nll": { + "bp": 0.1983191967010498, + "fa": 0.07540652751922616, + "dfa": 0.11086342334747323, + "clean_kp": 0.051169395446777344, + "sdil": 0.0511782169342041 + }, + "summary": "All 15 trajectories are finite. Reciprocal KP/SDIL show the smallest short-budget depth degradation, but the clean-task KP/SDIL identity is not innovation evidence." +} |