Upload multi_student_5/checkpoint-1800/trainer_state.json with huggingface_hub
Browse files
multi_student_5/checkpoint-1800/trainer_state.json
ADDED
|
@@ -0,0 +1,1294 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0118043844856661,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1800,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.005621135469364812,
|
| 14 |
+
"grad_norm": 19.072066040018505,
|
| 15 |
+
"learning_rate": 7.223113964686999e-07,
|
| 16 |
+
"loss": 8.2549,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.011242270938729624,
|
| 21 |
+
"grad_norm": 19.552394701284502,
|
| 22 |
+
"learning_rate": 1.5248796147672553e-06,
|
| 23 |
+
"loss": 8.1791,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.016863406408094434,
|
| 28 |
+
"grad_norm": 9.93392906727828,
|
| 29 |
+
"learning_rate": 2.3274478330658108e-06,
|
| 30 |
+
"loss": 7.6759,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.022484541877459248,
|
| 35 |
+
"grad_norm": 7.426576059473498,
|
| 36 |
+
"learning_rate": 3.1300160513643665e-06,
|
| 37 |
+
"loss": 6.913,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.028105677346824058,
|
| 42 |
+
"grad_norm": 3.5646171014288366,
|
| 43 |
+
"learning_rate": 3.932584269662922e-06,
|
| 44 |
+
"loss": 6.3931,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03372681281618887,
|
| 49 |
+
"grad_norm": 2.8779163477648666,
|
| 50 |
+
"learning_rate": 4.735152487961477e-06,
|
| 51 |
+
"loss": 6.0533,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.03934794828555368,
|
| 56 |
+
"grad_norm": 3.3997232487978883,
|
| 57 |
+
"learning_rate": 5.537720706260032e-06,
|
| 58 |
+
"loss": 5.7246,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.044969083754918496,
|
| 63 |
+
"grad_norm": 4.059106771548473,
|
| 64 |
+
"learning_rate": 6.340288924558588e-06,
|
| 65 |
+
"loss": 5.2465,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.050590219224283306,
|
| 70 |
+
"grad_norm": 4.7307302237777895,
|
| 71 |
+
"learning_rate": 7.142857142857143e-06,
|
| 72 |
+
"loss": 4.6393,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.056211354693648116,
|
| 77 |
+
"grad_norm": 5.242698651813807,
|
| 78 |
+
"learning_rate": 7.9454253611557e-06,
|
| 79 |
+
"loss": 4.0362,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.061832490163012926,
|
| 84 |
+
"grad_norm": 4.723996169899083,
|
| 85 |
+
"learning_rate": 8.747993579454254e-06,
|
| 86 |
+
"loss": 3.549,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06745362563237774,
|
| 91 |
+
"grad_norm": 2.885143750378497,
|
| 92 |
+
"learning_rate": 9.550561797752809e-06,
|
| 93 |
+
"loss": 3.2171,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07307476110174255,
|
| 98 |
+
"grad_norm": 6.825878781249646,
|
| 99 |
+
"learning_rate": 1.0353130016051364e-05,
|
| 100 |
+
"loss": 3.027,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.07869589657110736,
|
| 105 |
+
"grad_norm": 6.737230278162123,
|
| 106 |
+
"learning_rate": 1.115569823434992e-05,
|
| 107 |
+
"loss": 2.9055,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.08431703204047218,
|
| 112 |
+
"grad_norm": 6.716587811023919,
|
| 113 |
+
"learning_rate": 1.1958266452648475e-05,
|
| 114 |
+
"loss": 2.8089,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.08993816750983699,
|
| 119 |
+
"grad_norm": 7.201415197148535,
|
| 120 |
+
"learning_rate": 1.276083467094703e-05,
|
| 121 |
+
"loss": 2.7296,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.0955593029792018,
|
| 126 |
+
"grad_norm": 7.301380585480233,
|
| 127 |
+
"learning_rate": 1.3563402889245585e-05,
|
| 128 |
+
"loss": 2.6609,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10118043844856661,
|
| 133 |
+
"grad_norm": 11.575326761105941,
|
| 134 |
+
"learning_rate": 1.4365971107544143e-05,
|
| 135 |
+
"loss": 2.5979,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.10680157391793142,
|
| 140 |
+
"grad_norm": 8.619354697957048,
|
| 141 |
+
"learning_rate": 1.5168539325842698e-05,
|
| 142 |
+
"loss": 2.5341,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.11242270938729623,
|
| 147 |
+
"grad_norm": 7.390257991673745,
|
| 148 |
+
"learning_rate": 1.5971107544141253e-05,
|
| 149 |
+
"loss": 2.4767,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.11804384485666104,
|
| 154 |
+
"grad_norm": 13.39500588081804,
|
| 155 |
+
"learning_rate": 1.6773675762439807e-05,
|
| 156 |
+
"loss": 2.4379,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.12366498032602585,
|
| 161 |
+
"grad_norm": 27.117334976761317,
|
| 162 |
+
"learning_rate": 1.7576243980738362e-05,
|
| 163 |
+
"loss": 2.3867,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.12928611579539068,
|
| 168 |
+
"grad_norm": 11.695868441564258,
|
| 169 |
+
"learning_rate": 1.8378812199036917e-05,
|
| 170 |
+
"loss": 2.3207,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.13490725126475547,
|
| 175 |
+
"grad_norm": 10.785553590242758,
|
| 176 |
+
"learning_rate": 1.9181380417335472e-05,
|
| 177 |
+
"loss": 2.2455,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1405283867341203,
|
| 182 |
+
"grad_norm": 25.384076838053424,
|
| 183 |
+
"learning_rate": 1.998394863563403e-05,
|
| 184 |
+
"loss": 2.1723,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1461495222034851,
|
| 189 |
+
"grad_norm": 16.42978149349179,
|
| 190 |
+
"learning_rate": 2.0786516853932585e-05,
|
| 191 |
+
"loss": 2.1322,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.15177065767284992,
|
| 196 |
+
"grad_norm": 15.299053481717623,
|
| 197 |
+
"learning_rate": 2.158908507223114e-05,
|
| 198 |
+
"loss": 2.0356,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.15739179314221471,
|
| 203 |
+
"grad_norm": 16.598966653865332,
|
| 204 |
+
"learning_rate": 2.2391653290529695e-05,
|
| 205 |
+
"loss": 1.9785,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.16301292861157954,
|
| 210 |
+
"grad_norm": 8.000521357822016,
|
| 211 |
+
"learning_rate": 2.319422150882825e-05,
|
| 212 |
+
"loss": 1.8885,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.16863406408094436,
|
| 217 |
+
"grad_norm": 12.337240959757104,
|
| 218 |
+
"learning_rate": 2.3996789727126808e-05,
|
| 219 |
+
"loss": 1.7929,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.17425519955030916,
|
| 224 |
+
"grad_norm": 9.313580975453032,
|
| 225 |
+
"learning_rate": 2.4799357945425362e-05,
|
| 226 |
+
"loss": 1.7252,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.17987633501967398,
|
| 231 |
+
"grad_norm": 9.713971076745432,
|
| 232 |
+
"learning_rate": 2.5601926163723917e-05,
|
| 233 |
+
"loss": 1.6366,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.18549747048903878,
|
| 238 |
+
"grad_norm": 6.604247664050388,
|
| 239 |
+
"learning_rate": 2.6404494382022472e-05,
|
| 240 |
+
"loss": 1.5532,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.1911186059584036,
|
| 245 |
+
"grad_norm": 8.06631111933664,
|
| 246 |
+
"learning_rate": 2.7207062600321027e-05,
|
| 247 |
+
"loss": 1.5014,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.1967397414277684,
|
| 252 |
+
"grad_norm": 5.369237324186266,
|
| 253 |
+
"learning_rate": 2.8009630818619585e-05,
|
| 254 |
+
"loss": 1.4407,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.20236087689713322,
|
| 259 |
+
"grad_norm": 6.309483928633083,
|
| 260 |
+
"learning_rate": 2.881219903691814e-05,
|
| 261 |
+
"loss": 1.398,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.20798201236649802,
|
| 266 |
+
"grad_norm": 6.173807170921828,
|
| 267 |
+
"learning_rate": 2.9614767255216698e-05,
|
| 268 |
+
"loss": 1.3462,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.21360314783586284,
|
| 273 |
+
"grad_norm": 7.913665828353373,
|
| 274 |
+
"learning_rate": 3.0417335473515253e-05,
|
| 275 |
+
"loss": 1.2942,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.21922428330522767,
|
| 280 |
+
"grad_norm": 6.320887435288008,
|
| 281 |
+
"learning_rate": 3.1219903691813805e-05,
|
| 282 |
+
"loss": 1.2581,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.22484541877459246,
|
| 287 |
+
"grad_norm": 4.001174978157139,
|
| 288 |
+
"learning_rate": 3.202247191011236e-05,
|
| 289 |
+
"loss": 1.2068,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2304665542439573,
|
| 294 |
+
"grad_norm": 5.833384870789841,
|
| 295 |
+
"learning_rate": 3.2825040128410914e-05,
|
| 296 |
+
"loss": 1.1774,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.23608768971332209,
|
| 301 |
+
"grad_norm": 5.418298506902902,
|
| 302 |
+
"learning_rate": 3.362760834670947e-05,
|
| 303 |
+
"loss": 1.1285,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2417088251826869,
|
| 308 |
+
"grad_norm": 6.822807280672466,
|
| 309 |
+
"learning_rate": 3.4430176565008024e-05,
|
| 310 |
+
"loss": 1.0971,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2473299606520517,
|
| 315 |
+
"grad_norm": 5.1633853492959725,
|
| 316 |
+
"learning_rate": 3.523274478330658e-05,
|
| 317 |
+
"loss": 1.0631,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.25295109612141653,
|
| 322 |
+
"grad_norm": 4.794453623656116,
|
| 323 |
+
"learning_rate": 3.6035313001605133e-05,
|
| 324 |
+
"loss": 1.0233,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.25857223159078135,
|
| 329 |
+
"grad_norm": 6.493147309773226,
|
| 330 |
+
"learning_rate": 3.683788121990369e-05,
|
| 331 |
+
"loss": 1.0093,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2641933670601461,
|
| 336 |
+
"grad_norm": 7.38273852542595,
|
| 337 |
+
"learning_rate": 3.764044943820225e-05,
|
| 338 |
+
"loss": 0.9745,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.26981450252951095,
|
| 343 |
+
"grad_norm": 5.953745454074908,
|
| 344 |
+
"learning_rate": 3.8443017656500805e-05,
|
| 345 |
+
"loss": 0.9609,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.27543563799887577,
|
| 350 |
+
"grad_norm": 5.92674703602574,
|
| 351 |
+
"learning_rate": 3.924558587479936e-05,
|
| 352 |
+
"loss": 0.9295,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2810567734682406,
|
| 357 |
+
"grad_norm": 4.500282226524396,
|
| 358 |
+
"learning_rate": 4.0048154093097914e-05,
|
| 359 |
+
"loss": 0.8992,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.2866779089376054,
|
| 364 |
+
"grad_norm": 4.320018263048497,
|
| 365 |
+
"learning_rate": 4.085072231139647e-05,
|
| 366 |
+
"loss": 0.8855,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.2922990444069702,
|
| 371 |
+
"grad_norm": 6.008360095853525,
|
| 372 |
+
"learning_rate": 4.1653290529695024e-05,
|
| 373 |
+
"loss": 0.8589,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.297920179876335,
|
| 378 |
+
"grad_norm": 5.252075871228091,
|
| 379 |
+
"learning_rate": 4.245585874799358e-05,
|
| 380 |
+
"loss": 0.8272,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.30354131534569984,
|
| 385 |
+
"grad_norm": 6.427285353419691,
|
| 386 |
+
"learning_rate": 4.3258426966292134e-05,
|
| 387 |
+
"loss": 0.808,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.30916245081506466,
|
| 392 |
+
"grad_norm": 5.854800753365741,
|
| 393 |
+
"learning_rate": 4.4060995184590695e-05,
|
| 394 |
+
"loss": 0.7901,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.31478358628442943,
|
| 399 |
+
"grad_norm": 7.046341886366144,
|
| 400 |
+
"learning_rate": 4.486356340288925e-05,
|
| 401 |
+
"loss": 0.7701,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.32040472175379425,
|
| 406 |
+
"grad_norm": 5.594440118602691,
|
| 407 |
+
"learning_rate": 4.5666131621187805e-05,
|
| 408 |
+
"loss": 0.7462,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3260258572231591,
|
| 413 |
+
"grad_norm": 5.653552049437021,
|
| 414 |
+
"learning_rate": 4.646869983948636e-05,
|
| 415 |
+
"loss": 0.7284,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3316469926925239,
|
| 420 |
+
"grad_norm": 5.196665356454571,
|
| 421 |
+
"learning_rate": 4.7271268057784915e-05,
|
| 422 |
+
"loss": 0.712,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3372681281618887,
|
| 427 |
+
"grad_norm": 6.1381022060931025,
|
| 428 |
+
"learning_rate": 4.807383627608347e-05,
|
| 429 |
+
"loss": 0.6868,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3428892636312535,
|
| 434 |
+
"grad_norm": 5.052546164435208,
|
| 435 |
+
"learning_rate": 4.8876404494382024e-05,
|
| 436 |
+
"loss": 0.6672,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3485103991006183,
|
| 441 |
+
"grad_norm": 4.916726449591455,
|
| 442 |
+
"learning_rate": 4.9678972712680586e-05,
|
| 443 |
+
"loss": 0.6501,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.35413153456998314,
|
| 448 |
+
"grad_norm": 4.832632041097204,
|
| 449 |
+
"learning_rate": 5.0481540930979134e-05,
|
| 450 |
+
"loss": 0.6313,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.35975267003934797,
|
| 455 |
+
"grad_norm": 4.751573877646501,
|
| 456 |
+
"learning_rate": 5.1284109149277696e-05,
|
| 457 |
+
"loss": 0.6165,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.36537380550871273,
|
| 462 |
+
"grad_norm": 5.912237963361695,
|
| 463 |
+
"learning_rate": 5.2086677367576244e-05,
|
| 464 |
+
"loss": 0.5907,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.37099494097807756,
|
| 469 |
+
"grad_norm": 4.086167788806448,
|
| 470 |
+
"learning_rate": 5.2889245585874805e-05,
|
| 471 |
+
"loss": 0.5768,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3766160764474424,
|
| 476 |
+
"grad_norm": 4.582982284092926,
|
| 477 |
+
"learning_rate": 5.369181380417335e-05,
|
| 478 |
+
"loss": 0.5579,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3822372119168072,
|
| 483 |
+
"grad_norm": 4.620402866105723,
|
| 484 |
+
"learning_rate": 5.4494382022471915e-05,
|
| 485 |
+
"loss": 0.5379,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.38785834738617203,
|
| 490 |
+
"grad_norm": 3.8404903506330834,
|
| 491 |
+
"learning_rate": 5.5296950240770476e-05,
|
| 492 |
+
"loss": 0.5112,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.3934794828555368,
|
| 497 |
+
"grad_norm": 3.2956728880451784,
|
| 498 |
+
"learning_rate": 5.6099518459069025e-05,
|
| 499 |
+
"loss": 0.502,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3991006183249016,
|
| 504 |
+
"grad_norm": 4.455851011958456,
|
| 505 |
+
"learning_rate": 5.6902086677367586e-05,
|
| 506 |
+
"loss": 0.4889,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.40472175379426645,
|
| 511 |
+
"grad_norm": 3.3081515499225693,
|
| 512 |
+
"learning_rate": 5.7704654895666134e-05,
|
| 513 |
+
"loss": 0.4703,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.41034288926363127,
|
| 518 |
+
"grad_norm": 3.5766535302434037,
|
| 519 |
+
"learning_rate": 5.8507223113964696e-05,
|
| 520 |
+
"loss": 0.4591,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.41596402473299604,
|
| 525 |
+
"grad_norm": 3.1816986592066154,
|
| 526 |
+
"learning_rate": 5.9309791332263244e-05,
|
| 527 |
+
"loss": 0.437,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.42158516020236086,
|
| 532 |
+
"grad_norm": 2.746088460027587,
|
| 533 |
+
"learning_rate": 6.0112359550561805e-05,
|
| 534 |
+
"loss": 0.429,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.4272062956717257,
|
| 539 |
+
"grad_norm": 2.7638950111486777,
|
| 540 |
+
"learning_rate": 6.0914927768860353e-05,
|
| 541 |
+
"loss": 0.4056,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4328274311410905,
|
| 546 |
+
"grad_norm": 2.621320210943319,
|
| 547 |
+
"learning_rate": 6.171749598715892e-05,
|
| 548 |
+
"loss": 0.3951,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.43844856661045534,
|
| 553 |
+
"grad_norm": 2.9039501807988635,
|
| 554 |
+
"learning_rate": 6.252006420545746e-05,
|
| 555 |
+
"loss": 0.3752,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.4440697020798201,
|
| 560 |
+
"grad_norm": 2.672708782303941,
|
| 561 |
+
"learning_rate": 6.332263242375602e-05,
|
| 562 |
+
"loss": 0.3732,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.44969083754918493,
|
| 567 |
+
"grad_norm": 2.534243622485374,
|
| 568 |
+
"learning_rate": 6.412520064205457e-05,
|
| 569 |
+
"loss": 0.3544,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.45531197301854975,
|
| 574 |
+
"grad_norm": 2.2060854376474297,
|
| 575 |
+
"learning_rate": 6.492776886035313e-05,
|
| 576 |
+
"loss": 0.337,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4609331084879146,
|
| 581 |
+
"grad_norm": 2.8950080552663353,
|
| 582 |
+
"learning_rate": 6.573033707865169e-05,
|
| 583 |
+
"loss": 0.3278,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.46655424395727935,
|
| 588 |
+
"grad_norm": 2.8077239251031183,
|
| 589 |
+
"learning_rate": 6.653290529695024e-05,
|
| 590 |
+
"loss": 0.3188,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.47217537942664417,
|
| 595 |
+
"grad_norm": 2.4109713129957173,
|
| 596 |
+
"learning_rate": 6.73354735152488e-05,
|
| 597 |
+
"loss": 0.3023,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.477796514896009,
|
| 602 |
+
"grad_norm": 2.4108771548493864,
|
| 603 |
+
"learning_rate": 6.813804173354735e-05,
|
| 604 |
+
"loss": 0.2907,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4834176503653738,
|
| 609 |
+
"grad_norm": 2.381688936764209,
|
| 610 |
+
"learning_rate": 6.894060995184591e-05,
|
| 611 |
+
"loss": 0.28,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.48903878583473864,
|
| 616 |
+
"grad_norm": 2.226240645538486,
|
| 617 |
+
"learning_rate": 6.974317817014446e-05,
|
| 618 |
+
"loss": 0.265,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4946599213041034,
|
| 623 |
+
"grad_norm": 2.36477519460133,
|
| 624 |
+
"learning_rate": 7.054574638844302e-05,
|
| 625 |
+
"loss": 0.2563,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.5002810567734682,
|
| 630 |
+
"grad_norm": 2.1877534753089325,
|
| 631 |
+
"learning_rate": 7.134831460674157e-05,
|
| 632 |
+
"loss": 0.2458,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5059021922428331,
|
| 637 |
+
"grad_norm": 1.9910110516875703,
|
| 638 |
+
"learning_rate": 7.215088282504013e-05,
|
| 639 |
+
"loss": 0.237,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.5115233277121979,
|
| 644 |
+
"grad_norm": 2.6541637478760203,
|
| 645 |
+
"learning_rate": 7.295345104333868e-05,
|
| 646 |
+
"loss": 0.2274,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.5171444631815627,
|
| 651 |
+
"grad_norm": 2.1481697620944273,
|
| 652 |
+
"learning_rate": 7.375601926163724e-05,
|
| 653 |
+
"loss": 0.2125,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5227655986509275,
|
| 658 |
+
"grad_norm": 2.2194085230343283,
|
| 659 |
+
"learning_rate": 7.455858747993579e-05,
|
| 660 |
+
"loss": 0.2086,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5283867341202922,
|
| 665 |
+
"grad_norm": 1.8687719324053211,
|
| 666 |
+
"learning_rate": 7.536115569823435e-05,
|
| 667 |
+
"loss": 0.2002,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5340078695896571,
|
| 672 |
+
"grad_norm": 1.809218029026303,
|
| 673 |
+
"learning_rate": 7.616372391653292e-05,
|
| 674 |
+
"loss": 0.19,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5396290050590219,
|
| 679 |
+
"grad_norm": 1.544075304452749,
|
| 680 |
+
"learning_rate": 7.696629213483147e-05,
|
| 681 |
+
"loss": 0.1824,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.5452501405283867,
|
| 686 |
+
"grad_norm": 1.7029729768183128,
|
| 687 |
+
"learning_rate": 7.776886035313003e-05,
|
| 688 |
+
"loss": 0.1766,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5508712759977515,
|
| 693 |
+
"grad_norm": 1.9321640691371915,
|
| 694 |
+
"learning_rate": 7.857142857142858e-05,
|
| 695 |
+
"loss": 0.1674,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.5564924114671164,
|
| 700 |
+
"grad_norm": 1.6567308102186273,
|
| 701 |
+
"learning_rate": 7.937399678972713e-05,
|
| 702 |
+
"loss": 0.1616,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.5621135469364812,
|
| 707 |
+
"grad_norm": 1.75914588466869,
|
| 708 |
+
"learning_rate": 8.017656500802569e-05,
|
| 709 |
+
"loss": 0.1496,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.567734682405846,
|
| 714 |
+
"grad_norm": 1.6033430089086067,
|
| 715 |
+
"learning_rate": 8.097913322632424e-05,
|
| 716 |
+
"loss": 0.1411,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5733558178752108,
|
| 721 |
+
"grad_norm": 1.5592637787510577,
|
| 722 |
+
"learning_rate": 8.17817014446228e-05,
|
| 723 |
+
"loss": 0.1349,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5789769533445756,
|
| 728 |
+
"grad_norm": 1.490862273444445,
|
| 729 |
+
"learning_rate": 8.258426966292135e-05,
|
| 730 |
+
"loss": 0.1321,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5845980888139404,
|
| 735 |
+
"grad_norm": 1.103403083662274,
|
| 736 |
+
"learning_rate": 8.338683788121991e-05,
|
| 737 |
+
"loss": 0.1237,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5902192242833052,
|
| 742 |
+
"grad_norm": 1.3322879027309737,
|
| 743 |
+
"learning_rate": 8.418940609951846e-05,
|
| 744 |
+
"loss": 0.1175,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.59584035975267,
|
| 749 |
+
"grad_norm": 1.1302428773890956,
|
| 750 |
+
"learning_rate": 8.499197431781702e-05,
|
| 751 |
+
"loss": 0.1134,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6014614952220348,
|
| 756 |
+
"grad_norm": 1.043524404705153,
|
| 757 |
+
"learning_rate": 8.579454253611557e-05,
|
| 758 |
+
"loss": 0.1079,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6070826306913997,
|
| 763 |
+
"grad_norm": 1.0807715777082445,
|
| 764 |
+
"learning_rate": 8.659711075441413e-05,
|
| 765 |
+
"loss": 0.1027,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6127037661607645,
|
| 770 |
+
"grad_norm": 1.0613560487233085,
|
| 771 |
+
"learning_rate": 8.739967897271268e-05,
|
| 772 |
+
"loss": 0.1014,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6183249016301293,
|
| 777 |
+
"grad_norm": 1.7970050911102724,
|
| 778 |
+
"learning_rate": 8.820224719101124e-05,
|
| 779 |
+
"loss": 0.0956,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.6239460370994941,
|
| 784 |
+
"grad_norm": 1.2518666492303272,
|
| 785 |
+
"learning_rate": 8.900481540930979e-05,
|
| 786 |
+
"loss": 0.0934,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6295671725688589,
|
| 791 |
+
"grad_norm": 1.1953063290224126,
|
| 792 |
+
"learning_rate": 8.980738362760836e-05,
|
| 793 |
+
"loss": 0.1004,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.6351883080382237,
|
| 798 |
+
"grad_norm": 0.9616154833979669,
|
| 799 |
+
"learning_rate": 9.06099518459069e-05,
|
| 800 |
+
"loss": 0.0843,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.6408094435075885,
|
| 805 |
+
"grad_norm": 0.8282021805405166,
|
| 806 |
+
"learning_rate": 9.141252006420547e-05,
|
| 807 |
+
"loss": 0.0788,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6464305789769533,
|
| 812 |
+
"grad_norm": 0.7645874993850531,
|
| 813 |
+
"learning_rate": 9.221508828250401e-05,
|
| 814 |
+
"loss": 0.0741,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6520517144463182,
|
| 819 |
+
"grad_norm": 1.0333771317137108,
|
| 820 |
+
"learning_rate": 9.301765650080258e-05,
|
| 821 |
+
"loss": 0.0708,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.657672849915683,
|
| 826 |
+
"grad_norm": 0.6189773458607581,
|
| 827 |
+
"learning_rate": 9.382022471910112e-05,
|
| 828 |
+
"loss": 0.0722,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.6632939853850478,
|
| 833 |
+
"grad_norm": 0.5553259399184006,
|
| 834 |
+
"learning_rate": 9.462279293739969e-05,
|
| 835 |
+
"loss": 0.0654,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.6689151208544126,
|
| 840 |
+
"grad_norm": 0.6418699826913336,
|
| 841 |
+
"learning_rate": 9.542536115569823e-05,
|
| 842 |
+
"loss": 0.0625,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.6745362563237775,
|
| 847 |
+
"grad_norm": 0.692560675949567,
|
| 848 |
+
"learning_rate": 9.62279293739968e-05,
|
| 849 |
+
"loss": 0.0591,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6801573917931422,
|
| 854 |
+
"grad_norm": 0.4228488541088058,
|
| 855 |
+
"learning_rate": 9.703049759229535e-05,
|
| 856 |
+
"loss": 0.0565,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.685778527262507,
|
| 861 |
+
"grad_norm": 0.4504043674769784,
|
| 862 |
+
"learning_rate": 9.783306581059391e-05,
|
| 863 |
+
"loss": 0.0546,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6913996627318718,
|
| 868 |
+
"grad_norm": 0.44916766960088017,
|
| 869 |
+
"learning_rate": 9.863563402889246e-05,
|
| 870 |
+
"loss": 0.0518,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6970207982012366,
|
| 875 |
+
"grad_norm": 0.46844229890271083,
|
| 876 |
+
"learning_rate": 9.943820224719102e-05,
|
| 877 |
+
"loss": 0.0504,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.7026419336706015,
|
| 882 |
+
"grad_norm": 0.4911174405862922,
|
| 883 |
+
"learning_rate": 9.999998231913558e-05,
|
| 884 |
+
"loss": 0.0482,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7082630691399663,
|
| 889 |
+
"grad_norm": 0.3927574276223742,
|
| 890 |
+
"learning_rate": 9.999966799300477e-05,
|
| 891 |
+
"loss": 0.0469,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.7138842046093311,
|
| 896 |
+
"grad_norm": 0.3446568196906315,
|
| 897 |
+
"learning_rate": 9.999896076161874e-05,
|
| 898 |
+
"loss": 0.0447,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7195053400786959,
|
| 903 |
+
"grad_norm": 0.6708037379046565,
|
| 904 |
+
"learning_rate": 9.999786063053502e-05,
|
| 905 |
+
"loss": 0.0437,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.7251264755480608,
|
| 910 |
+
"grad_norm": 0.5265622160752478,
|
| 911 |
+
"learning_rate": 9.999636760839859e-05,
|
| 912 |
+
"loss": 0.0428,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.7307476110174255,
|
| 917 |
+
"grad_norm": 0.3741304206840176,
|
| 918 |
+
"learning_rate": 9.999448170694188e-05,
|
| 919 |
+
"loss": 0.0418,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.7363687464867903,
|
| 924 |
+
"grad_norm": 0.3488892086457768,
|
| 925 |
+
"learning_rate": 9.999220294098459e-05,
|
| 926 |
+
"loss": 0.0401,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.7419898819561551,
|
| 931 |
+
"grad_norm": 0.364156745530358,
|
| 932 |
+
"learning_rate": 9.998953132843362e-05,
|
| 933 |
+
"loss": 0.0392,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.7476110174255199,
|
| 938 |
+
"grad_norm": 0.316984179052284,
|
| 939 |
+
"learning_rate": 9.998646689028292e-05,
|
| 940 |
+
"loss": 0.038,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.7532321528948848,
|
| 945 |
+
"grad_norm": 0.4122286636140672,
|
| 946 |
+
"learning_rate": 9.998300965061335e-05,
|
| 947 |
+
"loss": 0.0369,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.7588532883642496,
|
| 952 |
+
"grad_norm": 0.2828777162977292,
|
| 953 |
+
"learning_rate": 9.997915963659241e-05,
|
| 954 |
+
"loss": 0.0358,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7644744238336144,
|
| 959 |
+
"grad_norm": 0.35038412611293795,
|
| 960 |
+
"learning_rate": 9.997491687847413e-05,
|
| 961 |
+
"loss": 0.0344,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.7700955593029792,
|
| 966 |
+
"grad_norm": 0.3359568867412438,
|
| 967 |
+
"learning_rate": 9.997028140959878e-05,
|
| 968 |
+
"loss": 0.0348,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.7757166947723441,
|
| 973 |
+
"grad_norm": 0.3008434778248411,
|
| 974 |
+
"learning_rate": 9.99652532663926e-05,
|
| 975 |
+
"loss": 0.0331,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.7813378302417088,
|
| 980 |
+
"grad_norm": 0.24375407955159972,
|
| 981 |
+
"learning_rate": 9.995983248836749e-05,
|
| 982 |
+
"loss": 0.0321,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.7869589657110736,
|
| 987 |
+
"grad_norm": 0.2580356766574075,
|
| 988 |
+
"learning_rate": 9.995401911812084e-05,
|
| 989 |
+
"loss": 0.0305,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7925801011804384,
|
| 994 |
+
"grad_norm": 0.22377682987678044,
|
| 995 |
+
"learning_rate": 9.994781320133499e-05,
|
| 996 |
+
"loss": 0.0297,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.7982012366498032,
|
| 1001 |
+
"grad_norm": 0.3192038760895579,
|
| 1002 |
+
"learning_rate": 9.9941214786777e-05,
|
| 1003 |
+
"loss": 0.0295,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.8038223721191681,
|
| 1008 |
+
"grad_norm": 0.42209606301924135,
|
| 1009 |
+
"learning_rate": 9.99342239262983e-05,
|
| 1010 |
+
"loss": 0.0283,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.8094435075885329,
|
| 1015 |
+
"grad_norm": 0.2740277370912408,
|
| 1016 |
+
"learning_rate": 9.992684067483411e-05,
|
| 1017 |
+
"loss": 0.0276,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.8150646430578977,
|
| 1022 |
+
"grad_norm": 0.2403536361378785,
|
| 1023 |
+
"learning_rate": 9.991906509040323e-05,
|
| 1024 |
+
"loss": 0.027,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.8206857785272625,
|
| 1029 |
+
"grad_norm": 0.25177256736722925,
|
| 1030 |
+
"learning_rate": 9.99108972341074e-05,
|
| 1031 |
+
"loss": 0.026,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.8263069139966274,
|
| 1036 |
+
"grad_norm": 0.2611127538329995,
|
| 1037 |
+
"learning_rate": 9.990233717013092e-05,
|
| 1038 |
+
"loss": 0.0254,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.8319280494659921,
|
| 1043 |
+
"grad_norm": 0.26600699519251914,
|
| 1044 |
+
"learning_rate": 9.989338496574013e-05,
|
| 1045 |
+
"loss": 0.0246,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.8375491849353569,
|
| 1050 |
+
"grad_norm": 0.2511518661704276,
|
| 1051 |
+
"learning_rate": 9.988404069128286e-05,
|
| 1052 |
+
"loss": 0.0241,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.8431703204047217,
|
| 1057 |
+
"grad_norm": 0.24275092866635561,
|
| 1058 |
+
"learning_rate": 9.987430442018787e-05,
|
| 1059 |
+
"loss": 0.0231,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.8487914558740866,
|
| 1064 |
+
"grad_norm": 0.2117797837985066,
|
| 1065 |
+
"learning_rate": 9.986417622896433e-05,
|
| 1066 |
+
"loss": 0.0222,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.8544125913434514,
|
| 1071 |
+
"grad_norm": 0.23706464693081078,
|
| 1072 |
+
"learning_rate": 9.985365619720114e-05,
|
| 1073 |
+
"loss": 0.0224,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.8600337268128162,
|
| 1078 |
+
"grad_norm": 0.21681891631644465,
|
| 1079 |
+
"learning_rate": 9.984274440756639e-05,
|
| 1080 |
+
"loss": 0.0216,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.865654862282181,
|
| 1085 |
+
"grad_norm": 0.21895203687881423,
|
| 1086 |
+
"learning_rate": 9.983144094580662e-05,
|
| 1087 |
+
"loss": 0.0205,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.8712759977515458,
|
| 1092 |
+
"grad_norm": 0.20958525636292016,
|
| 1093 |
+
"learning_rate": 9.981974590074623e-05,
|
| 1094 |
+
"loss": 0.0197,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.8768971332209107,
|
| 1099 |
+
"grad_norm": 0.2092568671749629,
|
| 1100 |
+
"learning_rate": 9.980765936428668e-05,
|
| 1101 |
+
"loss": 0.0192,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.8825182686902754,
|
| 1106 |
+
"grad_norm": 0.20242771809585056,
|
| 1107 |
+
"learning_rate": 9.979518143140591e-05,
|
| 1108 |
+
"loss": 0.0186,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8881394041596402,
|
| 1113 |
+
"grad_norm": 0.2067518737526698,
|
| 1114 |
+
"learning_rate": 9.97823122001575e-05,
|
| 1115 |
+
"loss": 0.0184,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.893760539629005,
|
| 1120 |
+
"grad_norm": 0.2131831404574167,
|
| 1121 |
+
"learning_rate": 9.976905177166984e-05,
|
| 1122 |
+
"loss": 0.0179,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8993816750983699,
|
| 1127 |
+
"grad_norm": 0.19975090191123349,
|
| 1128 |
+
"learning_rate": 9.975540025014551e-05,
|
| 1129 |
+
"loss": 0.0171,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.9050028105677347,
|
| 1134 |
+
"grad_norm": 0.1840388312874417,
|
| 1135 |
+
"learning_rate": 9.974135774286027e-05,
|
| 1136 |
+
"loss": 0.0167,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.9106239460370995,
|
| 1141 |
+
"grad_norm": 0.16080844246323525,
|
| 1142 |
+
"learning_rate": 9.972692436016237e-05,
|
| 1143 |
+
"loss": 0.0162,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.9162450815064643,
|
| 1148 |
+
"grad_norm": 0.7406842420136835,
|
| 1149 |
+
"learning_rate": 9.971210021547163e-05,
|
| 1150 |
+
"loss": 0.0164,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.9218662169758292,
|
| 1155 |
+
"grad_norm": 0.6842280048869254,
|
| 1156 |
+
"learning_rate": 9.969688542527847e-05,
|
| 1157 |
+
"loss": 0.0185,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.927487352445194,
|
| 1162 |
+
"grad_norm": 1.580555984265757,
|
| 1163 |
+
"learning_rate": 9.968128010914313e-05,
|
| 1164 |
+
"loss": 0.0316,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.9331084879145587,
|
| 1169 |
+
"grad_norm": 0.37860426611015985,
|
| 1170 |
+
"learning_rate": 9.966528438969462e-05,
|
| 1171 |
+
"loss": 0.0249,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.9387296233839235,
|
| 1176 |
+
"grad_norm": 0.23262955055081527,
|
| 1177 |
+
"learning_rate": 9.964889839262983e-05,
|
| 1178 |
+
"loss": 0.0186,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.9443507588532883,
|
| 1183 |
+
"grad_norm": 0.20273477480045216,
|
| 1184 |
+
"learning_rate": 9.963212224671249e-05,
|
| 1185 |
+
"loss": 0.0161,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.9499718943226532,
|
| 1190 |
+
"grad_norm": 0.13881800111147724,
|
| 1191 |
+
"learning_rate": 9.961495608377221e-05,
|
| 1192 |
+
"loss": 0.0148,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.955593029792018,
|
| 1197 |
+
"grad_norm": 0.17108269648499383,
|
| 1198 |
+
"learning_rate": 9.959740003870339e-05,
|
| 1199 |
+
"loss": 0.0144,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.9612141652613828,
|
| 1204 |
+
"grad_norm": 0.16536831484992337,
|
| 1205 |
+
"learning_rate": 9.95794542494642e-05,
|
| 1206 |
+
"loss": 0.0137,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.9668353007307476,
|
| 1211 |
+
"grad_norm": 0.17400543803234356,
|
| 1212 |
+
"learning_rate": 9.956111885707546e-05,
|
| 1213 |
+
"loss": 0.0131,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.9724564362001125,
|
| 1218 |
+
"grad_norm": 0.14754960398432485,
|
| 1219 |
+
"learning_rate": 9.954239400561958e-05,
|
| 1220 |
+
"loss": 0.0126,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.9780775716694773,
|
| 1225 |
+
"grad_norm": 0.13872283525841225,
|
| 1226 |
+
"learning_rate": 9.952327984223938e-05,
|
| 1227 |
+
"loss": 0.012,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.983698707138842,
|
| 1232 |
+
"grad_norm": 0.15345071148869308,
|
| 1233 |
+
"learning_rate": 9.950377651713698e-05,
|
| 1234 |
+
"loss": 0.0119,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.9893198426082068,
|
| 1239 |
+
"grad_norm": 0.15868096773505136,
|
| 1240 |
+
"learning_rate": 9.948388418357255e-05,
|
| 1241 |
+
"loss": 0.0117,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.9949409780775716,
|
| 1246 |
+
"grad_norm": 0.13964728528048598,
|
| 1247 |
+
"learning_rate": 9.94636029978632e-05,
|
| 1248 |
+
"loss": 0.0111,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.0005621135469365,
|
| 1253 |
+
"grad_norm": 0.12990131986586648,
|
| 1254 |
+
"learning_rate": 9.944293311938164e-05,
|
| 1255 |
+
"loss": 0.0111,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.0061832490163012,
|
| 1260 |
+
"grad_norm": 0.15109117991919174,
|
| 1261 |
+
"learning_rate": 9.942187471055507e-05,
|
| 1262 |
+
"loss": 0.0106,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.0118043844856661,
|
| 1267 |
+
"grad_norm": 0.15023868490530576,
|
| 1268 |
+
"learning_rate": 9.940042793686378e-05,
|
| 1269 |
+
"loss": 0.0101,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
}
|
| 1272 |
+
],
|
| 1273 |
+
"logging_steps": 10,
|
| 1274 |
+
"max_steps": 12453,
|
| 1275 |
+
"num_input_tokens_seen": 0,
|
| 1276 |
+
"num_train_epochs": 7,
|
| 1277 |
+
"save_steps": 1800,
|
| 1278 |
+
"stateful_callbacks": {
|
| 1279 |
+
"TrainerControl": {
|
| 1280 |
+
"args": {
|
| 1281 |
+
"should_epoch_stop": false,
|
| 1282 |
+
"should_evaluate": false,
|
| 1283 |
+
"should_log": false,
|
| 1284 |
+
"should_save": true,
|
| 1285 |
+
"should_training_stop": false
|
| 1286 |
+
},
|
| 1287 |
+
"attributes": {}
|
| 1288 |
+
}
|
| 1289 |
+
},
|
| 1290 |
+
"total_flos": 146813294739456.0,
|
| 1291 |
+
"train_batch_size": 128,
|
| 1292 |
+
"trial_name": null,
|
| 1293 |
+
"trial_params": null
|
| 1294 |
+
}
|