Upload artifacts/stage4_run_metadata.json with huggingface_hub
Browse files
artifacts/stage4_run_metadata.json
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"stage": 4,
|
| 3 |
+
"dataset_id": "THemidli/applied-ner-stage4-improved",
|
| 4 |
+
"base_model": "google/bert_uncased_L-2_H-128_A-2",
|
| 5 |
+
"seed": 20260802,
|
| 6 |
+
"labels": [
|
| 7 |
+
"O",
|
| 8 |
+
"B-PERSON",
|
| 9 |
+
"I-PERSON",
|
| 10 |
+
"B-ORGANIZATION",
|
| 11 |
+
"I-ORGANIZATION",
|
| 12 |
+
"B-LOCATION",
|
| 13 |
+
"I-LOCATION",
|
| 14 |
+
"B-TIMEDATE",
|
| 15 |
+
"I-TIMEDATE",
|
| 16 |
+
"B-PRODUCT",
|
| 17 |
+
"I-PRODUCT",
|
| 18 |
+
"B-WORKOFART",
|
| 19 |
+
"I-WORKOFART",
|
| 20 |
+
"B-JOB",
|
| 21 |
+
"I-JOB",
|
| 22 |
+
"B-AMOUNT",
|
| 23 |
+
"I-AMOUNT"
|
| 24 |
+
],
|
| 25 |
+
"parameter_count": 4371601,
|
| 26 |
+
"trainable_parameter_count": 4371601,
|
| 27 |
+
"device": "mps",
|
| 28 |
+
"platform": "macOS-27.0-arm64-arm-64bit",
|
| 29 |
+
"torch_version": "2.13.0",
|
| 30 |
+
"train_records": 841,
|
| 31 |
+
"test_records": 159,
|
| 32 |
+
"train_chunks": 865,
|
| 33 |
+
"test_chunks": 165,
|
| 34 |
+
"max_length": 256,
|
| 35 |
+
"wall_seconds": 14.937,
|
| 36 |
+
"trainer_metrics": {
|
| 37 |
+
"train_runtime": 14.8851,
|
| 38 |
+
"train_samples_per_second": 929.789,
|
| 39 |
+
"train_steps_per_second": 30.097,
|
| 40 |
+
"total_flos": 6246600094602.0,
|
| 41 |
+
"train_loss": 0.4432598276595984,
|
| 42 |
+
"epoch": 16.0
|
| 43 |
+
},
|
| 44 |
+
"hyperparameters": {
|
| 45 |
+
"epochs": 16,
|
| 46 |
+
"learning_rate": 0.0005,
|
| 47 |
+
"train_batch_size": 32,
|
| 48 |
+
"eval_batch_size": 64,
|
| 49 |
+
"weight_decay": 0.02,
|
| 50 |
+
"label_smoothing_factor": 0.0,
|
| 51 |
+
"warmup_steps": 45,
|
| 52 |
+
"scheduler": "linear",
|
| 53 |
+
"hidden_dropout": 0.1,
|
| 54 |
+
"attention_dropout": 0.1,
|
| 55 |
+
"classifier_dropout": 0.1
|
| 56 |
+
},
|
| 57 |
+
"train_overall": {
|
| 58 |
+
"overall_precision": 0.9539951573849879,
|
| 59 |
+
"overall_recall": 0.9708781362007168,
|
| 60 |
+
"overall_f1": 0.9623626068613301,
|
| 61 |
+
"overall_accuracy": 0.9948478767829058
|
| 62 |
+
},
|
| 63 |
+
"test_overall": {
|
| 64 |
+
"overall_precision": 0.42610652663165793,
|
| 65 |
+
"overall_recall": 0.5264133456904542,
|
| 66 |
+
"overall_f1": 0.4709784411276949,
|
| 67 |
+
"overall_accuracy": 0.8332408742926883
|
| 68 |
+
},
|
| 69 |
+
"test_f1_change_vs_stage3": 0.002472883284961258
|
| 70 |
+
}
|