Upload multi_student_1/checkpoint-1800/trainer_state.json with huggingface_hub
Browse files
multi_student_1/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": 2.8532910388580492,
|
| 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": 0.01586042823156225,
|
| 14 |
+
"grad_norm": 19.139528850030874,
|
| 15 |
+
"learning_rate": 8.910891089108911e-07,
|
| 16 |
+
"loss": 8.2466,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0317208564631245,
|
| 21 |
+
"grad_norm": 17.8179869039751,
|
| 22 |
+
"learning_rate": 1.8811881188118813e-06,
|
| 23 |
+
"loss": 8.1206,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.047581284694686754,
|
| 28 |
+
"grad_norm": 9.164020050000067,
|
| 29 |
+
"learning_rate": 2.8712871287128717e-06,
|
| 30 |
+
"loss": 7.4762,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.063441712926249,
|
| 35 |
+
"grad_norm": 4.906293902424142,
|
| 36 |
+
"learning_rate": 3.861386138613861e-06,
|
| 37 |
+
"loss": 6.7446,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07930214115781126,
|
| 42 |
+
"grad_norm": 3.052162672695787,
|
| 43 |
+
"learning_rate": 4.851485148514852e-06,
|
| 44 |
+
"loss": 6.2761,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09516256938937351,
|
| 49 |
+
"grad_norm": 2.98736461961841,
|
| 50 |
+
"learning_rate": 5.841584158415842e-06,
|
| 51 |
+
"loss": 5.929,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.11102299762093576,
|
| 56 |
+
"grad_norm": 3.707337444231957,
|
| 57 |
+
"learning_rate": 6.831683168316833e-06,
|
| 58 |
+
"loss": 5.5255,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.126883425852498,
|
| 63 |
+
"grad_norm": 4.5149941785829,
|
| 64 |
+
"learning_rate": 7.821782178217822e-06,
|
| 65 |
+
"loss": 4.9205,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.14274385408406026,
|
| 70 |
+
"grad_norm": 4.194108732375239,
|
| 71 |
+
"learning_rate": 8.811881188118812e-06,
|
| 72 |
+
"loss": 4.2318,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.1586042823156225,
|
| 77 |
+
"grad_norm": 3.8369189131135757,
|
| 78 |
+
"learning_rate": 9.801980198019804e-06,
|
| 79 |
+
"loss": 3.6352,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.17446471054718476,
|
| 84 |
+
"grad_norm": 3.7340050812724255,
|
| 85 |
+
"learning_rate": 1.0792079207920792e-05,
|
| 86 |
+
"loss": 3.2605,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.19032513877874702,
|
| 91 |
+
"grad_norm": 3.9254417022343637,
|
| 92 |
+
"learning_rate": 1.1782178217821782e-05,
|
| 93 |
+
"loss": 3.0544,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.20618556701030927,
|
| 98 |
+
"grad_norm": 4.840686832835472,
|
| 99 |
+
"learning_rate": 1.2772277227722773e-05,
|
| 100 |
+
"loss": 2.9196,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.22204599524187152,
|
| 105 |
+
"grad_norm": 13.780201154926932,
|
| 106 |
+
"learning_rate": 1.3762376237623763e-05,
|
| 107 |
+
"loss": 2.8095,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.23790642347343377,
|
| 112 |
+
"grad_norm": 6.344054558352859,
|
| 113 |
+
"learning_rate": 1.4752475247524752e-05,
|
| 114 |
+
"loss": 2.7136,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.253766851704996,
|
| 119 |
+
"grad_norm": 4.982092617175093,
|
| 120 |
+
"learning_rate": 1.5742574257425743e-05,
|
| 121 |
+
"loss": 2.6392,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2696272799365583,
|
| 126 |
+
"grad_norm": 8.162264297232719,
|
| 127 |
+
"learning_rate": 1.6732673267326735e-05,
|
| 128 |
+
"loss": 2.5676,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.2854877081681205,
|
| 133 |
+
"grad_norm": 9.068632822505595,
|
| 134 |
+
"learning_rate": 1.7722772277227723e-05,
|
| 135 |
+
"loss": 2.4886,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3013481363996828,
|
| 140 |
+
"grad_norm": 21.522422147460635,
|
| 141 |
+
"learning_rate": 1.8712871287128715e-05,
|
| 142 |
+
"loss": 2.424,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.317208564631245,
|
| 147 |
+
"grad_norm": 20.83191234754218,
|
| 148 |
+
"learning_rate": 1.9702970297029703e-05,
|
| 149 |
+
"loss": 2.3979,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3330689928628073,
|
| 154 |
+
"grad_norm": 11.966941640205993,
|
| 155 |
+
"learning_rate": 2.069306930693069e-05,
|
| 156 |
+
"loss": 2.3199,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.34892942109436953,
|
| 161 |
+
"grad_norm": 13.845469765092606,
|
| 162 |
+
"learning_rate": 2.1683168316831683e-05,
|
| 163 |
+
"loss": 2.2642,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.3647898493259318,
|
| 168 |
+
"grad_norm": 10.896989301572372,
|
| 169 |
+
"learning_rate": 2.2673267326732675e-05,
|
| 170 |
+
"loss": 2.1938,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.38065027755749403,
|
| 175 |
+
"grad_norm": 8.461441256784877,
|
| 176 |
+
"learning_rate": 2.3663366336633663e-05,
|
| 177 |
+
"loss": 2.1093,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3965107057890563,
|
| 182 |
+
"grad_norm": 15.437949044441957,
|
| 183 |
+
"learning_rate": 2.4653465346534655e-05,
|
| 184 |
+
"loss": 2.0244,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.41237113402061853,
|
| 189 |
+
"grad_norm": 6.9324747811801215,
|
| 190 |
+
"learning_rate": 2.5643564356435646e-05,
|
| 191 |
+
"loss": 1.9328,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.4282315622521808,
|
| 196 |
+
"grad_norm": 6.539929054412363,
|
| 197 |
+
"learning_rate": 2.6633663366336638e-05,
|
| 198 |
+
"loss": 1.8107,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.44409199048374304,
|
| 203 |
+
"grad_norm": 9.546138483110752,
|
| 204 |
+
"learning_rate": 2.7623762376237626e-05,
|
| 205 |
+
"loss": 1.719,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.4599524187153053,
|
| 210 |
+
"grad_norm": 6.921145854402886,
|
| 211 |
+
"learning_rate": 2.8613861386138614e-05,
|
| 212 |
+
"loss": 1.64,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.47581284694686754,
|
| 217 |
+
"grad_norm": 5.464388935314663,
|
| 218 |
+
"learning_rate": 2.9603960396039603e-05,
|
| 219 |
+
"loss": 1.5848,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.4916732751784298,
|
| 224 |
+
"grad_norm": 5.644631505546815,
|
| 225 |
+
"learning_rate": 3.0594059405940594e-05,
|
| 226 |
+
"loss": 1.515,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.507533703409992,
|
| 231 |
+
"grad_norm": 7.749393965323654,
|
| 232 |
+
"learning_rate": 3.158415841584158e-05,
|
| 233 |
+
"loss": 1.4565,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5233941316415543,
|
| 238 |
+
"grad_norm": 7.011857430438995,
|
| 239 |
+
"learning_rate": 3.257425742574258e-05,
|
| 240 |
+
"loss": 1.4069,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5392545598731165,
|
| 245 |
+
"grad_norm": 4.774306275517458,
|
| 246 |
+
"learning_rate": 3.3564356435643566e-05,
|
| 247 |
+
"loss": 1.3559,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5551149881046789,
|
| 252 |
+
"grad_norm": 4.827841991876933,
|
| 253 |
+
"learning_rate": 3.4554455445544554e-05,
|
| 254 |
+
"loss": 1.3113,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.570975416336241,
|
| 259 |
+
"grad_norm": 6.409703044827469,
|
| 260 |
+
"learning_rate": 3.554455445544554e-05,
|
| 261 |
+
"loss": 1.272,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5868358445678034,
|
| 266 |
+
"grad_norm": 7.223403220564374,
|
| 267 |
+
"learning_rate": 3.653465346534654e-05,
|
| 268 |
+
"loss": 1.2188,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.6026962727993656,
|
| 273 |
+
"grad_norm": 5.777859827384273,
|
| 274 |
+
"learning_rate": 3.7524752475247526e-05,
|
| 275 |
+
"loss": 1.1884,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6185567010309279,
|
| 280 |
+
"grad_norm": 6.4063926115137875,
|
| 281 |
+
"learning_rate": 3.851485148514852e-05,
|
| 282 |
+
"loss": 1.1581,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.63441712926249,
|
| 287 |
+
"grad_norm": 6.651129225778287,
|
| 288 |
+
"learning_rate": 3.950495049504951e-05,
|
| 289 |
+
"loss": 1.1047,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.6502775574940524,
|
| 294 |
+
"grad_norm": 5.6870828905793065,
|
| 295 |
+
"learning_rate": 4.04950495049505e-05,
|
| 296 |
+
"loss": 1.0823,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6661379857256146,
|
| 301 |
+
"grad_norm": 5.667978471416249,
|
| 302 |
+
"learning_rate": 4.1485148514851486e-05,
|
| 303 |
+
"loss": 1.0494,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6819984139571769,
|
| 308 |
+
"grad_norm": 4.237807026037709,
|
| 309 |
+
"learning_rate": 4.2475247524752474e-05,
|
| 310 |
+
"loss": 1.0155,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6978588421887391,
|
| 315 |
+
"grad_norm": 4.255500663396079,
|
| 316 |
+
"learning_rate": 4.346534653465347e-05,
|
| 317 |
+
"loss": 0.9875,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.7137192704203014,
|
| 322 |
+
"grad_norm": 5.185306420438398,
|
| 323 |
+
"learning_rate": 4.445544554455446e-05,
|
| 324 |
+
"loss": 0.9639,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7295796986518636,
|
| 329 |
+
"grad_norm": 5.272706809075813,
|
| 330 |
+
"learning_rate": 4.544554455445545e-05,
|
| 331 |
+
"loss": 0.9387,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7454401268834259,
|
| 336 |
+
"grad_norm": 6.077622808418557,
|
| 337 |
+
"learning_rate": 4.643564356435644e-05,
|
| 338 |
+
"loss": 0.921,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7613005551149881,
|
| 343 |
+
"grad_norm": 4.6482898197595235,
|
| 344 |
+
"learning_rate": 4.742574257425743e-05,
|
| 345 |
+
"loss": 0.8856,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7771609833465504,
|
| 350 |
+
"grad_norm": 5.2219845727325955,
|
| 351 |
+
"learning_rate": 4.841584158415842e-05,
|
| 352 |
+
"loss": 0.8562,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7930214115781126,
|
| 357 |
+
"grad_norm": 5.064442866871307,
|
| 358 |
+
"learning_rate": 4.9405940594059405e-05,
|
| 359 |
+
"loss": 0.8393,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.8088818398096749,
|
| 364 |
+
"grad_norm": 4.238944041111026,
|
| 365 |
+
"learning_rate": 5.03960396039604e-05,
|
| 366 |
+
"loss": 0.8133,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.8247422680412371,
|
| 371 |
+
"grad_norm": 5.703536274698318,
|
| 372 |
+
"learning_rate": 5.138613861386139e-05,
|
| 373 |
+
"loss": 0.8012,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8406026962727994,
|
| 378 |
+
"grad_norm": 5.296419707622884,
|
| 379 |
+
"learning_rate": 5.2376237623762384e-05,
|
| 380 |
+
"loss": 0.767,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8564631245043616,
|
| 385 |
+
"grad_norm": 5.699284324132107,
|
| 386 |
+
"learning_rate": 5.336633663366337e-05,
|
| 387 |
+
"loss": 0.7484,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8723235527359239,
|
| 392 |
+
"grad_norm": 6.425639332496651,
|
| 393 |
+
"learning_rate": 5.435643564356436e-05,
|
| 394 |
+
"loss": 0.7292,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8881839809674861,
|
| 399 |
+
"grad_norm": 5.045854763956847,
|
| 400 |
+
"learning_rate": 5.5346534653465355e-05,
|
| 401 |
+
"loss": 0.6976,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.9040444091990484,
|
| 406 |
+
"grad_norm": 5.0868187480748315,
|
| 407 |
+
"learning_rate": 5.633663366336634e-05,
|
| 408 |
+
"loss": 0.6719,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.9199048374306106,
|
| 413 |
+
"grad_norm": 6.760967897420004,
|
| 414 |
+
"learning_rate": 5.7326732673267325e-05,
|
| 415 |
+
"loss": 0.6573,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.9357652656621729,
|
| 420 |
+
"grad_norm": 4.427355231592498,
|
| 421 |
+
"learning_rate": 5.831683168316831e-05,
|
| 422 |
+
"loss": 0.6471,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.9516256938937351,
|
| 427 |
+
"grad_norm": 3.9839664702745634,
|
| 428 |
+
"learning_rate": 5.930693069306931e-05,
|
| 429 |
+
"loss": 0.6221,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9674861221252974,
|
| 434 |
+
"grad_norm": 3.2513703063334196,
|
| 435 |
+
"learning_rate": 6.02970297029703e-05,
|
| 436 |
+
"loss": 0.5913,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9833465503568596,
|
| 441 |
+
"grad_norm": 3.485916478339426,
|
| 442 |
+
"learning_rate": 6.128712871287129e-05,
|
| 443 |
+
"loss": 0.5775,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.9992069785884219,
|
| 448 |
+
"grad_norm": 4.08651811102019,
|
| 449 |
+
"learning_rate": 6.227722772277228e-05,
|
| 450 |
+
"loss": 0.5598,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.014274385408406,
|
| 455 |
+
"grad_norm": 3.438814479162103,
|
| 456 |
+
"learning_rate": 6.326732673267327e-05,
|
| 457 |
+
"loss": 0.5349,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.0301348136399682,
|
| 462 |
+
"grad_norm": 4.44087122057166,
|
| 463 |
+
"learning_rate": 6.425742574257426e-05,
|
| 464 |
+
"loss": 0.5194,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.0459952418715306,
|
| 469 |
+
"grad_norm": 3.5256251500611384,
|
| 470 |
+
"learning_rate": 6.524752475247526e-05,
|
| 471 |
+
"loss": 0.506,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0618556701030928,
|
| 476 |
+
"grad_norm": 3.422611919607941,
|
| 477 |
+
"learning_rate": 6.623762376237625e-05,
|
| 478 |
+
"loss": 0.4861,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.077716098334655,
|
| 483 |
+
"grad_norm": 3.133948265271591,
|
| 484 |
+
"learning_rate": 6.722772277227723e-05,
|
| 485 |
+
"loss": 0.4602,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0935765265662174,
|
| 490 |
+
"grad_norm": 3.6082454324392534,
|
| 491 |
+
"learning_rate": 6.821782178217822e-05,
|
| 492 |
+
"loss": 0.4496,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.1094369547977796,
|
| 497 |
+
"grad_norm": 2.9254028901121027,
|
| 498 |
+
"learning_rate": 6.920792079207921e-05,
|
| 499 |
+
"loss": 0.4271,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.1252973830293418,
|
| 504 |
+
"grad_norm": 2.873534663709151,
|
| 505 |
+
"learning_rate": 7.01980198019802e-05,
|
| 506 |
+
"loss": 0.408,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.141157811260904,
|
| 511 |
+
"grad_norm": 2.4876771741690122,
|
| 512 |
+
"learning_rate": 7.118811881188119e-05,
|
| 513 |
+
"loss": 0.3908,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1570182394924662,
|
| 518 |
+
"grad_norm": 2.7097832052411737,
|
| 519 |
+
"learning_rate": 7.217821782178218e-05,
|
| 520 |
+
"loss": 0.382,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.1728786677240286,
|
| 525 |
+
"grad_norm": 2.9660438369371884,
|
| 526 |
+
"learning_rate": 7.316831683168316e-05,
|
| 527 |
+
"loss": 0.3582,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1887390959555908,
|
| 532 |
+
"grad_norm": 2.5000910168051895,
|
| 533 |
+
"learning_rate": 7.415841584158417e-05,
|
| 534 |
+
"loss": 0.3497,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.204599524187153,
|
| 539 |
+
"grad_norm": 2.633285101346317,
|
| 540 |
+
"learning_rate": 7.514851485148515e-05,
|
| 541 |
+
"loss": 0.3334,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2204599524187154,
|
| 546 |
+
"grad_norm": 2.60732404875358,
|
| 547 |
+
"learning_rate": 7.613861386138614e-05,
|
| 548 |
+
"loss": 0.3184,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2363203806502776,
|
| 553 |
+
"grad_norm": 2.0629946739238005,
|
| 554 |
+
"learning_rate": 7.712871287128713e-05,
|
| 555 |
+
"loss": 0.3067,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.2521808088818398,
|
| 560 |
+
"grad_norm": 2.4100126998703253,
|
| 561 |
+
"learning_rate": 7.811881188118812e-05,
|
| 562 |
+
"loss": 0.2875,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.268041237113402,
|
| 567 |
+
"grad_norm": 2.798339817744955,
|
| 568 |
+
"learning_rate": 7.910891089108912e-05,
|
| 569 |
+
"loss": 0.277,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2839016653449642,
|
| 574 |
+
"grad_norm": 2.806131384717819,
|
| 575 |
+
"learning_rate": 8.009900990099011e-05,
|
| 576 |
+
"loss": 0.2701,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.2997620935765266,
|
| 581 |
+
"grad_norm": 1.9986997398098798,
|
| 582 |
+
"learning_rate": 8.10891089108911e-05,
|
| 583 |
+
"loss": 0.2559,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.3156225218080888,
|
| 588 |
+
"grad_norm": 2.359590651524235,
|
| 589 |
+
"learning_rate": 8.207920792079207e-05,
|
| 590 |
+
"loss": 0.2419,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.331482950039651,
|
| 595 |
+
"grad_norm": 2.331111900194689,
|
| 596 |
+
"learning_rate": 8.306930693069307e-05,
|
| 597 |
+
"loss": 0.2294,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.3473433782712134,
|
| 602 |
+
"grad_norm": 1.8818075156301755,
|
| 603 |
+
"learning_rate": 8.405940594059406e-05,
|
| 604 |
+
"loss": 0.2141,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.3632038065027756,
|
| 609 |
+
"grad_norm": 1.634132799610872,
|
| 610 |
+
"learning_rate": 8.504950495049505e-05,
|
| 611 |
+
"loss": 0.2069,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3790642347343378,
|
| 616 |
+
"grad_norm": 1.5120836080111373,
|
| 617 |
+
"learning_rate": 8.603960396039604e-05,
|
| 618 |
+
"loss": 0.1966,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3949246629659,
|
| 623 |
+
"grad_norm": 1.6982638535321994,
|
| 624 |
+
"learning_rate": 8.702970297029703e-05,
|
| 625 |
+
"loss": 0.1882,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4107850911974622,
|
| 630 |
+
"grad_norm": 1.7108735595543574,
|
| 631 |
+
"learning_rate": 8.801980198019803e-05,
|
| 632 |
+
"loss": 0.1779,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4266455194290246,
|
| 637 |
+
"grad_norm": 1.7296122709014143,
|
| 638 |
+
"learning_rate": 8.900990099009902e-05,
|
| 639 |
+
"loss": 0.1676,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.4425059476605868,
|
| 644 |
+
"grad_norm": 1.8189109064576499,
|
| 645 |
+
"learning_rate": 9e-05,
|
| 646 |
+
"loss": 0.1601,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.458366375892149,
|
| 651 |
+
"grad_norm": 2.095375586550601,
|
| 652 |
+
"learning_rate": 9.0990099009901e-05,
|
| 653 |
+
"loss": 0.1499,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.4742268041237114,
|
| 658 |
+
"grad_norm": 2.6273043704027312,
|
| 659 |
+
"learning_rate": 9.198019801980198e-05,
|
| 660 |
+
"loss": 0.1455,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.4900872323552736,
|
| 665 |
+
"grad_norm": 1.1840652988818152,
|
| 666 |
+
"learning_rate": 9.297029702970298e-05,
|
| 667 |
+
"loss": 0.1434,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5059476605868358,
|
| 672 |
+
"grad_norm": 1.102318462015262,
|
| 673 |
+
"learning_rate": 9.396039603960396e-05,
|
| 674 |
+
"loss": 0.1322,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.521808088818398,
|
| 679 |
+
"grad_norm": 1.249035070323803,
|
| 680 |
+
"learning_rate": 9.495049504950495e-05,
|
| 681 |
+
"loss": 0.1232,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.5376685170499602,
|
| 686 |
+
"grad_norm": 0.8585709469514784,
|
| 687 |
+
"learning_rate": 9.594059405940594e-05,
|
| 688 |
+
"loss": 0.1175,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.5535289452815226,
|
| 693 |
+
"grad_norm": 0.959556178843799,
|
| 694 |
+
"learning_rate": 9.693069306930694e-05,
|
| 695 |
+
"loss": 0.1119,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.5693893735130848,
|
| 700 |
+
"grad_norm": 0.9203530254113947,
|
| 701 |
+
"learning_rate": 9.792079207920793e-05,
|
| 702 |
+
"loss": 0.1083,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.5852498017446472,
|
| 707 |
+
"grad_norm": 0.8563247229051851,
|
| 708 |
+
"learning_rate": 9.891089108910891e-05,
|
| 709 |
+
"loss": 0.1009,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6011102299762094,
|
| 714 |
+
"grad_norm": 1.4810040972537613,
|
| 715 |
+
"learning_rate": 9.99009900990099e-05,
|
| 716 |
+
"loss": 0.0954,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.6169706582077716,
|
| 721 |
+
"grad_norm": 0.6927338195573747,
|
| 722 |
+
"learning_rate": 9.999975790882544e-05,
|
| 723 |
+
"loss": 0.0907,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.6328310864393338,
|
| 728 |
+
"grad_norm": 0.6353525161914889,
|
| 729 |
+
"learning_rate": 9.999892105345399e-05,
|
| 730 |
+
"loss": 0.0856,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.648691514670896,
|
| 735 |
+
"grad_norm": 1.0066185794390339,
|
| 736 |
+
"learning_rate": 9.999748645510806e-05,
|
| 737 |
+
"loss": 0.0904,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.6645519429024582,
|
| 742 |
+
"grad_norm": 0.5499242841223613,
|
| 743 |
+
"learning_rate": 9.999545413093843e-05,
|
| 744 |
+
"loss": 0.0817,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.6804123711340206,
|
| 749 |
+
"grad_norm": 0.6026955934647209,
|
| 750 |
+
"learning_rate": 9.999282410524176e-05,
|
| 751 |
+
"loss": 0.076,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6962727993655828,
|
| 756 |
+
"grad_norm": 0.5189336537822474,
|
| 757 |
+
"learning_rate": 9.998959640946033e-05,
|
| 758 |
+
"loss": 0.0706,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.7121332275971453,
|
| 763 |
+
"grad_norm": 0.5108915419488772,
|
| 764 |
+
"learning_rate": 9.998577108218162e-05,
|
| 765 |
+
"loss": 0.0685,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.7279936558287075,
|
| 770 |
+
"grad_norm": 0.4759581343462929,
|
| 771 |
+
"learning_rate": 9.998134816913788e-05,
|
| 772 |
+
"loss": 0.0642,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.7438540840602696,
|
| 777 |
+
"grad_norm": 0.44750114998963825,
|
| 778 |
+
"learning_rate": 9.997632772320553e-05,
|
| 779 |
+
"loss": 0.0615,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.7597145122918318,
|
| 784 |
+
"grad_norm": 0.4682540236602841,
|
| 785 |
+
"learning_rate": 9.997070980440462e-05,
|
| 786 |
+
"loss": 0.0587,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.775574940523394,
|
| 791 |
+
"grad_norm": 0.46100974574978826,
|
| 792 |
+
"learning_rate": 9.9964494479898e-05,
|
| 793 |
+
"loss": 0.0571,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.7914353687549562,
|
| 798 |
+
"grad_norm": 0.4802771820307019,
|
| 799 |
+
"learning_rate": 9.995768182399062e-05,
|
| 800 |
+
"loss": 0.0542,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.8072957969865187,
|
| 805 |
+
"grad_norm": 0.402517941066615,
|
| 806 |
+
"learning_rate": 9.995027191812859e-05,
|
| 807 |
+
"loss": 0.0526,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.8231562252180809,
|
| 812 |
+
"grad_norm": 0.36529022302452335,
|
| 813 |
+
"learning_rate": 9.99422648508982e-05,
|
| 814 |
+
"loss": 0.0502,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.8390166534496433,
|
| 819 |
+
"grad_norm": 0.43348415026389686,
|
| 820 |
+
"learning_rate": 9.993366071802484e-05,
|
| 821 |
+
"loss": 0.0478,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.8548770816812055,
|
| 826 |
+
"grad_norm": 0.32832066422044764,
|
| 827 |
+
"learning_rate": 9.9924459622372e-05,
|
| 828 |
+
"loss": 0.0469,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.8707375099127677,
|
| 833 |
+
"grad_norm": 0.36783738330763543,
|
| 834 |
+
"learning_rate": 9.991466167393981e-05,
|
| 835 |
+
"loss": 0.0458,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.8865979381443299,
|
| 840 |
+
"grad_norm": 2.3188302609712563,
|
| 841 |
+
"learning_rate": 9.99042669898639e-05,
|
| 842 |
+
"loss": 0.0459,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.902458366375892,
|
| 847 |
+
"grad_norm": 0.5436195698507995,
|
| 848 |
+
"learning_rate": 9.989327569441394e-05,
|
| 849 |
+
"loss": 0.0464,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.9183187946074542,
|
| 854 |
+
"grad_norm": 0.3263229164097963,
|
| 855 |
+
"learning_rate": 9.988168791899218e-05,
|
| 856 |
+
"loss": 0.0436,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.9341792228390167,
|
| 861 |
+
"grad_norm": 0.3022710473853306,
|
| 862 |
+
"learning_rate": 9.986950380213182e-05,
|
| 863 |
+
"loss": 0.0416,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.9500396510705789,
|
| 868 |
+
"grad_norm": 0.30154108928053425,
|
| 869 |
+
"learning_rate": 9.98567234894954e-05,
|
| 870 |
+
"loss": 0.0406,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.9659000793021413,
|
| 875 |
+
"grad_norm": 0.3196021212235642,
|
| 876 |
+
"learning_rate": 9.984334713387308e-05,
|
| 877 |
+
"loss": 0.0396,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.9817605075337035,
|
| 882 |
+
"grad_norm": 0.3236037547731566,
|
| 883 |
+
"learning_rate": 9.982937489518075e-05,
|
| 884 |
+
"loss": 0.0389,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.9976209357652657,
|
| 889 |
+
"grad_norm": 0.3711446498733101,
|
| 890 |
+
"learning_rate": 9.981480694045816e-05,
|
| 891 |
+
"loss": 0.038,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.0126883425852498,
|
| 896 |
+
"grad_norm": 0.3063522932516602,
|
| 897 |
+
"learning_rate": 9.979964344386692e-05,
|
| 898 |
+
"loss": 0.0359,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.028548770816812,
|
| 903 |
+
"grad_norm": 0.30361076290584094,
|
| 904 |
+
"learning_rate": 9.978388458668842e-05,
|
| 905 |
+
"loss": 0.0343,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.044409199048374,
|
| 910 |
+
"grad_norm": 0.258513333686689,
|
| 911 |
+
"learning_rate": 9.976753055732166e-05,
|
| 912 |
+
"loss": 0.0338,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.0602696272799363,
|
| 917 |
+
"grad_norm": 0.25186209885924893,
|
| 918 |
+
"learning_rate": 9.975058155128097e-05,
|
| 919 |
+
"loss": 0.0332,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.076130055511499,
|
| 924 |
+
"grad_norm": 0.2667716434965371,
|
| 925 |
+
"learning_rate": 9.97330377711937e-05,
|
| 926 |
+
"loss": 0.0325,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.091990483743061,
|
| 931 |
+
"grad_norm": 0.25831881157696407,
|
| 932 |
+
"learning_rate": 9.97148994267978e-05,
|
| 933 |
+
"loss": 0.0322,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.1078509119746234,
|
| 938 |
+
"grad_norm": 0.2422521939116205,
|
| 939 |
+
"learning_rate": 9.969616673493932e-05,
|
| 940 |
+
"loss": 0.0312,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.1237113402061856,
|
| 945 |
+
"grad_norm": 0.25167325235563537,
|
| 946 |
+
"learning_rate": 9.96768399195698e-05,
|
| 947 |
+
"loss": 0.0311,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.1395717684377478,
|
| 952 |
+
"grad_norm": 0.2648060208843995,
|
| 953 |
+
"learning_rate": 9.965691921174357e-05,
|
| 954 |
+
"loss": 0.0308,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.15543219666931,
|
| 959 |
+
"grad_norm": 0.26050946392333785,
|
| 960 |
+
"learning_rate": 9.963640484961505e-05,
|
| 961 |
+
"loss": 0.03,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.171292624900872,
|
| 966 |
+
"grad_norm": 0.2384998360178266,
|
| 967 |
+
"learning_rate": 9.961529707843583e-05,
|
| 968 |
+
"loss": 0.0297,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.187153053132435,
|
| 973 |
+
"grad_norm": 0.25739858905812224,
|
| 974 |
+
"learning_rate": 9.95935961505518e-05,
|
| 975 |
+
"loss": 0.029,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.203013481363997,
|
| 980 |
+
"grad_norm": 0.23794124732997246,
|
| 981 |
+
"learning_rate": 9.95713023254001e-05,
|
| 982 |
+
"loss": 0.0286,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.218873909595559,
|
| 987 |
+
"grad_norm": 0.2301971290512307,
|
| 988 |
+
"learning_rate": 9.954841586950606e-05,
|
| 989 |
+
"loss": 0.0286,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.2347343378271214,
|
| 994 |
+
"grad_norm": 0.2290063026328564,
|
| 995 |
+
"learning_rate": 9.95249370564799e-05,
|
| 996 |
+
"loss": 0.0276,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.2505947660586836,
|
| 1001 |
+
"grad_norm": 0.23440055002640264,
|
| 1002 |
+
"learning_rate": 9.95008661670136e-05,
|
| 1003 |
+
"loss": 0.0273,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.266455194290246,
|
| 1008 |
+
"grad_norm": 0.34512157189164844,
|
| 1009 |
+
"learning_rate": 9.947620348887749e-05,
|
| 1010 |
+
"loss": 0.0266,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.282315622521808,
|
| 1015 |
+
"grad_norm": 0.25726085487236555,
|
| 1016 |
+
"learning_rate": 9.945094931691678e-05,
|
| 1017 |
+
"loss": 0.0268,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.29817605075337,
|
| 1022 |
+
"grad_norm": 0.3637610324377928,
|
| 1023 |
+
"learning_rate": 9.942510395304805e-05,
|
| 1024 |
+
"loss": 0.0259,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.3140364789849324,
|
| 1029 |
+
"grad_norm": 1.1472903509254984,
|
| 1030 |
+
"learning_rate": 9.939866770625567e-05,
|
| 1031 |
+
"loss": 0.0275,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.329896907216495,
|
| 1036 |
+
"grad_norm": 2.5053341106388856,
|
| 1037 |
+
"learning_rate": 9.937164089258809e-05,
|
| 1038 |
+
"loss": 0.0304,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.345757335448057,
|
| 1043 |
+
"grad_norm": 1.0767870929445287,
|
| 1044 |
+
"learning_rate": 9.9344023835154e-05,
|
| 1045 |
+
"loss": 0.0454,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.3616177636796194,
|
| 1050 |
+
"grad_norm": 0.43278952742259996,
|
| 1051 |
+
"learning_rate": 9.931581686411859e-05,
|
| 1052 |
+
"loss": 0.0362,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.3774781919111816,
|
| 1057 |
+
"grad_norm": 0.2745476890309345,
|
| 1058 |
+
"learning_rate": 9.928702031669954e-05,
|
| 1059 |
+
"loss": 0.0294,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.393338620142744,
|
| 1064 |
+
"grad_norm": 0.23151635915708174,
|
| 1065 |
+
"learning_rate": 9.92576345371629e-05,
|
| 1066 |
+
"loss": 0.0267,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.409199048374306,
|
| 1071 |
+
"grad_norm": 0.2528640123475777,
|
| 1072 |
+
"learning_rate": 9.922765987681917e-05,
|
| 1073 |
+
"loss": 0.0254,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.425059476605868,
|
| 1078 |
+
"grad_norm": 0.22322528832884078,
|
| 1079 |
+
"learning_rate": 9.919709669401891e-05,
|
| 1080 |
+
"loss": 0.0241,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.440919904837431,
|
| 1085 |
+
"grad_norm": 0.25767429518855794,
|
| 1086 |
+
"learning_rate": 9.916594535414856e-05,
|
| 1087 |
+
"loss": 0.0236,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.456780333068993,
|
| 1092 |
+
"grad_norm": 0.20859014169787804,
|
| 1093 |
+
"learning_rate": 9.913420622962606e-05,
|
| 1094 |
+
"loss": 0.0231,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.472640761300555,
|
| 1099 |
+
"grad_norm": 0.20097276712386156,
|
| 1100 |
+
"learning_rate": 9.910187969989635e-05,
|
| 1101 |
+
"loss": 0.0227,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.4885011895321174,
|
| 1106 |
+
"grad_norm": 0.19502715852546781,
|
| 1107 |
+
"learning_rate": 9.90689661514269e-05,
|
| 1108 |
+
"loss": 0.0223,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.5043616177636796,
|
| 1113 |
+
"grad_norm": 0.18926852564659236,
|
| 1114 |
+
"learning_rate": 9.903546597770305e-05,
|
| 1115 |
+
"loss": 0.0217,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.520222045995242,
|
| 1120 |
+
"grad_norm": 0.18977619403937718,
|
| 1121 |
+
"learning_rate": 9.900137957922333e-05,
|
| 1122 |
+
"loss": 0.0213,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.536082474226804,
|
| 1127 |
+
"grad_norm": 0.19098675458895795,
|
| 1128 |
+
"learning_rate": 9.896670736349462e-05,
|
| 1129 |
+
"loss": 0.0208,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.551942902458366,
|
| 1134 |
+
"grad_norm": 0.2009242073740019,
|
| 1135 |
+
"learning_rate": 9.893144974502733e-05,
|
| 1136 |
+
"loss": 0.0205,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.5678033306899284,
|
| 1141 |
+
"grad_norm": 0.1882743338134484,
|
| 1142 |
+
"learning_rate": 9.889560714533042e-05,
|
| 1143 |
+
"loss": 0.0209,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.583663758921491,
|
| 1148 |
+
"grad_norm": 0.20524223857112522,
|
| 1149 |
+
"learning_rate": 9.885917999290641e-05,
|
| 1150 |
+
"loss": 0.0202,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.5995241871530532,
|
| 1155 |
+
"grad_norm": 0.17581923873322422,
|
| 1156 |
+
"learning_rate": 9.882216872324617e-05,
|
| 1157 |
+
"loss": 0.02,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.6153846153846154,
|
| 1162 |
+
"grad_norm": 0.21569051031935538,
|
| 1163 |
+
"learning_rate": 9.878457377882376e-05,
|
| 1164 |
+
"loss": 0.0201,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.6312450436161776,
|
| 1169 |
+
"grad_norm": 0.1891563070911822,
|
| 1170 |
+
"learning_rate": 9.874639560909117e-05,
|
| 1171 |
+
"loss": 0.0196,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.64710547184774,
|
| 1176 |
+
"grad_norm": 0.19956230575208342,
|
| 1177 |
+
"learning_rate": 9.870763467047293e-05,
|
| 1178 |
+
"loss": 0.019,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.662965900079302,
|
| 1183 |
+
"grad_norm": 0.19799721546110283,
|
| 1184 |
+
"learning_rate": 9.866829142636058e-05,
|
| 1185 |
+
"loss": 0.0189,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.6788263283108646,
|
| 1190 |
+
"grad_norm": 0.19531667504057218,
|
| 1191 |
+
"learning_rate": 9.862836634710724e-05,
|
| 1192 |
+
"loss": 0.0186,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.694686756542427,
|
| 1197 |
+
"grad_norm": 0.18324610519548817,
|
| 1198 |
+
"learning_rate": 9.858785991002192e-05,
|
| 1199 |
+
"loss": 0.0185,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.710547184773989,
|
| 1204 |
+
"grad_norm": 0.19516552428750053,
|
| 1205 |
+
"learning_rate": 9.854677259936383e-05,
|
| 1206 |
+
"loss": 0.0179,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.7264076130055512,
|
| 1211 |
+
"grad_norm": 0.20561324061869957,
|
| 1212 |
+
"learning_rate": 9.85051049063366e-05,
|
| 1213 |
+
"loss": 0.0182,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.7422680412371134,
|
| 1218 |
+
"grad_norm": 0.1844230160943975,
|
| 1219 |
+
"learning_rate": 9.846285732908238e-05,
|
| 1220 |
+
"loss": 0.0177,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.7581284694686756,
|
| 1225 |
+
"grad_norm": 0.1890633342236241,
|
| 1226 |
+
"learning_rate": 9.842003037267593e-05,
|
| 1227 |
+
"loss": 0.0174,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.773988897700238,
|
| 1232 |
+
"grad_norm": 0.1661243506045367,
|
| 1233 |
+
"learning_rate": 9.837662454911855e-05,
|
| 1234 |
+
"loss": 0.0171,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.7898493259318,
|
| 1239 |
+
"grad_norm": 0.1603089791676492,
|
| 1240 |
+
"learning_rate": 9.833264037733198e-05,
|
| 1241 |
+
"loss": 0.0167,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.805709754163362,
|
| 1246 |
+
"grad_norm": 0.198488901552779,
|
| 1247 |
+
"learning_rate": 9.82880783831521e-05,
|
| 1248 |
+
"loss": 0.0167,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.8215701823949244,
|
| 1253 |
+
"grad_norm": 0.17631212946503003,
|
| 1254 |
+
"learning_rate": 9.824293909932285e-05,
|
| 1255 |
+
"loss": 0.0163,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.837430610626487,
|
| 1260 |
+
"grad_norm": 0.16861160104838285,
|
| 1261 |
+
"learning_rate": 9.819722306548962e-05,
|
| 1262 |
+
"loss": 0.0161,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.8532910388580492,
|
| 1267 |
+
"grad_norm": 0.19230483320622835,
|
| 1268 |
+
"learning_rate": 9.8150930828193e-05,
|
| 1269 |
+
"loss": 0.0158,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
}
|
| 1272 |
+
],
|
| 1273 |
+
"logging_steps": 10,
|
| 1274 |
+
"max_steps": 10096,
|
| 1275 |
+
"num_input_tokens_seen": 0,
|
| 1276 |
+
"num_train_epochs": 16,
|
| 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": 129469208592384.0,
|
| 1291 |
+
"train_batch_size": 128,
|
| 1292 |
+
"trial_name": null,
|
| 1293 |
+
"trial_params": null
|
| 1294 |
+
}
|