Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1205 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1205 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1660,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.060240963855421686,
|
| 14 |
+
"grad_norm": 4.8845930099487305,
|
| 15 |
+
"learning_rate": 1.0843373493975904e-05,
|
| 16 |
+
"loss": 1.0727,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.12048192771084337,
|
| 21 |
+
"grad_norm": 1.764939546585083,
|
| 22 |
+
"learning_rate": 2.289156626506024e-05,
|
| 23 |
+
"loss": 0.3548,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.18072289156626506,
|
| 28 |
+
"grad_norm": 1.1481366157531738,
|
| 29 |
+
"learning_rate": 3.4939759036144585e-05,
|
| 30 |
+
"loss": 0.2915,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.24096385542168675,
|
| 35 |
+
"grad_norm": 1.9749256372451782,
|
| 36 |
+
"learning_rate": 4.698795180722892e-05,
|
| 37 |
+
"loss": 0.2527,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.30120481927710846,
|
| 42 |
+
"grad_norm": 1.3480041027069092,
|
| 43 |
+
"learning_rate": 5.903614457831326e-05,
|
| 44 |
+
"loss": 0.2602,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3614457831325301,
|
| 49 |
+
"grad_norm": 1.096863031387329,
|
| 50 |
+
"learning_rate": 7.108433734939759e-05,
|
| 51 |
+
"loss": 0.2011,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.42168674698795183,
|
| 56 |
+
"grad_norm": 1.8047171831130981,
|
| 57 |
+
"learning_rate": 8.313253012048194e-05,
|
| 58 |
+
"loss": 0.1829,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.4819277108433735,
|
| 63 |
+
"grad_norm": 1.5872362852096558,
|
| 64 |
+
"learning_rate": 9.518072289156626e-05,
|
| 65 |
+
"loss": 0.1247,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5421686746987951,
|
| 70 |
+
"grad_norm": 1.5146158933639526,
|
| 71 |
+
"learning_rate": 9.99964283104959e-05,
|
| 72 |
+
"loss": 0.111,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6024096385542169,
|
| 77 |
+
"grad_norm": 1.3115952014923096,
|
| 78 |
+
"learning_rate": 9.997460316697722e-05,
|
| 79 |
+
"loss": 0.1124,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.6626506024096386,
|
| 84 |
+
"grad_norm": 0.850968062877655,
|
| 85 |
+
"learning_rate": 9.993294580260754e-05,
|
| 86 |
+
"loss": 0.0974,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.7228915662650602,
|
| 91 |
+
"grad_norm": 1.0239537954330444,
|
| 92 |
+
"learning_rate": 9.987147274894482e-05,
|
| 93 |
+
"loss": 0.0929,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7831325301204819,
|
| 98 |
+
"grad_norm": 1.0559345483779907,
|
| 99 |
+
"learning_rate": 9.979020840132389e-05,
|
| 100 |
+
"loss": 0.0745,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8433734939759037,
|
| 105 |
+
"grad_norm": 0.9941114783287048,
|
| 106 |
+
"learning_rate": 9.968918500917519e-05,
|
| 107 |
+
"loss": 0.0897,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.9036144578313253,
|
| 112 |
+
"grad_norm": 1.5254260301589966,
|
| 113 |
+
"learning_rate": 9.956844266322679e-05,
|
| 114 |
+
"loss": 0.0881,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.963855421686747,
|
| 119 |
+
"grad_norm": 0.8886228203773499,
|
| 120 |
+
"learning_rate": 9.942802927959443e-05,
|
| 121 |
+
"loss": 0.07,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.0240963855421688,
|
| 126 |
+
"grad_norm": 0.8230854272842407,
|
| 127 |
+
"learning_rate": 9.926800058076631e-05,
|
| 128 |
+
"loss": 0.057,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0843373493975903,
|
| 133 |
+
"grad_norm": 0.7548713684082031,
|
| 134 |
+
"learning_rate": 9.90884200734898e-05,
|
| 135 |
+
"loss": 0.0763,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.144578313253012,
|
| 140 |
+
"grad_norm": 1.0724797248840332,
|
| 141 |
+
"learning_rate": 9.8889359023569e-05,
|
| 142 |
+
"loss": 0.0806,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.2048192771084336,
|
| 147 |
+
"grad_norm": 1.020551323890686,
|
| 148 |
+
"learning_rate": 9.867089642758328e-05,
|
| 149 |
+
"loss": 0.0705,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.2650602409638554,
|
| 154 |
+
"grad_norm": 1.1815987825393677,
|
| 155 |
+
"learning_rate": 9.843311898153775e-05,
|
| 156 |
+
"loss": 0.0724,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.3253012048192772,
|
| 161 |
+
"grad_norm": 0.9071843028068542,
|
| 162 |
+
"learning_rate": 9.81761210464583e-05,
|
| 163 |
+
"loss": 0.0706,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.3855421686746987,
|
| 168 |
+
"grad_norm": 1.3225083351135254,
|
| 169 |
+
"learning_rate": 9.790000461094483e-05,
|
| 170 |
+
"loss": 0.0661,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.4457831325301205,
|
| 175 |
+
"grad_norm": 0.8698434233665466,
|
| 176 |
+
"learning_rate": 9.76048792506974e-05,
|
| 177 |
+
"loss": 0.0689,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.5060240963855422,
|
| 182 |
+
"grad_norm": 0.7452718615531921,
|
| 183 |
+
"learning_rate": 9.729086208503174e-05,
|
| 184 |
+
"loss": 0.0596,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.5662650602409638,
|
| 189 |
+
"grad_norm": 1.1432762145996094,
|
| 190 |
+
"learning_rate": 9.695807773040066e-05,
|
| 191 |
+
"loss": 0.0559,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.6265060240963856,
|
| 196 |
+
"grad_norm": 1.346431851387024,
|
| 197 |
+
"learning_rate": 9.660665825094075e-05,
|
| 198 |
+
"loss": 0.0592,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.6867469879518073,
|
| 203 |
+
"grad_norm": 1.0724352598190308,
|
| 204 |
+
"learning_rate": 9.62367431060631e-05,
|
| 205 |
+
"loss": 0.0559,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.7469879518072289,
|
| 210 |
+
"grad_norm": 0.4916505813598633,
|
| 211 |
+
"learning_rate": 9.584847909510946e-05,
|
| 212 |
+
"loss": 0.0555,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.8072289156626506,
|
| 217 |
+
"grad_norm": 0.8980555534362793,
|
| 218 |
+
"learning_rate": 9.544202029909552e-05,
|
| 219 |
+
"loss": 0.0652,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.8674698795180724,
|
| 224 |
+
"grad_norm": 1.0622588396072388,
|
| 225 |
+
"learning_rate": 9.501752801956442e-05,
|
| 226 |
+
"loss": 0.0618,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.927710843373494,
|
| 231 |
+
"grad_norm": 0.9151955842971802,
|
| 232 |
+
"learning_rate": 9.457517071457493e-05,
|
| 233 |
+
"loss": 0.0633,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.9879518072289155,
|
| 238 |
+
"grad_norm": 0.8452358841896057,
|
| 239 |
+
"learning_rate": 9.411512393184954e-05,
|
| 240 |
+
"loss": 0.0631,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.0481927710843375,
|
| 245 |
+
"grad_norm": 0.5913276672363281,
|
| 246 |
+
"learning_rate": 9.363757023910901e-05,
|
| 247 |
+
"loss": 0.0535,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.108433734939759,
|
| 252 |
+
"grad_norm": 0.7085587978363037,
|
| 253 |
+
"learning_rate": 9.314269915162114e-05,
|
| 254 |
+
"loss": 0.0636,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.1686746987951806,
|
| 259 |
+
"grad_norm": 0.6347533464431763,
|
| 260 |
+
"learning_rate": 9.263070705699242e-05,
|
| 261 |
+
"loss": 0.0575,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.2289156626506026,
|
| 266 |
+
"grad_norm": 0.8668927550315857,
|
| 267 |
+
"learning_rate": 9.21017971372323e-05,
|
| 268 |
+
"loss": 0.0508,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.289156626506024,
|
| 273 |
+
"grad_norm": 0.5759926438331604,
|
| 274 |
+
"learning_rate": 9.155617928812135e-05,
|
| 275 |
+
"loss": 0.0527,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.3493975903614457,
|
| 280 |
+
"grad_norm": 1.3119134902954102,
|
| 281 |
+
"learning_rate": 9.099407003591484e-05,
|
| 282 |
+
"loss": 0.0618,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.4096385542168672,
|
| 287 |
+
"grad_norm": 0.8716314435005188,
|
| 288 |
+
"learning_rate": 9.041569245141528e-05,
|
| 289 |
+
"loss": 0.0634,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.4698795180722892,
|
| 294 |
+
"grad_norm": 1.0534164905548096,
|
| 295 |
+
"learning_rate": 8.982127606144757e-05,
|
| 296 |
+
"loss": 0.0563,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.5301204819277108,
|
| 301 |
+
"grad_norm": 0.9702185392379761,
|
| 302 |
+
"learning_rate": 8.921105675777224e-05,
|
| 303 |
+
"loss": 0.0535,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.5903614457831328,
|
| 308 |
+
"grad_norm": 0.7751500010490417,
|
| 309 |
+
"learning_rate": 8.858527670347276e-05,
|
| 310 |
+
"loss": 0.0515,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.6506024096385543,
|
| 315 |
+
"grad_norm": 0.7694882750511169,
|
| 316 |
+
"learning_rate": 8.794418423685408e-05,
|
| 317 |
+
"loss": 0.0607,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.710843373493976,
|
| 322 |
+
"grad_norm": 0.9385339617729187,
|
| 323 |
+
"learning_rate": 8.728803377289061e-05,
|
| 324 |
+
"loss": 0.0585,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.7710843373493974,
|
| 329 |
+
"grad_norm": 0.5949042439460754,
|
| 330 |
+
"learning_rate": 8.661708570226252e-05,
|
| 331 |
+
"loss": 0.0607,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.8313253012048194,
|
| 336 |
+
"grad_norm": 0.8255975246429443,
|
| 337 |
+
"learning_rate": 8.593160628802093e-05,
|
| 338 |
+
"loss": 0.0588,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.891566265060241,
|
| 343 |
+
"grad_norm": 0.7478345632553101,
|
| 344 |
+
"learning_rate": 8.523186755992238e-05,
|
| 345 |
+
"loss": 0.049,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.9518072289156625,
|
| 350 |
+
"grad_norm": 0.7347597479820251,
|
| 351 |
+
"learning_rate": 8.451814720647488e-05,
|
| 352 |
+
"loss": 0.0536,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.0120481927710845,
|
| 357 |
+
"grad_norm": 0.5678183436393738,
|
| 358 |
+
"learning_rate": 8.379072846473845e-05,
|
| 359 |
+
"loss": 0.0524,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 3.072289156626506,
|
| 364 |
+
"grad_norm": 1.0103727579116821,
|
| 365 |
+
"learning_rate": 8.304990000792353e-05,
|
| 366 |
+
"loss": 0.0519,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 3.1325301204819276,
|
| 371 |
+
"grad_norm": 0.6050271391868591,
|
| 372 |
+
"learning_rate": 8.229595583083223e-05,
|
| 373 |
+
"loss": 0.0484,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.1927710843373496,
|
| 378 |
+
"grad_norm": 0.6075813174247742,
|
| 379 |
+
"learning_rate": 8.152919513318767e-05,
|
| 380 |
+
"loss": 0.055,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 3.253012048192771,
|
| 385 |
+
"grad_norm": 0.7674846053123474,
|
| 386 |
+
"learning_rate": 8.074992220089769e-05,
|
| 387 |
+
"loss": 0.0446,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.3132530120481927,
|
| 392 |
+
"grad_norm": 0.6162508130073547,
|
| 393 |
+
"learning_rate": 7.995844628530032e-05,
|
| 394 |
+
"loss": 0.0516,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.3734939759036147,
|
| 399 |
+
"grad_norm": 0.6467682123184204,
|
| 400 |
+
"learning_rate": 7.915508148043858e-05,
|
| 401 |
+
"loss": 0.0549,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.433734939759036,
|
| 406 |
+
"grad_norm": 0.7080559730529785,
|
| 407 |
+
"learning_rate": 7.834014659841349e-05,
|
| 408 |
+
"loss": 0.0558,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.4939759036144578,
|
| 413 |
+
"grad_norm": 0.4286852180957794,
|
| 414 |
+
"learning_rate": 7.751396504286482e-05,
|
| 415 |
+
"loss": 0.0438,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.5542168674698793,
|
| 420 |
+
"grad_norm": 0.6678372025489807,
|
| 421 |
+
"learning_rate": 7.66768646806296e-05,
|
| 422 |
+
"loss": 0.0523,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.6144578313253013,
|
| 427 |
+
"grad_norm": 0.7281410694122314,
|
| 428 |
+
"learning_rate": 7.58291777116295e-05,
|
| 429 |
+
"loss": 0.0456,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.674698795180723,
|
| 434 |
+
"grad_norm": 0.8437229990959167,
|
| 435 |
+
"learning_rate": 7.497124053703868e-05,
|
| 436 |
+
"loss": 0.0515,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.734939759036145,
|
| 441 |
+
"grad_norm": 0.6051187515258789,
|
| 442 |
+
"learning_rate": 7.410339362578422e-05,
|
| 443 |
+
"loss": 0.0504,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.7951807228915664,
|
| 448 |
+
"grad_norm": 0.942037045955658,
|
| 449 |
+
"learning_rate": 7.322598137943249e-05,
|
| 450 |
+
"loss": 0.0476,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.855421686746988,
|
| 455 |
+
"grad_norm": 1.020855188369751,
|
| 456 |
+
"learning_rate": 7.23393519955147e-05,
|
| 457 |
+
"loss": 0.0437,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.9156626506024095,
|
| 462 |
+
"grad_norm": 0.48984652757644653,
|
| 463 |
+
"learning_rate": 7.144385732934617e-05,
|
| 464 |
+
"loss": 0.0443,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.9759036144578315,
|
| 469 |
+
"grad_norm": 1.2586121559143066,
|
| 470 |
+
"learning_rate": 7.053985275439394e-05,
|
| 471 |
+
"loss": 0.0465,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.036144578313253,
|
| 476 |
+
"grad_norm": 0.6487300992012024,
|
| 477 |
+
"learning_rate": 6.962769702124819e-05,
|
| 478 |
+
"loss": 0.0455,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 4.096385542168675,
|
| 483 |
+
"grad_norm": 0.9266290664672852,
|
| 484 |
+
"learning_rate": 6.870775211525356e-05,
|
| 485 |
+
"loss": 0.0474,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 4.156626506024097,
|
| 490 |
+
"grad_norm": 0.5156567692756653,
|
| 491 |
+
"learning_rate": 6.778038311285666e-05,
|
| 492 |
+
"loss": 0.0397,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 4.216867469879518,
|
| 497 |
+
"grad_norm": 0.7056227326393127,
|
| 498 |
+
"learning_rate": 6.684595803672705e-05,
|
| 499 |
+
"loss": 0.0485,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 4.27710843373494,
|
| 504 |
+
"grad_norm": 0.6396136283874512,
|
| 505 |
+
"learning_rate": 6.59048477097088e-05,
|
| 506 |
+
"loss": 0.0396,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 4.337349397590361,
|
| 511 |
+
"grad_norm": 0.47223004698753357,
|
| 512 |
+
"learning_rate": 6.495742560766101e-05,
|
| 513 |
+
"loss": 0.0358,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.397590361445783,
|
| 518 |
+
"grad_norm": 0.5486592054367065,
|
| 519 |
+
"learning_rate": 6.400406771124536e-05,
|
| 520 |
+
"loss": 0.0415,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.457831325301205,
|
| 525 |
+
"grad_norm": 0.8488519191741943,
|
| 526 |
+
"learning_rate": 6.304515235671976e-05,
|
| 527 |
+
"loss": 0.0469,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.518072289156627,
|
| 532 |
+
"grad_norm": 0.4704228639602661,
|
| 533 |
+
"learning_rate": 6.208106008579706e-05,
|
| 534 |
+
"loss": 0.0418,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 4.578313253012048,
|
| 539 |
+
"grad_norm": 0.7474722862243652,
|
| 540 |
+
"learning_rate": 6.111217349462864e-05,
|
| 541 |
+
"loss": 0.0428,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.63855421686747,
|
| 546 |
+
"grad_norm": 0.537147045135498,
|
| 547 |
+
"learning_rate": 6.013887708197261e-05,
|
| 548 |
+
"loss": 0.0376,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.698795180722891,
|
| 553 |
+
"grad_norm": 0.8423475623130798,
|
| 554 |
+
"learning_rate": 5.916155709660707e-05,
|
| 555 |
+
"loss": 0.0439,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.759036144578313,
|
| 560 |
+
"grad_norm": 0.6123393774032593,
|
| 561 |
+
"learning_rate": 5.818060138404882e-05,
|
| 562 |
+
"loss": 0.0427,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.8192771084337345,
|
| 567 |
+
"grad_norm": 0.49415451288223267,
|
| 568 |
+
"learning_rate": 5.719639923263841e-05,
|
| 569 |
+
"loss": 0.0413,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 4.879518072289157,
|
| 574 |
+
"grad_norm": 0.6448298692703247,
|
| 575 |
+
"learning_rate": 5.620934121905262e-05,
|
| 576 |
+
"loss": 0.0404,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 4.9397590361445785,
|
| 581 |
+
"grad_norm": 0.6740503311157227,
|
| 582 |
+
"learning_rate": 5.521981905330572e-05,
|
| 583 |
+
"loss": 0.0474,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 5.0,
|
| 588 |
+
"grad_norm": 0.7011913657188416,
|
| 589 |
+
"learning_rate": 5.422822542330084e-05,
|
| 590 |
+
"loss": 0.0427,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 5.0602409638554215,
|
| 595 |
+
"grad_norm": 0.3385421931743622,
|
| 596 |
+
"learning_rate": 5.3234953838993364e-05,
|
| 597 |
+
"loss": 0.0457,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 5.120481927710843,
|
| 602 |
+
"grad_norm": 0.6901074051856995,
|
| 603 |
+
"learning_rate": 5.224039847622821e-05,
|
| 604 |
+
"loss": 0.0422,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 5.180722891566265,
|
| 609 |
+
"grad_norm": 0.5770816206932068,
|
| 610 |
+
"learning_rate": 5.124495402031255e-05,
|
| 611 |
+
"loss": 0.0408,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 5.240963855421687,
|
| 616 |
+
"grad_norm": 0.6947376132011414,
|
| 617 |
+
"learning_rate": 5.024901550938664e-05,
|
| 618 |
+
"loss": 0.0385,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 5.301204819277109,
|
| 623 |
+
"grad_norm": 0.4837009310722351,
|
| 624 |
+
"learning_rate": 4.925297817765447e-05,
|
| 625 |
+
"loss": 0.0369,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 5.36144578313253,
|
| 630 |
+
"grad_norm": 0.4218069911003113,
|
| 631 |
+
"learning_rate": 4.8257237298536665e-05,
|
| 632 |
+
"loss": 0.0318,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 5.421686746987952,
|
| 637 |
+
"grad_norm": 0.4409538805484772,
|
| 638 |
+
"learning_rate": 4.726218802780786e-05,
|
| 639 |
+
"loss": 0.0338,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.481927710843373,
|
| 644 |
+
"grad_norm": 0.48165932297706604,
|
| 645 |
+
"learning_rate": 4.6268225246780597e-05,
|
| 646 |
+
"loss": 0.0311,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 5.542168674698795,
|
| 651 |
+
"grad_norm": 0.5170968770980835,
|
| 652 |
+
"learning_rate": 4.527574340559844e-05,
|
| 653 |
+
"loss": 0.0344,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 5.602409638554217,
|
| 658 |
+
"grad_norm": 0.40369731187820435,
|
| 659 |
+
"learning_rate": 4.4285136366699834e-05,
|
| 660 |
+
"loss": 0.0331,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 5.662650602409639,
|
| 665 |
+
"grad_norm": 0.5428269505500793,
|
| 666 |
+
"learning_rate": 4.3296797248515476e-05,
|
| 667 |
+
"loss": 0.0365,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 5.72289156626506,
|
| 672 |
+
"grad_norm": 0.5662912726402283,
|
| 673 |
+
"learning_rate": 4.231111826946085e-05,
|
| 674 |
+
"loss": 0.0317,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 5.783132530120482,
|
| 679 |
+
"grad_norm": 0.5741373300552368,
|
| 680 |
+
"learning_rate": 4.132849059228583e-05,
|
| 681 |
+
"loss": 0.0342,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 5.843373493975903,
|
| 686 |
+
"grad_norm": 0.5292026400566101,
|
| 687 |
+
"learning_rate": 4.034930416884346e-05,
|
| 688 |
+
"loss": 0.0418,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 5.903614457831325,
|
| 693 |
+
"grad_norm": 0.7288020253181458,
|
| 694 |
+
"learning_rate": 3.9373947585339e-05,
|
| 695 |
+
"loss": 0.0341,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.9638554216867465,
|
| 700 |
+
"grad_norm": 0.5807807445526123,
|
| 701 |
+
"learning_rate": 3.840280790812106e-05,
|
| 702 |
+
"loss": 0.0389,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 6.024096385542169,
|
| 707 |
+
"grad_norm": 0.8824964761734009,
|
| 708 |
+
"learning_rate": 3.7436270530075974e-05,
|
| 709 |
+
"loss": 0.0379,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 6.0843373493975905,
|
| 714 |
+
"grad_norm": 0.43787458539009094,
|
| 715 |
+
"learning_rate": 3.647471901768601e-05,
|
| 716 |
+
"loss": 0.0362,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 6.144578313253012,
|
| 721 |
+
"grad_norm": 0.6869735717773438,
|
| 722 |
+
"learning_rate": 3.551853495881273e-05,
|
| 723 |
+
"loss": 0.0335,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 6.204819277108434,
|
| 728 |
+
"grad_norm": 0.4364345669746399,
|
| 729 |
+
"learning_rate": 3.456809781126528e-05,
|
| 730 |
+
"loss": 0.0342,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 6.265060240963855,
|
| 735 |
+
"grad_norm": 0.6566846966743469,
|
| 736 |
+
"learning_rate": 3.362378475221404e-05,
|
| 737 |
+
"loss": 0.0373,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 6.325301204819277,
|
| 742 |
+
"grad_norm": 0.35175663232803345,
|
| 743 |
+
"learning_rate": 3.268597052850948e-05,
|
| 744 |
+
"loss": 0.0305,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 6.385542168674699,
|
| 749 |
+
"grad_norm": 0.6008738279342651,
|
| 750 |
+
"learning_rate": 3.1755027307965314e-05,
|
| 751 |
+
"loss": 0.0314,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 6.445783132530121,
|
| 756 |
+
"grad_norm": 0.3923781216144562,
|
| 757 |
+
"learning_rate": 3.083132453166519e-05,
|
| 758 |
+
"loss": 0.0275,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 6.506024096385542,
|
| 763 |
+
"grad_norm": 0.5415867567062378,
|
| 764 |
+
"learning_rate": 2.991522876735154e-05,
|
| 765 |
+
"loss": 0.0319,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 6.566265060240964,
|
| 770 |
+
"grad_norm": 0.4087677597999573,
|
| 771 |
+
"learning_rate": 2.9007103563954523e-05,
|
| 772 |
+
"loss": 0.0307,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 6.626506024096385,
|
| 777 |
+
"grad_norm": 0.5138642191886902,
|
| 778 |
+
"learning_rate": 2.810730930731923e-05,
|
| 779 |
+
"loss": 0.0275,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 6.686746987951807,
|
| 784 |
+
"grad_norm": 0.39721858501434326,
|
| 785 |
+
"learning_rate": 2.721620307718793e-05,
|
| 786 |
+
"loss": 0.0288,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 6.746987951807229,
|
| 791 |
+
"grad_norm": 0.47175243496894836,
|
| 792 |
+
"learning_rate": 2.6334138505494355e-05,
|
| 793 |
+
"loss": 0.0316,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 6.807228915662651,
|
| 798 |
+
"grad_norm": 0.4157463610172272,
|
| 799 |
+
"learning_rate": 2.546146563602644e-05,
|
| 800 |
+
"loss": 0.0324,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 6.867469879518072,
|
| 805 |
+
"grad_norm": 0.4235219657421112,
|
| 806 |
+
"learning_rate": 2.4598530785512664e-05,
|
| 807 |
+
"loss": 0.0334,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 6.927710843373494,
|
| 812 |
+
"grad_norm": 0.5735675692558289,
|
| 813 |
+
"learning_rate": 2.37456764061878e-05,
|
| 814 |
+
"loss": 0.0327,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 6.9879518072289155,
|
| 819 |
+
"grad_norm": 0.3093161880970001,
|
| 820 |
+
"learning_rate": 2.290324094989205e-05,
|
| 821 |
+
"loss": 0.0332,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 7.048192771084337,
|
| 826 |
+
"grad_norm": 0.491948664188385,
|
| 827 |
+
"learning_rate": 2.2071558733757756e-05,
|
| 828 |
+
"loss": 0.0331,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 7.108433734939759,
|
| 833 |
+
"grad_norm": 0.34060969948768616,
|
| 834 |
+
"learning_rate": 2.125095980753708e-05,
|
| 835 |
+
"loss": 0.0359,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.168674698795181,
|
| 840 |
+
"grad_norm": 0.3169781267642975,
|
| 841 |
+
"learning_rate": 2.0441769822622965e-05,
|
| 842 |
+
"loss": 0.0287,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.228915662650603,
|
| 847 |
+
"grad_norm": 0.3545713424682617,
|
| 848 |
+
"learning_rate": 1.9644309902815815e-05,
|
| 849 |
+
"loss": 0.0223,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.289156626506024,
|
| 854 |
+
"grad_norm": 0.34389132261276245,
|
| 855 |
+
"learning_rate": 1.885889651688671e-05,
|
| 856 |
+
"loss": 0.0286,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.349397590361446,
|
| 861 |
+
"grad_norm": 0.49160459637641907,
|
| 862 |
+
"learning_rate": 1.8085841352988043e-05,
|
| 863 |
+
"loss": 0.029,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.409638554216867,
|
| 868 |
+
"grad_norm": 0.3772972524166107,
|
| 869 |
+
"learning_rate": 1.7325451194961412e-05,
|
| 870 |
+
"loss": 0.0296,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.469879518072289,
|
| 875 |
+
"grad_norm": 0.5732247829437256,
|
| 876 |
+
"learning_rate": 1.6578027800591578e-05,
|
| 877 |
+
"loss": 0.0221,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.530120481927711,
|
| 882 |
+
"grad_norm": 0.3436269164085388,
|
| 883 |
+
"learning_rate": 1.584386778185517e-05,
|
| 884 |
+
"loss": 0.0319,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 7.590361445783133,
|
| 889 |
+
"grad_norm": 0.2605295181274414,
|
| 890 |
+
"learning_rate": 1.5123262487211404e-05,
|
| 891 |
+
"loss": 0.0292,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 7.650602409638554,
|
| 896 |
+
"grad_norm": 0.25415462255477905,
|
| 897 |
+
"learning_rate": 1.441649788598154e-05,
|
| 898 |
+
"loss": 0.0306,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 7.710843373493976,
|
| 903 |
+
"grad_norm": 0.3771422207355499,
|
| 904 |
+
"learning_rate": 1.3723854454863239e-05,
|
| 905 |
+
"loss": 0.0297,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 7.771084337349397,
|
| 910 |
+
"grad_norm": 0.895290195941925,
|
| 911 |
+
"learning_rate": 1.304560706662436e-05,
|
| 912 |
+
"loss": 0.03,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 7.831325301204819,
|
| 917 |
+
"grad_norm": 0.5237590074539185,
|
| 918 |
+
"learning_rate": 1.2382024881020937e-05,
|
| 919 |
+
"loss": 0.0316,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 7.891566265060241,
|
| 924 |
+
"grad_norm": 0.48082613945007324,
|
| 925 |
+
"learning_rate": 1.1733371237982132e-05,
|
| 926 |
+
"loss": 0.0225,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 7.951807228915663,
|
| 931 |
+
"grad_norm": 0.5486697554588318,
|
| 932 |
+
"learning_rate": 1.1099903553104834e-05,
|
| 933 |
+
"loss": 0.0236,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 8.012048192771084,
|
| 938 |
+
"grad_norm": 0.5749785304069519,
|
| 939 |
+
"learning_rate": 1.0481873215499378e-05,
|
| 940 |
+
"loss": 0.0341,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 8.072289156626505,
|
| 945 |
+
"grad_norm": 0.47802257537841797,
|
| 946 |
+
"learning_rate": 9.879525488026704e-06,
|
| 947 |
+
"loss": 0.0285,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 8.132530120481928,
|
| 952 |
+
"grad_norm": 0.39925578236579895,
|
| 953 |
+
"learning_rate": 9.293099409966843e-06,
|
| 954 |
+
"loss": 0.0226,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 8.19277108433735,
|
| 959 |
+
"grad_norm": 0.3280559182167053,
|
| 960 |
+
"learning_rate": 8.722827702157104e-06,
|
| 961 |
+
"loss": 0.0225,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 8.25301204819277,
|
| 966 |
+
"grad_norm": 0.47502222657203674,
|
| 967 |
+
"learning_rate": 8.168936674637728e-06,
|
| 968 |
+
"loss": 0.0237,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 8.313253012048193,
|
| 973 |
+
"grad_norm": 0.2550809681415558,
|
| 974 |
+
"learning_rate": 7.631646136841702e-06,
|
| 975 |
+
"loss": 0.022,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 8.373493975903614,
|
| 980 |
+
"grad_norm": 0.32360807061195374,
|
| 981 |
+
"learning_rate": 7.111169310364202e-06,
|
| 982 |
+
"loss": 0.0225,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 8.433734939759036,
|
| 987 |
+
"grad_norm": 0.41676631569862366,
|
| 988 |
+
"learning_rate": 6.607712744346539e-06,
|
| 989 |
+
"loss": 0.023,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 8.493975903614459,
|
| 994 |
+
"grad_norm": 0.36988356709480286,
|
| 995 |
+
"learning_rate": 6.1214762335079e-06,
|
| 996 |
+
"loss": 0.028,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 8.55421686746988,
|
| 1001 |
+
"grad_norm": 0.4012046456336975,
|
| 1002 |
+
"learning_rate": 5.652652738857628e-06,
|
| 1003 |
+
"loss": 0.0233,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 8.614457831325302,
|
| 1008 |
+
"grad_norm": 0.36382830142974854,
|
| 1009 |
+
"learning_rate": 5.201428311119416e-06,
|
| 1010 |
+
"loss": 0.0239,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 8.674698795180722,
|
| 1015 |
+
"grad_norm": 0.38201379776000977,
|
| 1016 |
+
"learning_rate": 4.767982016897854e-06,
|
| 1017 |
+
"loss": 0.0235,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 8.734939759036145,
|
| 1022 |
+
"grad_norm": 0.437858521938324,
|
| 1023 |
+
"learning_rate": 4.352485867616524e-06,
|
| 1024 |
+
"loss": 0.0188,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 8.795180722891565,
|
| 1029 |
+
"grad_norm": 0.36419281363487244,
|
| 1030 |
+
"learning_rate": 3.955104751255995e-06,
|
| 1031 |
+
"loss": 0.0224,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 8.855421686746988,
|
| 1036 |
+
"grad_norm": 0.2299613207578659,
|
| 1037 |
+
"learning_rate": 3.575996366918638e-06,
|
| 1038 |
+
"loss": 0.0217,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 8.91566265060241,
|
| 1043 |
+
"grad_norm": 0.347234845161438,
|
| 1044 |
+
"learning_rate": 3.215311162246426e-06,
|
| 1045 |
+
"loss": 0.0272,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 8.975903614457831,
|
| 1050 |
+
"grad_norm": 0.29449528455734253,
|
| 1051 |
+
"learning_rate": 2.8731922737163685e-06,
|
| 1052 |
+
"loss": 0.024,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 9.036144578313253,
|
| 1057 |
+
"grad_norm": 0.3682246804237366,
|
| 1058 |
+
"learning_rate": 2.5497754698373454e-06,
|
| 1059 |
+
"loss": 0.0232,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 9.096385542168674,
|
| 1064 |
+
"grad_norm": 0.7771542072296143,
|
| 1065 |
+
"learning_rate": 2.2451890972709775e-06,
|
| 1066 |
+
"loss": 0.0234,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 9.156626506024097,
|
| 1071 |
+
"grad_norm": 0.44349297881126404,
|
| 1072 |
+
"learning_rate": 1.959554029897703e-06,
|
| 1073 |
+
"loss": 0.0191,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 9.216867469879517,
|
| 1078 |
+
"grad_norm": 0.5158273577690125,
|
| 1079 |
+
"learning_rate": 1.6929836208485206e-06,
|
| 1080 |
+
"loss": 0.026,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 9.27710843373494,
|
| 1085 |
+
"grad_norm": 0.2851838171482086,
|
| 1086 |
+
"learning_rate": 1.4455836575212157e-06,
|
| 1087 |
+
"loss": 0.0231,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 9.337349397590362,
|
| 1092 |
+
"grad_norm": 0.2680720388889313,
|
| 1093 |
+
"learning_rate": 1.2174523195990773e-06,
|
| 1094 |
+
"loss": 0.0194,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 9.397590361445783,
|
| 1099 |
+
"grad_norm": 0.40615180134773254,
|
| 1100 |
+
"learning_rate": 1.0086801400886747e-06,
|
| 1101 |
+
"loss": 0.0207,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 9.457831325301205,
|
| 1106 |
+
"grad_norm": 0.35740238428115845,
|
| 1107 |
+
"learning_rate": 8.193499693921858e-07,
|
| 1108 |
+
"loss": 0.0295,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 9.518072289156626,
|
| 1113 |
+
"grad_norm": 0.3606247007846832,
|
| 1114 |
+
"learning_rate": 6.495369424285747e-07,
|
| 1115 |
+
"loss": 0.0202,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 9.578313253012048,
|
| 1120 |
+
"grad_norm": 0.41667816042900085,
|
| 1121 |
+
"learning_rate": 4.993084488165933e-07,
|
| 1122 |
+
"loss": 0.0214,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 9.638554216867469,
|
| 1127 |
+
"grad_norm": 0.3058539032936096,
|
| 1128 |
+
"learning_rate": 3.687241061314461e-07,
|
| 1129 |
+
"loss": 0.0274,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 9.698795180722891,
|
| 1134 |
+
"grad_norm": 0.4347921907901764,
|
| 1135 |
+
"learning_rate": 2.578357362458539e-07,
|
| 1136 |
+
"loss": 0.0229,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 9.759036144578314,
|
| 1141 |
+
"grad_norm": 0.187905415892601,
|
| 1142 |
+
"learning_rate": 1.6668734476468773e-07,
|
| 1143 |
+
"loss": 0.0214,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 9.819277108433734,
|
| 1148 |
+
"grad_norm": 0.24851736426353455,
|
| 1149 |
+
"learning_rate": 9.53151035615829e-08,
|
| 1150 |
+
"loss": 0.0251,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 9.879518072289157,
|
| 1155 |
+
"grad_norm": 0.2686261832714081,
|
| 1156 |
+
"learning_rate": 4.3747336424226905e-08,
|
| 1157 |
+
"loss": 0.0242,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 9.939759036144578,
|
| 1162 |
+
"grad_norm": 0.2994416356086731,
|
| 1163 |
+
"learning_rate": 1.2004507814200594e-08,
|
| 1164 |
+
"loss": 0.0212,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 10.0,
|
| 1169 |
+
"grad_norm": 0.47449934482574463,
|
| 1170 |
+
"learning_rate": 9.921474574681177e-11,
|
| 1171 |
+
"loss": 0.019,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 10.0,
|
| 1176 |
+
"step": 1660,
|
| 1177 |
+
"total_flos": 0.0,
|
| 1178 |
+
"train_loss": 0.05688802816243057,
|
| 1179 |
+
"train_runtime": 1674.5284,
|
| 1180 |
+
"train_samples_per_second": 48.378,
|
| 1181 |
+
"train_steps_per_second": 0.991
|
| 1182 |
+
}
|
| 1183 |
+
],
|
| 1184 |
+
"logging_steps": 10,
|
| 1185 |
+
"max_steps": 1660,
|
| 1186 |
+
"num_input_tokens_seen": 0,
|
| 1187 |
+
"num_train_epochs": 10,
|
| 1188 |
+
"save_steps": 3000,
|
| 1189 |
+
"stateful_callbacks": {
|
| 1190 |
+
"TrainerControl": {
|
| 1191 |
+
"args": {
|
| 1192 |
+
"should_epoch_stop": false,
|
| 1193 |
+
"should_evaluate": false,
|
| 1194 |
+
"should_log": false,
|
| 1195 |
+
"should_save": true,
|
| 1196 |
+
"should_training_stop": true
|
| 1197 |
+
},
|
| 1198 |
+
"attributes": {}
|
| 1199 |
+
}
|
| 1200 |
+
},
|
| 1201 |
+
"total_flos": 0.0,
|
| 1202 |
+
"train_batch_size": 49,
|
| 1203 |
+
"trial_name": null,
|
| 1204 |
+
"trial_params": null
|
| 1205 |
+
}
|