Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1309 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1309 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.89010989010989,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1810,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.10989010989010989,
|
| 13 |
+
"grad_norm": 6.35939884185791,
|
| 14 |
+
"learning_rate": 2.1978021978021977e-05,
|
| 15 |
+
"loss": 1.3933,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.21978021978021978,
|
| 20 |
+
"grad_norm": 3.3019938468933105,
|
| 21 |
+
"learning_rate": 4.3956043956043955e-05,
|
| 22 |
+
"loss": 0.6013,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.32967032967032966,
|
| 27 |
+
"grad_norm": 1.7010605335235596,
|
| 28 |
+
"learning_rate": 6.593406593406594e-05,
|
| 29 |
+
"loss": 0.3661,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.43956043956043955,
|
| 34 |
+
"grad_norm": 1.9770668745040894,
|
| 35 |
+
"learning_rate": 8.791208791208791e-05,
|
| 36 |
+
"loss": 0.2746,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5494505494505495,
|
| 41 |
+
"grad_norm": 1.2615512609481812,
|
| 42 |
+
"learning_rate": 0.0001098901098901099,
|
| 43 |
+
"loss": 0.2108,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6593406593406593,
|
| 48 |
+
"grad_norm": 0.9937906265258789,
|
| 49 |
+
"learning_rate": 0.00013186813186813188,
|
| 50 |
+
"loss": 0.1888,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7692307692307693,
|
| 55 |
+
"grad_norm": 0.7049073576927185,
|
| 56 |
+
"learning_rate": 0.00015384615384615385,
|
| 57 |
+
"loss": 0.166,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.8791208791208791,
|
| 62 |
+
"grad_norm": 0.6901200413703918,
|
| 63 |
+
"learning_rate": 0.00017582417582417582,
|
| 64 |
+
"loss": 0.1526,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.989010989010989,
|
| 69 |
+
"grad_norm": 1.3578896522521973,
|
| 70 |
+
"learning_rate": 0.0001978021978021978,
|
| 71 |
+
"loss": 0.1373,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.098901098901099,
|
| 76 |
+
"grad_norm": 0.9918624758720398,
|
| 77 |
+
"learning_rate": 0.00019998647325745995,
|
| 78 |
+
"loss": 0.1303,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.2087912087912087,
|
| 83 |
+
"grad_norm": 0.9448220133781433,
|
| 84 |
+
"learning_rate": 0.00019993971884561357,
|
| 85 |
+
"loss": 0.1236,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.3186813186813187,
|
| 90 |
+
"grad_norm": 0.5491310358047485,
|
| 91 |
+
"learning_rate": 0.00019985958537951462,
|
| 92 |
+
"loss": 0.1164,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.4285714285714286,
|
| 97 |
+
"grad_norm": 0.6634742617607117,
|
| 98 |
+
"learning_rate": 0.00019974609962308986,
|
| 99 |
+
"loss": 0.1024,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.5384615384615383,
|
| 104 |
+
"grad_norm": 0.652184784412384,
|
| 105 |
+
"learning_rate": 0.00019959929947966,
|
| 106 |
+
"loss": 0.108,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.6483516483516483,
|
| 111 |
+
"grad_norm": 0.42199188470840454,
|
| 112 |
+
"learning_rate": 0.00019941923397928047,
|
| 113 |
+
"loss": 0.0963,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.7582417582417582,
|
| 118 |
+
"grad_norm": 0.7065568566322327,
|
| 119 |
+
"learning_rate": 0.0001992059632623657,
|
| 120 |
+
"loss": 0.099,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.8681318681318682,
|
| 125 |
+
"grad_norm": 0.6244102120399475,
|
| 126 |
+
"learning_rate": 0.0001989595585596026,
|
| 127 |
+
"loss": 0.0897,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.978021978021978,
|
| 132 |
+
"grad_norm": 0.6415707468986511,
|
| 133 |
+
"learning_rate": 0.00019868010216816034,
|
| 134 |
+
"loss": 0.0781,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.087912087912088,
|
| 139 |
+
"grad_norm": 0.3453947603702545,
|
| 140 |
+
"learning_rate": 0.00019836768742420352,
|
| 141 |
+
"loss": 0.079,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.197802197802198,
|
| 146 |
+
"grad_norm": 0.37823784351348877,
|
| 147 |
+
"learning_rate": 0.00019802241867171878,
|
| 148 |
+
"loss": 0.075,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.3076923076923075,
|
| 153 |
+
"grad_norm": 0.8650999069213867,
|
| 154 |
+
"learning_rate": 0.00019764441122766476,
|
| 155 |
+
"loss": 0.0796,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.4175824175824174,
|
| 160 |
+
"grad_norm": 0.4968554973602295,
|
| 161 |
+
"learning_rate": 0.000197233791343457,
|
| 162 |
+
"loss": 0.0754,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.5274725274725274,
|
| 167 |
+
"grad_norm": 0.5782051682472229,
|
| 168 |
+
"learning_rate": 0.00019679069616280133,
|
| 169 |
+
"loss": 0.076,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.6373626373626373,
|
| 174 |
+
"grad_norm": 0.4181021749973297,
|
| 175 |
+
"learning_rate": 0.00019631527367588864,
|
| 176 |
+
"loss": 0.0769,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.7472527472527473,
|
| 181 |
+
"grad_norm": 0.41538622975349426,
|
| 182 |
+
"learning_rate": 0.00019580768266996756,
|
| 183 |
+
"loss": 0.0653,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.857142857142857,
|
| 188 |
+
"grad_norm": 0.3886262774467468,
|
| 189 |
+
"learning_rate": 0.00019526809267631073,
|
| 190 |
+
"loss": 0.0723,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.967032967032967,
|
| 195 |
+
"grad_norm": 0.466183602809906,
|
| 196 |
+
"learning_rate": 0.0001946966839135928,
|
| 197 |
+
"loss": 0.0702,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.076923076923077,
|
| 202 |
+
"grad_norm": 0.34900644421577454,
|
| 203 |
+
"learning_rate": 0.00019409364722769882,
|
| 204 |
+
"loss": 0.0638,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.186813186813187,
|
| 209 |
+
"grad_norm": 0.5098071694374084,
|
| 210 |
+
"learning_rate": 0.00019345918402798314,
|
| 211 |
+
"loss": 0.0645,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.2967032967032965,
|
| 216 |
+
"grad_norm": 0.45593976974487305,
|
| 217 |
+
"learning_rate": 0.00019279350622000054,
|
| 218 |
+
"loss": 0.0691,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.4065934065934065,
|
| 223 |
+
"grad_norm": 0.3932492733001709,
|
| 224 |
+
"learning_rate": 0.0001920968361347314,
|
| 225 |
+
"loss": 0.0706,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.5164835164835164,
|
| 230 |
+
"grad_norm": 0.3770773708820343,
|
| 231 |
+
"learning_rate": 0.00019136940645432494,
|
| 232 |
+
"loss": 0.0646,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.6263736263736264,
|
| 237 |
+
"grad_norm": 0.432841420173645,
|
| 238 |
+
"learning_rate": 0.00019061146013438527,
|
| 239 |
+
"loss": 0.0557,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.7362637362637363,
|
| 244 |
+
"grad_norm": 0.5559947490692139,
|
| 245 |
+
"learning_rate": 0.00018982325032282615,
|
| 246 |
+
"loss": 0.0729,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.8461538461538463,
|
| 251 |
+
"grad_norm": 0.5822933912277222,
|
| 252 |
+
"learning_rate": 0.00018900504027532185,
|
| 253 |
+
"loss": 0.0632,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.956043956043956,
|
| 258 |
+
"grad_norm": 0.5372587442398071,
|
| 259 |
+
"learning_rate": 0.0001881571032673816,
|
| 260 |
+
"loss": 0.0666,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.065934065934066,
|
| 265 |
+
"grad_norm": 0.4164665937423706,
|
| 266 |
+
"learning_rate": 0.000187279722503078,
|
| 267 |
+
"loss": 0.0566,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.175824175824176,
|
| 272 |
+
"grad_norm": 0.570026695728302,
|
| 273 |
+
"learning_rate": 0.00018637319102045912,
|
| 274 |
+
"loss": 0.0621,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.285714285714286,
|
| 279 |
+
"grad_norm": 0.3998757302761078,
|
| 280 |
+
"learning_rate": 0.00018543781159367607,
|
| 281 |
+
"loss": 0.0587,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.395604395604396,
|
| 286 |
+
"grad_norm": 0.3779265880584717,
|
| 287 |
+
"learning_rate": 0.00018447389663185906,
|
| 288 |
+
"loss": 0.0568,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.5054945054945055,
|
| 293 |
+
"grad_norm": 0.4073057174682617,
|
| 294 |
+
"learning_rate": 0.0001834817680747751,
|
| 295 |
+
"loss": 0.0615,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.615384615384615,
|
| 300 |
+
"grad_norm": 0.3638319969177246,
|
| 301 |
+
"learning_rate": 0.00018246175728530296,
|
| 302 |
+
"loss": 0.0548,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.725274725274725,
|
| 307 |
+
"grad_norm": 0.2937782108783722,
|
| 308 |
+
"learning_rate": 0.00018141420493876035,
|
| 309 |
+
"loss": 0.0601,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 4.835164835164835,
|
| 314 |
+
"grad_norm": 0.4016660749912262,
|
| 315 |
+
"learning_rate": 0.00018033946090912148,
|
| 316 |
+
"loss": 0.0563,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 4.945054945054945,
|
| 321 |
+
"grad_norm": 0.39952585101127625,
|
| 322 |
+
"learning_rate": 0.00017923788415216175,
|
| 323 |
+
"loss": 0.0554,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.054945054945055,
|
| 328 |
+
"grad_norm": 0.3408401906490326,
|
| 329 |
+
"learning_rate": 0.00017810984258556957,
|
| 330 |
+
"loss": 0.0515,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.164835164835165,
|
| 335 |
+
"grad_norm": 0.3611335754394531,
|
| 336 |
+
"learning_rate": 0.00017695571296606465,
|
| 337 |
+
"loss": 0.0511,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.274725274725275,
|
| 342 |
+
"grad_norm": 0.24698808789253235,
|
| 343 |
+
"learning_rate": 0.00017577588076356465,
|
| 344 |
+
"loss": 0.0531,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.384615384615385,
|
| 349 |
+
"grad_norm": 0.40450191497802734,
|
| 350 |
+
"learning_rate": 0.000174570740032441,
|
| 351 |
+
"loss": 0.048,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.4945054945054945,
|
| 356 |
+
"grad_norm": 0.44147250056266785,
|
| 357 |
+
"learning_rate": 0.00017334069327990816,
|
| 358 |
+
"loss": 0.0545,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.604395604395604,
|
| 363 |
+
"grad_norm": 0.45513996481895447,
|
| 364 |
+
"learning_rate": 0.00017208615133158946,
|
| 365 |
+
"loss": 0.0505,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.714285714285714,
|
| 370 |
+
"grad_norm": 0.3807731568813324,
|
| 371 |
+
"learning_rate": 0.00017080753319430452,
|
| 372 |
+
"loss": 0.0509,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.824175824175824,
|
| 377 |
+
"grad_norm": 0.288381427526474,
|
| 378 |
+
"learning_rate": 0.00016950526591612462,
|
| 379 |
+
"loss": 0.0493,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.934065934065934,
|
| 384 |
+
"grad_norm": 0.234198197722435,
|
| 385 |
+
"learning_rate": 0.00016817978444374194,
|
| 386 |
+
"loss": 0.0552,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.043956043956044,
|
| 391 |
+
"grad_norm": 0.33030226826667786,
|
| 392 |
+
"learning_rate": 0.00016683153147720097,
|
| 393 |
+
"loss": 0.0453,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.153846153846154,
|
| 398 |
+
"grad_norm": 0.30306366086006165,
|
| 399 |
+
"learning_rate": 0.00016546095732204022,
|
| 400 |
+
"loss": 0.0521,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 6.263736263736264,
|
| 405 |
+
"grad_norm": 0.32483500242233276,
|
| 406 |
+
"learning_rate": 0.00016406851973889392,
|
| 407 |
+
"loss": 0.0473,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 6.373626373626374,
|
| 412 |
+
"grad_norm": 0.3211219310760498,
|
| 413 |
+
"learning_rate": 0.00016265468379060365,
|
| 414 |
+
"loss": 0.0527,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 6.483516483516484,
|
| 419 |
+
"grad_norm": 0.2360507994890213,
|
| 420 |
+
"learning_rate": 0.00016121992168689108,
|
| 421 |
+
"loss": 0.0449,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 6.593406593406593,
|
| 426 |
+
"grad_norm": 0.3557351231575012,
|
| 427 |
+
"learning_rate": 0.00015976471262664382,
|
| 428 |
+
"loss": 0.0471,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 6.7032967032967035,
|
| 433 |
+
"grad_norm": 0.3993636667728424,
|
| 434 |
+
"learning_rate": 0.00015828954263786687,
|
| 435 |
+
"loss": 0.0461,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 6.813186813186813,
|
| 440 |
+
"grad_norm": 0.4391331076622009,
|
| 441 |
+
"learning_rate": 0.00015679490441535324,
|
| 442 |
+
"loss": 0.0431,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 6.923076923076923,
|
| 447 |
+
"grad_norm": 0.311146080493927,
|
| 448 |
+
"learning_rate": 0.0001552812971561278,
|
| 449 |
+
"loss": 0.0497,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.032967032967033,
|
| 454 |
+
"grad_norm": 0.29637405276298523,
|
| 455 |
+
"learning_rate": 0.00015374922639271963,
|
| 456 |
+
"loss": 0.0466,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.142857142857143,
|
| 461 |
+
"grad_norm": 0.24265693128108978,
|
| 462 |
+
"learning_rate": 0.0001521992038243182,
|
| 463 |
+
"loss": 0.048,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.252747252747253,
|
| 468 |
+
"grad_norm": 0.3331180214881897,
|
| 469 |
+
"learning_rate": 0.0001506317471458701,
|
| 470 |
+
"loss": 0.0444,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.362637362637362,
|
| 475 |
+
"grad_norm": 0.29099053144454956,
|
| 476 |
+
"learning_rate": 0.00014904737987517293,
|
| 477 |
+
"loss": 0.0446,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.472527472527473,
|
| 482 |
+
"grad_norm": 0.26169735193252563,
|
| 483 |
+
"learning_rate": 0.00014744663117802475,
|
| 484 |
+
"loss": 0.0442,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.582417582417582,
|
| 489 |
+
"grad_norm": 0.27879634499549866,
|
| 490 |
+
"learning_rate": 0.00014583003569148712,
|
| 491 |
+
"loss": 0.0424,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.6923076923076925,
|
| 496 |
+
"grad_norm": 0.3208893835544586,
|
| 497 |
+
"learning_rate": 0.00014419813334532036,
|
| 498 |
+
"loss": 0.0421,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.802197802197802,
|
| 503 |
+
"grad_norm": 0.2651744484901428,
|
| 504 |
+
"learning_rate": 0.0001425514691816516,
|
| 505 |
+
"loss": 0.046,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.912087912087912,
|
| 510 |
+
"grad_norm": 0.30851149559020996,
|
| 511 |
+
"learning_rate": 0.00014089059317293524,
|
| 512 |
+
"loss": 0.046,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.021978021978022,
|
| 517 |
+
"grad_norm": 0.32210397720336914,
|
| 518 |
+
"learning_rate": 0.00013921606003826627,
|
| 519 |
+
"loss": 0.0448,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 8.131868131868131,
|
| 524 |
+
"grad_norm": 0.49959078431129456,
|
| 525 |
+
"learning_rate": 0.00013752842905810896,
|
| 526 |
+
"loss": 0.0447,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 8.241758241758241,
|
| 531 |
+
"grad_norm": 0.5785067677497864,
|
| 532 |
+
"learning_rate": 0.00013582826388750153,
|
| 533 |
+
"loss": 0.0397,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.351648351648352,
|
| 538 |
+
"grad_norm": 0.28836700320243835,
|
| 539 |
+
"learning_rate": 0.00013411613236779995,
|
| 540 |
+
"loss": 0.0493,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.461538461538462,
|
| 545 |
+
"grad_norm": 0.26861655712127686,
|
| 546 |
+
"learning_rate": 0.00013239260633702364,
|
| 547 |
+
"loss": 0.0442,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 8.571428571428571,
|
| 552 |
+
"grad_norm": 0.3270936608314514,
|
| 553 |
+
"learning_rate": 0.00013065826143886616,
|
| 554 |
+
"loss": 0.0489,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 8.68131868131868,
|
| 559 |
+
"grad_norm": 0.24692483246326447,
|
| 560 |
+
"learning_rate": 0.00012891367693043478,
|
| 561 |
+
"loss": 0.0441,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 8.791208791208792,
|
| 566 |
+
"grad_norm": 0.32261839509010315,
|
| 567 |
+
"learning_rate": 0.00012715943548878355,
|
| 568 |
+
"loss": 0.042,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 8.901098901098901,
|
| 573 |
+
"grad_norm": 0.22817495465278625,
|
| 574 |
+
"learning_rate": 0.00012539612301630377,
|
| 575 |
+
"loss": 0.0415,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 9.010989010989011,
|
| 580 |
+
"grad_norm": 0.16383874416351318,
|
| 581 |
+
"learning_rate": 0.00012362432844503725,
|
| 582 |
+
"loss": 0.0366,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 9.12087912087912,
|
| 587 |
+
"grad_norm": 0.35975322127342224,
|
| 588 |
+
"learning_rate": 0.00012184464353997796,
|
| 589 |
+
"loss": 0.0424,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 9.23076923076923,
|
| 594 |
+
"grad_norm": 0.22110863029956818,
|
| 595 |
+
"learning_rate": 0.00012005766270142724,
|
| 596 |
+
"loss": 0.0334,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 9.340659340659341,
|
| 601 |
+
"grad_norm": 0.26277679204940796,
|
| 602 |
+
"learning_rate": 0.00011826398276646897,
|
| 603 |
+
"loss": 0.041,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 9.45054945054945,
|
| 608 |
+
"grad_norm": 0.29648274183273315,
|
| 609 |
+
"learning_rate": 0.00011646420280963081,
|
| 610 |
+
"loss": 0.0358,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 9.56043956043956,
|
| 615 |
+
"grad_norm": 0.27250102162361145,
|
| 616 |
+
"learning_rate": 0.00011465892394279814,
|
| 617 |
+
"loss": 0.0402,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 9.67032967032967,
|
| 622 |
+
"grad_norm": 0.24232593178749084,
|
| 623 |
+
"learning_rate": 0.00011284874911444763,
|
| 624 |
+
"loss": 0.0372,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 9.780219780219781,
|
| 629 |
+
"grad_norm": 0.25964832305908203,
|
| 630 |
+
"learning_rate": 0.00011103428290826736,
|
| 631 |
+
"loss": 0.0427,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 9.89010989010989,
|
| 636 |
+
"grad_norm": 0.31330835819244385,
|
| 637 |
+
"learning_rate": 0.00010921613134123068,
|
| 638 |
+
"loss": 0.0397,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 10.0,
|
| 643 |
+
"grad_norm": 0.3337639570236206,
|
| 644 |
+
"learning_rate": 0.00010739490166119155,
|
| 645 |
+
"loss": 0.0363,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 10.10989010989011,
|
| 650 |
+
"grad_norm": 0.1793065071105957,
|
| 651 |
+
"learning_rate": 0.00010557120214406875,
|
| 652 |
+
"loss": 0.0341,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 10.219780219780219,
|
| 657 |
+
"grad_norm": 0.32348737120628357,
|
| 658 |
+
"learning_rate": 0.00010374564189068641,
|
| 659 |
+
"loss": 0.0381,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 10.32967032967033,
|
| 664 |
+
"grad_norm": 0.3930850625038147,
|
| 665 |
+
"learning_rate": 0.00010191883062333965,
|
| 666 |
+
"loss": 0.0398,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 10.43956043956044,
|
| 671 |
+
"grad_norm": 0.2554318308830261,
|
| 672 |
+
"learning_rate": 0.00010009137848215195,
|
| 673 |
+
"loss": 0.034,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 10.54945054945055,
|
| 678 |
+
"grad_norm": 0.2729535698890686,
|
| 679 |
+
"learning_rate": 9.826389582129352e-05,
|
| 680 |
+
"loss": 0.0408,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 10.659340659340659,
|
| 685 |
+
"grad_norm": 0.2841566503047943,
|
| 686 |
+
"learning_rate": 9.64369930051278e-05,
|
| 687 |
+
"loss": 0.0373,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 10.76923076923077,
|
| 692 |
+
"grad_norm": 0.18831486999988556,
|
| 693 |
+
"learning_rate": 9.4611280204355e-05,
|
| 694 |
+
"loss": 0.0354,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 10.87912087912088,
|
| 699 |
+
"grad_norm": 0.25787901878356934,
|
| 700 |
+
"learning_rate": 9.278736719221964e-05,
|
| 701 |
+
"loss": 0.0357,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 10.989010989010989,
|
| 706 |
+
"grad_norm": 0.268319696187973,
|
| 707 |
+
"learning_rate": 9.096586314085162e-05,
|
| 708 |
+
"loss": 0.0357,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 11.098901098901099,
|
| 713 |
+
"grad_norm": 0.29258257150650024,
|
| 714 |
+
"learning_rate": 8.914737641780738e-05,
|
| 715 |
+
"loss": 0.0298,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 11.208791208791208,
|
| 720 |
+
"grad_norm": 0.26910606026649475,
|
| 721 |
+
"learning_rate": 8.733251438288032e-05,
|
| 722 |
+
"loss": 0.0385,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 11.31868131868132,
|
| 727 |
+
"grad_norm": 0.275666207075119,
|
| 728 |
+
"learning_rate": 8.552188318524737e-05,
|
| 729 |
+
"loss": 0.0371,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 11.428571428571429,
|
| 734 |
+
"grad_norm": 0.3032730221748352,
|
| 735 |
+
"learning_rate": 8.371608756102028e-05,
|
| 736 |
+
"loss": 0.0298,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 11.538461538461538,
|
| 741 |
+
"grad_norm": 0.19398105144500732,
|
| 742 |
+
"learning_rate": 8.191573063126842e-05,
|
| 743 |
+
"loss": 0.032,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 11.648351648351648,
|
| 748 |
+
"grad_norm": 0.23560473322868347,
|
| 749 |
+
"learning_rate": 8.012141370058151e-05,
|
| 750 |
+
"loss": 0.0294,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 11.758241758241759,
|
| 755 |
+
"grad_norm": 0.24311786890029907,
|
| 756 |
+
"learning_rate": 7.833373605623855e-05,
|
| 757 |
+
"loss": 0.034,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 11.868131868131869,
|
| 762 |
+
"grad_norm": 0.1695002168416977,
|
| 763 |
+
"learning_rate": 7.655329476805095e-05,
|
| 764 |
+
"loss": 0.0345,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 11.978021978021978,
|
| 769 |
+
"grad_norm": 0.1906553953886032,
|
| 770 |
+
"learning_rate": 7.478068448894577e-05,
|
| 771 |
+
"loss": 0.03,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.087912087912088,
|
| 776 |
+
"grad_norm": 0.18722419440746307,
|
| 777 |
+
"learning_rate": 7.301649725635689e-05,
|
| 778 |
+
"loss": 0.0327,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.197802197802197,
|
| 783 |
+
"grad_norm": 0.2109653502702713,
|
| 784 |
+
"learning_rate": 7.126132229448883e-05,
|
| 785 |
+
"loss": 0.03,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.307692307692308,
|
| 790 |
+
"grad_norm": 0.3317616879940033,
|
| 791 |
+
"learning_rate": 6.95157458175211e-05,
|
| 792 |
+
"loss": 0.0289,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.417582417582418,
|
| 797 |
+
"grad_norm": 0.18462032079696655,
|
| 798 |
+
"learning_rate": 6.778035083381708e-05,
|
| 799 |
+
"loss": 0.0336,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 12.527472527472527,
|
| 804 |
+
"grad_norm": 0.17815662920475006,
|
| 805 |
+
"learning_rate": 6.60557169512042e-05,
|
| 806 |
+
"loss": 0.0317,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 12.637362637362637,
|
| 811 |
+
"grad_norm": 0.19909383356571198,
|
| 812 |
+
"learning_rate": 6.434242018338947e-05,
|
| 813 |
+
"loss": 0.0317,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 12.747252747252748,
|
| 818 |
+
"grad_norm": 0.21948197484016418,
|
| 819 |
+
"learning_rate": 6.264103275757578e-05,
|
| 820 |
+
"loss": 0.0288,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 12.857142857142858,
|
| 825 |
+
"grad_norm": 0.30106085538864136,
|
| 826 |
+
"learning_rate": 6.095212292334232e-05,
|
| 827 |
+
"loss": 0.0298,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 12.967032967032967,
|
| 832 |
+
"grad_norm": 0.3014363944530487,
|
| 833 |
+
"learning_rate": 5.927625476285426e-05,
|
| 834 |
+
"loss": 0.0318,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 13.076923076923077,
|
| 839 |
+
"grad_norm": 0.5095160007476807,
|
| 840 |
+
"learning_rate": 5.761398800246354e-05,
|
| 841 |
+
"loss": 0.0272,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 13.186813186813186,
|
| 846 |
+
"grad_norm": 0.19524632394313812,
|
| 847 |
+
"learning_rate": 5.596587782576509e-05,
|
| 848 |
+
"loss": 0.0294,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 13.296703296703297,
|
| 853 |
+
"grad_norm": 0.20751602947711945,
|
| 854 |
+
"learning_rate": 5.433247468816977e-05,
|
| 855 |
+
"loss": 0.0295,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 13.406593406593407,
|
| 860 |
+
"grad_norm": 0.12134165316820145,
|
| 861 |
+
"learning_rate": 5.271432413305687e-05,
|
| 862 |
+
"loss": 0.0258,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 13.516483516483516,
|
| 867 |
+
"grad_norm": 0.2958245277404785,
|
| 868 |
+
"learning_rate": 5.111196660956703e-05,
|
| 869 |
+
"loss": 0.0328,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 13.626373626373626,
|
| 874 |
+
"grad_norm": 0.1860036551952362,
|
| 875 |
+
"learning_rate": 4.952593729209671e-05,
|
| 876 |
+
"loss": 0.028,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 13.736263736263737,
|
| 881 |
+
"grad_norm": 0.16853293776512146,
|
| 882 |
+
"learning_rate": 4.7956765901554065e-05,
|
| 883 |
+
"loss": 0.0259,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 13.846153846153847,
|
| 888 |
+
"grad_norm": 0.22942283749580383,
|
| 889 |
+
"learning_rate": 4.640497652843672e-05,
|
| 890 |
+
"loss": 0.0303,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 13.956043956043956,
|
| 895 |
+
"grad_norm": 0.1646496206521988,
|
| 896 |
+
"learning_rate": 4.4871087457789584e-05,
|
| 897 |
+
"loss": 0.0263,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 14.065934065934066,
|
| 902 |
+
"grad_norm": 0.22191093862056732,
|
| 903 |
+
"learning_rate": 4.3355610996102e-05,
|
| 904 |
+
"loss": 0.033,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 14.175824175824175,
|
| 909 |
+
"grad_norm": 0.19444125890731812,
|
| 910 |
+
"learning_rate": 4.185905330020143e-05,
|
| 911 |
+
"loss": 0.0268,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 14.285714285714286,
|
| 916 |
+
"grad_norm": 0.16687805950641632,
|
| 917 |
+
"learning_rate": 4.038191420820139e-05,
|
| 918 |
+
"loss": 0.0293,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 14.395604395604396,
|
| 923 |
+
"grad_norm": 0.21555913984775543,
|
| 924 |
+
"learning_rate": 3.892468707255923e-05,
|
| 925 |
+
"loss": 0.0269,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 14.505494505494505,
|
| 930 |
+
"grad_norm": 0.1728578358888626,
|
| 931 |
+
"learning_rate": 3.7487858595300884e-05,
|
| 932 |
+
"loss": 0.0222,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 14.615384615384615,
|
| 937 |
+
"grad_norm": 0.25217944383621216,
|
| 938 |
+
"learning_rate": 3.607190866546578e-05,
|
| 939 |
+
"loss": 0.023,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 14.725274725274724,
|
| 944 |
+
"grad_norm": 0.45261967182159424,
|
| 945 |
+
"learning_rate": 3.467731019882838e-05,
|
| 946 |
+
"loss": 0.0366,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 14.835164835164836,
|
| 951 |
+
"grad_norm": 0.19587543606758118,
|
| 952 |
+
"learning_rate": 3.330452897994773e-05,
|
| 953 |
+
"loss": 0.0249,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 14.945054945054945,
|
| 958 |
+
"grad_norm": 0.16712340712547302,
|
| 959 |
+
"learning_rate": 3.195402350659945e-05,
|
| 960 |
+
"loss": 0.0291,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 15.054945054945055,
|
| 965 |
+
"grad_norm": 0.3442732095718384,
|
| 966 |
+
"learning_rate": 3.0626244836640993e-05,
|
| 967 |
+
"loss": 0.0267,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 15.164835164835164,
|
| 972 |
+
"grad_norm": 0.20348604023456573,
|
| 973 |
+
"learning_rate": 2.9321636437362253e-05,
|
| 974 |
+
"loss": 0.0255,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 15.274725274725276,
|
| 979 |
+
"grad_norm": 0.18801550567150116,
|
| 980 |
+
"learning_rate": 2.8040634037370728e-05,
|
| 981 |
+
"loss": 0.0247,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 15.384615384615385,
|
| 986 |
+
"grad_norm": 0.2544891834259033,
|
| 987 |
+
"learning_rate": 2.6783665481062113e-05,
|
| 988 |
+
"loss": 0.0275,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 15.494505494505495,
|
| 993 |
+
"grad_norm": 0.3431093394756317,
|
| 994 |
+
"learning_rate": 2.5551150585723415e-05,
|
| 995 |
+
"loss": 0.0238,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 15.604395604395604,
|
| 1000 |
+
"grad_norm": 0.17770768702030182,
|
| 1001 |
+
"learning_rate": 2.4343501001317605e-05,
|
| 1002 |
+
"loss": 0.0251,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 15.714285714285714,
|
| 1007 |
+
"grad_norm": 0.170986145734787,
|
| 1008 |
+
"learning_rate": 2.3161120072995757e-05,
|
| 1009 |
+
"loss": 0.0243,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 15.824175824175825,
|
| 1014 |
+
"grad_norm": 0.19961877167224884,
|
| 1015 |
+
"learning_rate": 2.2004402706383043e-05,
|
| 1016 |
+
"loss": 0.0237,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 15.934065934065934,
|
| 1021 |
+
"grad_norm": 0.38295048475265503,
|
| 1022 |
+
"learning_rate": 2.0873735235683535e-05,
|
| 1023 |
+
"loss": 0.0239,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 16.043956043956044,
|
| 1028 |
+
"grad_norm": 0.15348948538303375,
|
| 1029 |
+
"learning_rate": 1.976949529464771e-05,
|
| 1030 |
+
"loss": 0.0282,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 16.153846153846153,
|
| 1035 |
+
"grad_norm": 0.3392385244369507,
|
| 1036 |
+
"learning_rate": 1.8692051690445743e-05,
|
| 1037 |
+
"loss": 0.0283,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 16.263736263736263,
|
| 1042 |
+
"grad_norm": 0.15572497248649597,
|
| 1043 |
+
"learning_rate": 1.764176428048908e-05,
|
| 1044 |
+
"loss": 0.0245,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 16.373626373626372,
|
| 1049 |
+
"grad_norm": 0.2224312424659729,
|
| 1050 |
+
"learning_rate": 1.661898385224079e-05,
|
| 1051 |
+
"loss": 0.0258,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 16.483516483516482,
|
| 1056 |
+
"grad_norm": 0.17235592007637024,
|
| 1057 |
+
"learning_rate": 1.5624052006055544e-05,
|
| 1058 |
+
"loss": 0.0248,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 16.593406593406595,
|
| 1063 |
+
"grad_norm": 0.17149613797664642,
|
| 1064 |
+
"learning_rate": 1.4657301041087812e-05,
|
| 1065 |
+
"loss": 0.0243,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 16.703296703296704,
|
| 1070 |
+
"grad_norm": 0.18543320894241333,
|
| 1071 |
+
"learning_rate": 1.3719053844306706e-05,
|
| 1072 |
+
"loss": 0.0272,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 16.813186813186814,
|
| 1077 |
+
"grad_norm": 0.1908874362707138,
|
| 1078 |
+
"learning_rate": 1.2809623782654223e-05,
|
| 1079 |
+
"loss": 0.0235,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 16.923076923076923,
|
| 1084 |
+
"grad_norm": 0.18022581934928894,
|
| 1085 |
+
"learning_rate": 1.1929314598383423e-05,
|
| 1086 |
+
"loss": 0.0263,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 17.032967032967033,
|
| 1091 |
+
"grad_norm": 0.1982187032699585,
|
| 1092 |
+
"learning_rate": 1.1078420307610783e-05,
|
| 1093 |
+
"loss": 0.0252,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 17.142857142857142,
|
| 1098 |
+
"grad_norm": 0.1451161950826645,
|
| 1099 |
+
"learning_rate": 1.0257225102117363e-05,
|
| 1100 |
+
"loss": 0.025,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 17.252747252747252,
|
| 1105 |
+
"grad_norm": 0.1547311544418335,
|
| 1106 |
+
"learning_rate": 9.466003254430934e-06,
|
| 1107 |
+
"loss": 0.024,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 17.36263736263736,
|
| 1112 |
+
"grad_norm": 0.20238442718982697,
|
| 1113 |
+
"learning_rate": 8.705019026221317e-06,
|
| 1114 |
+
"loss": 0.0249,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 17.47252747252747,
|
| 1119 |
+
"grad_norm": 0.18659718334674835,
|
| 1120 |
+
"learning_rate": 7.974526580038988e-06,
|
| 1121 |
+
"loss": 0.0257,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 17.582417582417584,
|
| 1126 |
+
"grad_norm": 0.2173370122909546,
|
| 1127 |
+
"learning_rate": 7.2747698944269915e-06,
|
| 1128 |
+
"loss": 0.0248,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 17.692307692307693,
|
| 1133 |
+
"grad_norm": 0.12479448318481445,
|
| 1134 |
+
"learning_rate": 6.605982682433953e-06,
|
| 1135 |
+
"loss": 0.0263,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 17.802197802197803,
|
| 1140 |
+
"grad_norm": 0.17071430385112762,
|
| 1141 |
+
"learning_rate": 5.968388313555895e-06,
|
| 1142 |
+
"loss": 0.0225,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 17.912087912087912,
|
| 1147 |
+
"grad_norm": 0.2871854901313782,
|
| 1148 |
+
"learning_rate": 5.3621997391326565e-06,
|
| 1149 |
+
"loss": 0.0191,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 18.021978021978022,
|
| 1154 |
+
"grad_norm": 0.39435771107673645,
|
| 1155 |
+
"learning_rate": 4.7876194212238434e-06,
|
| 1156 |
+
"loss": 0.0291,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 18.13186813186813,
|
| 1161 |
+
"grad_norm": 0.18047714233398438,
|
| 1162 |
+
"learning_rate": 4.24483926498812e-06,
|
| 1163 |
+
"loss": 0.0207,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 18.24175824175824,
|
| 1168 |
+
"grad_norm": 0.19172823429107666,
|
| 1169 |
+
"learning_rate": 3.734040554588514e-06,
|
| 1170 |
+
"loss": 0.0214,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 18.35164835164835,
|
| 1175 |
+
"grad_norm": 0.2549908757209778,
|
| 1176 |
+
"learning_rate": 3.255393892644909e-06,
|
| 1177 |
+
"loss": 0.0205,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 18.46153846153846,
|
| 1182 |
+
"grad_norm": 0.21838967502117157,
|
| 1183 |
+
"learning_rate": 2.8090591432542e-06,
|
| 1184 |
+
"loss": 0.0214,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 18.571428571428573,
|
| 1189 |
+
"grad_norm": 0.2188936024904251,
|
| 1190 |
+
"learning_rate": 2.3951853785969537e-06,
|
| 1191 |
+
"loss": 0.0227,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 18.681318681318682,
|
| 1196 |
+
"grad_norm": 0.15396256744861603,
|
| 1197 |
+
"learning_rate": 2.0139108291485574e-06,
|
| 1198 |
+
"loss": 0.0286,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 18.791208791208792,
|
| 1203 |
+
"grad_norm": 0.22776196897029877,
|
| 1204 |
+
"learning_rate": 1.6653628375112972e-06,
|
| 1205 |
+
"loss": 0.0248,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 18.9010989010989,
|
| 1210 |
+
"grad_norm": 0.17969872057437897,
|
| 1211 |
+
"learning_rate": 1.349657815883032e-06,
|
| 1212 |
+
"loss": 0.0227,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 19.01098901098901,
|
| 1217 |
+
"grad_norm": 0.22254760563373566,
|
| 1218 |
+
"learning_rate": 1.0669012071764073e-06,
|
| 1219 |
+
"loss": 0.0197,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 19.12087912087912,
|
| 1224 |
+
"grad_norm": 0.1676974892616272,
|
| 1225 |
+
"learning_rate": 8.171874498018039e-07,
|
| 1226 |
+
"loss": 0.0264,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 19.23076923076923,
|
| 1231 |
+
"grad_norm": 0.21863488852977753,
|
| 1232 |
+
"learning_rate": 6.005999461256684e-07,
|
| 1233 |
+
"loss": 0.0224,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 19.34065934065934,
|
| 1238 |
+
"grad_norm": 0.13759975135326385,
|
| 1239 |
+
"learning_rate": 4.172110346148506e-07,
|
| 1240 |
+
"loss": 0.0265,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 19.45054945054945,
|
| 1245 |
+
"grad_norm": 0.16070395708084106,
|
| 1246 |
+
"learning_rate": 2.670819656760526e-07,
|
| 1247 |
+
"loss": 0.0233,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 19.560439560439562,
|
| 1252 |
+
"grad_norm": 0.2525140643119812,
|
| 1253 |
+
"learning_rate": 1.502628811987483e-07,
|
| 1254 |
+
"loss": 0.0241,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 19.67032967032967,
|
| 1259 |
+
"grad_norm": 0.16685707867145538,
|
| 1260 |
+
"learning_rate": 6.679279780821279e-08,
|
| 1261 |
+
"loss": 0.0248,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 19.78021978021978,
|
| 1266 |
+
"grad_norm": 0.17456871271133423,
|
| 1267 |
+
"learning_rate": 1.6699593834224036e-08,
|
| 1268 |
+
"loss": 0.0259,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 19.89010989010989,
|
| 1273 |
+
"grad_norm": 0.18164990842342377,
|
| 1274 |
+
"learning_rate": 0.0,
|
| 1275 |
+
"loss": 0.0233,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 19.89010989010989,
|
| 1280 |
+
"step": 1810,
|
| 1281 |
+
"total_flos": 2.654273592491328e+17,
|
| 1282 |
+
"train_loss": 0.05974875727246479,
|
| 1283 |
+
"train_runtime": 2083.4502,
|
| 1284 |
+
"train_samples_per_second": 55.6,
|
| 1285 |
+
"train_steps_per_second": 0.869
|
| 1286 |
+
}
|
| 1287 |
+
],
|
| 1288 |
+
"logging_steps": 10,
|
| 1289 |
+
"max_steps": 1810,
|
| 1290 |
+
"num_input_tokens_seen": 0,
|
| 1291 |
+
"num_train_epochs": 20,
|
| 1292 |
+
"save_steps": 10000,
|
| 1293 |
+
"stateful_callbacks": {
|
| 1294 |
+
"TrainerControl": {
|
| 1295 |
+
"args": {
|
| 1296 |
+
"should_epoch_stop": false,
|
| 1297 |
+
"should_evaluate": false,
|
| 1298 |
+
"should_log": false,
|
| 1299 |
+
"should_save": true,
|
| 1300 |
+
"should_training_stop": true
|
| 1301 |
+
},
|
| 1302 |
+
"attributes": {}
|
| 1303 |
+
}
|
| 1304 |
+
},
|
| 1305 |
+
"total_flos": 2.654273592491328e+17,
|
| 1306 |
+
"train_batch_size": 64,
|
| 1307 |
+
"trial_name": null,
|
| 1308 |
+
"trial_params": null
|
| 1309 |
+
}
|