Upload training_log.json with huggingface_hub
Browse files- training_log.json +45 -0
training_log.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"experiment_id": "bert_dual_head_20260213_060359",
|
| 3 |
+
"config": {
|
| 4 |
+
"model_name": "bert-base-uncased",
|
| 5 |
+
"num_category_classes": 7,
|
| 6 |
+
"num_severity_classes": 4,
|
| 7 |
+
"batch_size": 16,
|
| 8 |
+
"learning_rate": 0.0003,
|
| 9 |
+
"num_epochs": 3,
|
| 10 |
+
"max_length": 128,
|
| 11 |
+
"lora_r": 8,
|
| 12 |
+
"lora_alpha": 16,
|
| 13 |
+
"lora_dropout": 0.1,
|
| 14 |
+
"seed": 42
|
| 15 |
+
},
|
| 16 |
+
"device": "cuda",
|
| 17 |
+
"epochs": [
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1,
|
| 20 |
+
"train_loss": 2.1082094073295594,
|
| 21 |
+
"val_loss": 1.4672169902107932,
|
| 22 |
+
"val_category_acc": 0.6914285714285714,
|
| 23 |
+
"val_severity_acc": 0.7457142857142857
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2,
|
| 27 |
+
"train_loss": 1.4187353157997131,
|
| 28 |
+
"val_loss": 1.252626198259267,
|
| 29 |
+
"val_category_acc": 0.7628571428571429,
|
| 30 |
+
"val_severity_acc": 0.7771428571428571
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 3,
|
| 34 |
+
"train_loss": 1.2215006688662937,
|
| 35 |
+
"val_loss": 1.1916442256082187,
|
| 36 |
+
"val_category_acc": 0.7628571428571429,
|
| 37 |
+
"val_severity_acc": 0.7828571428571428
|
| 38 |
+
}
|
| 39 |
+
],
|
| 40 |
+
"final_test_metrics": {
|
| 41 |
+
"loss": 1.1840462291782552,
|
| 42 |
+
"category_accuracy": 0.7628571428571429,
|
| 43 |
+
"severity_accuracy": 0.7828571428571428
|
| 44 |
+
}
|
| 45 |
+
}
|