Training checkpoint at step 1000
Browse files- trainer_state.json +1891 -10
trainer_state.json
CHANGED
|
@@ -1,19 +1,1900 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric":
|
| 4 |
-
"best_model_checkpoint":
|
| 5 |
-
"epoch": 0.
|
| 6 |
-
"eval_steps":
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
"logging_steps": 25,
|
| 13 |
"max_steps": 50000,
|
| 14 |
"num_input_tokens_seen": 0,
|
| 15 |
"num_train_epochs": 9223372036854775807,
|
| 16 |
-
"save_steps":
|
| 17 |
"stateful_callbacks": {
|
| 18 |
"TrainerControl": {
|
| 19 |
"args": {
|
|
@@ -26,8 +1907,8 @@
|
|
| 26 |
"attributes": {}
|
| 27 |
}
|
| 28 |
},
|
| 29 |
-
"total_flos":
|
| 30 |
-
"train_batch_size":
|
| 31 |
"trial_name": null,
|
| 32 |
"trial_params": null
|
| 33 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 1000,
|
| 3 |
+
"best_metric": 2.488457202911377,
|
| 4 |
+
"best_model_checkpoint": "../runs/karpathy/fineweb-edu-100b-shuffle/meta-llama/Llama-3.2-1B/linear_adamw_wd1e-03_7x1024_mem32_bs64_hf_armt_dmem64/run_20/checkpoint-1000",
|
| 5 |
+
"epoch": 0.02,
|
| 6 |
+
"eval_steps": 5,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0001,
|
| 14 |
+
"eval_loss": 3.320133686065674,
|
| 15 |
+
"eval_runtime": 33.1817,
|
| 16 |
+
"eval_samples_per_second": 3.526,
|
| 17 |
+
"eval_steps_per_second": 1.778,
|
| 18 |
+
"step": 5
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.0002,
|
| 22 |
+
"eval_loss": 3.319335460662842,
|
| 23 |
+
"eval_runtime": 33.1229,
|
| 24 |
+
"eval_samples_per_second": 3.532,
|
| 25 |
+
"eval_steps_per_second": 1.781,
|
| 26 |
+
"step": 10
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.0003,
|
| 30 |
+
"eval_loss": 3.318042516708374,
|
| 31 |
+
"eval_runtime": 33.3382,
|
| 32 |
+
"eval_samples_per_second": 3.509,
|
| 33 |
+
"eval_steps_per_second": 1.77,
|
| 34 |
+
"step": 15
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.0004,
|
| 38 |
+
"eval_loss": 3.31443190574646,
|
| 39 |
+
"eval_runtime": 33.2423,
|
| 40 |
+
"eval_samples_per_second": 3.52,
|
| 41 |
+
"eval_steps_per_second": 1.775,
|
| 42 |
+
"step": 20
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.0005,
|
| 46 |
+
"grad_norm": 0.8831791054097137,
|
| 47 |
+
"learning_rate": 4.8e-08,
|
| 48 |
+
"loss": 3.4942,
|
| 49 |
+
"step": 25
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.0005,
|
| 53 |
+
"eval_loss": 3.3073768615722656,
|
| 54 |
+
"eval_runtime": 33.3914,
|
| 55 |
+
"eval_samples_per_second": 3.504,
|
| 56 |
+
"eval_steps_per_second": 1.767,
|
| 57 |
+
"step": 25
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.0006,
|
| 61 |
+
"eval_loss": 3.299119472503662,
|
| 62 |
+
"eval_runtime": 33.4042,
|
| 63 |
+
"eval_samples_per_second": 3.503,
|
| 64 |
+
"eval_steps_per_second": 1.766,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0007,
|
| 69 |
+
"eval_loss": 3.2837445735931396,
|
| 70 |
+
"eval_runtime": 33.3171,
|
| 71 |
+
"eval_samples_per_second": 3.512,
|
| 72 |
+
"eval_steps_per_second": 1.771,
|
| 73 |
+
"step": 35
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.0008,
|
| 77 |
+
"eval_loss": 3.26920747756958,
|
| 78 |
+
"eval_runtime": 33.2887,
|
| 79 |
+
"eval_samples_per_second": 3.515,
|
| 80 |
+
"eval_steps_per_second": 1.772,
|
| 81 |
+
"step": 40
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.0009,
|
| 85 |
+
"eval_loss": 3.2481868267059326,
|
| 86 |
+
"eval_runtime": 33.3291,
|
| 87 |
+
"eval_samples_per_second": 3.51,
|
| 88 |
+
"eval_steps_per_second": 1.77,
|
| 89 |
+
"step": 45
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.001,
|
| 93 |
+
"grad_norm": 0.5545255682809549,
|
| 94 |
+
"learning_rate": 9.8e-08,
|
| 95 |
+
"loss": 3.4174,
|
| 96 |
+
"step": 50
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.001,
|
| 100 |
+
"eval_loss": 3.2263057231903076,
|
| 101 |
+
"eval_runtime": 33.3242,
|
| 102 |
+
"eval_samples_per_second": 3.511,
|
| 103 |
+
"eval_steps_per_second": 1.77,
|
| 104 |
+
"step": 50
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.0011,
|
| 108 |
+
"eval_loss": 3.2074711322784424,
|
| 109 |
+
"eval_runtime": 33.3412,
|
| 110 |
+
"eval_samples_per_second": 3.509,
|
| 111 |
+
"eval_steps_per_second": 1.77,
|
| 112 |
+
"step": 55
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.0012,
|
| 116 |
+
"eval_loss": 3.1877729892730713,
|
| 117 |
+
"eval_runtime": 33.5109,
|
| 118 |
+
"eval_samples_per_second": 3.491,
|
| 119 |
+
"eval_steps_per_second": 1.761,
|
| 120 |
+
"step": 60
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.0013,
|
| 124 |
+
"eval_loss": 3.153503894805908,
|
| 125 |
+
"eval_runtime": 33.4747,
|
| 126 |
+
"eval_samples_per_second": 3.495,
|
| 127 |
+
"eval_steps_per_second": 1.763,
|
| 128 |
+
"step": 65
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0014,
|
| 132 |
+
"eval_loss": 3.1214191913604736,
|
| 133 |
+
"eval_runtime": 33.5956,
|
| 134 |
+
"eval_samples_per_second": 3.483,
|
| 135 |
+
"eval_steps_per_second": 1.756,
|
| 136 |
+
"step": 70
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.0015,
|
| 140 |
+
"grad_norm": 0.5083106511895727,
|
| 141 |
+
"learning_rate": 1.4800000000000003e-07,
|
| 142 |
+
"loss": 3.2951,
|
| 143 |
+
"step": 75
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.0015,
|
| 147 |
+
"eval_loss": 3.101821184158325,
|
| 148 |
+
"eval_runtime": 33.6,
|
| 149 |
+
"eval_samples_per_second": 3.482,
|
| 150 |
+
"eval_steps_per_second": 1.756,
|
| 151 |
+
"step": 75
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.0016,
|
| 155 |
+
"eval_loss": 3.0797102451324463,
|
| 156 |
+
"eval_runtime": 33.5302,
|
| 157 |
+
"eval_samples_per_second": 3.489,
|
| 158 |
+
"eval_steps_per_second": 1.76,
|
| 159 |
+
"step": 80
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.0017,
|
| 163 |
+
"eval_loss": 3.0523691177368164,
|
| 164 |
+
"eval_runtime": 33.5031,
|
| 165 |
+
"eval_samples_per_second": 3.492,
|
| 166 |
+
"eval_steps_per_second": 1.761,
|
| 167 |
+
"step": 85
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.0018,
|
| 171 |
+
"eval_loss": 3.022620677947998,
|
| 172 |
+
"eval_runtime": 33.6265,
|
| 173 |
+
"eval_samples_per_second": 3.479,
|
| 174 |
+
"eval_steps_per_second": 1.755,
|
| 175 |
+
"step": 90
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.0019,
|
| 179 |
+
"eval_loss": 2.991481065750122,
|
| 180 |
+
"eval_runtime": 33.5519,
|
| 181 |
+
"eval_samples_per_second": 3.487,
|
| 182 |
+
"eval_steps_per_second": 1.758,
|
| 183 |
+
"step": 95
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.002,
|
| 187 |
+
"grad_norm": 0.28367624064943,
|
| 188 |
+
"learning_rate": 1.9800000000000003e-07,
|
| 189 |
+
"loss": 3.1531,
|
| 190 |
+
"step": 100
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.002,
|
| 194 |
+
"eval_loss": 2.9630048274993896,
|
| 195 |
+
"eval_runtime": 33.734,
|
| 196 |
+
"eval_samples_per_second": 3.468,
|
| 197 |
+
"eval_steps_per_second": 1.749,
|
| 198 |
+
"step": 100
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.0021,
|
| 202 |
+
"eval_loss": 2.93916916847229,
|
| 203 |
+
"eval_runtime": 33.4897,
|
| 204 |
+
"eval_samples_per_second": 3.494,
|
| 205 |
+
"eval_steps_per_second": 1.762,
|
| 206 |
+
"step": 105
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.0022,
|
| 210 |
+
"eval_loss": 2.9186832904815674,
|
| 211 |
+
"eval_runtime": 33.5154,
|
| 212 |
+
"eval_samples_per_second": 3.491,
|
| 213 |
+
"eval_steps_per_second": 1.76,
|
| 214 |
+
"step": 110
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.0023,
|
| 218 |
+
"eval_loss": 2.8985302448272705,
|
| 219 |
+
"eval_runtime": 33.5846,
|
| 220 |
+
"eval_samples_per_second": 3.484,
|
| 221 |
+
"eval_steps_per_second": 1.757,
|
| 222 |
+
"step": 115
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.0024,
|
| 226 |
+
"eval_loss": 2.8786001205444336,
|
| 227 |
+
"eval_runtime": 33.5482,
|
| 228 |
+
"eval_samples_per_second": 3.488,
|
| 229 |
+
"eval_steps_per_second": 1.759,
|
| 230 |
+
"step": 120
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.0025,
|
| 234 |
+
"grad_norm": 0.19615444236413476,
|
| 235 |
+
"learning_rate": 2.48e-07,
|
| 236 |
+
"loss": 3.0101,
|
| 237 |
+
"step": 125
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.0025,
|
| 241 |
+
"eval_loss": 2.860034704208374,
|
| 242 |
+
"eval_runtime": 33.5143,
|
| 243 |
+
"eval_samples_per_second": 3.491,
|
| 244 |
+
"eval_steps_per_second": 1.76,
|
| 245 |
+
"step": 125
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.0026,
|
| 249 |
+
"eval_loss": 2.843663454055786,
|
| 250 |
+
"eval_runtime": 33.5082,
|
| 251 |
+
"eval_samples_per_second": 3.492,
|
| 252 |
+
"eval_steps_per_second": 1.761,
|
| 253 |
+
"step": 130
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.0027,
|
| 257 |
+
"eval_loss": 2.82882022857666,
|
| 258 |
+
"eval_runtime": 33.4921,
|
| 259 |
+
"eval_samples_per_second": 3.493,
|
| 260 |
+
"eval_steps_per_second": 1.762,
|
| 261 |
+
"step": 135
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.0028,
|
| 265 |
+
"eval_loss": 2.8154728412628174,
|
| 266 |
+
"eval_runtime": 33.6656,
|
| 267 |
+
"eval_samples_per_second": 3.475,
|
| 268 |
+
"eval_steps_per_second": 1.753,
|
| 269 |
+
"step": 140
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.0029,
|
| 273 |
+
"eval_loss": 2.801098346710205,
|
| 274 |
+
"eval_runtime": 33.5229,
|
| 275 |
+
"eval_samples_per_second": 3.49,
|
| 276 |
+
"eval_steps_per_second": 1.76,
|
| 277 |
+
"step": 145
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.003,
|
| 281 |
+
"grad_norm": 0.5710572013823593,
|
| 282 |
+
"learning_rate": 2.9800000000000005e-07,
|
| 283 |
+
"loss": 2.876,
|
| 284 |
+
"step": 150
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.003,
|
| 288 |
+
"eval_loss": 2.789198160171509,
|
| 289 |
+
"eval_runtime": 33.4535,
|
| 290 |
+
"eval_samples_per_second": 3.497,
|
| 291 |
+
"eval_steps_per_second": 1.764,
|
| 292 |
+
"step": 150
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.0031,
|
| 296 |
+
"eval_loss": 2.7789695262908936,
|
| 297 |
+
"eval_runtime": 33.6409,
|
| 298 |
+
"eval_samples_per_second": 3.478,
|
| 299 |
+
"eval_steps_per_second": 1.754,
|
| 300 |
+
"step": 155
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.0032,
|
| 304 |
+
"eval_loss": 2.7694201469421387,
|
| 305 |
+
"eval_runtime": 33.4266,
|
| 306 |
+
"eval_samples_per_second": 3.5,
|
| 307 |
+
"eval_steps_per_second": 1.765,
|
| 308 |
+
"step": 160
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.0033,
|
| 312 |
+
"eval_loss": 2.7600762844085693,
|
| 313 |
+
"eval_runtime": 33.4725,
|
| 314 |
+
"eval_samples_per_second": 3.495,
|
| 315 |
+
"eval_steps_per_second": 1.763,
|
| 316 |
+
"step": 165
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.0034,
|
| 320 |
+
"eval_loss": 2.7517828941345215,
|
| 321 |
+
"eval_runtime": 33.6223,
|
| 322 |
+
"eval_samples_per_second": 3.48,
|
| 323 |
+
"eval_steps_per_second": 1.755,
|
| 324 |
+
"step": 170
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.0035,
|
| 328 |
+
"grad_norm": 0.151307501186972,
|
| 329 |
+
"learning_rate": 3.48e-07,
|
| 330 |
+
"loss": 2.811,
|
| 331 |
+
"step": 175
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.0035,
|
| 335 |
+
"eval_loss": 2.743870258331299,
|
| 336 |
+
"eval_runtime": 33.5221,
|
| 337 |
+
"eval_samples_per_second": 3.49,
|
| 338 |
+
"eval_steps_per_second": 1.76,
|
| 339 |
+
"step": 175
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.0036,
|
| 343 |
+
"eval_loss": 2.7366557121276855,
|
| 344 |
+
"eval_runtime": 33.5448,
|
| 345 |
+
"eval_samples_per_second": 3.488,
|
| 346 |
+
"eval_steps_per_second": 1.759,
|
| 347 |
+
"step": 180
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.0037,
|
| 351 |
+
"eval_loss": 2.7298200130462646,
|
| 352 |
+
"eval_runtime": 33.5428,
|
| 353 |
+
"eval_samples_per_second": 3.488,
|
| 354 |
+
"eval_steps_per_second": 1.759,
|
| 355 |
+
"step": 185
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.0038,
|
| 359 |
+
"eval_loss": 2.722888708114624,
|
| 360 |
+
"eval_runtime": 33.6302,
|
| 361 |
+
"eval_samples_per_second": 3.479,
|
| 362 |
+
"eval_steps_per_second": 1.754,
|
| 363 |
+
"step": 190
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.0039,
|
| 367 |
+
"eval_loss": 2.714289426803589,
|
| 368 |
+
"eval_runtime": 33.5594,
|
| 369 |
+
"eval_samples_per_second": 3.486,
|
| 370 |
+
"eval_steps_per_second": 1.758,
|
| 371 |
+
"step": 195
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.004,
|
| 375 |
+
"grad_norm": 0.10362348542700331,
|
| 376 |
+
"learning_rate": 3.9800000000000004e-07,
|
| 377 |
+
"loss": 2.7606,
|
| 378 |
+
"step": 200
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.004,
|
| 382 |
+
"eval_loss": 2.7078425884246826,
|
| 383 |
+
"eval_runtime": 33.6447,
|
| 384 |
+
"eval_samples_per_second": 3.478,
|
| 385 |
+
"eval_steps_per_second": 1.754,
|
| 386 |
+
"step": 200
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.0041,
|
| 390 |
+
"eval_loss": 2.7014663219451904,
|
| 391 |
+
"eval_runtime": 33.565,
|
| 392 |
+
"eval_samples_per_second": 3.486,
|
| 393 |
+
"eval_steps_per_second": 1.758,
|
| 394 |
+
"step": 205
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.0042,
|
| 398 |
+
"eval_loss": 2.6956119537353516,
|
| 399 |
+
"eval_runtime": 33.5938,
|
| 400 |
+
"eval_samples_per_second": 3.483,
|
| 401 |
+
"eval_steps_per_second": 1.756,
|
| 402 |
+
"step": 210
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.0043,
|
| 406 |
+
"eval_loss": 2.6901819705963135,
|
| 407 |
+
"eval_runtime": 33.5009,
|
| 408 |
+
"eval_samples_per_second": 3.492,
|
| 409 |
+
"eval_steps_per_second": 1.761,
|
| 410 |
+
"step": 215
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.0044,
|
| 414 |
+
"eval_loss": 2.684842824935913,
|
| 415 |
+
"eval_runtime": 33.5857,
|
| 416 |
+
"eval_samples_per_second": 3.484,
|
| 417 |
+
"eval_steps_per_second": 1.757,
|
| 418 |
+
"step": 220
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.0045,
|
| 422 |
+
"grad_norm": 0.08395542059093342,
|
| 423 |
+
"learning_rate": 4.4800000000000004e-07,
|
| 424 |
+
"loss": 2.727,
|
| 425 |
+
"step": 225
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.0045,
|
| 429 |
+
"eval_loss": 2.679893732070923,
|
| 430 |
+
"eval_runtime": 33.5333,
|
| 431 |
+
"eval_samples_per_second": 3.489,
|
| 432 |
+
"eval_steps_per_second": 1.759,
|
| 433 |
+
"step": 225
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.0046,
|
| 437 |
+
"eval_loss": 2.6749234199523926,
|
| 438 |
+
"eval_runtime": 33.6847,
|
| 439 |
+
"eval_samples_per_second": 3.473,
|
| 440 |
+
"eval_steps_per_second": 1.752,
|
| 441 |
+
"step": 230
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.0047,
|
| 445 |
+
"eval_loss": 2.670543670654297,
|
| 446 |
+
"eval_runtime": 33.5814,
|
| 447 |
+
"eval_samples_per_second": 3.484,
|
| 448 |
+
"eval_steps_per_second": 1.757,
|
| 449 |
+
"step": 235
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.0048,
|
| 453 |
+
"eval_loss": 2.6663973331451416,
|
| 454 |
+
"eval_runtime": 33.5943,
|
| 455 |
+
"eval_samples_per_second": 3.483,
|
| 456 |
+
"eval_steps_per_second": 1.756,
|
| 457 |
+
"step": 240
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.0049,
|
| 461 |
+
"eval_loss": 2.662304162979126,
|
| 462 |
+
"eval_runtime": 33.5309,
|
| 463 |
+
"eval_samples_per_second": 3.489,
|
| 464 |
+
"eval_steps_per_second": 1.76,
|
| 465 |
+
"step": 245
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.005,
|
| 469 |
+
"grad_norm": 0.06968304771462097,
|
| 470 |
+
"learning_rate": 4.98e-07,
|
| 471 |
+
"loss": 2.6931,
|
| 472 |
+
"step": 250
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.005,
|
| 476 |
+
"eval_loss": 2.65859317779541,
|
| 477 |
+
"eval_runtime": 33.4663,
|
| 478 |
+
"eval_samples_per_second": 3.496,
|
| 479 |
+
"eval_steps_per_second": 1.763,
|
| 480 |
+
"step": 250
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.0051,
|
| 484 |
+
"eval_loss": 2.654831886291504,
|
| 485 |
+
"eval_runtime": 33.5962,
|
| 486 |
+
"eval_samples_per_second": 3.483,
|
| 487 |
+
"eval_steps_per_second": 1.756,
|
| 488 |
+
"step": 255
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.0052,
|
| 492 |
+
"eval_loss": 2.6509766578674316,
|
| 493 |
+
"eval_runtime": 33.5064,
|
| 494 |
+
"eval_samples_per_second": 3.492,
|
| 495 |
+
"eval_steps_per_second": 1.761,
|
| 496 |
+
"step": 260
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.0053,
|
| 500 |
+
"eval_loss": 2.6467387676239014,
|
| 501 |
+
"eval_runtime": 33.5346,
|
| 502 |
+
"eval_samples_per_second": 3.489,
|
| 503 |
+
"eval_steps_per_second": 1.759,
|
| 504 |
+
"step": 265
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.0054,
|
| 508 |
+
"eval_loss": 2.6428205966949463,
|
| 509 |
+
"eval_runtime": 33.5418,
|
| 510 |
+
"eval_samples_per_second": 3.488,
|
| 511 |
+
"eval_steps_per_second": 1.759,
|
| 512 |
+
"step": 270
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.0055,
|
| 516 |
+
"grad_norm": 0.05704195405230526,
|
| 517 |
+
"learning_rate": 5.480000000000001e-07,
|
| 518 |
+
"loss": 2.674,
|
| 519 |
+
"step": 275
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.0055,
|
| 523 |
+
"eval_loss": 2.6392645835876465,
|
| 524 |
+
"eval_runtime": 33.6509,
|
| 525 |
+
"eval_samples_per_second": 3.477,
|
| 526 |
+
"eval_steps_per_second": 1.753,
|
| 527 |
+
"step": 275
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.0056,
|
| 531 |
+
"eval_loss": 2.6361024379730225,
|
| 532 |
+
"eval_runtime": 33.6973,
|
| 533 |
+
"eval_samples_per_second": 3.472,
|
| 534 |
+
"eval_steps_per_second": 1.751,
|
| 535 |
+
"step": 280
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.0057,
|
| 539 |
+
"eval_loss": 2.6328718662261963,
|
| 540 |
+
"eval_runtime": 33.5639,
|
| 541 |
+
"eval_samples_per_second": 3.486,
|
| 542 |
+
"eval_steps_per_second": 1.758,
|
| 543 |
+
"step": 285
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.0058,
|
| 547 |
+
"eval_loss": 2.629871129989624,
|
| 548 |
+
"eval_runtime": 33.5243,
|
| 549 |
+
"eval_samples_per_second": 3.49,
|
| 550 |
+
"eval_steps_per_second": 1.76,
|
| 551 |
+
"step": 290
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.0059,
|
| 555 |
+
"eval_loss": 2.6271257400512695,
|
| 556 |
+
"eval_runtime": 33.6427,
|
| 557 |
+
"eval_samples_per_second": 3.478,
|
| 558 |
+
"eval_steps_per_second": 1.754,
|
| 559 |
+
"step": 295
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.006,
|
| 563 |
+
"grad_norm": 0.05013991368613539,
|
| 564 |
+
"learning_rate": 5.98e-07,
|
| 565 |
+
"loss": 2.6504,
|
| 566 |
+
"step": 300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.006,
|
| 570 |
+
"eval_loss": 2.6243784427642822,
|
| 571 |
+
"eval_runtime": 33.5815,
|
| 572 |
+
"eval_samples_per_second": 3.484,
|
| 573 |
+
"eval_steps_per_second": 1.757,
|
| 574 |
+
"step": 300
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 0.0061,
|
| 578 |
+
"eval_loss": 2.621882915496826,
|
| 579 |
+
"eval_runtime": 33.7331,
|
| 580 |
+
"eval_samples_per_second": 3.468,
|
| 581 |
+
"eval_steps_per_second": 1.749,
|
| 582 |
+
"step": 305
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.0062,
|
| 586 |
+
"eval_loss": 2.6194233894348145,
|
| 587 |
+
"eval_runtime": 33.594,
|
| 588 |
+
"eval_samples_per_second": 3.483,
|
| 589 |
+
"eval_steps_per_second": 1.756,
|
| 590 |
+
"step": 310
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.0063,
|
| 594 |
+
"eval_loss": 2.6167914867401123,
|
| 595 |
+
"eval_runtime": 33.5521,
|
| 596 |
+
"eval_samples_per_second": 3.487,
|
| 597 |
+
"eval_steps_per_second": 1.758,
|
| 598 |
+
"step": 315
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.0064,
|
| 602 |
+
"eval_loss": 2.6143040657043457,
|
| 603 |
+
"eval_runtime": 33.58,
|
| 604 |
+
"eval_samples_per_second": 3.484,
|
| 605 |
+
"eval_steps_per_second": 1.757,
|
| 606 |
+
"step": 320
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.0065,
|
| 610 |
+
"grad_norm": 0.04696879401248375,
|
| 611 |
+
"learning_rate": 6.48e-07,
|
| 612 |
+
"loss": 2.6372,
|
| 613 |
+
"step": 325
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.0065,
|
| 617 |
+
"eval_loss": 2.611804246902466,
|
| 618 |
+
"eval_runtime": 33.5371,
|
| 619 |
+
"eval_samples_per_second": 3.489,
|
| 620 |
+
"eval_steps_per_second": 1.759,
|
| 621 |
+
"step": 325
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.0066,
|
| 625 |
+
"eval_loss": 2.6093685626983643,
|
| 626 |
+
"eval_runtime": 33.8057,
|
| 627 |
+
"eval_samples_per_second": 3.461,
|
| 628 |
+
"eval_steps_per_second": 1.745,
|
| 629 |
+
"step": 330
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.0067,
|
| 633 |
+
"eval_loss": 2.607069492340088,
|
| 634 |
+
"eval_runtime": 33.5819,
|
| 635 |
+
"eval_samples_per_second": 3.484,
|
| 636 |
+
"eval_steps_per_second": 1.757,
|
| 637 |
+
"step": 335
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 0.0068,
|
| 641 |
+
"eval_loss": 2.604562520980835,
|
| 642 |
+
"eval_runtime": 33.5971,
|
| 643 |
+
"eval_samples_per_second": 3.482,
|
| 644 |
+
"eval_steps_per_second": 1.756,
|
| 645 |
+
"step": 340
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.0069,
|
| 649 |
+
"eval_loss": 2.6024069786071777,
|
| 650 |
+
"eval_runtime": 33.5107,
|
| 651 |
+
"eval_samples_per_second": 3.491,
|
| 652 |
+
"eval_steps_per_second": 1.761,
|
| 653 |
+
"step": 345
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.007,
|
| 657 |
+
"grad_norm": 0.04335213523196003,
|
| 658 |
+
"learning_rate": 6.98e-07,
|
| 659 |
+
"loss": 2.6173,
|
| 660 |
+
"step": 350
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.007,
|
| 664 |
+
"eval_loss": 2.6002795696258545,
|
| 665 |
+
"eval_runtime": 33.6194,
|
| 666 |
+
"eval_samples_per_second": 3.48,
|
| 667 |
+
"eval_steps_per_second": 1.755,
|
| 668 |
+
"step": 350
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.0071,
|
| 672 |
+
"eval_loss": 2.598109245300293,
|
| 673 |
+
"eval_runtime": 33.807,
|
| 674 |
+
"eval_samples_per_second": 3.461,
|
| 675 |
+
"eval_steps_per_second": 1.745,
|
| 676 |
+
"step": 355
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.0072,
|
| 680 |
+
"eval_loss": 2.596126079559326,
|
| 681 |
+
"eval_runtime": 33.5287,
|
| 682 |
+
"eval_samples_per_second": 3.49,
|
| 683 |
+
"eval_steps_per_second": 1.76,
|
| 684 |
+
"step": 360
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.0073,
|
| 688 |
+
"eval_loss": 2.5941832065582275,
|
| 689 |
+
"eval_runtime": 33.5456,
|
| 690 |
+
"eval_samples_per_second": 3.488,
|
| 691 |
+
"eval_steps_per_second": 1.759,
|
| 692 |
+
"step": 365
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.0074,
|
| 696 |
+
"eval_loss": 2.592336893081665,
|
| 697 |
+
"eval_runtime": 33.6972,
|
| 698 |
+
"eval_samples_per_second": 3.472,
|
| 699 |
+
"eval_steps_per_second": 1.751,
|
| 700 |
+
"step": 370
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 0.0075,
|
| 704 |
+
"grad_norm": 0.04553004087145917,
|
| 705 |
+
"learning_rate": 7.480000000000001e-07,
|
| 706 |
+
"loss": 2.608,
|
| 707 |
+
"step": 375
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.0075,
|
| 711 |
+
"eval_loss": 2.590573310852051,
|
| 712 |
+
"eval_runtime": 33.6132,
|
| 713 |
+
"eval_samples_per_second": 3.481,
|
| 714 |
+
"eval_steps_per_second": 1.755,
|
| 715 |
+
"step": 375
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.0076,
|
| 719 |
+
"eval_loss": 2.5888302326202393,
|
| 720 |
+
"eval_runtime": 33.6363,
|
| 721 |
+
"eval_samples_per_second": 3.478,
|
| 722 |
+
"eval_steps_per_second": 1.754,
|
| 723 |
+
"step": 380
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.0077,
|
| 727 |
+
"eval_loss": 2.5870487689971924,
|
| 728 |
+
"eval_runtime": 33.6309,
|
| 729 |
+
"eval_samples_per_second": 3.479,
|
| 730 |
+
"eval_steps_per_second": 1.754,
|
| 731 |
+
"step": 385
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.0078,
|
| 735 |
+
"eval_loss": 2.5851986408233643,
|
| 736 |
+
"eval_runtime": 33.5237,
|
| 737 |
+
"eval_samples_per_second": 3.49,
|
| 738 |
+
"eval_steps_per_second": 1.76,
|
| 739 |
+
"step": 390
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.0079,
|
| 743 |
+
"eval_loss": 2.583341598510742,
|
| 744 |
+
"eval_runtime": 33.4914,
|
| 745 |
+
"eval_samples_per_second": 3.493,
|
| 746 |
+
"eval_steps_per_second": 1.762,
|
| 747 |
+
"step": 395
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.008,
|
| 751 |
+
"grad_norm": 0.04067489002091025,
|
| 752 |
+
"learning_rate": 7.98e-07,
|
| 753 |
+
"loss": 2.6034,
|
| 754 |
+
"step": 400
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.008,
|
| 758 |
+
"eval_loss": 2.5816242694854736,
|
| 759 |
+
"eval_runtime": 33.6305,
|
| 760 |
+
"eval_samples_per_second": 3.479,
|
| 761 |
+
"eval_steps_per_second": 1.754,
|
| 762 |
+
"step": 400
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.0081,
|
| 766 |
+
"eval_loss": 2.5800209045410156,
|
| 767 |
+
"eval_runtime": 33.9049,
|
| 768 |
+
"eval_samples_per_second": 3.451,
|
| 769 |
+
"eval_steps_per_second": 1.74,
|
| 770 |
+
"step": 405
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.0082,
|
| 774 |
+
"eval_loss": 2.5783472061157227,
|
| 775 |
+
"eval_runtime": 33.6847,
|
| 776 |
+
"eval_samples_per_second": 3.473,
|
| 777 |
+
"eval_steps_per_second": 1.752,
|
| 778 |
+
"step": 410
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.0083,
|
| 782 |
+
"eval_loss": 2.5765581130981445,
|
| 783 |
+
"eval_runtime": 33.5467,
|
| 784 |
+
"eval_samples_per_second": 3.488,
|
| 785 |
+
"eval_steps_per_second": 1.759,
|
| 786 |
+
"step": 415
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.0084,
|
| 790 |
+
"eval_loss": 2.574805974960327,
|
| 791 |
+
"eval_runtime": 33.6837,
|
| 792 |
+
"eval_samples_per_second": 3.473,
|
| 793 |
+
"eval_steps_per_second": 1.752,
|
| 794 |
+
"step": 420
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.0085,
|
| 798 |
+
"grad_norm": 0.03957021300313725,
|
| 799 |
+
"learning_rate": 8.480000000000001e-07,
|
| 800 |
+
"loss": 2.5881,
|
| 801 |
+
"step": 425
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.0085,
|
| 805 |
+
"eval_loss": 2.5732243061065674,
|
| 806 |
+
"eval_runtime": 33.6883,
|
| 807 |
+
"eval_samples_per_second": 3.473,
|
| 808 |
+
"eval_steps_per_second": 1.751,
|
| 809 |
+
"step": 425
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.0086,
|
| 813 |
+
"eval_loss": 2.5712339878082275,
|
| 814 |
+
"eval_runtime": 34.0087,
|
| 815 |
+
"eval_samples_per_second": 3.44,
|
| 816 |
+
"eval_steps_per_second": 1.735,
|
| 817 |
+
"step": 430
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.0087,
|
| 821 |
+
"eval_loss": 2.5696043968200684,
|
| 822 |
+
"eval_runtime": 33.5522,
|
| 823 |
+
"eval_samples_per_second": 3.487,
|
| 824 |
+
"eval_steps_per_second": 1.758,
|
| 825 |
+
"step": 435
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.0088,
|
| 829 |
+
"eval_loss": 2.568011522293091,
|
| 830 |
+
"eval_runtime": 33.7026,
|
| 831 |
+
"eval_samples_per_second": 3.472,
|
| 832 |
+
"eval_steps_per_second": 1.751,
|
| 833 |
+
"step": 440
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.0089,
|
| 837 |
+
"eval_loss": 2.5661723613739014,
|
| 838 |
+
"eval_runtime": 33.7143,
|
| 839 |
+
"eval_samples_per_second": 3.47,
|
| 840 |
+
"eval_steps_per_second": 1.75,
|
| 841 |
+
"step": 445
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.009,
|
| 845 |
+
"grad_norm": 0.04518058243135632,
|
| 846 |
+
"learning_rate": 8.980000000000001e-07,
|
| 847 |
+
"loss": 2.577,
|
| 848 |
+
"step": 450
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.009,
|
| 852 |
+
"eval_loss": 2.5647170543670654,
|
| 853 |
+
"eval_runtime": 33.6066,
|
| 854 |
+
"eval_samples_per_second": 3.481,
|
| 855 |
+
"eval_steps_per_second": 1.756,
|
| 856 |
+
"step": 450
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.0091,
|
| 860 |
+
"eval_loss": 2.5629138946533203,
|
| 861 |
+
"eval_runtime": 33.695,
|
| 862 |
+
"eval_samples_per_second": 3.472,
|
| 863 |
+
"eval_steps_per_second": 1.751,
|
| 864 |
+
"step": 455
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.0092,
|
| 868 |
+
"eval_loss": 2.561223268508911,
|
| 869 |
+
"eval_runtime": 33.7639,
|
| 870 |
+
"eval_samples_per_second": 3.465,
|
| 871 |
+
"eval_steps_per_second": 1.747,
|
| 872 |
+
"step": 460
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.0093,
|
| 876 |
+
"eval_loss": 2.559941053390503,
|
| 877 |
+
"eval_runtime": 33.5726,
|
| 878 |
+
"eval_samples_per_second": 3.485,
|
| 879 |
+
"eval_steps_per_second": 1.757,
|
| 880 |
+
"step": 465
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.0094,
|
| 884 |
+
"eval_loss": 2.5585126876831055,
|
| 885 |
+
"eval_runtime": 33.5393,
|
| 886 |
+
"eval_samples_per_second": 3.488,
|
| 887 |
+
"eval_steps_per_second": 1.759,
|
| 888 |
+
"step": 470
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.0095,
|
| 892 |
+
"grad_norm": 0.04841685552742973,
|
| 893 |
+
"learning_rate": 9.480000000000001e-07,
|
| 894 |
+
"loss": 2.5614,
|
| 895 |
+
"step": 475
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 0.0095,
|
| 899 |
+
"eval_loss": 2.557070732116699,
|
| 900 |
+
"eval_runtime": 33.5396,
|
| 901 |
+
"eval_samples_per_second": 3.488,
|
| 902 |
+
"eval_steps_per_second": 1.759,
|
| 903 |
+
"step": 475
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.0096,
|
| 907 |
+
"eval_loss": 2.5551016330718994,
|
| 908 |
+
"eval_runtime": 33.8951,
|
| 909 |
+
"eval_samples_per_second": 3.452,
|
| 910 |
+
"eval_steps_per_second": 1.741,
|
| 911 |
+
"step": 480
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.0097,
|
| 915 |
+
"eval_loss": 2.553600311279297,
|
| 916 |
+
"eval_runtime": 33.6678,
|
| 917 |
+
"eval_samples_per_second": 3.475,
|
| 918 |
+
"eval_steps_per_second": 1.752,
|
| 919 |
+
"step": 485
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.0098,
|
| 923 |
+
"eval_loss": 2.5523183345794678,
|
| 924 |
+
"eval_runtime": 33.6551,
|
| 925 |
+
"eval_samples_per_second": 3.476,
|
| 926 |
+
"eval_steps_per_second": 1.753,
|
| 927 |
+
"step": 490
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.0099,
|
| 931 |
+
"eval_loss": 2.5510056018829346,
|
| 932 |
+
"eval_runtime": 33.6214,
|
| 933 |
+
"eval_samples_per_second": 3.48,
|
| 934 |
+
"eval_steps_per_second": 1.755,
|
| 935 |
+
"step": 495
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.01,
|
| 939 |
+
"grad_norm": 0.043993000876628545,
|
| 940 |
+
"learning_rate": 9.98e-07,
|
| 941 |
+
"loss": 2.5613,
|
| 942 |
+
"step": 500
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.01,
|
| 946 |
+
"eval_loss": 2.5498273372650146,
|
| 947 |
+
"eval_runtime": 33.6069,
|
| 948 |
+
"eval_samples_per_second": 3.481,
|
| 949 |
+
"eval_steps_per_second": 1.756,
|
| 950 |
+
"step": 500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.0101,
|
| 954 |
+
"eval_loss": 2.548828601837158,
|
| 955 |
+
"eval_runtime": 33.7909,
|
| 956 |
+
"eval_samples_per_second": 3.462,
|
| 957 |
+
"eval_steps_per_second": 1.746,
|
| 958 |
+
"step": 505
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 0.0102,
|
| 962 |
+
"eval_loss": 2.5474376678466797,
|
| 963 |
+
"eval_runtime": 33.543,
|
| 964 |
+
"eval_samples_per_second": 3.488,
|
| 965 |
+
"eval_steps_per_second": 1.759,
|
| 966 |
+
"step": 510
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.0103,
|
| 970 |
+
"eval_loss": 2.5464441776275635,
|
| 971 |
+
"eval_runtime": 33.6579,
|
| 972 |
+
"eval_samples_per_second": 3.476,
|
| 973 |
+
"eval_steps_per_second": 1.753,
|
| 974 |
+
"step": 515
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.0104,
|
| 978 |
+
"eval_loss": 2.5453498363494873,
|
| 979 |
+
"eval_runtime": 33.4841,
|
| 980 |
+
"eval_samples_per_second": 3.494,
|
| 981 |
+
"eval_steps_per_second": 1.762,
|
| 982 |
+
"step": 520
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.0105,
|
| 986 |
+
"grad_norm": 0.04663602312001795,
|
| 987 |
+
"learning_rate": 1.0480000000000002e-06,
|
| 988 |
+
"loss": 2.5521,
|
| 989 |
+
"step": 525
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.0105,
|
| 993 |
+
"eval_loss": 2.5442492961883545,
|
| 994 |
+
"eval_runtime": 33.5915,
|
| 995 |
+
"eval_samples_per_second": 3.483,
|
| 996 |
+
"eval_steps_per_second": 1.756,
|
| 997 |
+
"step": 525
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.0106,
|
| 1001 |
+
"eval_loss": 2.5432002544403076,
|
| 1002 |
+
"eval_runtime": 33.6717,
|
| 1003 |
+
"eval_samples_per_second": 3.475,
|
| 1004 |
+
"eval_steps_per_second": 1.752,
|
| 1005 |
+
"step": 530
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.0107,
|
| 1009 |
+
"eval_loss": 2.542072057723999,
|
| 1010 |
+
"eval_runtime": 33.6153,
|
| 1011 |
+
"eval_samples_per_second": 3.481,
|
| 1012 |
+
"eval_steps_per_second": 1.755,
|
| 1013 |
+
"step": 535
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.0108,
|
| 1017 |
+
"eval_loss": 2.541541814804077,
|
| 1018 |
+
"eval_runtime": 34.4505,
|
| 1019 |
+
"eval_samples_per_second": 3.396,
|
| 1020 |
+
"eval_steps_per_second": 1.713,
|
| 1021 |
+
"step": 540
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 0.0109,
|
| 1025 |
+
"eval_loss": 2.540494203567505,
|
| 1026 |
+
"eval_runtime": 33.6369,
|
| 1027 |
+
"eval_samples_per_second": 3.478,
|
| 1028 |
+
"eval_steps_per_second": 1.754,
|
| 1029 |
+
"step": 545
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.011,
|
| 1033 |
+
"grad_norm": 0.044473565671350655,
|
| 1034 |
+
"learning_rate": 1.0980000000000001e-06,
|
| 1035 |
+
"loss": 2.5433,
|
| 1036 |
+
"step": 550
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.011,
|
| 1040 |
+
"eval_loss": 2.539369821548462,
|
| 1041 |
+
"eval_runtime": 33.5742,
|
| 1042 |
+
"eval_samples_per_second": 3.485,
|
| 1043 |
+
"eval_steps_per_second": 1.757,
|
| 1044 |
+
"step": 550
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.0111,
|
| 1048 |
+
"eval_loss": 2.5384223461151123,
|
| 1049 |
+
"eval_runtime": 33.9094,
|
| 1050 |
+
"eval_samples_per_second": 3.45,
|
| 1051 |
+
"eval_steps_per_second": 1.74,
|
| 1052 |
+
"step": 555
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.0112,
|
| 1056 |
+
"eval_loss": 2.5375945568084717,
|
| 1057 |
+
"eval_runtime": 33.6016,
|
| 1058 |
+
"eval_samples_per_second": 3.482,
|
| 1059 |
+
"eval_steps_per_second": 1.756,
|
| 1060 |
+
"step": 560
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.0113,
|
| 1064 |
+
"eval_loss": 2.536487340927124,
|
| 1065 |
+
"eval_runtime": 34.3561,
|
| 1066 |
+
"eval_samples_per_second": 3.406,
|
| 1067 |
+
"eval_steps_per_second": 1.717,
|
| 1068 |
+
"step": 565
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.0114,
|
| 1072 |
+
"eval_loss": 2.5356836318969727,
|
| 1073 |
+
"eval_runtime": 34.5074,
|
| 1074 |
+
"eval_samples_per_second": 3.391,
|
| 1075 |
+
"eval_steps_per_second": 1.71,
|
| 1076 |
+
"step": 570
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.0115,
|
| 1080 |
+
"grad_norm": 0.04668528198599521,
|
| 1081 |
+
"learning_rate": 1.148e-06,
|
| 1082 |
+
"loss": 2.5496,
|
| 1083 |
+
"step": 575
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.0115,
|
| 1087 |
+
"eval_loss": 2.5347819328308105,
|
| 1088 |
+
"eval_runtime": 33.5932,
|
| 1089 |
+
"eval_samples_per_second": 3.483,
|
| 1090 |
+
"eval_steps_per_second": 1.756,
|
| 1091 |
+
"step": 575
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.0116,
|
| 1095 |
+
"eval_loss": 2.534010410308838,
|
| 1096 |
+
"eval_runtime": 33.8124,
|
| 1097 |
+
"eval_samples_per_second": 3.46,
|
| 1098 |
+
"eval_steps_per_second": 1.745,
|
| 1099 |
+
"step": 580
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 0.0117,
|
| 1103 |
+
"eval_loss": 2.5331332683563232,
|
| 1104 |
+
"eval_runtime": 33.5617,
|
| 1105 |
+
"eval_samples_per_second": 3.486,
|
| 1106 |
+
"eval_steps_per_second": 1.758,
|
| 1107 |
+
"step": 585
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.0118,
|
| 1111 |
+
"eval_loss": 2.5322561264038086,
|
| 1112 |
+
"eval_runtime": 33.8081,
|
| 1113 |
+
"eval_samples_per_second": 3.461,
|
| 1114 |
+
"eval_steps_per_second": 1.745,
|
| 1115 |
+
"step": 590
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.0119,
|
| 1119 |
+
"eval_loss": 2.5314669609069824,
|
| 1120 |
+
"eval_runtime": 33.7053,
|
| 1121 |
+
"eval_samples_per_second": 3.471,
|
| 1122 |
+
"eval_steps_per_second": 1.75,
|
| 1123 |
+
"step": 595
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.012,
|
| 1127 |
+
"grad_norm": 0.043769011241975755,
|
| 1128 |
+
"learning_rate": 1.1980000000000002e-06,
|
| 1129 |
+
"loss": 2.5455,
|
| 1130 |
+
"step": 600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.012,
|
| 1134 |
+
"eval_loss": 2.5307207107543945,
|
| 1135 |
+
"eval_runtime": 33.6848,
|
| 1136 |
+
"eval_samples_per_second": 3.473,
|
| 1137 |
+
"eval_steps_per_second": 1.752,
|
| 1138 |
+
"step": 600
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.0121,
|
| 1142 |
+
"eval_loss": 2.530006170272827,
|
| 1143 |
+
"eval_runtime": 33.686,
|
| 1144 |
+
"eval_samples_per_second": 3.473,
|
| 1145 |
+
"eval_steps_per_second": 1.751,
|
| 1146 |
+
"step": 605
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.0122,
|
| 1150 |
+
"eval_loss": 2.529109239578247,
|
| 1151 |
+
"eval_runtime": 33.7013,
|
| 1152 |
+
"eval_samples_per_second": 3.472,
|
| 1153 |
+
"eval_steps_per_second": 1.751,
|
| 1154 |
+
"step": 610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.0123,
|
| 1158 |
+
"eval_loss": 2.5284457206726074,
|
| 1159 |
+
"eval_runtime": 33.6733,
|
| 1160 |
+
"eval_samples_per_second": 3.475,
|
| 1161 |
+
"eval_steps_per_second": 1.752,
|
| 1162 |
+
"step": 615
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 0.0124,
|
| 1166 |
+
"eval_loss": 2.5276710987091064,
|
| 1167 |
+
"eval_runtime": 33.624,
|
| 1168 |
+
"eval_samples_per_second": 3.48,
|
| 1169 |
+
"eval_steps_per_second": 1.755,
|
| 1170 |
+
"step": 620
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 0.0125,
|
| 1174 |
+
"grad_norm": 0.04196307636615052,
|
| 1175 |
+
"learning_rate": 1.248e-06,
|
| 1176 |
+
"loss": 2.5273,
|
| 1177 |
+
"step": 625
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.0125,
|
| 1181 |
+
"eval_loss": 2.526918411254883,
|
| 1182 |
+
"eval_runtime": 33.5952,
|
| 1183 |
+
"eval_samples_per_second": 3.483,
|
| 1184 |
+
"eval_steps_per_second": 1.756,
|
| 1185 |
+
"step": 625
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.0126,
|
| 1189 |
+
"eval_loss": 2.5262696743011475,
|
| 1190 |
+
"eval_runtime": 33.7522,
|
| 1191 |
+
"eval_samples_per_second": 3.466,
|
| 1192 |
+
"eval_steps_per_second": 1.748,
|
| 1193 |
+
"step": 630
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.0127,
|
| 1197 |
+
"eval_loss": 2.5255067348480225,
|
| 1198 |
+
"eval_runtime": 33.7929,
|
| 1199 |
+
"eval_samples_per_second": 3.462,
|
| 1200 |
+
"eval_steps_per_second": 1.746,
|
| 1201 |
+
"step": 635
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.0128,
|
| 1205 |
+
"eval_loss": 2.524789810180664,
|
| 1206 |
+
"eval_runtime": 33.7139,
|
| 1207 |
+
"eval_samples_per_second": 3.47,
|
| 1208 |
+
"eval_steps_per_second": 1.75,
|
| 1209 |
+
"step": 640
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.0129,
|
| 1213 |
+
"eval_loss": 2.524181604385376,
|
| 1214 |
+
"eval_runtime": 33.7772,
|
| 1215 |
+
"eval_samples_per_second": 3.464,
|
| 1216 |
+
"eval_steps_per_second": 1.747,
|
| 1217 |
+
"step": 645
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.013,
|
| 1221 |
+
"grad_norm": 0.04719575571491393,
|
| 1222 |
+
"learning_rate": 1.2980000000000001e-06,
|
| 1223 |
+
"loss": 2.5226,
|
| 1224 |
+
"step": 650
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.013,
|
| 1228 |
+
"eval_loss": 2.5235090255737305,
|
| 1229 |
+
"eval_runtime": 33.6972,
|
| 1230 |
+
"eval_samples_per_second": 3.472,
|
| 1231 |
+
"eval_steps_per_second": 1.751,
|
| 1232 |
+
"step": 650
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.0131,
|
| 1236 |
+
"eval_loss": 2.5227887630462646,
|
| 1237 |
+
"eval_runtime": 33.8073,
|
| 1238 |
+
"eval_samples_per_second": 3.461,
|
| 1239 |
+
"eval_steps_per_second": 1.745,
|
| 1240 |
+
"step": 655
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 0.0132,
|
| 1244 |
+
"eval_loss": 2.522101402282715,
|
| 1245 |
+
"eval_runtime": 33.7192,
|
| 1246 |
+
"eval_samples_per_second": 3.47,
|
| 1247 |
+
"eval_steps_per_second": 1.75,
|
| 1248 |
+
"step": 660
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.0133,
|
| 1252 |
+
"eval_loss": 2.5215632915496826,
|
| 1253 |
+
"eval_runtime": 33.7966,
|
| 1254 |
+
"eval_samples_per_second": 3.462,
|
| 1255 |
+
"eval_steps_per_second": 1.746,
|
| 1256 |
+
"step": 665
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.0134,
|
| 1260 |
+
"eval_loss": 2.5208749771118164,
|
| 1261 |
+
"eval_runtime": 33.7485,
|
| 1262 |
+
"eval_samples_per_second": 3.467,
|
| 1263 |
+
"eval_steps_per_second": 1.748,
|
| 1264 |
+
"step": 670
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.0135,
|
| 1268 |
+
"grad_norm": 0.044734235617461586,
|
| 1269 |
+
"learning_rate": 1.348e-06,
|
| 1270 |
+
"loss": 2.5273,
|
| 1271 |
+
"step": 675
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.0135,
|
| 1275 |
+
"eval_loss": 2.5201478004455566,
|
| 1276 |
+
"eval_runtime": 33.8972,
|
| 1277 |
+
"eval_samples_per_second": 3.452,
|
| 1278 |
+
"eval_steps_per_second": 1.741,
|
| 1279 |
+
"step": 675
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.0136,
|
| 1283 |
+
"eval_loss": 2.5197227001190186,
|
| 1284 |
+
"eval_runtime": 33.6652,
|
| 1285 |
+
"eval_samples_per_second": 3.475,
|
| 1286 |
+
"eval_steps_per_second": 1.753,
|
| 1287 |
+
"step": 680
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.0137,
|
| 1291 |
+
"eval_loss": 2.519151449203491,
|
| 1292 |
+
"eval_runtime": 33.6031,
|
| 1293 |
+
"eval_samples_per_second": 3.482,
|
| 1294 |
+
"eval_steps_per_second": 1.756,
|
| 1295 |
+
"step": 685
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 0.0138,
|
| 1299 |
+
"eval_loss": 2.5185396671295166,
|
| 1300 |
+
"eval_runtime": 33.6292,
|
| 1301 |
+
"eval_samples_per_second": 3.479,
|
| 1302 |
+
"eval_steps_per_second": 1.754,
|
| 1303 |
+
"step": 690
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 0.0139,
|
| 1307 |
+
"eval_loss": 2.517947196960449,
|
| 1308 |
+
"eval_runtime": 33.5987,
|
| 1309 |
+
"eval_samples_per_second": 3.482,
|
| 1310 |
+
"eval_steps_per_second": 1.756,
|
| 1311 |
+
"step": 695
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.014,
|
| 1315 |
+
"grad_norm": 0.04124740305893712,
|
| 1316 |
+
"learning_rate": 1.3980000000000002e-06,
|
| 1317 |
+
"loss": 2.5214,
|
| 1318 |
+
"step": 700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.014,
|
| 1322 |
+
"eval_loss": 2.5173356533050537,
|
| 1323 |
+
"eval_runtime": 33.6657,
|
| 1324 |
+
"eval_samples_per_second": 3.475,
|
| 1325 |
+
"eval_steps_per_second": 1.753,
|
| 1326 |
+
"step": 700
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.0141,
|
| 1330 |
+
"eval_loss": 2.5167977809906006,
|
| 1331 |
+
"eval_runtime": 33.5728,
|
| 1332 |
+
"eval_samples_per_second": 3.485,
|
| 1333 |
+
"eval_steps_per_second": 1.757,
|
| 1334 |
+
"step": 705
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.0142,
|
| 1338 |
+
"eval_loss": 2.5162267684936523,
|
| 1339 |
+
"eval_runtime": 33.2779,
|
| 1340 |
+
"eval_samples_per_second": 3.516,
|
| 1341 |
+
"eval_steps_per_second": 1.773,
|
| 1342 |
+
"step": 710
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.0143,
|
| 1346 |
+
"eval_loss": 2.5155909061431885,
|
| 1347 |
+
"eval_runtime": 33.4627,
|
| 1348 |
+
"eval_samples_per_second": 3.496,
|
| 1349 |
+
"eval_steps_per_second": 1.763,
|
| 1350 |
+
"step": 715
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.0144,
|
| 1354 |
+
"eval_loss": 2.515427589416504,
|
| 1355 |
+
"eval_runtime": 33.439,
|
| 1356 |
+
"eval_samples_per_second": 3.499,
|
| 1357 |
+
"eval_steps_per_second": 1.764,
|
| 1358 |
+
"step": 720
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.0145,
|
| 1362 |
+
"grad_norm": 0.04140679897915697,
|
| 1363 |
+
"learning_rate": 1.4480000000000002e-06,
|
| 1364 |
+
"loss": 2.5192,
|
| 1365 |
+
"step": 725
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.0145,
|
| 1369 |
+
"eval_loss": 2.514657735824585,
|
| 1370 |
+
"eval_runtime": 33.3527,
|
| 1371 |
+
"eval_samples_per_second": 3.508,
|
| 1372 |
+
"eval_steps_per_second": 1.769,
|
| 1373 |
+
"step": 725
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.0146,
|
| 1377 |
+
"eval_loss": 2.5141184329986572,
|
| 1378 |
+
"eval_runtime": 33.3623,
|
| 1379 |
+
"eval_samples_per_second": 3.507,
|
| 1380 |
+
"eval_steps_per_second": 1.768,
|
| 1381 |
+
"step": 730
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.0147,
|
| 1385 |
+
"eval_loss": 2.5135021209716797,
|
| 1386 |
+
"eval_runtime": 36.2875,
|
| 1387 |
+
"eval_samples_per_second": 3.224,
|
| 1388 |
+
"eval_steps_per_second": 1.626,
|
| 1389 |
+
"step": 735
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.0148,
|
| 1393 |
+
"eval_loss": 2.5130276679992676,
|
| 1394 |
+
"eval_runtime": 33.3738,
|
| 1395 |
+
"eval_samples_per_second": 3.506,
|
| 1396 |
+
"eval_steps_per_second": 1.768,
|
| 1397 |
+
"step": 740
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 0.0149,
|
| 1401 |
+
"eval_loss": 2.5123140811920166,
|
| 1402 |
+
"eval_runtime": 33.7458,
|
| 1403 |
+
"eval_samples_per_second": 3.467,
|
| 1404 |
+
"eval_steps_per_second": 1.748,
|
| 1405 |
+
"step": 745
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.015,
|
| 1409 |
+
"grad_norm": 0.03921746155872101,
|
| 1410 |
+
"learning_rate": 1.498e-06,
|
| 1411 |
+
"loss": 2.5077,
|
| 1412 |
+
"step": 750
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.015,
|
| 1416 |
+
"eval_loss": 2.5117204189300537,
|
| 1417 |
+
"eval_runtime": 33.3164,
|
| 1418 |
+
"eval_samples_per_second": 3.512,
|
| 1419 |
+
"eval_steps_per_second": 1.771,
|
| 1420 |
+
"step": 750
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.0151,
|
| 1424 |
+
"eval_loss": 2.5113115310668945,
|
| 1425 |
+
"eval_runtime": 33.464,
|
| 1426 |
+
"eval_samples_per_second": 3.496,
|
| 1427 |
+
"eval_steps_per_second": 1.763,
|
| 1428 |
+
"step": 755
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.0152,
|
| 1432 |
+
"eval_loss": 2.510754108428955,
|
| 1433 |
+
"eval_runtime": 33.426,
|
| 1434 |
+
"eval_samples_per_second": 3.5,
|
| 1435 |
+
"eval_steps_per_second": 1.765,
|
| 1436 |
+
"step": 760
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.0153,
|
| 1440 |
+
"eval_loss": 2.510148525238037,
|
| 1441 |
+
"eval_runtime": 33.5135,
|
| 1442 |
+
"eval_samples_per_second": 3.491,
|
| 1443 |
+
"eval_steps_per_second": 1.76,
|
| 1444 |
+
"step": 765
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.0154,
|
| 1448 |
+
"eval_loss": 2.5096797943115234,
|
| 1449 |
+
"eval_runtime": 33.5467,
|
| 1450 |
+
"eval_samples_per_second": 3.488,
|
| 1451 |
+
"eval_steps_per_second": 1.759,
|
| 1452 |
+
"step": 770
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.0155,
|
| 1456 |
+
"grad_norm": 0.038161493704092234,
|
| 1457 |
+
"learning_rate": 1.548e-06,
|
| 1458 |
+
"loss": 2.5127,
|
| 1459 |
+
"step": 775
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.0155,
|
| 1463 |
+
"eval_loss": 2.5091397762298584,
|
| 1464 |
+
"eval_runtime": 33.6296,
|
| 1465 |
+
"eval_samples_per_second": 3.479,
|
| 1466 |
+
"eval_steps_per_second": 1.754,
|
| 1467 |
+
"step": 775
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.0156,
|
| 1471 |
+
"eval_loss": 2.5085766315460205,
|
| 1472 |
+
"eval_runtime": 33.6417,
|
| 1473 |
+
"eval_samples_per_second": 3.478,
|
| 1474 |
+
"eval_steps_per_second": 1.754,
|
| 1475 |
+
"step": 780
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.0157,
|
| 1479 |
+
"eval_loss": 2.5081799030303955,
|
| 1480 |
+
"eval_runtime": 33.5831,
|
| 1481 |
+
"eval_samples_per_second": 3.484,
|
| 1482 |
+
"eval_steps_per_second": 1.757,
|
| 1483 |
+
"step": 785
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.0158,
|
| 1487 |
+
"eval_loss": 2.5075252056121826,
|
| 1488 |
+
"eval_runtime": 33.5806,
|
| 1489 |
+
"eval_samples_per_second": 3.484,
|
| 1490 |
+
"eval_steps_per_second": 1.757,
|
| 1491 |
+
"step": 790
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.0159,
|
| 1495 |
+
"eval_loss": 2.5069563388824463,
|
| 1496 |
+
"eval_runtime": 33.6257,
|
| 1497 |
+
"eval_samples_per_second": 3.479,
|
| 1498 |
+
"eval_steps_per_second": 1.755,
|
| 1499 |
+
"step": 795
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.016,
|
| 1503 |
+
"grad_norm": 0.04372605860022339,
|
| 1504 |
+
"learning_rate": 1.5980000000000002e-06,
|
| 1505 |
+
"loss": 2.5019,
|
| 1506 |
+
"step": 800
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.016,
|
| 1510 |
+
"eval_loss": 2.5065925121307373,
|
| 1511 |
+
"eval_runtime": 33.6041,
|
| 1512 |
+
"eval_samples_per_second": 3.482,
|
| 1513 |
+
"eval_steps_per_second": 1.756,
|
| 1514 |
+
"step": 800
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.0161,
|
| 1518 |
+
"eval_loss": 2.5059759616851807,
|
| 1519 |
+
"eval_runtime": 33.6116,
|
| 1520 |
+
"eval_samples_per_second": 3.481,
|
| 1521 |
+
"eval_steps_per_second": 1.755,
|
| 1522 |
+
"step": 805
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.0162,
|
| 1526 |
+
"eval_loss": 2.505453109741211,
|
| 1527 |
+
"eval_runtime": 33.5794,
|
| 1528 |
+
"eval_samples_per_second": 3.484,
|
| 1529 |
+
"eval_steps_per_second": 1.757,
|
| 1530 |
+
"step": 810
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.0163,
|
| 1534 |
+
"eval_loss": 2.505023241043091,
|
| 1535 |
+
"eval_runtime": 33.461,
|
| 1536 |
+
"eval_samples_per_second": 3.497,
|
| 1537 |
+
"eval_steps_per_second": 1.763,
|
| 1538 |
+
"step": 815
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.0164,
|
| 1542 |
+
"eval_loss": 2.5042824745178223,
|
| 1543 |
+
"eval_runtime": 33.5988,
|
| 1544 |
+
"eval_samples_per_second": 3.482,
|
| 1545 |
+
"eval_steps_per_second": 1.756,
|
| 1546 |
+
"step": 820
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.0165,
|
| 1550 |
+
"grad_norm": 0.041497520045134684,
|
| 1551 |
+
"learning_rate": 1.6480000000000001e-06,
|
| 1552 |
+
"loss": 2.4977,
|
| 1553 |
+
"step": 825
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.0165,
|
| 1557 |
+
"eval_loss": 2.5039255619049072,
|
| 1558 |
+
"eval_runtime": 33.6107,
|
| 1559 |
+
"eval_samples_per_second": 3.481,
|
| 1560 |
+
"eval_steps_per_second": 1.755,
|
| 1561 |
+
"step": 825
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.0166,
|
| 1565 |
+
"eval_loss": 2.503436803817749,
|
| 1566 |
+
"eval_runtime": 33.6213,
|
| 1567 |
+
"eval_samples_per_second": 3.48,
|
| 1568 |
+
"eval_steps_per_second": 1.755,
|
| 1569 |
+
"step": 830
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.0167,
|
| 1573 |
+
"eval_loss": 2.5028321743011475,
|
| 1574 |
+
"eval_runtime": 33.5009,
|
| 1575 |
+
"eval_samples_per_second": 3.492,
|
| 1576 |
+
"eval_steps_per_second": 1.761,
|
| 1577 |
+
"step": 835
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.0168,
|
| 1581 |
+
"eval_loss": 2.5022666454315186,
|
| 1582 |
+
"eval_runtime": 33.6392,
|
| 1583 |
+
"eval_samples_per_second": 3.478,
|
| 1584 |
+
"eval_steps_per_second": 1.754,
|
| 1585 |
+
"step": 840
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.0169,
|
| 1589 |
+
"eval_loss": 2.5018374919891357,
|
| 1590 |
+
"eval_runtime": 33.5928,
|
| 1591 |
+
"eval_samples_per_second": 3.483,
|
| 1592 |
+
"eval_steps_per_second": 1.756,
|
| 1593 |
+
"step": 845
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.017,
|
| 1597 |
+
"grad_norm": 0.040226840781059835,
|
| 1598 |
+
"learning_rate": 1.6980000000000003e-06,
|
| 1599 |
+
"loss": 2.4968,
|
| 1600 |
+
"step": 850
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.017,
|
| 1604 |
+
"eval_loss": 2.5012588500976562,
|
| 1605 |
+
"eval_runtime": 33.5216,
|
| 1606 |
+
"eval_samples_per_second": 3.49,
|
| 1607 |
+
"eval_steps_per_second": 1.76,
|
| 1608 |
+
"step": 850
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.0171,
|
| 1612 |
+
"eval_loss": 2.5006515979766846,
|
| 1613 |
+
"eval_runtime": 33.4799,
|
| 1614 |
+
"eval_samples_per_second": 3.495,
|
| 1615 |
+
"eval_steps_per_second": 1.762,
|
| 1616 |
+
"step": 855
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.0172,
|
| 1620 |
+
"eval_loss": 2.5001821517944336,
|
| 1621 |
+
"eval_runtime": 33.6067,
|
| 1622 |
+
"eval_samples_per_second": 3.481,
|
| 1623 |
+
"eval_steps_per_second": 1.756,
|
| 1624 |
+
"step": 860
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 0.0173,
|
| 1628 |
+
"eval_loss": 2.499708652496338,
|
| 1629 |
+
"eval_runtime": 33.5478,
|
| 1630 |
+
"eval_samples_per_second": 3.488,
|
| 1631 |
+
"eval_steps_per_second": 1.759,
|
| 1632 |
+
"step": 865
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.0174,
|
| 1636 |
+
"eval_loss": 2.4992101192474365,
|
| 1637 |
+
"eval_runtime": 33.3608,
|
| 1638 |
+
"eval_samples_per_second": 3.507,
|
| 1639 |
+
"eval_steps_per_second": 1.769,
|
| 1640 |
+
"step": 870
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.0175,
|
| 1644 |
+
"grad_norm": 0.043360400185163274,
|
| 1645 |
+
"learning_rate": 1.7480000000000002e-06,
|
| 1646 |
+
"loss": 2.4947,
|
| 1647 |
+
"step": 875
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.0175,
|
| 1651 |
+
"eval_loss": 2.49912428855896,
|
| 1652 |
+
"eval_runtime": 33.3782,
|
| 1653 |
+
"eval_samples_per_second": 3.505,
|
| 1654 |
+
"eval_steps_per_second": 1.768,
|
| 1655 |
+
"step": 875
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.0176,
|
| 1659 |
+
"eval_loss": 2.498539686203003,
|
| 1660 |
+
"eval_runtime": 33.4271,
|
| 1661 |
+
"eval_samples_per_second": 3.5,
|
| 1662 |
+
"eval_steps_per_second": 1.765,
|
| 1663 |
+
"step": 880
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.0177,
|
| 1667 |
+
"eval_loss": 2.4980475902557373,
|
| 1668 |
+
"eval_runtime": 33.508,
|
| 1669 |
+
"eval_samples_per_second": 3.492,
|
| 1670 |
+
"eval_steps_per_second": 1.761,
|
| 1671 |
+
"step": 885
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.0178,
|
| 1675 |
+
"eval_loss": 2.4972891807556152,
|
| 1676 |
+
"eval_runtime": 33.5801,
|
| 1677 |
+
"eval_samples_per_second": 3.484,
|
| 1678 |
+
"eval_steps_per_second": 1.757,
|
| 1679 |
+
"step": 890
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.0179,
|
| 1683 |
+
"eval_loss": 2.496943473815918,
|
| 1684 |
+
"eval_runtime": 33.4984,
|
| 1685 |
+
"eval_samples_per_second": 3.493,
|
| 1686 |
+
"eval_steps_per_second": 1.761,
|
| 1687 |
+
"step": 895
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 0.018,
|
| 1691 |
+
"grad_norm": 0.040565773819723885,
|
| 1692 |
+
"learning_rate": 1.798e-06,
|
| 1693 |
+
"loss": 2.4878,
|
| 1694 |
+
"step": 900
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.018,
|
| 1698 |
+
"eval_loss": 2.496464252471924,
|
| 1699 |
+
"eval_runtime": 33.6538,
|
| 1700 |
+
"eval_samples_per_second": 3.477,
|
| 1701 |
+
"eval_steps_per_second": 1.753,
|
| 1702 |
+
"step": 900
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 0.0181,
|
| 1706 |
+
"eval_loss": 2.496126890182495,
|
| 1707 |
+
"eval_runtime": 33.6415,
|
| 1708 |
+
"eval_samples_per_second": 3.478,
|
| 1709 |
+
"eval_steps_per_second": 1.754,
|
| 1710 |
+
"step": 905
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.0182,
|
| 1714 |
+
"eval_loss": 2.4957361221313477,
|
| 1715 |
+
"eval_runtime": 33.7646,
|
| 1716 |
+
"eval_samples_per_second": 3.465,
|
| 1717 |
+
"eval_steps_per_second": 1.747,
|
| 1718 |
+
"step": 910
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.0183,
|
| 1722 |
+
"eval_loss": 2.4954254627227783,
|
| 1723 |
+
"eval_runtime": 33.5639,
|
| 1724 |
+
"eval_samples_per_second": 3.486,
|
| 1725 |
+
"eval_steps_per_second": 1.758,
|
| 1726 |
+
"step": 915
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.0184,
|
| 1730 |
+
"eval_loss": 2.4948976039886475,
|
| 1731 |
+
"eval_runtime": 33.6038,
|
| 1732 |
+
"eval_samples_per_second": 3.482,
|
| 1733 |
+
"eval_steps_per_second": 1.756,
|
| 1734 |
+
"step": 920
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 0.0185,
|
| 1738 |
+
"grad_norm": 0.039370814834696136,
|
| 1739 |
+
"learning_rate": 1.8480000000000001e-06,
|
| 1740 |
+
"loss": 2.4986,
|
| 1741 |
+
"step": 925
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.0185,
|
| 1745 |
+
"eval_loss": 2.494521379470825,
|
| 1746 |
+
"eval_runtime": 33.7082,
|
| 1747 |
+
"eval_samples_per_second": 3.471,
|
| 1748 |
+
"eval_steps_per_second": 1.75,
|
| 1749 |
+
"step": 925
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.0186,
|
| 1753 |
+
"eval_loss": 2.4939730167388916,
|
| 1754 |
+
"eval_runtime": 33.6147,
|
| 1755 |
+
"eval_samples_per_second": 3.481,
|
| 1756 |
+
"eval_steps_per_second": 1.755,
|
| 1757 |
+
"step": 930
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.0187,
|
| 1761 |
+
"eval_loss": 2.49343204498291,
|
| 1762 |
+
"eval_runtime": 35.1502,
|
| 1763 |
+
"eval_samples_per_second": 3.329,
|
| 1764 |
+
"eval_steps_per_second": 1.679,
|
| 1765 |
+
"step": 935
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 0.0188,
|
| 1769 |
+
"eval_loss": 2.493082046508789,
|
| 1770 |
+
"eval_runtime": 33.6381,
|
| 1771 |
+
"eval_samples_per_second": 3.478,
|
| 1772 |
+
"eval_steps_per_second": 1.754,
|
| 1773 |
+
"step": 940
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.0189,
|
| 1777 |
+
"eval_loss": 2.492797374725342,
|
| 1778 |
+
"eval_runtime": 33.7089,
|
| 1779 |
+
"eval_samples_per_second": 3.471,
|
| 1780 |
+
"eval_steps_per_second": 1.75,
|
| 1781 |
+
"step": 945
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.019,
|
| 1785 |
+
"grad_norm": 0.04019472793080496,
|
| 1786 |
+
"learning_rate": 1.898e-06,
|
| 1787 |
+
"loss": 2.481,
|
| 1788 |
+
"step": 950
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.019,
|
| 1792 |
+
"eval_loss": 2.4925599098205566,
|
| 1793 |
+
"eval_runtime": 33.5096,
|
| 1794 |
+
"eval_samples_per_second": 3.492,
|
| 1795 |
+
"eval_steps_per_second": 1.761,
|
| 1796 |
+
"step": 950
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.0191,
|
| 1800 |
+
"eval_loss": 2.4918878078460693,
|
| 1801 |
+
"eval_runtime": 33.4921,
|
| 1802 |
+
"eval_samples_per_second": 3.493,
|
| 1803 |
+
"eval_steps_per_second": 1.762,
|
| 1804 |
+
"step": 955
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.0192,
|
| 1808 |
+
"eval_loss": 2.4916608333587646,
|
| 1809 |
+
"eval_runtime": 33.5126,
|
| 1810 |
+
"eval_samples_per_second": 3.491,
|
| 1811 |
+
"eval_steps_per_second": 1.761,
|
| 1812 |
+
"step": 960
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 0.0193,
|
| 1816 |
+
"eval_loss": 2.491708517074585,
|
| 1817 |
+
"eval_runtime": 33.6466,
|
| 1818 |
+
"eval_samples_per_second": 3.477,
|
| 1819 |
+
"eval_steps_per_second": 1.754,
|
| 1820 |
+
"step": 965
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.0194,
|
| 1824 |
+
"eval_loss": 2.4911839962005615,
|
| 1825 |
+
"eval_runtime": 33.6119,
|
| 1826 |
+
"eval_samples_per_second": 3.481,
|
| 1827 |
+
"eval_steps_per_second": 1.755,
|
| 1828 |
+
"step": 970
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 0.0195,
|
| 1832 |
+
"grad_norm": 0.04683912756161822,
|
| 1833 |
+
"learning_rate": 1.9480000000000002e-06,
|
| 1834 |
+
"loss": 2.4879,
|
| 1835 |
+
"step": 975
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 0.0195,
|
| 1839 |
+
"eval_loss": 2.490492343902588,
|
| 1840 |
+
"eval_runtime": 33.4389,
|
| 1841 |
+
"eval_samples_per_second": 3.499,
|
| 1842 |
+
"eval_steps_per_second": 1.764,
|
| 1843 |
+
"step": 975
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.0196,
|
| 1847 |
+
"eval_loss": 2.490133285522461,
|
| 1848 |
+
"eval_runtime": 33.361,
|
| 1849 |
+
"eval_samples_per_second": 3.507,
|
| 1850 |
+
"eval_steps_per_second": 1.769,
|
| 1851 |
+
"step": 980
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.0197,
|
| 1855 |
+
"eval_loss": 2.4896316528320312,
|
| 1856 |
+
"eval_runtime": 33.5863,
|
| 1857 |
+
"eval_samples_per_second": 3.484,
|
| 1858 |
+
"eval_steps_per_second": 1.757,
|
| 1859 |
+
"step": 985
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.0198,
|
| 1863 |
+
"eval_loss": 2.489122152328491,
|
| 1864 |
+
"eval_runtime": 33.6173,
|
| 1865 |
+
"eval_samples_per_second": 3.48,
|
| 1866 |
+
"eval_steps_per_second": 1.755,
|
| 1867 |
+
"step": 990
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.0199,
|
| 1871 |
+
"eval_loss": 2.488906145095825,
|
| 1872 |
+
"eval_runtime": 33.6531,
|
| 1873 |
+
"eval_samples_per_second": 3.477,
|
| 1874 |
+
"eval_steps_per_second": 1.753,
|
| 1875 |
+
"step": 995
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 0.02,
|
| 1879 |
+
"grad_norm": 0.047671496052023164,
|
| 1880 |
+
"learning_rate": 1.998e-06,
|
| 1881 |
+
"loss": 2.4879,
|
| 1882 |
+
"step": 1000
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.02,
|
| 1886 |
+
"eval_loss": 2.488457202911377,
|
| 1887 |
+
"eval_runtime": 33.7763,
|
| 1888 |
+
"eval_samples_per_second": 3.464,
|
| 1889 |
+
"eval_steps_per_second": 1.747,
|
| 1890 |
+
"step": 1000
|
| 1891 |
+
}
|
| 1892 |
+
],
|
| 1893 |
"logging_steps": 25,
|
| 1894 |
"max_steps": 50000,
|
| 1895 |
"num_input_tokens_seen": 0,
|
| 1896 |
"num_train_epochs": 9223372036854775807,
|
| 1897 |
+
"save_steps": 1000,
|
| 1898 |
"stateful_callbacks": {
|
| 1899 |
"TrainerControl": {
|
| 1900 |
"args": {
|
|
|
|
| 1907 |
"attributes": {}
|
| 1908 |
}
|
| 1909 |
},
|
| 1910 |
+
"total_flos": 2.785301757633233e+18,
|
| 1911 |
+
"train_batch_size": 1,
|
| 1912 |
"trial_name": null,
|
| 1913 |
"trial_params": null
|
| 1914 |
}
|