End of training
Browse files- trainer_state.json +1317 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1317 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 17094,
|
| 3 |
+
"best_metric": 0.9845099349219312,
|
| 4 |
+
"best_model_checkpoint": "./xlmr-language-identification/checkpoint-17094",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 2500,
|
| 7 |
+
"global_step": 17094,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011701304695473546,
|
| 14 |
+
"grad_norm": 0.9905334711074829,
|
| 15 |
+
"learning_rate": 4.971042471042471e-05,
|
| 16 |
+
"loss": 16.156219482421875,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023402609390947092,
|
| 21 |
+
"grad_norm": 1.069915533065796,
|
| 22 |
+
"learning_rate": 4.941792441792442e-05,
|
| 23 |
+
"loss": 5.045557556152343,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.035103914086420634,
|
| 28 |
+
"grad_norm": 1.8720062971115112,
|
| 29 |
+
"learning_rate": 4.912542412542413e-05,
|
| 30 |
+
"loss": 4.602024841308594,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.046805218781894184,
|
| 35 |
+
"grad_norm": 2.246952772140503,
|
| 36 |
+
"learning_rate": 4.883292383292383e-05,
|
| 37 |
+
"loss": 2.9484735107421876,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.058506523477367726,
|
| 42 |
+
"grad_norm": 1.7273783683776855,
|
| 43 |
+
"learning_rate": 4.8540423540423544e-05,
|
| 44 |
+
"loss": 1.8604541015625,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07020782817284127,
|
| 49 |
+
"grad_norm": 1.871302843093872,
|
| 50 |
+
"learning_rate": 4.824792324792325e-05,
|
| 51 |
+
"loss": 1.2988777160644531,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.08190913286831482,
|
| 56 |
+
"grad_norm": 1.9515539407730103,
|
| 57 |
+
"learning_rate": 4.7955422955422954e-05,
|
| 58 |
+
"loss": 0.9935964965820312,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09361043756378837,
|
| 63 |
+
"grad_norm": 1.469983696937561,
|
| 64 |
+
"learning_rate": 4.7662922662922666e-05,
|
| 65 |
+
"loss": 0.803764877319336,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1053117422592619,
|
| 70 |
+
"grad_norm": 1.6788742542266846,
|
| 71 |
+
"learning_rate": 4.737042237042237e-05,
|
| 72 |
+
"loss": 0.6818081665039063,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11701304695473545,
|
| 77 |
+
"grad_norm": 1.4318715333938599,
|
| 78 |
+
"learning_rate": 4.707792207792208e-05,
|
| 79 |
+
"loss": 0.5820805740356445,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.128714351650209,
|
| 84 |
+
"grad_norm": 1.9161852598190308,
|
| 85 |
+
"learning_rate": 4.678542178542179e-05,
|
| 86 |
+
"loss": 0.5121846389770508,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.14041565634568254,
|
| 91 |
+
"grad_norm": 1.6554639339447021,
|
| 92 |
+
"learning_rate": 4.6492921492921494e-05,
|
| 93 |
+
"loss": 0.4561897277832031,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1521169610411561,
|
| 98 |
+
"grad_norm": 1.839687466621399,
|
| 99 |
+
"learning_rate": 4.62004212004212e-05,
|
| 100 |
+
"loss": 0.4188547134399414,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.16381826573662964,
|
| 105 |
+
"grad_norm": 1.735112190246582,
|
| 106 |
+
"learning_rate": 4.590792090792091e-05,
|
| 107 |
+
"loss": 0.3874264144897461,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.17551957043210317,
|
| 112 |
+
"grad_norm": 2.1364924907684326,
|
| 113 |
+
"learning_rate": 4.5615420615420616e-05,
|
| 114 |
+
"loss": 0.36660758972167967,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.18722087512757674,
|
| 119 |
+
"grad_norm": 1.5190038681030273,
|
| 120 |
+
"learning_rate": 4.532292032292033e-05,
|
| 121 |
+
"loss": 0.3320538330078125,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.19892217982305027,
|
| 126 |
+
"grad_norm": 1.4698094129562378,
|
| 127 |
+
"learning_rate": 4.503042003042003e-05,
|
| 128 |
+
"loss": 0.3180769729614258,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.2106234845185238,
|
| 133 |
+
"grad_norm": 1.4362446069717407,
|
| 134 |
+
"learning_rate": 4.4737919737919745e-05,
|
| 135 |
+
"loss": 0.3039104461669922,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.22232478921399737,
|
| 140 |
+
"grad_norm": 1.5884244441986084,
|
| 141 |
+
"learning_rate": 4.444541944541944e-05,
|
| 142 |
+
"loss": 0.2853180694580078,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2340260939094709,
|
| 147 |
+
"grad_norm": 1.14271879196167,
|
| 148 |
+
"learning_rate": 4.4152919152919155e-05,
|
| 149 |
+
"loss": 0.2749899673461914,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.24572739860494444,
|
| 154 |
+
"grad_norm": 1.3998445272445679,
|
| 155 |
+
"learning_rate": 4.386041886041886e-05,
|
| 156 |
+
"loss": 0.2584847450256348,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.257428703300418,
|
| 161 |
+
"grad_norm": 1.126080870628357,
|
| 162 |
+
"learning_rate": 4.356791856791857e-05,
|
| 163 |
+
"loss": 0.2443878746032715,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.26913000799589154,
|
| 168 |
+
"grad_norm": 1.6860474348068237,
|
| 169 |
+
"learning_rate": 4.327541827541828e-05,
|
| 170 |
+
"loss": 0.23169761657714844,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2808313126913651,
|
| 175 |
+
"grad_norm": 1.8752473592758179,
|
| 176 |
+
"learning_rate": 4.298291798291799e-05,
|
| 177 |
+
"loss": 0.22600336074829103,
|
| 178 |
+
"step": 2400
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2925326173868386,
|
| 182 |
+
"grad_norm": 1.6066241264343262,
|
| 183 |
+
"learning_rate": 4.2690417690417694e-05,
|
| 184 |
+
"loss": 0.2185952568054199,
|
| 185 |
+
"step": 2500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.2925326173868386,
|
| 189 |
+
"eval_accuracy": 0.8560363636363636,
|
| 190 |
+
"eval_f1": 0.9651478556858426,
|
| 191 |
+
"eval_loss": 0.039476945996284485,
|
| 192 |
+
"eval_precision": 0.9777755239169846,
|
| 193 |
+
"eval_recall": 0.9528421936966432,
|
| 194 |
+
"eval_runtime": 165.9297,
|
| 195 |
+
"eval_samples_per_second": 165.733,
|
| 196 |
+
"eval_steps_per_second": 5.183,
|
| 197 |
+
"step": 2500
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.3042339220823122,
|
| 201 |
+
"grad_norm": 1.2747637033462524,
|
| 202 |
+
"learning_rate": 4.2397917397917406e-05,
|
| 203 |
+
"loss": 0.22376758575439454,
|
| 204 |
+
"step": 2600
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.31593522677778574,
|
| 208 |
+
"grad_norm": 1.5281429290771484,
|
| 209 |
+
"learning_rate": 4.2105417105417104e-05,
|
| 210 |
+
"loss": 0.21567138671875,
|
| 211 |
+
"step": 2700
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.3276365314732593,
|
| 215 |
+
"grad_norm": 1.0241914987564087,
|
| 216 |
+
"learning_rate": 4.1812916812916816e-05,
|
| 217 |
+
"loss": 0.20220571517944336,
|
| 218 |
+
"step": 2800
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.3393378361687328,
|
| 222 |
+
"grad_norm": 1.9281812906265259,
|
| 223 |
+
"learning_rate": 4.152041652041652e-05,
|
| 224 |
+
"loss": 0.19900651931762695,
|
| 225 |
+
"step": 2900
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.35103914086420634,
|
| 229 |
+
"grad_norm": 1.0900094509124756,
|
| 230 |
+
"learning_rate": 4.122791622791623e-05,
|
| 231 |
+
"loss": 0.19108434677124023,
|
| 232 |
+
"step": 3000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.3627404455596799,
|
| 236 |
+
"grad_norm": 1.9392260313034058,
|
| 237 |
+
"learning_rate": 4.093541593541594e-05,
|
| 238 |
+
"loss": 0.1862166404724121,
|
| 239 |
+
"step": 3100
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.37444175025515347,
|
| 243 |
+
"grad_norm": 1.4907201528549194,
|
| 244 |
+
"learning_rate": 4.064291564291564e-05,
|
| 245 |
+
"loss": 0.18224323272705079,
|
| 246 |
+
"step": 3200
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.386143054950627,
|
| 250 |
+
"grad_norm": 1.0142017602920532,
|
| 251 |
+
"learning_rate": 4.0350415350415355e-05,
|
| 252 |
+
"loss": 0.17916433334350587,
|
| 253 |
+
"step": 3300
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.39784435964610054,
|
| 257 |
+
"grad_norm": 1.3136364221572876,
|
| 258 |
+
"learning_rate": 4.005791505791506e-05,
|
| 259 |
+
"loss": 0.17482011795043945,
|
| 260 |
+
"step": 3400
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.4095456643415741,
|
| 264 |
+
"grad_norm": 1.7887938022613525,
|
| 265 |
+
"learning_rate": 3.9765414765414765e-05,
|
| 266 |
+
"loss": 0.16918930053710937,
|
| 267 |
+
"step": 3500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.4212469690370476,
|
| 271 |
+
"grad_norm": 1.368762731552124,
|
| 272 |
+
"learning_rate": 3.947291447291447e-05,
|
| 273 |
+
"loss": 0.1643056869506836,
|
| 274 |
+
"step": 3600
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.43294827373252115,
|
| 278 |
+
"grad_norm": 1.666864275932312,
|
| 279 |
+
"learning_rate": 3.918041418041418e-05,
|
| 280 |
+
"loss": 0.16255182266235352,
|
| 281 |
+
"step": 3700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.44464957842799474,
|
| 285 |
+
"grad_norm": 1.4107720851898193,
|
| 286 |
+
"learning_rate": 3.888791388791389e-05,
|
| 287 |
+
"loss": 0.15760359764099122,
|
| 288 |
+
"step": 3800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.4563508831234683,
|
| 292 |
+
"grad_norm": 1.6475497484207153,
|
| 293 |
+
"learning_rate": 3.85954135954136e-05,
|
| 294 |
+
"loss": 0.1564300537109375,
|
| 295 |
+
"step": 3900
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.4680521878189418,
|
| 299 |
+
"grad_norm": 1.3566536903381348,
|
| 300 |
+
"learning_rate": 3.8302913302913305e-05,
|
| 301 |
+
"loss": 0.15231061935424806,
|
| 302 |
+
"step": 4000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.47975349251441535,
|
| 306 |
+
"grad_norm": 0.9237350225448608,
|
| 307 |
+
"learning_rate": 3.801041301041301e-05,
|
| 308 |
+
"loss": 0.14665995597839354,
|
| 309 |
+
"step": 4100
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.4914547972098889,
|
| 313 |
+
"grad_norm": 2.0160303115844727,
|
| 314 |
+
"learning_rate": 3.7717912717912715e-05,
|
| 315 |
+
"loss": 0.1439968776702881,
|
| 316 |
+
"step": 4200
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.5031561019053624,
|
| 320 |
+
"grad_norm": 1.495821237564087,
|
| 321 |
+
"learning_rate": 3.742541242541243e-05,
|
| 322 |
+
"loss": 0.14349424362182617,
|
| 323 |
+
"step": 4300
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 0.514857406600836,
|
| 327 |
+
"grad_norm": 1.1643633842468262,
|
| 328 |
+
"learning_rate": 3.713291213291213e-05,
|
| 329 |
+
"loss": 0.1439281463623047,
|
| 330 |
+
"step": 4400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.5265587112963095,
|
| 334 |
+
"grad_norm": 1.2707788944244385,
|
| 335 |
+
"learning_rate": 3.6840411840411844e-05,
|
| 336 |
+
"loss": 0.14324712753295898,
|
| 337 |
+
"step": 4500
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.5382600159917831,
|
| 341 |
+
"grad_norm": 0.9467172026634216,
|
| 342 |
+
"learning_rate": 3.654791154791155e-05,
|
| 343 |
+
"loss": 0.136934118270874,
|
| 344 |
+
"step": 4600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.5499613206872567,
|
| 348 |
+
"grad_norm": 1.4451957941055298,
|
| 349 |
+
"learning_rate": 3.625541125541126e-05,
|
| 350 |
+
"loss": 0.13631460189819336,
|
| 351 |
+
"step": 4700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.5616626253827302,
|
| 355 |
+
"grad_norm": 1.1038591861724854,
|
| 356 |
+
"learning_rate": 3.5962910962910966e-05,
|
| 357 |
+
"loss": 0.13634085655212402,
|
| 358 |
+
"step": 4800
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.5733639300782037,
|
| 362 |
+
"grad_norm": 1.2149685621261597,
|
| 363 |
+
"learning_rate": 3.567041067041067e-05,
|
| 364 |
+
"loss": 0.13158455848693848,
|
| 365 |
+
"step": 4900
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.5850652347736772,
|
| 369 |
+
"grad_norm": 1.421316146850586,
|
| 370 |
+
"learning_rate": 3.5377910377910376e-05,
|
| 371 |
+
"loss": 0.13308174133300782,
|
| 372 |
+
"step": 5000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.5850652347736772,
|
| 376 |
+
"eval_accuracy": 0.9069818181818182,
|
| 377 |
+
"eval_f1": 0.9759635668726578,
|
| 378 |
+
"eval_loss": 0.02319982275366783,
|
| 379 |
+
"eval_precision": 0.9802819359100597,
|
| 380 |
+
"eval_recall": 0.971683077797408,
|
| 381 |
+
"eval_runtime": 164.1216,
|
| 382 |
+
"eval_samples_per_second": 167.559,
|
| 383 |
+
"eval_steps_per_second": 5.24,
|
| 384 |
+
"step": 5000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.5967665394691508,
|
| 388 |
+
"grad_norm": 1.9884291887283325,
|
| 389 |
+
"learning_rate": 3.508541008541009e-05,
|
| 390 |
+
"loss": 0.12942585945129395,
|
| 391 |
+
"step": 5100
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.6084678441646244,
|
| 395 |
+
"grad_norm": 1.1202914714813232,
|
| 396 |
+
"learning_rate": 3.479290979290979e-05,
|
| 397 |
+
"loss": 0.12913961410522462,
|
| 398 |
+
"step": 5200
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.6201691488600979,
|
| 402 |
+
"grad_norm": 1.470561146736145,
|
| 403 |
+
"learning_rate": 3.4500409500409505e-05,
|
| 404 |
+
"loss": 0.12901023864746095,
|
| 405 |
+
"step": 5300
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.6318704535555715,
|
| 409 |
+
"grad_norm": 0.930363655090332,
|
| 410 |
+
"learning_rate": 3.420790920790921e-05,
|
| 411 |
+
"loss": 0.12278788566589355,
|
| 412 |
+
"step": 5400
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.643571758251045,
|
| 416 |
+
"grad_norm": 1.7242929935455322,
|
| 417 |
+
"learning_rate": 3.391540891540892e-05,
|
| 418 |
+
"loss": 0.12592774391174316,
|
| 419 |
+
"step": 5500
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.6552730629465185,
|
| 423 |
+
"grad_norm": 1.8954977989196777,
|
| 424 |
+
"learning_rate": 3.362290862290862e-05,
|
| 425 |
+
"loss": 0.12447975158691406,
|
| 426 |
+
"step": 5600
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.666974367641992,
|
| 430 |
+
"grad_norm": 1.352124571800232,
|
| 431 |
+
"learning_rate": 3.333040833040833e-05,
|
| 432 |
+
"loss": 0.1232685947418213,
|
| 433 |
+
"step": 5700
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.6786756723374656,
|
| 437 |
+
"grad_norm": 1.2466479539871216,
|
| 438 |
+
"learning_rate": 3.303790803790804e-05,
|
| 439 |
+
"loss": 0.11894026756286621,
|
| 440 |
+
"step": 5800
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.6903769770329392,
|
| 444 |
+
"grad_norm": 1.429012656211853,
|
| 445 |
+
"learning_rate": 3.274540774540775e-05,
|
| 446 |
+
"loss": 0.11655033111572266,
|
| 447 |
+
"step": 5900
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.7020782817284127,
|
| 451 |
+
"grad_norm": 1.4804608821868896,
|
| 452 |
+
"learning_rate": 3.2452907452907454e-05,
|
| 453 |
+
"loss": 0.11837745666503906,
|
| 454 |
+
"step": 6000
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.7137795864238863,
|
| 458 |
+
"grad_norm": 1.3687437772750854,
|
| 459 |
+
"learning_rate": 3.2160407160407166e-05,
|
| 460 |
+
"loss": 0.11587305068969726,
|
| 461 |
+
"step": 6100
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.7254808911193598,
|
| 465 |
+
"grad_norm": 1.4826337099075317,
|
| 466 |
+
"learning_rate": 3.186790686790687e-05,
|
| 467 |
+
"loss": 0.11750144958496093,
|
| 468 |
+
"step": 6200
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.7371821958148334,
|
| 472 |
+
"grad_norm": 0.959021270275116,
|
| 473 |
+
"learning_rate": 3.1575406575406577e-05,
|
| 474 |
+
"loss": 0.11341402053833008,
|
| 475 |
+
"step": 6300
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.7488835005103069,
|
| 479 |
+
"grad_norm": 0.8908078670501709,
|
| 480 |
+
"learning_rate": 3.128290628290628e-05,
|
| 481 |
+
"loss": 0.11524638175964355,
|
| 482 |
+
"step": 6400
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.7605848052057804,
|
| 486 |
+
"grad_norm": 0.9967349171638489,
|
| 487 |
+
"learning_rate": 3.0990405990405994e-05,
|
| 488 |
+
"loss": 0.1124759292602539,
|
| 489 |
+
"step": 6500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.772286109901254,
|
| 493 |
+
"grad_norm": 1.4200743436813354,
|
| 494 |
+
"learning_rate": 3.06979056979057e-05,
|
| 495 |
+
"loss": 0.10863906860351563,
|
| 496 |
+
"step": 6600
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.7839874145967275,
|
| 500 |
+
"grad_norm": 1.6694557666778564,
|
| 501 |
+
"learning_rate": 3.0405405405405407e-05,
|
| 502 |
+
"loss": 0.11165953636169433,
|
| 503 |
+
"step": 6700
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.7956887192922011,
|
| 507 |
+
"grad_norm": 1.0781575441360474,
|
| 508 |
+
"learning_rate": 3.0112905112905116e-05,
|
| 509 |
+
"loss": 0.1116695213317871,
|
| 510 |
+
"step": 6800
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.8073900239876747,
|
| 514 |
+
"grad_norm": 1.3060591220855713,
|
| 515 |
+
"learning_rate": 2.9820404820404824e-05,
|
| 516 |
+
"loss": 0.1075587272644043,
|
| 517 |
+
"step": 6900
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 0.8190913286831482,
|
| 521 |
+
"grad_norm": 1.448541283607483,
|
| 522 |
+
"learning_rate": 2.9527904527904533e-05,
|
| 523 |
+
"loss": 0.11105222702026367,
|
| 524 |
+
"step": 7000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.8307926333786217,
|
| 528 |
+
"grad_norm": 1.8385331630706787,
|
| 529 |
+
"learning_rate": 2.9235404235404234e-05,
|
| 530 |
+
"loss": 0.10690691947937012,
|
| 531 |
+
"step": 7100
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.8424939380740952,
|
| 535 |
+
"grad_norm": 1.4097715616226196,
|
| 536 |
+
"learning_rate": 2.8942903942903943e-05,
|
| 537 |
+
"loss": 0.10494094848632812,
|
| 538 |
+
"step": 7200
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.8541952427695688,
|
| 542 |
+
"grad_norm": 1.7280794382095337,
|
| 543 |
+
"learning_rate": 2.865040365040365e-05,
|
| 544 |
+
"loss": 0.10908853530883789,
|
| 545 |
+
"step": 7300
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.8658965474650423,
|
| 549 |
+
"grad_norm": 1.0693798065185547,
|
| 550 |
+
"learning_rate": 2.835790335790336e-05,
|
| 551 |
+
"loss": 0.10775123596191406,
|
| 552 |
+
"step": 7400
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.8775978521605159,
|
| 556 |
+
"grad_norm": 1.5537015199661255,
|
| 557 |
+
"learning_rate": 2.806540306540307e-05,
|
| 558 |
+
"loss": 0.10442270278930664,
|
| 559 |
+
"step": 7500
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.8775978521605159,
|
| 563 |
+
"eval_accuracy": 0.9217818181818181,
|
| 564 |
+
"eval_f1": 0.9801159592655914,
|
| 565 |
+
"eval_loss": 0.017201833426952362,
|
| 566 |
+
"eval_precision": 0.982831231753525,
|
| 567 |
+
"eval_recall": 0.9774156484352801,
|
| 568 |
+
"eval_runtime": 185.5814,
|
| 569 |
+
"eval_samples_per_second": 148.183,
|
| 570 |
+
"eval_steps_per_second": 4.634,
|
| 571 |
+
"step": 7500
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.8892991568559895,
|
| 575 |
+
"grad_norm": 1.0237928628921509,
|
| 576 |
+
"learning_rate": 2.7772902772902777e-05,
|
| 577 |
+
"loss": 0.10509194374084473,
|
| 578 |
+
"step": 7600
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.901000461551463,
|
| 582 |
+
"grad_norm": 1.5233043432235718,
|
| 583 |
+
"learning_rate": 2.7480402480402486e-05,
|
| 584 |
+
"loss": 0.10594425201416016,
|
| 585 |
+
"step": 7700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.9127017662469366,
|
| 589 |
+
"grad_norm": 0.8517800569534302,
|
| 590 |
+
"learning_rate": 2.7187902187902187e-05,
|
| 591 |
+
"loss": 0.10539012908935547,
|
| 592 |
+
"step": 7800
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.92440307094241,
|
| 596 |
+
"grad_norm": 1.3082163333892822,
|
| 597 |
+
"learning_rate": 2.6895401895401896e-05,
|
| 598 |
+
"loss": 0.09960749626159668,
|
| 599 |
+
"step": 7900
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.9361043756378836,
|
| 603 |
+
"grad_norm": 0.8803576827049255,
|
| 604 |
+
"learning_rate": 2.6602901602901604e-05,
|
| 605 |
+
"loss": 0.10325809478759766,
|
| 606 |
+
"step": 8000
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.9478056803333572,
|
| 610 |
+
"grad_norm": 1.5242176055908203,
|
| 611 |
+
"learning_rate": 2.6310401310401313e-05,
|
| 612 |
+
"loss": 0.10169939041137695,
|
| 613 |
+
"step": 8100
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.9595069850288307,
|
| 617 |
+
"grad_norm": 1.2510749101638794,
|
| 618 |
+
"learning_rate": 2.601790101790102e-05,
|
| 619 |
+
"loss": 0.100986909866333,
|
| 620 |
+
"step": 8200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.9712082897243043,
|
| 624 |
+
"grad_norm": 1.4342858791351318,
|
| 625 |
+
"learning_rate": 2.5725400725400726e-05,
|
| 626 |
+
"loss": 0.09906567573547363,
|
| 627 |
+
"step": 8300
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.9829095944197778,
|
| 631 |
+
"grad_norm": 1.1943491697311401,
|
| 632 |
+
"learning_rate": 2.5432900432900435e-05,
|
| 633 |
+
"loss": 0.09981002807617187,
|
| 634 |
+
"step": 8400
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.9946108991152514,
|
| 638 |
+
"grad_norm": 1.0533969402313232,
|
| 639 |
+
"learning_rate": 2.5140400140400143e-05,
|
| 640 |
+
"loss": 0.10099024772644043,
|
| 641 |
+
"step": 8500
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.006201691488601,
|
| 645 |
+
"grad_norm": 0.952810525894165,
|
| 646 |
+
"learning_rate": 2.484789984789985e-05,
|
| 647 |
+
"loss": 0.09148699760437012,
|
| 648 |
+
"step": 8600
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.0179029961840744,
|
| 652 |
+
"grad_norm": 0.9051135182380676,
|
| 653 |
+
"learning_rate": 2.4555399555399554e-05,
|
| 654 |
+
"loss": 0.08493613243103028,
|
| 655 |
+
"step": 8700
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.0296043008795481,
|
| 659 |
+
"grad_norm": 1.4660611152648926,
|
| 660 |
+
"learning_rate": 2.4262899262899262e-05,
|
| 661 |
+
"loss": 0.08283645629882813,
|
| 662 |
+
"step": 8800
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.0413056055750216,
|
| 666 |
+
"grad_norm": 1.2308636903762817,
|
| 667 |
+
"learning_rate": 2.397039897039897e-05,
|
| 668 |
+
"loss": 0.08224470138549805,
|
| 669 |
+
"step": 8900
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.053006910270495,
|
| 673 |
+
"grad_norm": 1.180936574935913,
|
| 674 |
+
"learning_rate": 2.367789867789868e-05,
|
| 675 |
+
"loss": 0.08456890106201172,
|
| 676 |
+
"step": 9000
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.0647082149659688,
|
| 680 |
+
"grad_norm": 1.1681982278823853,
|
| 681 |
+
"learning_rate": 2.3385398385398384e-05,
|
| 682 |
+
"loss": 0.08683476448059083,
|
| 683 |
+
"step": 9100
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.0764095196614423,
|
| 687 |
+
"grad_norm": 1.1853655576705933,
|
| 688 |
+
"learning_rate": 2.3092898092898093e-05,
|
| 689 |
+
"loss": 0.08339482307434082,
|
| 690 |
+
"step": 9200
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.0881108243569158,
|
| 694 |
+
"grad_norm": 0.7993655204772949,
|
| 695 |
+
"learning_rate": 2.28003978003978e-05,
|
| 696 |
+
"loss": 0.08154037475585937,
|
| 697 |
+
"step": 9300
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.0998121290523892,
|
| 701 |
+
"grad_norm": 1.3015902042388916,
|
| 702 |
+
"learning_rate": 2.250789750789751e-05,
|
| 703 |
+
"loss": 0.0843494701385498,
|
| 704 |
+
"step": 9400
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.111513433747863,
|
| 708 |
+
"grad_norm": 0.927474319934845,
|
| 709 |
+
"learning_rate": 2.2215397215397215e-05,
|
| 710 |
+
"loss": 0.08107766151428222,
|
| 711 |
+
"step": 9500
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.1232147384433364,
|
| 715 |
+
"grad_norm": 1.2510508298873901,
|
| 716 |
+
"learning_rate": 2.1922896922896923e-05,
|
| 717 |
+
"loss": 0.08390594482421875,
|
| 718 |
+
"step": 9600
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.13491604313881,
|
| 722 |
+
"grad_norm": 1.012722373008728,
|
| 723 |
+
"learning_rate": 2.1630396630396632e-05,
|
| 724 |
+
"loss": 0.08134162902832032,
|
| 725 |
+
"step": 9700
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.1466173478342836,
|
| 729 |
+
"grad_norm": 1.8160477876663208,
|
| 730 |
+
"learning_rate": 2.1337896337896337e-05,
|
| 731 |
+
"loss": 0.08689288139343261,
|
| 732 |
+
"step": 9800
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.158318652529757,
|
| 736 |
+
"grad_norm": 1.6144487857818604,
|
| 737 |
+
"learning_rate": 2.1045396045396046e-05,
|
| 738 |
+
"loss": 0.08774213790893555,
|
| 739 |
+
"step": 9900
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.1700199572252306,
|
| 743 |
+
"grad_norm": 1.088478922843933,
|
| 744 |
+
"learning_rate": 2.0752895752895754e-05,
|
| 745 |
+
"loss": 0.08510435104370118,
|
| 746 |
+
"step": 10000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.1700199572252306,
|
| 750 |
+
"eval_accuracy": 0.9310545454545455,
|
| 751 |
+
"eval_f1": 0.9822452239157081,
|
| 752 |
+
"eval_loss": 0.0149807995185256,
|
| 753 |
+
"eval_precision": 0.9843658878562654,
|
| 754 |
+
"eval_recall": 0.9801336776170297,
|
| 755 |
+
"eval_runtime": 187.3637,
|
| 756 |
+
"eval_samples_per_second": 146.773,
|
| 757 |
+
"eval_steps_per_second": 4.59,
|
| 758 |
+
"step": 10000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.1817212619207043,
|
| 762 |
+
"grad_norm": 1.2410812377929688,
|
| 763 |
+
"learning_rate": 2.0460395460395463e-05,
|
| 764 |
+
"loss": 0.08356905937194824,
|
| 765 |
+
"step": 10100
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.1934225666161777,
|
| 769 |
+
"grad_norm": 0.9291555881500244,
|
| 770 |
+
"learning_rate": 2.0167895167895168e-05,
|
| 771 |
+
"loss": 0.08259629249572754,
|
| 772 |
+
"step": 10200
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.2051238713116512,
|
| 776 |
+
"grad_norm": 1.146966576576233,
|
| 777 |
+
"learning_rate": 1.9875394875394876e-05,
|
| 778 |
+
"loss": 0.08215347290039063,
|
| 779 |
+
"step": 10300
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2168251760071247,
|
| 783 |
+
"grad_norm": 0.9125510454177856,
|
| 784 |
+
"learning_rate": 1.9582894582894585e-05,
|
| 785 |
+
"loss": 0.08209040641784668,
|
| 786 |
+
"step": 10400
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2285264807025984,
|
| 790 |
+
"grad_norm": 0.906428337097168,
|
| 791 |
+
"learning_rate": 1.9290394290394293e-05,
|
| 792 |
+
"loss": 0.08041071891784668,
|
| 793 |
+
"step": 10500
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2402277853980719,
|
| 797 |
+
"grad_norm": 1.0091209411621094,
|
| 798 |
+
"learning_rate": 1.8997893997894e-05,
|
| 799 |
+
"loss": 0.07910086631774903,
|
| 800 |
+
"step": 10600
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2519290900935454,
|
| 804 |
+
"grad_norm": 0.9152646660804749,
|
| 805 |
+
"learning_rate": 1.8705393705393707e-05,
|
| 806 |
+
"loss": 0.08119074821472168,
|
| 807 |
+
"step": 10700
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.2636303947890188,
|
| 811 |
+
"grad_norm": 1.0613765716552734,
|
| 812 |
+
"learning_rate": 1.8412893412893415e-05,
|
| 813 |
+
"loss": 0.07813576221466065,
|
| 814 |
+
"step": 10800
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2753316994844925,
|
| 818 |
+
"grad_norm": 0.7284146547317505,
|
| 819 |
+
"learning_rate": 1.8120393120393124e-05,
|
| 820 |
+
"loss": 0.08263915061950683,
|
| 821 |
+
"step": 10900
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.287033004179966,
|
| 825 |
+
"grad_norm": 0.8245161771774292,
|
| 826 |
+
"learning_rate": 1.782789282789283e-05,
|
| 827 |
+
"loss": 0.07991621494293213,
|
| 828 |
+
"step": 11000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.2987343088754395,
|
| 832 |
+
"grad_norm": 0.9131597876548767,
|
| 833 |
+
"learning_rate": 1.7535392535392538e-05,
|
| 834 |
+
"loss": 0.08203693389892579,
|
| 835 |
+
"step": 11100
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.3104356135709132,
|
| 839 |
+
"grad_norm": 1.4919288158416748,
|
| 840 |
+
"learning_rate": 1.7242892242892246e-05,
|
| 841 |
+
"loss": 0.07703531742095947,
|
| 842 |
+
"step": 11200
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3221369182663867,
|
| 846 |
+
"grad_norm": 0.9880580306053162,
|
| 847 |
+
"learning_rate": 1.695039195039195e-05,
|
| 848 |
+
"loss": 0.07978516101837158,
|
| 849 |
+
"step": 11300
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3338382229618602,
|
| 853 |
+
"grad_norm": 1.483494758605957,
|
| 854 |
+
"learning_rate": 1.665789165789166e-05,
|
| 855 |
+
"loss": 0.0792800235748291,
|
| 856 |
+
"step": 11400
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3455395276573339,
|
| 860 |
+
"grad_norm": 1.2706217765808105,
|
| 861 |
+
"learning_rate": 1.6365391365391368e-05,
|
| 862 |
+
"loss": 0.08085798263549805,
|
| 863 |
+
"step": 11500
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3572408323528073,
|
| 867 |
+
"grad_norm": 0.9506115913391113,
|
| 868 |
+
"learning_rate": 1.6072891072891073e-05,
|
| 869 |
+
"loss": 0.0732752513885498,
|
| 870 |
+
"step": 11600
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3689421370482808,
|
| 874 |
+
"grad_norm": 0.7671661972999573,
|
| 875 |
+
"learning_rate": 1.5780390780390782e-05,
|
| 876 |
+
"loss": 0.07732769966125488,
|
| 877 |
+
"step": 11700
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.3806434417437545,
|
| 881 |
+
"grad_norm": 1.3610262870788574,
|
| 882 |
+
"learning_rate": 1.5487890487890487e-05,
|
| 883 |
+
"loss": 0.07633553028106689,
|
| 884 |
+
"step": 11800
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.392344746439228,
|
| 888 |
+
"grad_norm": 1.1439802646636963,
|
| 889 |
+
"learning_rate": 1.5195390195390197e-05,
|
| 890 |
+
"loss": 0.07713084220886231,
|
| 891 |
+
"step": 11900
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4040460511347015,
|
| 895 |
+
"grad_norm": 1.389791488647461,
|
| 896 |
+
"learning_rate": 1.4902889902889906e-05,
|
| 897 |
+
"loss": 0.0749136209487915,
|
| 898 |
+
"step": 12000
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4157473558301752,
|
| 902 |
+
"grad_norm": 0.6515536308288574,
|
| 903 |
+
"learning_rate": 1.461038961038961e-05,
|
| 904 |
+
"loss": 0.07383899688720703,
|
| 905 |
+
"step": 12100
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.4274486605256487,
|
| 909 |
+
"grad_norm": 0.7534876465797424,
|
| 910 |
+
"learning_rate": 1.431788931788932e-05,
|
| 911 |
+
"loss": 0.07853510379791259,
|
| 912 |
+
"step": 12200
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4391499652211222,
|
| 916 |
+
"grad_norm": 0.6730746626853943,
|
| 917 |
+
"learning_rate": 1.4025389025389026e-05,
|
| 918 |
+
"loss": 0.0697617483139038,
|
| 919 |
+
"step": 12300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4508512699165956,
|
| 923 |
+
"grad_norm": 1.8400371074676514,
|
| 924 |
+
"learning_rate": 1.3732888732888733e-05,
|
| 925 |
+
"loss": 0.07848044872283935,
|
| 926 |
+
"step": 12400
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.4625525746120691,
|
| 930 |
+
"grad_norm": 1.2114777565002441,
|
| 931 |
+
"learning_rate": 1.3440388440388441e-05,
|
| 932 |
+
"loss": 0.0783261775970459,
|
| 933 |
+
"step": 12500
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4625525746120691,
|
| 937 |
+
"eval_accuracy": 0.9353818181818182,
|
| 938 |
+
"eval_f1": 0.9834085799751036,
|
| 939 |
+
"eval_loss": 0.013636507093906403,
|
| 940 |
+
"eval_precision": 0.9859303596264654,
|
| 941 |
+
"eval_recall": 0.9808996676591591,
|
| 942 |
+
"eval_runtime": 167.3633,
|
| 943 |
+
"eval_samples_per_second": 164.313,
|
| 944 |
+
"eval_steps_per_second": 5.139,
|
| 945 |
+
"step": 12500
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1.4742538793075428,
|
| 949 |
+
"grad_norm": 1.7295293807983398,
|
| 950 |
+
"learning_rate": 1.3147888147888148e-05,
|
| 951 |
+
"loss": 0.074767746925354,
|
| 952 |
+
"step": 12600
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1.4859551840030163,
|
| 956 |
+
"grad_norm": 1.08072030544281,
|
| 957 |
+
"learning_rate": 1.2855387855387857e-05,
|
| 958 |
+
"loss": 0.07684030532836914,
|
| 959 |
+
"step": 12700
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.4976564886984898,
|
| 963 |
+
"grad_norm": 1.334511637687683,
|
| 964 |
+
"learning_rate": 1.2562887562887562e-05,
|
| 965 |
+
"loss": 0.07815152645111084,
|
| 966 |
+
"step": 12800
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 1.5093577933939635,
|
| 970 |
+
"grad_norm": 0.8732834458351135,
|
| 971 |
+
"learning_rate": 1.227038727038727e-05,
|
| 972 |
+
"loss": 0.07617097854614258,
|
| 973 |
+
"step": 12900
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 1.521059098089437,
|
| 977 |
+
"grad_norm": 1.038492202758789,
|
| 978 |
+
"learning_rate": 1.1977886977886979e-05,
|
| 979 |
+
"loss": 0.07482788562774659,
|
| 980 |
+
"step": 13000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.5327604027849104,
|
| 984 |
+
"grad_norm": 1.1699777841567993,
|
| 985 |
+
"learning_rate": 1.1685386685386686e-05,
|
| 986 |
+
"loss": 0.07500550746917725,
|
| 987 |
+
"step": 13100
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 1.5444617074803841,
|
| 991 |
+
"grad_norm": 1.1473757028579712,
|
| 992 |
+
"learning_rate": 1.1392886392886394e-05,
|
| 993 |
+
"loss": 0.07690254688262939,
|
| 994 |
+
"step": 13200
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1.5561630121758576,
|
| 998 |
+
"grad_norm": 1.6335910558700562,
|
| 999 |
+
"learning_rate": 1.1100386100386101e-05,
|
| 1000 |
+
"loss": 0.07375136375427246,
|
| 1001 |
+
"step": 13300
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.567864316871331,
|
| 1005 |
+
"grad_norm": 0.9858669638633728,
|
| 1006 |
+
"learning_rate": 1.0807885807885808e-05,
|
| 1007 |
+
"loss": 0.07300055027008057,
|
| 1008 |
+
"step": 13400
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 1.5795656215668048,
|
| 1012 |
+
"grad_norm": 1.1223580837249756,
|
| 1013 |
+
"learning_rate": 1.0515385515385516e-05,
|
| 1014 |
+
"loss": 0.07348923683166504,
|
| 1015 |
+
"step": 13500
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.5912669262622783,
|
| 1019 |
+
"grad_norm": 0.8121886253356934,
|
| 1020 |
+
"learning_rate": 1.0222885222885223e-05,
|
| 1021 |
+
"loss": 0.0721654510498047,
|
| 1022 |
+
"step": 13600
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.6029682309577518,
|
| 1026 |
+
"grad_norm": 1.125582218170166,
|
| 1027 |
+
"learning_rate": 9.930384930384932e-06,
|
| 1028 |
+
"loss": 0.07356025695800782,
|
| 1029 |
+
"step": 13700
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 1.6146695356532255,
|
| 1033 |
+
"grad_norm": 0.9660710692405701,
|
| 1034 |
+
"learning_rate": 9.637884637884638e-06,
|
| 1035 |
+
"loss": 0.07352369308471679,
|
| 1036 |
+
"step": 13800
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.6263708403486987,
|
| 1040 |
+
"grad_norm": 0.9546118378639221,
|
| 1041 |
+
"learning_rate": 9.345384345384347e-06,
|
| 1042 |
+
"loss": 0.07006223201751709,
|
| 1043 |
+
"step": 13900
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 1.6380721450441724,
|
| 1047 |
+
"grad_norm": 1.2664750814437866,
|
| 1048 |
+
"learning_rate": 9.052884052884054e-06,
|
| 1049 |
+
"loss": 0.07232100963592529,
|
| 1050 |
+
"step": 14000
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 1.6497734497396461,
|
| 1054 |
+
"grad_norm": 1.087778925895691,
|
| 1055 |
+
"learning_rate": 8.76038376038376e-06,
|
| 1056 |
+
"loss": 0.07518599510192871,
|
| 1057 |
+
"step": 14100
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 1.6614747544351194,
|
| 1061 |
+
"grad_norm": 1.0009450912475586,
|
| 1062 |
+
"learning_rate": 8.467883467883467e-06,
|
| 1063 |
+
"loss": 0.07395988464355469,
|
| 1064 |
+
"step": 14200
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.673176059130593,
|
| 1068 |
+
"grad_norm": 1.052869200706482,
|
| 1069 |
+
"learning_rate": 8.175383175383176e-06,
|
| 1070 |
+
"loss": 0.07117973327636719,
|
| 1071 |
+
"step": 14300
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.6848773638260666,
|
| 1075 |
+
"grad_norm": 0.9935372471809387,
|
| 1076 |
+
"learning_rate": 7.882882882882883e-06,
|
| 1077 |
+
"loss": 0.07157835006713867,
|
| 1078 |
+
"step": 14400
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 1.69657866852154,
|
| 1082 |
+
"grad_norm": 0.8762065768241882,
|
| 1083 |
+
"learning_rate": 7.59038259038259e-06,
|
| 1084 |
+
"loss": 0.07200119018554688,
|
| 1085 |
+
"step": 14500
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 1.7082799732170137,
|
| 1089 |
+
"grad_norm": 1.1860989332199097,
|
| 1090 |
+
"learning_rate": 7.297882297882298e-06,
|
| 1091 |
+
"loss": 0.06994849681854248,
|
| 1092 |
+
"step": 14600
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 1.7199812779124872,
|
| 1096 |
+
"grad_norm": 0.7438328862190247,
|
| 1097 |
+
"learning_rate": 7.005382005382005e-06,
|
| 1098 |
+
"loss": 0.06839815139770508,
|
| 1099 |
+
"step": 14700
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 1.7316825826079607,
|
| 1103 |
+
"grad_norm": 1.6703062057495117,
|
| 1104 |
+
"learning_rate": 6.712881712881713e-06,
|
| 1105 |
+
"loss": 0.07001969337463379,
|
| 1106 |
+
"step": 14800
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.7433838873034344,
|
| 1110 |
+
"grad_norm": 1.143515706062317,
|
| 1111 |
+
"learning_rate": 6.42038142038142e-06,
|
| 1112 |
+
"loss": 0.07166263103485107,
|
| 1113 |
+
"step": 14900
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 1.755085191998908,
|
| 1117 |
+
"grad_norm": 1.0320720672607422,
|
| 1118 |
+
"learning_rate": 6.127881127881129e-06,
|
| 1119 |
+
"loss": 0.0704725456237793,
|
| 1120 |
+
"step": 15000
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.755085191998908,
|
| 1124 |
+
"eval_accuracy": 0.9399272727272727,
|
| 1125 |
+
"eval_f1": 0.9843185306203277,
|
| 1126 |
+
"eval_loss": 0.012572239153087139,
|
| 1127 |
+
"eval_precision": 0.9860882072137978,
|
| 1128 |
+
"eval_recall": 0.9825551945244066,
|
| 1129 |
+
"eval_runtime": 207.4736,
|
| 1130 |
+
"eval_samples_per_second": 132.547,
|
| 1131 |
+
"eval_steps_per_second": 4.145,
|
| 1132 |
+
"step": 15000
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 1.7667864966943814,
|
| 1136 |
+
"grad_norm": 0.8305580019950867,
|
| 1137 |
+
"learning_rate": 5.8353808353808354e-06,
|
| 1138 |
+
"loss": 0.07156490802764892,
|
| 1139 |
+
"step": 15100
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.778487801389855,
|
| 1143 |
+
"grad_norm": 1.298708438873291,
|
| 1144 |
+
"learning_rate": 5.542880542880543e-06,
|
| 1145 |
+
"loss": 0.06973484516143799,
|
| 1146 |
+
"step": 15200
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 1.7901891060853286,
|
| 1150 |
+
"grad_norm": 0.8593277931213379,
|
| 1151 |
+
"learning_rate": 5.250380250380251e-06,
|
| 1152 |
+
"loss": 0.06879038333892823,
|
| 1153 |
+
"step": 15300
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 1.801890410780802,
|
| 1157 |
+
"grad_norm": 0.732769787311554,
|
| 1158 |
+
"learning_rate": 4.9578799578799576e-06,
|
| 1159 |
+
"loss": 0.06908240318298339,
|
| 1160 |
+
"step": 15400
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.8135917154762757,
|
| 1164 |
+
"grad_norm": 1.3583112955093384,
|
| 1165 |
+
"learning_rate": 4.665379665379665e-06,
|
| 1166 |
+
"loss": 0.07114370822906495,
|
| 1167 |
+
"step": 15500
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 1.825293020171749,
|
| 1171 |
+
"grad_norm": 0.9253562092781067,
|
| 1172 |
+
"learning_rate": 4.372879372879373e-06,
|
| 1173 |
+
"loss": 0.07082613945007324,
|
| 1174 |
+
"step": 15600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 1.8369943248672227,
|
| 1178 |
+
"grad_norm": 1.340505599975586,
|
| 1179 |
+
"learning_rate": 4.0803790803790806e-06,
|
| 1180 |
+
"loss": 0.06995931625366211,
|
| 1181 |
+
"step": 15700
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.8486956295626964,
|
| 1185 |
+
"grad_norm": 0.9078701734542847,
|
| 1186 |
+
"learning_rate": 3.7878787878787882e-06,
|
| 1187 |
+
"loss": 0.06670016288757324,
|
| 1188 |
+
"step": 15800
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.8603969342581697,
|
| 1192 |
+
"grad_norm": 1.2157623767852783,
|
| 1193 |
+
"learning_rate": 3.495378495378496e-06,
|
| 1194 |
+
"loss": 0.0690723991394043,
|
| 1195 |
+
"step": 15900
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.8720982389536434,
|
| 1199 |
+
"grad_norm": 1.7708772420883179,
|
| 1200 |
+
"learning_rate": 3.202878202878203e-06,
|
| 1201 |
+
"loss": 0.06770940780639649,
|
| 1202 |
+
"step": 16000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 1.8837995436491168,
|
| 1206 |
+
"grad_norm": 0.976672887802124,
|
| 1207 |
+
"learning_rate": 2.9103779103779103e-06,
|
| 1208 |
+
"loss": 0.0646241569519043,
|
| 1209 |
+
"step": 16100
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 1.8955008483445903,
|
| 1213 |
+
"grad_norm": 1.7383304834365845,
|
| 1214 |
+
"learning_rate": 2.617877617877618e-06,
|
| 1215 |
+
"loss": 0.06627600193023682,
|
| 1216 |
+
"step": 16200
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 1.907202153040064,
|
| 1220 |
+
"grad_norm": 0.8907257318496704,
|
| 1221 |
+
"learning_rate": 2.3253773253773257e-06,
|
| 1222 |
+
"loss": 0.06847190380096435,
|
| 1223 |
+
"step": 16300
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 1.9189034577355375,
|
| 1227 |
+
"grad_norm": 0.7403037548065186,
|
| 1228 |
+
"learning_rate": 2.032877032877033e-06,
|
| 1229 |
+
"loss": 0.06623115539550781,
|
| 1230 |
+
"step": 16400
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 1.930604762431011,
|
| 1234 |
+
"grad_norm": 1.0598572492599487,
|
| 1235 |
+
"learning_rate": 1.7403767403767406e-06,
|
| 1236 |
+
"loss": 0.06684101104736329,
|
| 1237 |
+
"step": 16500
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 1.9423060671264847,
|
| 1241 |
+
"grad_norm": 1.2067943811416626,
|
| 1242 |
+
"learning_rate": 1.447876447876448e-06,
|
| 1243 |
+
"loss": 0.06935319423675537,
|
| 1244 |
+
"step": 16600
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 1.9540073718219582,
|
| 1248 |
+
"grad_norm": 0.9936091899871826,
|
| 1249 |
+
"learning_rate": 1.1553761553761555e-06,
|
| 1250 |
+
"loss": 0.06889093399047852,
|
| 1251 |
+
"step": 16700
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 1.9657086765174316,
|
| 1255 |
+
"grad_norm": 1.5107425451278687,
|
| 1256 |
+
"learning_rate": 8.628758628758629e-07,
|
| 1257 |
+
"loss": 0.07209495544433593,
|
| 1258 |
+
"step": 16800
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 1.9774099812129053,
|
| 1262 |
+
"grad_norm": 1.0668072700500488,
|
| 1263 |
+
"learning_rate": 5.703755703755704e-07,
|
| 1264 |
+
"loss": 0.06988236904144288,
|
| 1265 |
+
"step": 16900
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 1.9891112859083788,
|
| 1269 |
+
"grad_norm": 1.113166093826294,
|
| 1270 |
+
"learning_rate": 2.7787527787527786e-07,
|
| 1271 |
+
"loss": 0.06924228668212891,
|
| 1272 |
+
"step": 17000
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 2.0,
|
| 1276 |
+
"eval_accuracy": 0.9411636363636363,
|
| 1277 |
+
"eval_f1": 0.9845099349219312,
|
| 1278 |
+
"eval_loss": 0.012292231433093548,
|
| 1279 |
+
"eval_precision": 0.9859127008710089,
|
| 1280 |
+
"eval_recall": 0.9831111550388555,
|
| 1281 |
+
"eval_runtime": 183.6292,
|
| 1282 |
+
"eval_samples_per_second": 149.758,
|
| 1283 |
+
"eval_steps_per_second": 4.683,
|
| 1284 |
+
"step": 17094
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.0,
|
| 1288 |
+
"step": 17094,
|
| 1289 |
+
"total_flos": 1.9402797892099978e+18,
|
| 1290 |
+
"train_loss": 0.08505657710317173,
|
| 1291 |
+
"train_runtime": 156175.5955,
|
| 1292 |
+
"train_samples_per_second": 63.038,
|
| 1293 |
+
"train_steps_per_second": 0.109
|
| 1294 |
+
}
|
| 1295 |
+
],
|
| 1296 |
+
"logging_steps": 100,
|
| 1297 |
+
"max_steps": 17094,
|
| 1298 |
+
"num_input_tokens_seen": 0,
|
| 1299 |
+
"num_train_epochs": 2,
|
| 1300 |
+
"save_steps": 2500,
|
| 1301 |
+
"stateful_callbacks": {
|
| 1302 |
+
"TrainerControl": {
|
| 1303 |
+
"args": {
|
| 1304 |
+
"should_epoch_stop": false,
|
| 1305 |
+
"should_evaluate": false,
|
| 1306 |
+
"should_log": false,
|
| 1307 |
+
"should_save": true,
|
| 1308 |
+
"should_training_stop": true
|
| 1309 |
+
},
|
| 1310 |
+
"attributes": {}
|
| 1311 |
+
}
|
| 1312 |
+
},
|
| 1313 |
+
"total_flos": 1.9402797892099978e+18,
|
| 1314 |
+
"train_batch_size": 32,
|
| 1315 |
+
"trial_name": null,
|
| 1316 |
+
"trial_params": null
|
| 1317 |
+
}
|