Upload multi_student/checkpoint-1800/trainer_state.json with huggingface_hub
Browse files
multi_student/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": 0.8547008547008547,
|
| 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": 0.004748338081671415,
|
| 14 |
+
"grad_norm": 19.59556146139992,
|
| 15 |
+
"learning_rate": 8.547008547008548e-07,
|
| 16 |
+
"loss": 8.2567,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.00949667616334283,
|
| 21 |
+
"grad_norm": 18.470526689140495,
|
| 22 |
+
"learning_rate": 1.8043684710351376e-06,
|
| 23 |
+
"loss": 8.1368,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.014245014245014245,
|
| 28 |
+
"grad_norm": 9.516482160543532,
|
| 29 |
+
"learning_rate": 2.7540360873694207e-06,
|
| 30 |
+
"loss": 7.4974,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.01899335232668566,
|
| 35 |
+
"grad_norm": 4.9847966260626215,
|
| 36 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 37 |
+
"loss": 6.7488,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.023741690408357077,
|
| 42 |
+
"grad_norm": 3.2443480486318945,
|
| 43 |
+
"learning_rate": 4.653371320037987e-06,
|
| 44 |
+
"loss": 6.2614,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.02849002849002849,
|
| 49 |
+
"grad_norm": 3.038128638423009,
|
| 50 |
+
"learning_rate": 5.60303893637227e-06,
|
| 51 |
+
"loss": 5.9181,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.03323836657169991,
|
| 56 |
+
"grad_norm": 3.7165566072111664,
|
| 57 |
+
"learning_rate": 6.5527065527065525e-06,
|
| 58 |
+
"loss": 5.5221,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.03798670465337132,
|
| 63 |
+
"grad_norm": 3.8807466365954384,
|
| 64 |
+
"learning_rate": 7.5023741690408365e-06,
|
| 65 |
+
"loss": 4.9324,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.042735042735042736,
|
| 70 |
+
"grad_norm": 4.403564124821985,
|
| 71 |
+
"learning_rate": 8.452041785375119e-06,
|
| 72 |
+
"loss": 4.2636,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.04748338081671415,
|
| 77 |
+
"grad_norm": 4.690397199283881,
|
| 78 |
+
"learning_rate": 9.401709401709402e-06,
|
| 79 |
+
"loss": 3.6723,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.052231718898385564,
|
| 84 |
+
"grad_norm": 4.093575668273963,
|
| 85 |
+
"learning_rate": 1.0351377018043685e-05,
|
| 86 |
+
"loss": 3.2591,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.05698005698005698,
|
| 91 |
+
"grad_norm": 2.8440866604729447,
|
| 92 |
+
"learning_rate": 1.1301044634377968e-05,
|
| 93 |
+
"loss": 3.0232,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.06172839506172839,
|
| 98 |
+
"grad_norm": 2.3787189109612714,
|
| 99 |
+
"learning_rate": 1.2250712250712251e-05,
|
| 100 |
+
"loss": 2.8761,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.06647673314339982,
|
| 105 |
+
"grad_norm": 5.294760512204114,
|
| 106 |
+
"learning_rate": 1.3200379867046534e-05,
|
| 107 |
+
"loss": 2.7696,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.07122507122507123,
|
| 112 |
+
"grad_norm": 7.938750274935964,
|
| 113 |
+
"learning_rate": 1.4150047483380819e-05,
|
| 114 |
+
"loss": 2.6909,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.07597340930674264,
|
| 119 |
+
"grad_norm": 4.651664460732956,
|
| 120 |
+
"learning_rate": 1.50997150997151e-05,
|
| 121 |
+
"loss": 2.6095,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08072174738841406,
|
| 126 |
+
"grad_norm": 9.256919687252832,
|
| 127 |
+
"learning_rate": 1.604938271604938e-05,
|
| 128 |
+
"loss": 2.5501,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.08547008547008547,
|
| 133 |
+
"grad_norm": 12.064856380603368,
|
| 134 |
+
"learning_rate": 1.6999050332383668e-05,
|
| 135 |
+
"loss": 2.4904,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.09021842355175688,
|
| 140 |
+
"grad_norm": 12.087773624352351,
|
| 141 |
+
"learning_rate": 1.794871794871795e-05,
|
| 142 |
+
"loss": 2.433,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.0949667616334283,
|
| 147 |
+
"grad_norm": 15.164867794392874,
|
| 148 |
+
"learning_rate": 1.889838556505223e-05,
|
| 149 |
+
"loss": 2.3694,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.09971509971509972,
|
| 154 |
+
"grad_norm": 13.822253776356119,
|
| 155 |
+
"learning_rate": 1.9848053181386517e-05,
|
| 156 |
+
"loss": 2.3272,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.10446343779677113,
|
| 161 |
+
"grad_norm": 16.631555759707965,
|
| 162 |
+
"learning_rate": 2.07977207977208e-05,
|
| 163 |
+
"loss": 2.2625,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.10921177587844255,
|
| 168 |
+
"grad_norm": 16.016081398374507,
|
| 169 |
+
"learning_rate": 2.174738841405508e-05,
|
| 170 |
+
"loss": 2.1999,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.11396011396011396,
|
| 175 |
+
"grad_norm": 22.615931913843276,
|
| 176 |
+
"learning_rate": 2.2697056030389367e-05,
|
| 177 |
+
"loss": 2.1498,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.11870845204178537,
|
| 182 |
+
"grad_norm": 19.23627606954933,
|
| 183 |
+
"learning_rate": 2.364672364672365e-05,
|
| 184 |
+
"loss": 2.0798,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.12345679012345678,
|
| 189 |
+
"grad_norm": 17.696013693555628,
|
| 190 |
+
"learning_rate": 2.459639126305793e-05,
|
| 191 |
+
"loss": 1.9921,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1282051282051282,
|
| 196 |
+
"grad_norm": 16.807227837452878,
|
| 197 |
+
"learning_rate": 2.5546058879392216e-05,
|
| 198 |
+
"loss": 1.9044,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.13295346628679963,
|
| 203 |
+
"grad_norm": 11.41571060893682,
|
| 204 |
+
"learning_rate": 2.64957264957265e-05,
|
| 205 |
+
"loss": 1.8066,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.13770180436847104,
|
| 210 |
+
"grad_norm": 6.5123968942805135,
|
| 211 |
+
"learning_rate": 2.744539411206078e-05,
|
| 212 |
+
"loss": 1.7012,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.14245014245014245,
|
| 217 |
+
"grad_norm": 6.283214265860335,
|
| 218 |
+
"learning_rate": 2.839506172839506e-05,
|
| 219 |
+
"loss": 1.6008,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.14719848053181386,
|
| 224 |
+
"grad_norm": 9.079823119870198,
|
| 225 |
+
"learning_rate": 2.9344729344729345e-05,
|
| 226 |
+
"loss": 1.5131,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.15194681861348527,
|
| 231 |
+
"grad_norm": 6.449231111194866,
|
| 232 |
+
"learning_rate": 3.029439696106363e-05,
|
| 233 |
+
"loss": 1.4408,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.15669515669515668,
|
| 238 |
+
"grad_norm": 7.974966858082437,
|
| 239 |
+
"learning_rate": 3.1244064577397914e-05,
|
| 240 |
+
"loss": 1.3754,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.16144349477682812,
|
| 245 |
+
"grad_norm": 4.822536153428492,
|
| 246 |
+
"learning_rate": 3.2193732193732194e-05,
|
| 247 |
+
"loss": 1.328,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.16619183285849953,
|
| 252 |
+
"grad_norm": 5.432798377694006,
|
| 253 |
+
"learning_rate": 3.314339981006648e-05,
|
| 254 |
+
"loss": 1.2678,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.17094017094017094,
|
| 259 |
+
"grad_norm": 5.187313579614613,
|
| 260 |
+
"learning_rate": 3.409306742640076e-05,
|
| 261 |
+
"loss": 1.2157,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.17568850902184235,
|
| 266 |
+
"grad_norm": 5.499195508916456,
|
| 267 |
+
"learning_rate": 3.504273504273504e-05,
|
| 268 |
+
"loss": 1.1706,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.18043684710351376,
|
| 273 |
+
"grad_norm": 4.926464577040076,
|
| 274 |
+
"learning_rate": 3.599240265906933e-05,
|
| 275 |
+
"loss": 1.1311,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.18518518518518517,
|
| 280 |
+
"grad_norm": 5.042820415005092,
|
| 281 |
+
"learning_rate": 3.694207027540361e-05,
|
| 282 |
+
"loss": 1.1026,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.1899335232668566,
|
| 287 |
+
"grad_norm": 5.458949172817995,
|
| 288 |
+
"learning_rate": 3.789173789173789e-05,
|
| 289 |
+
"loss": 1.0523,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.19468186134852802,
|
| 294 |
+
"grad_norm": 5.673599382737787,
|
| 295 |
+
"learning_rate": 3.884140550807218e-05,
|
| 296 |
+
"loss": 1.0205,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.19943019943019943,
|
| 301 |
+
"grad_norm": 3.4632351646502926,
|
| 302 |
+
"learning_rate": 3.979107312440646e-05,
|
| 303 |
+
"loss": 0.9893,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.20417853751187084,
|
| 308 |
+
"grad_norm": 6.070644448266208,
|
| 309 |
+
"learning_rate": 4.074074074074074e-05,
|
| 310 |
+
"loss": 0.9638,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.20892687559354226,
|
| 315 |
+
"grad_norm": 4.854679094572386,
|
| 316 |
+
"learning_rate": 4.169040835707503e-05,
|
| 317 |
+
"loss": 0.9357,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.21367521367521367,
|
| 322 |
+
"grad_norm": 4.529619524145627,
|
| 323 |
+
"learning_rate": 4.264007597340931e-05,
|
| 324 |
+
"loss": 0.9076,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2184235517568851,
|
| 329 |
+
"grad_norm": 4.157674076266944,
|
| 330 |
+
"learning_rate": 4.358974358974359e-05,
|
| 331 |
+
"loss": 0.8933,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.22317188983855651,
|
| 336 |
+
"grad_norm": 6.4050560210455165,
|
| 337 |
+
"learning_rate": 4.453941120607788e-05,
|
| 338 |
+
"loss": 0.868,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.22792022792022792,
|
| 343 |
+
"grad_norm": 5.032943336700719,
|
| 344 |
+
"learning_rate": 4.548907882241216e-05,
|
| 345 |
+
"loss": 0.842,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.23266856600189934,
|
| 350 |
+
"grad_norm": 6.013522469421973,
|
| 351 |
+
"learning_rate": 4.643874643874644e-05,
|
| 352 |
+
"loss": 0.8071,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.23741690408357075,
|
| 357 |
+
"grad_norm": 5.110412155585145,
|
| 358 |
+
"learning_rate": 4.738841405508073e-05,
|
| 359 |
+
"loss": 0.7809,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.24216524216524216,
|
| 364 |
+
"grad_norm": 5.7836833567564305,
|
| 365 |
+
"learning_rate": 4.833808167141501e-05,
|
| 366 |
+
"loss": 0.7605,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.24691358024691357,
|
| 371 |
+
"grad_norm": 4.319290309225294,
|
| 372 |
+
"learning_rate": 4.928774928774929e-05,
|
| 373 |
+
"loss": 0.7286,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.251661918328585,
|
| 378 |
+
"grad_norm": 5.40178280723504,
|
| 379 |
+
"learning_rate": 5.0237416904083576e-05,
|
| 380 |
+
"loss": 0.7105,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.2564102564102564,
|
| 385 |
+
"grad_norm": 4.71777919695988,
|
| 386 |
+
"learning_rate": 5.118708452041786e-05,
|
| 387 |
+
"loss": 0.6931,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.2611585944919278,
|
| 392 |
+
"grad_norm": 5.6291823771994665,
|
| 393 |
+
"learning_rate": 5.213675213675214e-05,
|
| 394 |
+
"loss": 0.6759,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.26590693257359926,
|
| 399 |
+
"grad_norm": 5.5347321476291835,
|
| 400 |
+
"learning_rate": 5.308641975308642e-05,
|
| 401 |
+
"loss": 0.6491,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2706552706552707,
|
| 406 |
+
"grad_norm": 5.257360790177073,
|
| 407 |
+
"learning_rate": 5.403608736942071e-05,
|
| 408 |
+
"loss": 0.6285,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2754036087369421,
|
| 413 |
+
"grad_norm": 5.511202010659421,
|
| 414 |
+
"learning_rate": 5.498575498575499e-05,
|
| 415 |
+
"loss": 0.6012,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.2801519468186135,
|
| 420 |
+
"grad_norm": 5.380510479349013,
|
| 421 |
+
"learning_rate": 5.5935422602089274e-05,
|
| 422 |
+
"loss": 0.5852,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.2849002849002849,
|
| 427 |
+
"grad_norm": 3.6898979087460755,
|
| 428 |
+
"learning_rate": 5.6885090218423554e-05,
|
| 429 |
+
"loss": 0.5671,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.2896486229819563,
|
| 434 |
+
"grad_norm": 4.727348568866969,
|
| 435 |
+
"learning_rate": 5.7834757834757834e-05,
|
| 436 |
+
"loss": 0.5514,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.2943969610636277,
|
| 441 |
+
"grad_norm": 4.668206154804636,
|
| 442 |
+
"learning_rate": 5.878442545109212e-05,
|
| 443 |
+
"loss": 0.5278,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.29914529914529914,
|
| 448 |
+
"grad_norm": 3.5726965671314566,
|
| 449 |
+
"learning_rate": 5.97340930674264e-05,
|
| 450 |
+
"loss": 0.5064,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.30389363722697055,
|
| 455 |
+
"grad_norm": 4.046849737286265,
|
| 456 |
+
"learning_rate": 6.068376068376068e-05,
|
| 457 |
+
"loss": 0.4812,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.30864197530864196,
|
| 462 |
+
"grad_norm": 4.042224830358077,
|
| 463 |
+
"learning_rate": 6.163342830009497e-05,
|
| 464 |
+
"loss": 0.4733,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.31339031339031337,
|
| 469 |
+
"grad_norm": 2.7104179759082787,
|
| 470 |
+
"learning_rate": 6.258309591642925e-05,
|
| 471 |
+
"loss": 0.4485,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3181386514719848,
|
| 476 |
+
"grad_norm": 3.530227033382703,
|
| 477 |
+
"learning_rate": 6.353276353276355e-05,
|
| 478 |
+
"loss": 0.436,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.32288698955365625,
|
| 483 |
+
"grad_norm": 4.347310302223748,
|
| 484 |
+
"learning_rate": 6.448243114909783e-05,
|
| 485 |
+
"loss": 0.4151,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.32763532763532766,
|
| 490 |
+
"grad_norm": 2.522321016680872,
|
| 491 |
+
"learning_rate": 6.54320987654321e-05,
|
| 492 |
+
"loss": 0.4114,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.33238366571699907,
|
| 497 |
+
"grad_norm": 3.7746726650422113,
|
| 498 |
+
"learning_rate": 6.638176638176638e-05,
|
| 499 |
+
"loss": 0.3887,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3371320037986705,
|
| 504 |
+
"grad_norm": 2.907400275302209,
|
| 505 |
+
"learning_rate": 6.733143399810066e-05,
|
| 506 |
+
"loss": 0.3768,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3418803418803419,
|
| 511 |
+
"grad_norm": 2.109887545655247,
|
| 512 |
+
"learning_rate": 6.828110161443494e-05,
|
| 513 |
+
"loss": 0.3621,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.3466286799620133,
|
| 518 |
+
"grad_norm": 2.597481793691592,
|
| 519 |
+
"learning_rate": 6.923076923076924e-05,
|
| 520 |
+
"loss": 0.3516,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3513770180436847,
|
| 525 |
+
"grad_norm": 2.6365690488571807,
|
| 526 |
+
"learning_rate": 7.018043684710352e-05,
|
| 527 |
+
"loss": 0.3339,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.3561253561253561,
|
| 532 |
+
"grad_norm": 2.8201086376128943,
|
| 533 |
+
"learning_rate": 7.11301044634378e-05,
|
| 534 |
+
"loss": 0.3198,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.36087369420702753,
|
| 539 |
+
"grad_norm": 2.396568559036651,
|
| 540 |
+
"learning_rate": 7.207977207977208e-05,
|
| 541 |
+
"loss": 0.3114,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.36562203228869894,
|
| 546 |
+
"grad_norm": 1.9490881875653105,
|
| 547 |
+
"learning_rate": 7.302943969610636e-05,
|
| 548 |
+
"loss": 0.295,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.37037037037037035,
|
| 553 |
+
"grad_norm": 3.0763488368388328,
|
| 554 |
+
"learning_rate": 7.397910731244065e-05,
|
| 555 |
+
"loss": 0.2809,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.37511870845204176,
|
| 560 |
+
"grad_norm": 2.094402852150148,
|
| 561 |
+
"learning_rate": 7.492877492877494e-05,
|
| 562 |
+
"loss": 0.2701,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3798670465337132,
|
| 567 |
+
"grad_norm": 2.274287566601407,
|
| 568 |
+
"learning_rate": 7.587844254510922e-05,
|
| 569 |
+
"loss": 0.2568,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.38461538461538464,
|
| 574 |
+
"grad_norm": 1.9068766015691088,
|
| 575 |
+
"learning_rate": 7.68281101614435e-05,
|
| 576 |
+
"loss": 0.2463,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.38936372269705605,
|
| 581 |
+
"grad_norm": 1.9600446391982316,
|
| 582 |
+
"learning_rate": 7.777777777777778e-05,
|
| 583 |
+
"loss": 0.2363,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.39411206077872746,
|
| 588 |
+
"grad_norm": 1.6629965521131573,
|
| 589 |
+
"learning_rate": 7.872744539411206e-05,
|
| 590 |
+
"loss": 0.2234,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.39886039886039887,
|
| 595 |
+
"grad_norm": 2.875635682058353,
|
| 596 |
+
"learning_rate": 7.967711301044634e-05,
|
| 597 |
+
"loss": 0.2145,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.4036087369420703,
|
| 602 |
+
"grad_norm": 1.7911106686775673,
|
| 603 |
+
"learning_rate": 8.062678062678063e-05,
|
| 604 |
+
"loss": 0.2081,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4083570750237417,
|
| 609 |
+
"grad_norm": 1.8738797366152795,
|
| 610 |
+
"learning_rate": 8.157644824311491e-05,
|
| 611 |
+
"loss": 0.1934,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.4131054131054131,
|
| 616 |
+
"grad_norm": 1.6699636137939626,
|
| 617 |
+
"learning_rate": 8.25261158594492e-05,
|
| 618 |
+
"loss": 0.1845,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4178537511870845,
|
| 623 |
+
"grad_norm": 2.1263011843663304,
|
| 624 |
+
"learning_rate": 8.347578347578347e-05,
|
| 625 |
+
"loss": 0.1815,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4226020892687559,
|
| 630 |
+
"grad_norm": 1.637253223509006,
|
| 631 |
+
"learning_rate": 8.442545109211775e-05,
|
| 632 |
+
"loss": 0.1677,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.42735042735042733,
|
| 637 |
+
"grad_norm": 1.6085444053060045,
|
| 638 |
+
"learning_rate": 8.537511870845205e-05,
|
| 639 |
+
"loss": 0.1617,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.43209876543209874,
|
| 644 |
+
"grad_norm": 1.3811533428380767,
|
| 645 |
+
"learning_rate": 8.632478632478634e-05,
|
| 646 |
+
"loss": 0.1536,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.4368471035137702,
|
| 651 |
+
"grad_norm": 1.5957142791343428,
|
| 652 |
+
"learning_rate": 8.727445394112062e-05,
|
| 653 |
+
"loss": 0.1471,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4415954415954416,
|
| 658 |
+
"grad_norm": 2.077322656782395,
|
| 659 |
+
"learning_rate": 8.82241215574549e-05,
|
| 660 |
+
"loss": 0.1367,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.44634377967711303,
|
| 665 |
+
"grad_norm": 1.3863777124344219,
|
| 666 |
+
"learning_rate": 8.917378917378918e-05,
|
| 667 |
+
"loss": 0.1305,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.45109211775878444,
|
| 672 |
+
"grad_norm": 1.379373447849942,
|
| 673 |
+
"learning_rate": 9.012345679012346e-05,
|
| 674 |
+
"loss": 0.1254,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.45584045584045585,
|
| 679 |
+
"grad_norm": 1.0225553190768473,
|
| 680 |
+
"learning_rate": 9.107312440645774e-05,
|
| 681 |
+
"loss": 0.1184,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.46058879392212726,
|
| 686 |
+
"grad_norm": 1.4701065784223963,
|
| 687 |
+
"learning_rate": 9.202279202279203e-05,
|
| 688 |
+
"loss": 0.1142,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.46533713200379867,
|
| 693 |
+
"grad_norm": 1.1083829513415684,
|
| 694 |
+
"learning_rate": 9.297245963912631e-05,
|
| 695 |
+
"loss": 0.1083,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.4700854700854701,
|
| 700 |
+
"grad_norm": 0.9116224133866325,
|
| 701 |
+
"learning_rate": 9.392212725546059e-05,
|
| 702 |
+
"loss": 0.1015,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.4748338081671415,
|
| 707 |
+
"grad_norm": 1.0732215966409744,
|
| 708 |
+
"learning_rate": 9.487179487179487e-05,
|
| 709 |
+
"loss": 0.0981,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.4795821462488129,
|
| 714 |
+
"grad_norm": 1.4448532002931684,
|
| 715 |
+
"learning_rate": 9.582146248812915e-05,
|
| 716 |
+
"loss": 0.0955,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.4843304843304843,
|
| 721 |
+
"grad_norm": 0.9727037977093907,
|
| 722 |
+
"learning_rate": 9.677113010446344e-05,
|
| 723 |
+
"loss": 0.0919,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.4890788224121557,
|
| 728 |
+
"grad_norm": 0.6538136654859616,
|
| 729 |
+
"learning_rate": 9.772079772079774e-05,
|
| 730 |
+
"loss": 0.0843,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.49382716049382713,
|
| 735 |
+
"grad_norm": 0.5485567949471478,
|
| 736 |
+
"learning_rate": 9.867046533713202e-05,
|
| 737 |
+
"loss": 0.0787,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.4985754985754986,
|
| 742 |
+
"grad_norm": 0.8261306762820796,
|
| 743 |
+
"learning_rate": 9.96201329534663e-05,
|
| 744 |
+
"loss": 0.0767,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.50332383665717,
|
| 749 |
+
"grad_norm": 0.639974727003365,
|
| 750 |
+
"learning_rate": 9.999990109907572e-05,
|
| 751 |
+
"loss": 0.0716,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5080721747388414,
|
| 756 |
+
"grad_norm": 0.5475316216026064,
|
| 757 |
+
"learning_rate": 9.999929670595522e-05,
|
| 758 |
+
"loss": 0.0686,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5128205128205128,
|
| 763 |
+
"grad_norm": 0.5409377296081475,
|
| 764 |
+
"learning_rate": 9.99981428713058e-05,
|
| 765 |
+
"loss": 0.0652,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5175688509021842,
|
| 770 |
+
"grad_norm": 0.7449946627107062,
|
| 771 |
+
"learning_rate": 9.999643960780693e-05,
|
| 772 |
+
"loss": 0.0612,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5223171889838556,
|
| 777 |
+
"grad_norm": 0.5311754725201671,
|
| 778 |
+
"learning_rate": 9.999418693417573e-05,
|
| 779 |
+
"loss": 0.059,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5270655270655271,
|
| 784 |
+
"grad_norm": 0.4708957925911922,
|
| 785 |
+
"learning_rate": 9.999138487516681e-05,
|
| 786 |
+
"loss": 0.0566,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5318138651471985,
|
| 791 |
+
"grad_norm": 0.5784265250336279,
|
| 792 |
+
"learning_rate": 9.998803346157195e-05,
|
| 793 |
+
"loss": 0.0564,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5365622032288699,
|
| 798 |
+
"grad_norm": 0.5168678159856918,
|
| 799 |
+
"learning_rate": 9.99841327302198e-05,
|
| 800 |
+
"loss": 0.0546,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5413105413105413,
|
| 805 |
+
"grad_norm": 0.32193609799540956,
|
| 806 |
+
"learning_rate": 9.997968272397542e-05,
|
| 807 |
+
"loss": 0.0512,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5460588793922128,
|
| 812 |
+
"grad_norm": 0.44104880195699325,
|
| 813 |
+
"learning_rate": 9.997468349173985e-05,
|
| 814 |
+
"loss": 0.0487,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5508072174738842,
|
| 819 |
+
"grad_norm": 0.4133398175893821,
|
| 820 |
+
"learning_rate": 9.99691350884496e-05,
|
| 821 |
+
"loss": 0.0477,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5555555555555556,
|
| 826 |
+
"grad_norm": 0.37749172837304285,
|
| 827 |
+
"learning_rate": 9.996303757507602e-05,
|
| 828 |
+
"loss": 0.0462,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.560303893637227,
|
| 833 |
+
"grad_norm": 0.4428973313011144,
|
| 834 |
+
"learning_rate": 9.995639101862453e-05,
|
| 835 |
+
"loss": 0.0458,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5650522317188984,
|
| 840 |
+
"grad_norm": 0.3162890941987217,
|
| 841 |
+
"learning_rate": 9.994919549213408e-05,
|
| 842 |
+
"loss": 0.0441,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5698005698005698,
|
| 847 |
+
"grad_norm": 0.3125288250812194,
|
| 848 |
+
"learning_rate": 9.99414510746762e-05,
|
| 849 |
+
"loss": 0.0425,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.5745489078822412,
|
| 854 |
+
"grad_norm": 0.3703137658941141,
|
| 855 |
+
"learning_rate": 9.993315785135416e-05,
|
| 856 |
+
"loss": 0.0419,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.5792972459639126,
|
| 861 |
+
"grad_norm": 0.31262957519265083,
|
| 862 |
+
"learning_rate": 9.99243159133021e-05,
|
| 863 |
+
"loss": 0.0406,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.584045584045584,
|
| 868 |
+
"grad_norm": 0.29796796936840436,
|
| 869 |
+
"learning_rate": 9.991492535768392e-05,
|
| 870 |
+
"loss": 0.0392,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.5887939221272555,
|
| 875 |
+
"grad_norm": 0.3318213355802759,
|
| 876 |
+
"learning_rate": 9.99049862876923e-05,
|
| 877 |
+
"loss": 0.0386,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.5935422602089269,
|
| 882 |
+
"grad_norm": 0.34217083103520574,
|
| 883 |
+
"learning_rate": 9.989449881254754e-05,
|
| 884 |
+
"loss": 0.0382,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.5982905982905983,
|
| 889 |
+
"grad_norm": 0.283802792484964,
|
| 890 |
+
"learning_rate": 9.988346304749635e-05,
|
| 891 |
+
"loss": 0.0366,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6030389363722697,
|
| 896 |
+
"grad_norm": 0.29414736488607734,
|
| 897 |
+
"learning_rate": 9.987187911381059e-05,
|
| 898 |
+
"loss": 0.0359,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6077872744539411,
|
| 903 |
+
"grad_norm": 0.2751603131267791,
|
| 904 |
+
"learning_rate": 9.985974713878592e-05,
|
| 905 |
+
"loss": 0.0347,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6125356125356125,
|
| 910 |
+
"grad_norm": 0.2925423079705012,
|
| 911 |
+
"learning_rate": 9.984706725574044e-05,
|
| 912 |
+
"loss": 0.0347,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6172839506172839,
|
| 917 |
+
"grad_norm": 0.27725866932297377,
|
| 918 |
+
"learning_rate": 9.983383960401319e-05,
|
| 919 |
+
"loss": 0.0336,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.6220322886989553,
|
| 924 |
+
"grad_norm": 0.7941618027767806,
|
| 925 |
+
"learning_rate": 9.982006432896266e-05,
|
| 926 |
+
"loss": 0.035,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.6267806267806267,
|
| 931 |
+
"grad_norm": 1.6840351403640696,
|
| 932 |
+
"learning_rate": 9.980574158196512e-05,
|
| 933 |
+
"loss": 0.0524,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.6315289648622981,
|
| 938 |
+
"grad_norm": 1.3554091307998244,
|
| 939 |
+
"learning_rate": 9.979087152041304e-05,
|
| 940 |
+
"loss": 0.0713,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6362773029439696,
|
| 945 |
+
"grad_norm": 0.36897356801036785,
|
| 946 |
+
"learning_rate": 9.977545430771332e-05,
|
| 947 |
+
"loss": 0.0417,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6410256410256411,
|
| 952 |
+
"grad_norm": 0.2904292474002773,
|
| 953 |
+
"learning_rate": 9.975949011328542e-05,
|
| 954 |
+
"loss": 0.0347,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6457739791073125,
|
| 959 |
+
"grad_norm": 0.2776192308462496,
|
| 960 |
+
"learning_rate": 9.97429791125597e-05,
|
| 961 |
+
"loss": 0.0322,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6505223171889839,
|
| 966 |
+
"grad_norm": 0.2598684841997891,
|
| 967 |
+
"learning_rate": 9.972592148697528e-05,
|
| 968 |
+
"loss": 0.0311,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6552706552706553,
|
| 973 |
+
"grad_norm": 0.23868060621698645,
|
| 974 |
+
"learning_rate": 9.970831742397815e-05,
|
| 975 |
+
"loss": 0.0302,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.6600189933523267,
|
| 980 |
+
"grad_norm": 0.22695749001395263,
|
| 981 |
+
"learning_rate": 9.969016711701908e-05,
|
| 982 |
+
"loss": 0.0295,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.6647673314339981,
|
| 987 |
+
"grad_norm": 0.23540766289324025,
|
| 988 |
+
"learning_rate": 9.967147076555157e-05,
|
| 989 |
+
"loss": 0.0289,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.6695156695156695,
|
| 994 |
+
"grad_norm": 0.22325211499568143,
|
| 995 |
+
"learning_rate": 9.965222857502951e-05,
|
| 996 |
+
"loss": 0.0277,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.674264007597341,
|
| 1001 |
+
"grad_norm": 0.20341994994410992,
|
| 1002 |
+
"learning_rate": 9.963244075690505e-05,
|
| 1003 |
+
"loss": 0.0267,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.6790123456790124,
|
| 1008 |
+
"grad_norm": 0.2082789878363781,
|
| 1009 |
+
"learning_rate": 9.961210752862626e-05,
|
| 1010 |
+
"loss": 0.0266,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.6837606837606838,
|
| 1015 |
+
"grad_norm": 0.1834817910047751,
|
| 1016 |
+
"learning_rate": 9.959122911363463e-05,
|
| 1017 |
+
"loss": 0.0258,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.6885090218423552,
|
| 1022 |
+
"grad_norm": 0.20370040107298587,
|
| 1023 |
+
"learning_rate": 9.956980574136281e-05,
|
| 1024 |
+
"loss": 0.0249,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.6932573599240266,
|
| 1029 |
+
"grad_norm": 0.20761457959687116,
|
| 1030 |
+
"learning_rate": 9.954783764723188e-05,
|
| 1031 |
+
"loss": 0.0245,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.698005698005698,
|
| 1036 |
+
"grad_norm": 0.19324113426502953,
|
| 1037 |
+
"learning_rate": 9.952532507264891e-05,
|
| 1038 |
+
"loss": 0.0235,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.7027540360873694,
|
| 1043 |
+
"grad_norm": 0.22152991183679277,
|
| 1044 |
+
"learning_rate": 9.950226826500426e-05,
|
| 1045 |
+
"loss": 0.0235,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7075023741690408,
|
| 1050 |
+
"grad_norm": 0.18291746764808686,
|
| 1051 |
+
"learning_rate": 9.947866747766882e-05,
|
| 1052 |
+
"loss": 0.0225,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.7122507122507122,
|
| 1057 |
+
"grad_norm": 0.18692740414379616,
|
| 1058 |
+
"learning_rate": 9.945452296999131e-05,
|
| 1059 |
+
"loss": 0.0221,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.7169990503323836,
|
| 1064 |
+
"grad_norm": 0.20635010095393333,
|
| 1065 |
+
"learning_rate": 9.942983500729536e-05,
|
| 1066 |
+
"loss": 0.0215,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.7217473884140551,
|
| 1071 |
+
"grad_norm": 0.22328566441681175,
|
| 1072 |
+
"learning_rate": 9.940460386087667e-05,
|
| 1073 |
+
"loss": 0.0238,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7264957264957265,
|
| 1078 |
+
"grad_norm": 0.22969926818749956,
|
| 1079 |
+
"learning_rate": 9.937882980799986e-05,
|
| 1080 |
+
"loss": 0.0238,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.7312440645773979,
|
| 1085 |
+
"grad_norm": 0.1891326951743918,
|
| 1086 |
+
"learning_rate": 9.935251313189564e-05,
|
| 1087 |
+
"loss": 0.0212,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7359924026590693,
|
| 1092 |
+
"grad_norm": 0.18894118096050505,
|
| 1093 |
+
"learning_rate": 9.93256541217576e-05,
|
| 1094 |
+
"loss": 0.0201,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7407407407407407,
|
| 1099 |
+
"grad_norm": 0.16839965478347466,
|
| 1100 |
+
"learning_rate": 9.929825307273898e-05,
|
| 1101 |
+
"loss": 0.0194,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.7454890788224121,
|
| 1106 |
+
"grad_norm": 0.1651031818463282,
|
| 1107 |
+
"learning_rate": 9.927031028594951e-05,
|
| 1108 |
+
"loss": 0.0188,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.7502374169040835,
|
| 1113 |
+
"grad_norm": 0.17864976798516016,
|
| 1114 |
+
"learning_rate": 9.924182606845204e-05,
|
| 1115 |
+
"loss": 0.0183,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.7549857549857549,
|
| 1120 |
+
"grad_norm": 0.15081354129337066,
|
| 1121 |
+
"learning_rate": 9.921280073325922e-05,
|
| 1122 |
+
"loss": 0.0177,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.7597340930674265,
|
| 1127 |
+
"grad_norm": 0.1492823322050948,
|
| 1128 |
+
"learning_rate": 9.918323459933005e-05,
|
| 1129 |
+
"loss": 0.0171,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.7644824311490979,
|
| 1134 |
+
"grad_norm": 0.19617624499767475,
|
| 1135 |
+
"learning_rate": 9.915312799156632e-05,
|
| 1136 |
+
"loss": 0.0168,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.7692307692307693,
|
| 1141 |
+
"grad_norm": 0.19442787986064758,
|
| 1142 |
+
"learning_rate": 9.91224812408091e-05,
|
| 1143 |
+
"loss": 0.0166,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.7739791073124407,
|
| 1148 |
+
"grad_norm": 0.15196414738372457,
|
| 1149 |
+
"learning_rate": 9.909129468383506e-05,
|
| 1150 |
+
"loss": 0.0161,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.7787274453941121,
|
| 1155 |
+
"grad_norm": 0.15107912995719286,
|
| 1156 |
+
"learning_rate": 9.905956866335286e-05,
|
| 1157 |
+
"loss": 0.0156,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.7834757834757835,
|
| 1162 |
+
"grad_norm": 0.18162048926922258,
|
| 1163 |
+
"learning_rate": 9.90273035279992e-05,
|
| 1164 |
+
"loss": 0.0153,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.7882241215574549,
|
| 1169 |
+
"grad_norm": 0.17872639214115862,
|
| 1170 |
+
"learning_rate": 9.899449963233523e-05,
|
| 1171 |
+
"loss": 0.015,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.7929724596391263,
|
| 1176 |
+
"grad_norm": 0.15227646692118818,
|
| 1177 |
+
"learning_rate": 9.896115733684244e-05,
|
| 1178 |
+
"loss": 0.0141,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.7977207977207977,
|
| 1183 |
+
"grad_norm": 0.14676815381097183,
|
| 1184 |
+
"learning_rate": 9.892727700791881e-05,
|
| 1185 |
+
"loss": 0.0143,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.8024691358024691,
|
| 1190 |
+
"grad_norm": 0.1461341101081371,
|
| 1191 |
+
"learning_rate": 9.889285901787482e-05,
|
| 1192 |
+
"loss": 0.0139,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.8072174738841406,
|
| 1197 |
+
"grad_norm": 0.144084266582809,
|
| 1198 |
+
"learning_rate": 9.885790374492921e-05,
|
| 1199 |
+
"loss": 0.0133,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.811965811965812,
|
| 1204 |
+
"grad_norm": 0.1524301640273114,
|
| 1205 |
+
"learning_rate": 9.882241157320497e-05,
|
| 1206 |
+
"loss": 0.0136,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.8167141500474834,
|
| 1211 |
+
"grad_norm": 0.142713391750164,
|
| 1212 |
+
"learning_rate": 9.878638289272507e-05,
|
| 1213 |
+
"loss": 0.013,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8214624881291548,
|
| 1218 |
+
"grad_norm": 0.1684868465091661,
|
| 1219 |
+
"learning_rate": 9.874981809940808e-05,
|
| 1220 |
+
"loss": 0.0133,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.8262108262108262,
|
| 1225 |
+
"grad_norm": 0.1495843385262352,
|
| 1226 |
+
"learning_rate": 9.871271759506404e-05,
|
| 1227 |
+
"loss": 0.013,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.8309591642924976,
|
| 1232 |
+
"grad_norm": 0.16249545483594732,
|
| 1233 |
+
"learning_rate": 9.86750817873898e-05,
|
| 1234 |
+
"loss": 0.0123,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.835707502374169,
|
| 1239 |
+
"grad_norm": 0.1484440363408758,
|
| 1240 |
+
"learning_rate": 9.863691108996471e-05,
|
| 1241 |
+
"loss": 0.0123,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.8404558404558404,
|
| 1246 |
+
"grad_norm": 0.13923187737177664,
|
| 1247 |
+
"learning_rate": 9.859820592224599e-05,
|
| 1248 |
+
"loss": 0.0118,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.8452041785375118,
|
| 1253 |
+
"grad_norm": 0.13420351938535,
|
| 1254 |
+
"learning_rate": 9.855896670956415e-05,
|
| 1255 |
+
"loss": 0.0115,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.8499525166191833,
|
| 1260 |
+
"grad_norm": 0.13630062969945986,
|
| 1261 |
+
"learning_rate": 9.851919388311832e-05,
|
| 1262 |
+
"loss": 0.0114,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.8547008547008547,
|
| 1267 |
+
"grad_norm": 0.1246353289517896,
|
| 1268 |
+
"learning_rate": 9.84788878799715e-05,
|
| 1269 |
+
"loss": 0.0112,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
}
|
| 1272 |
+
],
|
| 1273 |
+
"logging_steps": 10,
|
| 1274 |
+
"max_steps": 10530,
|
| 1275 |
+
"num_input_tokens_seen": 0,
|
| 1276 |
+
"num_train_epochs": 5,
|
| 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": 166234778763264.0,
|
| 1291 |
+
"train_batch_size": 128,
|
| 1292 |
+
"trial_name": null,
|
| 1293 |
+
"trial_params": null
|
| 1294 |
+
}
|