Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1204 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1204 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.821428571428573,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1665,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11904761904761904,
|
| 13 |
+
"grad_norm": 4.05410099029541,
|
| 14 |
+
"learning_rate": 2.380952380952381e-05,
|
| 15 |
+
"loss": 1.0474,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.23809523809523808,
|
| 20 |
+
"grad_norm": 1.3301610946655273,
|
| 21 |
+
"learning_rate": 4.761904761904762e-05,
|
| 22 |
+
"loss": 0.4554,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.35714285714285715,
|
| 27 |
+
"grad_norm": 1.262498378753662,
|
| 28 |
+
"learning_rate": 7.142857142857143e-05,
|
| 29 |
+
"loss": 0.3098,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.47619047619047616,
|
| 34 |
+
"grad_norm": 0.8813798427581787,
|
| 35 |
+
"learning_rate": 9.523809523809524e-05,
|
| 36 |
+
"loss": 0.2559,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5952380952380952,
|
| 41 |
+
"grad_norm": 1.2415088415145874,
|
| 42 |
+
"learning_rate": 0.00011904761904761905,
|
| 43 |
+
"loss": 0.2103,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.7142857142857143,
|
| 48 |
+
"grad_norm": 1.0791610479354858,
|
| 49 |
+
"learning_rate": 0.00014285714285714287,
|
| 50 |
+
"loss": 0.1988,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8333333333333334,
|
| 55 |
+
"grad_norm": 0.8729690313339233,
|
| 56 |
+
"learning_rate": 0.0001666666666666667,
|
| 57 |
+
"loss": 0.1797,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9523809523809523,
|
| 62 |
+
"grad_norm": 1.2665623426437378,
|
| 63 |
+
"learning_rate": 0.00019047619047619048,
|
| 64 |
+
"loss": 0.1687,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.0714285714285714,
|
| 69 |
+
"grad_norm": 0.9594415426254272,
|
| 70 |
+
"learning_rate": 0.00019999289272096886,
|
| 71 |
+
"loss": 0.1538,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.1904761904761905,
|
| 76 |
+
"grad_norm": 1.446765422821045,
|
| 77 |
+
"learning_rate": 0.00019994946300764274,
|
| 78 |
+
"loss": 0.145,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.3095238095238095,
|
| 83 |
+
"grad_norm": 0.7596405744552612,
|
| 84 |
+
"learning_rate": 0.00019986656919636397,
|
| 85 |
+
"loss": 0.1428,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.4285714285714286,
|
| 90 |
+
"grad_norm": 0.8046891093254089,
|
| 91 |
+
"learning_rate": 0.00019974424401696808,
|
| 92 |
+
"loss": 0.1262,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.5476190476190477,
|
| 97 |
+
"grad_norm": 0.3929666578769684,
|
| 98 |
+
"learning_rate": 0.00019958253576839256,
|
| 99 |
+
"loss": 0.117,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.6666666666666665,
|
| 104 |
+
"grad_norm": 1.0657896995544434,
|
| 105 |
+
"learning_rate": 0.00019938150829960634,
|
| 106 |
+
"loss": 0.1223,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.7857142857142856,
|
| 111 |
+
"grad_norm": 0.4697973132133484,
|
| 112 |
+
"learning_rate": 0.00019914124098439974,
|
| 113 |
+
"loss": 0.1152,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.9047619047619047,
|
| 118 |
+
"grad_norm": 0.40580397844314575,
|
| 119 |
+
"learning_rate": 0.00019886182869004445,
|
| 120 |
+
"loss": 0.104,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.0238095238095237,
|
| 125 |
+
"grad_norm": 0.5756620168685913,
|
| 126 |
+
"learning_rate": 0.00019854338173983614,
|
| 127 |
+
"loss": 0.0907,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.142857142857143,
|
| 132 |
+
"grad_norm": 0.33545899391174316,
|
| 133 |
+
"learning_rate": 0.00019818602586953415,
|
| 134 |
+
"loss": 0.0889,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.261904761904762,
|
| 139 |
+
"grad_norm": 0.7573866844177246,
|
| 140 |
+
"learning_rate": 0.00019778990217771621,
|
| 141 |
+
"loss": 0.0932,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.380952380952381,
|
| 146 |
+
"grad_norm": 0.45556336641311646,
|
| 147 |
+
"learning_rate": 0.00019735516707006676,
|
| 148 |
+
"loss": 0.0855,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.5,
|
| 153 |
+
"grad_norm": 0.346583753824234,
|
| 154 |
+
"learning_rate": 0.00019688199219762182,
|
| 155 |
+
"loss": 0.0853,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.619047619047619,
|
| 160 |
+
"grad_norm": 0.44082921743392944,
|
| 161 |
+
"learning_rate": 0.0001963705643889941,
|
| 162 |
+
"loss": 0.0818,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.738095238095238,
|
| 167 |
+
"grad_norm": 0.5446610450744629,
|
| 168 |
+
"learning_rate": 0.00019582108557660553,
|
| 169 |
+
"loss": 0.079,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.857142857142857,
|
| 174 |
+
"grad_norm": 0.5620890259742737,
|
| 175 |
+
"learning_rate": 0.0001952337727169561,
|
| 176 |
+
"loss": 0.0826,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.9761904761904763,
|
| 181 |
+
"grad_norm": 0.27951428294181824,
|
| 182 |
+
"learning_rate": 0.0001946088577049608,
|
| 183 |
+
"loss": 0.077,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.0952380952380953,
|
| 188 |
+
"grad_norm": 0.28334465622901917,
|
| 189 |
+
"learning_rate": 0.00019394658728238794,
|
| 190 |
+
"loss": 0.0791,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.2142857142857144,
|
| 195 |
+
"grad_norm": 0.33841413259506226,
|
| 196 |
+
"learning_rate": 0.00019324722294043558,
|
| 197 |
+
"loss": 0.0832,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.3333333333333335,
|
| 202 |
+
"grad_norm": 0.3847584128379822,
|
| 203 |
+
"learning_rate": 0.0001925110408164842,
|
| 204 |
+
"loss": 0.0716,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.4523809523809526,
|
| 209 |
+
"grad_norm": 0.277832567691803,
|
| 210 |
+
"learning_rate": 0.00019173833158506648,
|
| 211 |
+
"loss": 0.07,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.571428571428571,
|
| 216 |
+
"grad_norm": 0.33352115750312805,
|
| 217 |
+
"learning_rate": 0.00019092940034309722,
|
| 218 |
+
"loss": 0.0709,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.6904761904761907,
|
| 223 |
+
"grad_norm": 0.5264162421226501,
|
| 224 |
+
"learning_rate": 0.0001900845664894086,
|
| 225 |
+
"loss": 0.0731,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.8095238095238093,
|
| 230 |
+
"grad_norm": 0.3690634071826935,
|
| 231 |
+
"learning_rate": 0.00018920416359863887,
|
| 232 |
+
"loss": 0.0635,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.928571428571429,
|
| 237 |
+
"grad_norm": 0.33004167675971985,
|
| 238 |
+
"learning_rate": 0.0001882885392895232,
|
| 239 |
+
"loss": 0.0745,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.0476190476190474,
|
| 244 |
+
"grad_norm": 0.4185870587825775,
|
| 245 |
+
"learning_rate": 0.00018733805508764002,
|
| 246 |
+
"loss": 0.0802,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 4.166666666666667,
|
| 251 |
+
"grad_norm": 0.30809977650642395,
|
| 252 |
+
"learning_rate": 0.00018635308628266585,
|
| 253 |
+
"loss": 0.071,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.285714285714286,
|
| 258 |
+
"grad_norm": 0.3416142463684082,
|
| 259 |
+
"learning_rate": 0.00018533402178019594,
|
| 260 |
+
"loss": 0.0746,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.404761904761905,
|
| 265 |
+
"grad_norm": 0.3178706467151642,
|
| 266 |
+
"learning_rate": 0.0001842812639481884,
|
| 267 |
+
"loss": 0.0615,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.523809523809524,
|
| 272 |
+
"grad_norm": 0.2836775779724121,
|
| 273 |
+
"learning_rate": 0.00018319522845809306,
|
| 274 |
+
"loss": 0.0595,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.642857142857143,
|
| 279 |
+
"grad_norm": 0.3071931004524231,
|
| 280 |
+
"learning_rate": 0.00018207634412072764,
|
| 281 |
+
"loss": 0.0636,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.761904761904762,
|
| 286 |
+
"grad_norm": 0.4010617434978485,
|
| 287 |
+
"learning_rate": 0.0001809250527169658,
|
| 288 |
+
"loss": 0.0593,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.880952380952381,
|
| 293 |
+
"grad_norm": 0.2992096245288849,
|
| 294 |
+
"learning_rate": 0.00017974180882330412,
|
| 295 |
+
"loss": 0.0687,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.0,
|
| 300 |
+
"grad_norm": 0.7678210735321045,
|
| 301 |
+
"learning_rate": 0.0001785270796323769,
|
| 302 |
+
"loss": 0.0647,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.119047619047619,
|
| 307 |
+
"grad_norm": 0.2817954123020172,
|
| 308 |
+
"learning_rate": 0.00017728134476848966,
|
| 309 |
+
"loss": 0.0637,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.238095238095238,
|
| 314 |
+
"grad_norm": 0.3238624930381775,
|
| 315 |
+
"learning_rate": 0.00017600509609824388,
|
| 316 |
+
"loss": 0.0654,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.357142857142857,
|
| 321 |
+
"grad_norm": 0.4325568377971649,
|
| 322 |
+
"learning_rate": 0.00017469883753632817,
|
| 323 |
+
"loss": 0.0574,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.476190476190476,
|
| 328 |
+
"grad_norm": 0.3361673653125763,
|
| 329 |
+
"learning_rate": 0.0001733630848465525,
|
| 330 |
+
"loss": 0.0614,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.595238095238095,
|
| 335 |
+
"grad_norm": 0.2163662165403366,
|
| 336 |
+
"learning_rate": 0.00017199836543820357,
|
| 337 |
+
"loss": 0.0564,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.714285714285714,
|
| 342 |
+
"grad_norm": 0.3433023989200592,
|
| 343 |
+
"learning_rate": 0.00017060521815780223,
|
| 344 |
+
"loss": 0.0566,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.833333333333333,
|
| 349 |
+
"grad_norm": 0.2643953859806061,
|
| 350 |
+
"learning_rate": 0.0001691841930763453,
|
| 351 |
+
"loss": 0.0486,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.9523809523809526,
|
| 356 |
+
"grad_norm": 0.2475927174091339,
|
| 357 |
+
"learning_rate": 0.00016773585127211478,
|
| 358 |
+
"loss": 0.0567,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.071428571428571,
|
| 363 |
+
"grad_norm": 0.23853139579296112,
|
| 364 |
+
"learning_rate": 0.00016626076460914198,
|
| 365 |
+
"loss": 0.0516,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.190476190476191,
|
| 370 |
+
"grad_norm": 0.22467206418514252,
|
| 371 |
+
"learning_rate": 0.00016475951551141199,
|
| 372 |
+
"loss": 0.0515,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.309523809523809,
|
| 377 |
+
"grad_norm": 0.3674350380897522,
|
| 378 |
+
"learning_rate": 0.0001632326967328993,
|
| 379 |
+
"loss": 0.0513,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.428571428571429,
|
| 384 |
+
"grad_norm": 0.2875892221927643,
|
| 385 |
+
"learning_rate": 0.0001616809111235244,
|
| 386 |
+
"loss": 0.0546,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.5476190476190474,
|
| 391 |
+
"grad_norm": 0.27081480622291565,
|
| 392 |
+
"learning_rate": 0.0001601047713911244,
|
| 393 |
+
"loss": 0.0486,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.666666666666667,
|
| 398 |
+
"grad_norm": 0.29003506898880005,
|
| 399 |
+
"learning_rate": 0.00015850489985953076,
|
| 400 |
+
"loss": 0.0519,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 6.785714285714286,
|
| 405 |
+
"grad_norm": 0.267265647649765,
|
| 406 |
+
"learning_rate": 0.00015688192822285117,
|
| 407 |
+
"loss": 0.0534,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 6.904761904761905,
|
| 412 |
+
"grad_norm": 0.2007722705602646,
|
| 413 |
+
"learning_rate": 0.0001552364972960506,
|
| 414 |
+
"loss": 0.0525,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.023809523809524,
|
| 419 |
+
"grad_norm": 0.2781243324279785,
|
| 420 |
+
"learning_rate": 0.0001535692567619319,
|
| 421 |
+
"loss": 0.049,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.142857142857143,
|
| 426 |
+
"grad_norm": 0.302306592464447,
|
| 427 |
+
"learning_rate": 0.00015188086491461466,
|
| 428 |
+
"loss": 0.0479,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.261904761904762,
|
| 433 |
+
"grad_norm": 0.23352427780628204,
|
| 434 |
+
"learning_rate": 0.0001501719883996139,
|
| 435 |
+
"loss": 0.0522,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.380952380952381,
|
| 440 |
+
"grad_norm": 0.3669947385787964,
|
| 441 |
+
"learning_rate": 0.00014844330195062144,
|
| 442 |
+
"loss": 0.0498,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.5,
|
| 447 |
+
"grad_norm": 0.22105714678764343,
|
| 448 |
+
"learning_rate": 0.00014669548812309388,
|
| 449 |
+
"loss": 0.048,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.619047619047619,
|
| 454 |
+
"grad_norm": 0.2889458239078522,
|
| 455 |
+
"learning_rate": 0.00014492923702475182,
|
| 456 |
+
"loss": 0.0475,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.738095238095238,
|
| 461 |
+
"grad_norm": 0.22984769940376282,
|
| 462 |
+
"learning_rate": 0.00014314524604309748,
|
| 463 |
+
"loss": 0.0537,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.857142857142857,
|
| 468 |
+
"grad_norm": 0.22611601650714874,
|
| 469 |
+
"learning_rate": 0.00014134421957005775,
|
| 470 |
+
"loss": 0.048,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.976190476190476,
|
| 475 |
+
"grad_norm": 0.37012603878974915,
|
| 476 |
+
"learning_rate": 0.00013952686872386195,
|
| 477 |
+
"loss": 0.0489,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.095238095238095,
|
| 482 |
+
"grad_norm": 0.27276352047920227,
|
| 483 |
+
"learning_rate": 0.00013769391106826327,
|
| 484 |
+
"loss": 0.0479,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 8.214285714285714,
|
| 489 |
+
"grad_norm": 0.23436495661735535,
|
| 490 |
+
"learning_rate": 0.00013584607032921566,
|
| 491 |
+
"loss": 0.0433,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 8.333333333333334,
|
| 496 |
+
"grad_norm": 0.2781357169151306,
|
| 497 |
+
"learning_rate": 0.0001339840761091175,
|
| 498 |
+
"loss": 0.043,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 8.452380952380953,
|
| 503 |
+
"grad_norm": 0.2650757133960724,
|
| 504 |
+
"learning_rate": 0.00013210866359873505,
|
| 505 |
+
"loss": 0.0447,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 8.571428571428571,
|
| 510 |
+
"grad_norm": 0.3430781364440918,
|
| 511 |
+
"learning_rate": 0.00013022057328691914,
|
| 512 |
+
"loss": 0.0452,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.69047619047619,
|
| 517 |
+
"grad_norm": 0.21453280746936798,
|
| 518 |
+
"learning_rate": 0.00012832055066823038,
|
| 519 |
+
"loss": 0.0438,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 8.80952380952381,
|
| 524 |
+
"grad_norm": 0.29838255047798157,
|
| 525 |
+
"learning_rate": 0.00012640934594858774,
|
| 526 |
+
"loss": 0.0479,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 8.928571428571429,
|
| 531 |
+
"grad_norm": 0.19045886397361755,
|
| 532 |
+
"learning_rate": 0.00012448771374905655,
|
| 533 |
+
"loss": 0.0459,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 9.047619047619047,
|
| 538 |
+
"grad_norm": 0.4510950744152069,
|
| 539 |
+
"learning_rate": 0.00012255641280789386,
|
| 540 |
+
"loss": 0.0512,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 9.166666666666666,
|
| 545 |
+
"grad_norm": 0.24877306818962097,
|
| 546 |
+
"learning_rate": 0.0001206162056809676,
|
| 547 |
+
"loss": 0.0472,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 9.285714285714286,
|
| 552 |
+
"grad_norm": 0.2406831532716751,
|
| 553 |
+
"learning_rate": 0.00011866785844066883,
|
| 554 |
+
"loss": 0.0476,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.404761904761905,
|
| 559 |
+
"grad_norm": 0.2927370071411133,
|
| 560 |
+
"learning_rate": 0.00011671214037343514,
|
| 561 |
+
"loss": 0.0412,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 9.523809523809524,
|
| 566 |
+
"grad_norm": 0.20210541784763336,
|
| 567 |
+
"learning_rate": 0.00011474982367600525,
|
| 568 |
+
"loss": 0.0424,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 9.642857142857142,
|
| 573 |
+
"grad_norm": 0.2872091829776764,
|
| 574 |
+
"learning_rate": 0.00011278168315052445,
|
| 575 |
+
"loss": 0.0399,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 9.761904761904763,
|
| 580 |
+
"grad_norm": 0.290270060300827,
|
| 581 |
+
"learning_rate": 0.00011080849589862142,
|
| 582 |
+
"loss": 0.0357,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 9.880952380952381,
|
| 587 |
+
"grad_norm": 0.20899908244609833,
|
| 588 |
+
"learning_rate": 0.0001088310410145768,
|
| 589 |
+
"loss": 0.0401,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 10.0,
|
| 594 |
+
"grad_norm": 0.248155415058136,
|
| 595 |
+
"learning_rate": 0.00010685009927770542,
|
| 596 |
+
"loss": 0.035,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 10.119047619047619,
|
| 601 |
+
"grad_norm": 0.18403619527816772,
|
| 602 |
+
"learning_rate": 0.00010486645284407281,
|
| 603 |
+
"loss": 0.039,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 10.238095238095237,
|
| 608 |
+
"grad_norm": 0.16922874748706818,
|
| 609 |
+
"learning_rate": 0.00010288088493766845,
|
| 610 |
+
"loss": 0.0361,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 10.357142857142858,
|
| 615 |
+
"grad_norm": 0.19182036817073822,
|
| 616 |
+
"learning_rate": 0.00010089417954115714,
|
| 617 |
+
"loss": 0.0404,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 10.476190476190476,
|
| 622 |
+
"grad_norm": 0.2680899202823639,
|
| 623 |
+
"learning_rate": 9.890712108633076e-05,
|
| 624 |
+
"loss": 0.0366,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 10.595238095238095,
|
| 629 |
+
"grad_norm": 0.21088555455207825,
|
| 630 |
+
"learning_rate": 9.692049414438299e-05,
|
| 631 |
+
"loss": 0.0397,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 10.714285714285714,
|
| 636 |
+
"grad_norm": 0.27201998233795166,
|
| 637 |
+
"learning_rate": 9.493508311612874e-05,
|
| 638 |
+
"loss": 0.0411,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 10.833333333333334,
|
| 643 |
+
"grad_norm": 0.15937431156635284,
|
| 644 |
+
"learning_rate": 9.295167192229093e-05,
|
| 645 |
+
"loss": 0.0394,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 10.952380952380953,
|
| 650 |
+
"grad_norm": 0.20022912323474884,
|
| 651 |
+
"learning_rate": 9.097104369397682e-05,
|
| 652 |
+
"loss": 0.0361,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 11.071428571428571,
|
| 657 |
+
"grad_norm": 0.2788039743900299,
|
| 658 |
+
"learning_rate": 8.899398046346608e-05,
|
| 659 |
+
"loss": 0.0397,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 11.19047619047619,
|
| 664 |
+
"grad_norm": 0.2813689708709717,
|
| 665 |
+
"learning_rate": 8.702126285543286e-05,
|
| 666 |
+
"loss": 0.0374,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 11.30952380952381,
|
| 671 |
+
"grad_norm": 0.1632857620716095,
|
| 672 |
+
"learning_rate": 8.505366977872336e-05,
|
| 673 |
+
"loss": 0.0372,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 11.428571428571429,
|
| 678 |
+
"grad_norm": 0.16908283531665802,
|
| 679 |
+
"learning_rate": 8.309197811881127e-05,
|
| 680 |
+
"loss": 0.0361,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 11.547619047619047,
|
| 685 |
+
"grad_norm": 0.2518416941165924,
|
| 686 |
+
"learning_rate": 8.113696243105176e-05,
|
| 687 |
+
"loss": 0.0374,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 11.666666666666666,
|
| 692 |
+
"grad_norm": 0.21874170005321503,
|
| 693 |
+
"learning_rate": 7.918939463485568e-05,
|
| 694 |
+
"loss": 0.0363,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 11.785714285714286,
|
| 699 |
+
"grad_norm": 0.2551394999027252,
|
| 700 |
+
"learning_rate": 7.72500437089046e-05,
|
| 701 |
+
"loss": 0.0348,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 11.904761904761905,
|
| 706 |
+
"grad_norm": 0.2511221170425415,
|
| 707 |
+
"learning_rate": 7.531967538752656e-05,
|
| 708 |
+
"loss": 0.0412,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 12.023809523809524,
|
| 713 |
+
"grad_norm": 0.24874921143054962,
|
| 714 |
+
"learning_rate": 7.33990518583535e-05,
|
| 715 |
+
"loss": 0.0369,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 12.142857142857142,
|
| 720 |
+
"grad_norm": 0.25549542903900146,
|
| 721 |
+
"learning_rate": 7.148893146137852e-05,
|
| 722 |
+
"loss": 0.0334,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.261904761904763,
|
| 727 |
+
"grad_norm": 0.25841251015663147,
|
| 728 |
+
"learning_rate": 6.95900683895325e-05,
|
| 729 |
+
"loss": 0.0378,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 12.380952380952381,
|
| 734 |
+
"grad_norm": 0.2499120533466339,
|
| 735 |
+
"learning_rate": 6.770321239089826e-05,
|
| 736 |
+
"loss": 0.0314,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 12.5,
|
| 741 |
+
"grad_norm": 0.19628271460533142,
|
| 742 |
+
"learning_rate": 6.582910847267957e-05,
|
| 743 |
+
"loss": 0.0287,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.619047619047619,
|
| 748 |
+
"grad_norm": 0.18714749813079834,
|
| 749 |
+
"learning_rate": 6.396849660704205e-05,
|
| 750 |
+
"loss": 0.0302,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.738095238095237,
|
| 755 |
+
"grad_norm": 0.19048580527305603,
|
| 756 |
+
"learning_rate": 6.21221114389424e-05,
|
| 757 |
+
"loss": 0.0276,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.857142857142858,
|
| 762 |
+
"grad_norm": 0.21586266160011292,
|
| 763 |
+
"learning_rate": 6.0290681996060605e-05,
|
| 764 |
+
"loss": 0.0341,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.976190476190476,
|
| 769 |
+
"grad_norm": 0.22894449532032013,
|
| 770 |
+
"learning_rate": 5.847493140095029e-05,
|
| 771 |
+
"loss": 0.0315,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 13.095238095238095,
|
| 776 |
+
"grad_norm": 0.2722737491130829,
|
| 777 |
+
"learning_rate": 5.6675576585520786e-05,
|
| 778 |
+
"loss": 0.0288,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 13.214285714285714,
|
| 783 |
+
"grad_norm": 0.14149416983127594,
|
| 784 |
+
"learning_rate": 5.4893328007963094e-05,
|
| 785 |
+
"loss": 0.0302,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 13.333333333333334,
|
| 790 |
+
"grad_norm": 0.2547774314880371,
|
| 791 |
+
"learning_rate": 5.312888937223244e-05,
|
| 792 |
+
"loss": 0.0282,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 13.452380952380953,
|
| 797 |
+
"grad_norm": 0.1506737470626831,
|
| 798 |
+
"learning_rate": 5.1382957350197405e-05,
|
| 799 |
+
"loss": 0.0285,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.571428571428571,
|
| 804 |
+
"grad_norm": 0.19141684472560883,
|
| 805 |
+
"learning_rate": 4.965622130656551e-05,
|
| 806 |
+
"loss": 0.0323,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 13.69047619047619,
|
| 811 |
+
"grad_norm": 0.18162524700164795,
|
| 812 |
+
"learning_rate": 4.794936302669417e-05,
|
| 813 |
+
"loss": 0.0306,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 13.80952380952381,
|
| 818 |
+
"grad_norm": 0.20886124670505524,
|
| 819 |
+
"learning_rate": 4.6263056447394347e-05,
|
| 820 |
+
"loss": 0.0272,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 13.928571428571429,
|
| 825 |
+
"grad_norm": 0.14155210554599762,
|
| 826 |
+
"learning_rate": 4.459796739083274e-05,
|
| 827 |
+
"loss": 0.0299,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 14.047619047619047,
|
| 832 |
+
"grad_norm": 0.13192126154899597,
|
| 833 |
+
"learning_rate": 4.2954753301638315e-05,
|
| 834 |
+
"loss": 0.0337,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 14.166666666666666,
|
| 839 |
+
"grad_norm": 0.17799627780914307,
|
| 840 |
+
"learning_rate": 4.133406298731669e-05,
|
| 841 |
+
"loss": 0.0278,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 14.285714285714286,
|
| 846 |
+
"grad_norm": 0.1519002765417099,
|
| 847 |
+
"learning_rate": 3.973653636207437e-05,
|
| 848 |
+
"loss": 0.026,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 14.404761904761905,
|
| 853 |
+
"grad_norm": 0.24323873221874237,
|
| 854 |
+
"learning_rate": 3.8162804194154864e-05,
|
| 855 |
+
"loss": 0.0329,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 14.523809523809524,
|
| 860 |
+
"grad_norm": 0.25485578179359436,
|
| 861 |
+
"learning_rate": 3.661348785678574e-05,
|
| 862 |
+
"loss": 0.0282,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 14.642857142857142,
|
| 867 |
+
"grad_norm": 0.14652098715305328,
|
| 868 |
+
"learning_rate": 3.508919908283543e-05,
|
| 869 |
+
"loss": 0.0263,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 14.761904761904763,
|
| 874 |
+
"grad_norm": 0.26984453201293945,
|
| 875 |
+
"learning_rate": 3.3590539723276083e-05,
|
| 876 |
+
"loss": 0.029,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 14.880952380952381,
|
| 881 |
+
"grad_norm": 0.13968226313591003,
|
| 882 |
+
"learning_rate": 3.211810150954867e-05,
|
| 883 |
+
"loss": 0.0267,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 15.0,
|
| 888 |
+
"grad_norm": 0.29029059410095215,
|
| 889 |
+
"learning_rate": 3.067246581992321e-05,
|
| 890 |
+
"loss": 0.0255,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 15.119047619047619,
|
| 895 |
+
"grad_norm": 0.16652598977088928,
|
| 896 |
+
"learning_rate": 2.925420344994719e-05,
|
| 897 |
+
"loss": 0.0268,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 15.238095238095237,
|
| 902 |
+
"grad_norm": 0.27046117186546326,
|
| 903 |
+
"learning_rate": 2.786387438707231e-05,
|
| 904 |
+
"loss": 0.0245,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 15.357142857142858,
|
| 909 |
+
"grad_norm": 0.22596995532512665,
|
| 910 |
+
"learning_rate": 2.6502027589548862e-05,
|
| 911 |
+
"loss": 0.0288,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 15.476190476190476,
|
| 916 |
+
"grad_norm": 0.16162142157554626,
|
| 917 |
+
"learning_rate": 2.516920076967455e-05,
|
| 918 |
+
"loss": 0.0263,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 15.595238095238095,
|
| 923 |
+
"grad_norm": 0.1669321060180664,
|
| 924 |
+
"learning_rate": 2.3865920181484123e-05,
|
| 925 |
+
"loss": 0.0284,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 15.714285714285714,
|
| 930 |
+
"grad_norm": 0.20296935737133026,
|
| 931 |
+
"learning_rate": 2.2592700412962777e-05,
|
| 932 |
+
"loss": 0.026,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 15.833333333333334,
|
| 937 |
+
"grad_norm": 0.17937733232975006,
|
| 938 |
+
"learning_rate": 2.1350044182866025e-05,
|
| 939 |
+
"loss": 0.0217,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 15.952380952380953,
|
| 944 |
+
"grad_norm": 0.14423789083957672,
|
| 945 |
+
"learning_rate": 2.0138442142226e-05,
|
| 946 |
+
"loss": 0.0289,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 16.071428571428573,
|
| 951 |
+
"grad_norm": 0.21349196135997772,
|
| 952 |
+
"learning_rate": 1.895837268062256e-05,
|
| 953 |
+
"loss": 0.029,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 16.19047619047619,
|
| 958 |
+
"grad_norm": 0.17763833701610565,
|
| 959 |
+
"learning_rate": 1.7810301737295588e-05,
|
| 960 |
+
"loss": 0.0234,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 16.30952380952381,
|
| 965 |
+
"grad_norm": 0.1566387116909027,
|
| 966 |
+
"learning_rate": 1.6694682617173452e-05,
|
| 967 |
+
"loss": 0.025,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 16.428571428571427,
|
| 972 |
+
"grad_norm": 0.19837665557861328,
|
| 973 |
+
"learning_rate": 1.5611955811889644e-05,
|
| 974 |
+
"loss": 0.0254,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 16.547619047619047,
|
| 979 |
+
"grad_norm": 0.17018267512321472,
|
| 980 |
+
"learning_rate": 1.456254882585909e-05,
|
| 981 |
+
"loss": 0.0212,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 16.666666666666668,
|
| 986 |
+
"grad_norm": 0.11024856567382812,
|
| 987 |
+
"learning_rate": 1.3546876007481845e-05,
|
| 988 |
+
"loss": 0.0208,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 16.785714285714285,
|
| 993 |
+
"grad_norm": 0.16558444499969482,
|
| 994 |
+
"learning_rate": 1.2565338385541792e-05,
|
| 995 |
+
"loss": 0.0234,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 16.904761904761905,
|
| 1000 |
+
"grad_norm": 0.14750836789608002,
|
| 1001 |
+
"learning_rate": 1.161832351086396e-05,
|
| 1002 |
+
"loss": 0.0234,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 17.023809523809526,
|
| 1007 |
+
"grad_norm": 0.14175614714622498,
|
| 1008 |
+
"learning_rate": 1.0706205303294026e-05,
|
| 1009 |
+
"loss": 0.0233,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 17.142857142857142,
|
| 1014 |
+
"grad_norm": 0.128941610455513,
|
| 1015 |
+
"learning_rate": 9.82934390405934e-06,
|
| 1016 |
+
"loss": 0.0228,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 17.261904761904763,
|
| 1021 |
+
"grad_norm": 0.17659620940685272,
|
| 1022 |
+
"learning_rate": 8.988085533570833e-06,
|
| 1023 |
+
"loss": 0.0244,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 17.38095238095238,
|
| 1028 |
+
"grad_norm": 0.1548050194978714,
|
| 1029 |
+
"learning_rate": 8.182762354720985e-06,
|
| 1030 |
+
"loss": 0.0203,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 17.5,
|
| 1035 |
+
"grad_norm": 0.16272428631782532,
|
| 1036 |
+
"learning_rate": 7.413692341732581e-06,
|
| 1037 |
+
"loss": 0.0285,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 17.61904761904762,
|
| 1042 |
+
"grad_norm": 0.22217531502246857,
|
| 1043 |
+
"learning_rate": 6.681179154609462e-06,
|
| 1044 |
+
"loss": 0.0195,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 17.738095238095237,
|
| 1049 |
+
"grad_norm": 0.16655217111110687,
|
| 1050 |
+
"learning_rate": 5.985512019239392e-06,
|
| 1051 |
+
"loss": 0.0236,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 17.857142857142858,
|
| 1056 |
+
"grad_norm": 0.19496409595012665,
|
| 1057 |
+
"learning_rate": 5.326965613195867e-06,
|
| 1058 |
+
"loss": 0.0246,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 17.976190476190474,
|
| 1063 |
+
"grad_norm": 0.18502101302146912,
|
| 1064 |
+
"learning_rate": 4.705799957284351e-06,
|
| 1065 |
+
"loss": 0.0209,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 18.095238095238095,
|
| 1070 |
+
"grad_norm": 0.20573672652244568,
|
| 1071 |
+
"learning_rate": 4.122260312875437e-06,
|
| 1072 |
+
"loss": 0.023,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 18.214285714285715,
|
| 1077 |
+
"grad_norm": 0.18070857226848602,
|
| 1078 |
+
"learning_rate": 3.576577085065824e-06,
|
| 1079 |
+
"loss": 0.0246,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 18.333333333333332,
|
| 1084 |
+
"grad_norm": 0.15732447803020477,
|
| 1085 |
+
"learning_rate": 3.0689657317049204e-06,
|
| 1086 |
+
"loss": 0.0232,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 18.452380952380953,
|
| 1091 |
+
"grad_norm": 0.18765419721603394,
|
| 1092 |
+
"learning_rate": 2.5996266783235078e-06,
|
| 1093 |
+
"loss": 0.0238,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 18.571428571428573,
|
| 1098 |
+
"grad_norm": 0.12155826389789581,
|
| 1099 |
+
"learning_rate": 2.1687452389974826e-06,
|
| 1100 |
+
"loss": 0.0223,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 18.69047619047619,
|
| 1105 |
+
"grad_norm": 0.12302082777023315,
|
| 1106 |
+
"learning_rate": 1.7764915431784378e-06,
|
| 1107 |
+
"loss": 0.0267,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 18.80952380952381,
|
| 1112 |
+
"grad_norm": 0.1826854795217514,
|
| 1113 |
+
"learning_rate": 1.4230204685196203e-06,
|
| 1114 |
+
"loss": 0.0238,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 18.928571428571427,
|
| 1119 |
+
"grad_norm": 0.1735326200723648,
|
| 1120 |
+
"learning_rate": 1.1084715797239798e-06,
|
| 1121 |
+
"loss": 0.0222,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 19.047619047619047,
|
| 1126 |
+
"grad_norm": 0.17233559489250183,
|
| 1127 |
+
"learning_rate": 8.329690734383277e-07,
|
| 1128 |
+
"loss": 0.0245,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 19.166666666666668,
|
| 1133 |
+
"grad_norm": 0.14161868393421173,
|
| 1134 |
+
"learning_rate": 5.966217292155296e-07,
|
| 1135 |
+
"loss": 0.0212,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 19.285714285714285,
|
| 1140 |
+
"grad_norm": 0.16612544655799866,
|
| 1141 |
+
"learning_rate": 3.99522866563895e-07,
|
| 1142 |
+
"loss": 0.0245,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 19.404761904761905,
|
| 1147 |
+
"grad_norm": 0.08409969508647919,
|
| 1148 |
+
"learning_rate": 2.417503081008632e-07,
|
| 1149 |
+
"loss": 0.0213,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 19.523809523809526,
|
| 1154 |
+
"grad_norm": 0.17151199281215668,
|
| 1155 |
+
"learning_rate": 1.2336634882544884e-07,
|
| 1156 |
+
"loss": 0.0176,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 19.642857142857142,
|
| 1161 |
+
"grad_norm": 0.13334044814109802,
|
| 1162 |
+
"learning_rate": 4.4417731521717574e-08,
|
| 1163 |
+
"loss": 0.0263,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 19.761904761904763,
|
| 1168 |
+
"grad_norm": 0.07808782160282135,
|
| 1169 |
+
"learning_rate": 4.935628302760176e-09,
|
| 1170 |
+
"loss": 0.0211,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 19.821428571428573,
|
| 1175 |
+
"step": 1665,
|
| 1176 |
+
"total_flos": 2.314379607040872e+17,
|
| 1177 |
+
"train_loss": 0.06144854805401496,
|
| 1178 |
+
"train_runtime": 1922.2885,
|
| 1179 |
+
"train_samples_per_second": 55.434,
|
| 1180 |
+
"train_steps_per_second": 0.866
|
| 1181 |
+
}
|
| 1182 |
+
],
|
| 1183 |
+
"logging_steps": 10,
|
| 1184 |
+
"max_steps": 1665,
|
| 1185 |
+
"num_input_tokens_seen": 0,
|
| 1186 |
+
"num_train_epochs": 20,
|
| 1187 |
+
"save_steps": 10000,
|
| 1188 |
+
"stateful_callbacks": {
|
| 1189 |
+
"TrainerControl": {
|
| 1190 |
+
"args": {
|
| 1191 |
+
"should_epoch_stop": false,
|
| 1192 |
+
"should_evaluate": false,
|
| 1193 |
+
"should_log": false,
|
| 1194 |
+
"should_save": true,
|
| 1195 |
+
"should_training_stop": true
|
| 1196 |
+
},
|
| 1197 |
+
"attributes": {}
|
| 1198 |
+
}
|
| 1199 |
+
},
|
| 1200 |
+
"total_flos": 2.314379607040872e+17,
|
| 1201 |
+
"train_batch_size": 64,
|
| 1202 |
+
"trial_name": null,
|
| 1203 |
+
"trial_params": null
|
| 1204 |
+
}
|