Hugging Face's logo

jacob-valdez
/
tensorcode-chatbot-hotpot-001

Text Generation
Safetensors
tensorcode
text2text-generation
Model card Files
xet
 Community
tensorcode-chatbot-hotpot-001 / replacement-record.json
jacob-valdez's picture
jacob-valdez
Publish bounded-workspace retrain (TensorCode 0.4.0a4) with evaluation against the original revision
0f4b98f verified
Raw History Blame Contribute Delete
8.88 kB
{
"repository": "jacob-valdez/tensorcode-chatbot-hotpot-001",
"status": "published as a new commit on main of jacob-valdez/tensorcode-chatbot-hotpot-001 (2026-09-24); qualification below was recorded before upload",
"replaces_on_main": "d74b40142c6e416cdc096f54e6d9d8c1de465568 (kept; requires TensorCode 6607a8b)",
"source_commit": "090ebc4eddc4c4fbc62d0f0b3498fa32df5680ae (v0.4.0a4), clean tree, no library changes",
"artifact": "/home/brandonin/.cache/tensorcode/replacement-20260924/chatbot-hotpot/attempt-01/model",
"original": {
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.618829257786274
},
"after": {
"exact_match": 0.46875,
"correct": 30,
"token_f1": 0.6014880952380952,
"mean_batch_cross_entropy": 0.5702201151289046
},
"bypass": {
"exact_match": 0.46875,
"correct": 30,
"token_f1": 0.6014880952380952,
"mean_batch_cross_entropy": 0.5643442408181727
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.725341647863388
}
},
"qualification": {
"after_em_ge_30": true,
"after_f1_ge_0.6014881": true,
"zero_em_le_1": true,
"active_beats_bypass": "EM 31 vs 30 (one question), F1 0.6171 vs 0.6067; bypass CE lower (0.5546 vs 0.5819). Not claimed as workspace benefit.",
"roundtrip": {
"cuda": {
"model_dir": "/home/brandonin/.cache/tensorcode/replacement-20260924/chatbot-hotpot/attempt-01/model",
"device": "cuda",
"model_safetensors_sha256": "ac39f8b3eec57ac13b53e68846bb8df30efaab85325858bcf2db37b0c065aad7",
"torch": "2.14.0+cu130",
"configuration_exact_vs_file": true,
"memory_update": "relative_rms_bounded",
"stored_tensors_bitwise_equal_loaded": true,
"stored_tensor_count": 208,
"resave_state_dict_bitwise_equal": true,
"resave_state_dict_mismatches": [],
"resave_configuration_exact": true,
"resave_config_file_bytes_equal": true,
"after_generations_equal_recorded": true,
"after_generation_mismatches": [],
"bypass_generations_equal_recorded": true,
"bypass_generation_mismatches": [],
"zero_generations_equal_recorded": true,
"zero_generation_mismatches": []
},
"cpu": {
"model_dir": "/home/brandonin/.cache/tensorcode/replacement-20260924/chatbot-hotpot/attempt-01/model",
"device": "cpu",
"model_safetensors_sha256": "ac39f8b3eec57ac13b53e68846bb8df30efaab85325858bcf2db37b0c065aad7",
"torch": "2.14.0+cu130",
"configuration_exact_vs_file": true,
"memory_update": "relative_rms_bounded",
"stored_tensors_bitwise_equal_loaded": true,
"stored_tensor_count": 208,
"resave_state_dict_bitwise_equal": true,
"resave_state_dict_mismatches": [],
"resave_configuration_exact": true,
"resave_config_file_bytes_equal": true,
"after_generations_equal_recorded": true,
"after_generation_mismatches": [],
"bypass_generations_equal_recorded": true,
"bypass_generation_mismatches": [],
"zero_generations_equal_recorded": true,
"zero_generation_mismatches": []
}
},
"pytest": "924 passed at 090ebc4 (pytest.log)",
"resave_file_bytes_note": "Resaved model.safetensors differs only in JSON key order of the shared-tensor alias metadata; tensor headers/offsets and all tensors are identical."
},
"runs": {
"attempt-01": {
"role": "candidate",
"seed": 7,
"epoch_losses": [
0.7308491980656981,
0.5275313123711385,
0.4226272698142566
],
"memory_gate": {
"before": 0.009999999776482582,
"after": 0.046188369393348694
},
"elapsed_seconds": 32.5627572479425,
"model_safetensors_sha256": "ac39f8b3eec57ac13b53e68846bb8df30efaab85325858bcf2db37b0c065aad7",
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.6201377287507057
},
"after": {
"exact_match": 0.484375,
"correct": 31,
"token_f1": 0.6171130952380952,
"mean_batch_cross_entropy": 0.5819374099373817
},
"bypass": {
"exact_match": 0.46875,
"correct": 30,
"token_f1": 0.6066964285714286,
"mean_batch_cross_entropy": 0.5546418600715697
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.698553770780563
},
"reload_generation_equal_first_input": true
},
"characterization/repro-seed7": {
"role": "non-candidate determinism check",
"seed": 7,
"epoch_losses": [
0.7308492966694757,
0.5275313895544969,
0.4226295731787104
],
"memory_gate": {
"before": 0.009999999776482582,
"after": 0.04618694633245468
},
"elapsed_seconds": 32.35384709993377,
"model_safetensors_sha256": "c6bcdb56b77bc0cd838285758f0241f9e6a391c26ec009cae07077b7729bb655",
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.6201377287507057
},
"after": {
"exact_match": 0.484375,
"correct": 31,
"token_f1": 0.6171130952380952,
"mean_batch_cross_entropy": 0.5819375016726553
},
"bypass": {
"exact_match": 0.46875,
"correct": 30,
"token_f1": 0.6066964285714286,
"mean_batch_cross_entropy": 0.5546412523835897
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.6985540091991425
},
"reload_generation_equal_first_input": true
},
"characterization/seed-08": {
"role": "non-candidate seed characterization",
"seed": 8,
"epoch_losses": [
0.7829204483423382,
0.5179650034406222,
0.43421322538051754
],
"memory_gate": {
"before": 0.009999999776482582,
"after": 0.039295025169849396
},
"elapsed_seconds": 32.18801082205027,
"model_safetensors_sha256": "ea3ae2a27cc1dded75812aecdc2277e2861aaa36a95e138325a65305e80f6aa6",
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.6181754404678941
},
"after": {
"exact_match": 0.484375,
"correct": 31,
"token_f1": 0.6348214285714285,
"mean_batch_cross_entropy": 0.5763567574322224
},
"bypass": {
"exact_match": 0.453125,
"correct": 29,
"token_f1": 0.5986111111111111,
"mean_batch_cross_entropy": 0.5582287386059761
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.733439326286316
},
"reload_generation_equal_first_input": true
},
"characterization/seed-09": {
"role": "non-candidate seed characterization",
"seed": 9,
"epoch_losses": [
0.7821961748413742,
0.49335869401693344,
0.45526367297861725
],
"memory_gate": {
"before": 0.009999999776482582,
"after": 0.046238597482442856
},
"elapsed_seconds": 32.08428092300892,
"model_safetensors_sha256": "b520fdaf16da58d7fd56ba24ddb6fed4bdf980e608bd61ec77c45109bb917c0a",
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.6191949816420674
},
"after": {
"exact_match": 0.484375,
"correct": 31,
"token_f1": 0.6277777777777778,
"mean_batch_cross_entropy": 0.540726182051003
},
"bypass": {
"exact_match": 0.46875,
"correct": 30,
"token_f1": 0.6121527777777778,
"mean_batch_cross_entropy": 0.5309799006208777
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.7022994458675385
},
"reload_generation_equal_first_input": true
},
"characterization/seed-10": {
"role": "non-candidate seed characterization",
"seed": 10,
"epoch_losses": [
0.7260864099953324,
0.5392665465478785,
0.42150062968721613
],
"memory_gate": {
"before": 0.009999999776482582,
"after": 0.039352625608444214
},
"elapsed_seconds": 32.40097103291191,
"model_safetensors_sha256": "ef9d63b2c0df81f420fa3fd09dd8f3e9a59a0680c5fd2b914b2a2befe8696a82",
"before": {
"exact_match": 0.421875,
"correct": 27,
"token_f1": 0.5563492063492064,
"mean_batch_cross_entropy": 0.6173182902857661
},
"after": {
"exact_match": 0.515625,
"correct": 33,
"token_f1": 0.643154761904762,
"mean_batch_cross_entropy": 0.5656530242413282
},
"bypass": {
"exact_match": 0.515625,
"correct": 33,
"token_f1": 0.6486111111111111,
"mean_batch_cross_entropy": 0.5527161192148924
},
"zero": {
"exact_match": 0.0,
"correct": 0,
"token_f1": 0.0,
"mean_batch_cross_entropy": 5.713347971439362
},
"reload_generation_equal_first_input": true
}
},
"selection": "attempt-01 was declared before any result; characterization runs were declared after attempt-01 and cannot replace it (ATTEMPTS.md)"
}