Training in progress, step 375, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2200119864
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ff898de4f88712fb50a5522ea47c598e5a58b6eb9f332c668b6443b61696a09
|
| 3 |
size 2200119864
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1734,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -675,6 +675,1988 @@
|
|
| 675 |
"eval_samples_per_second": 130.347,
|
| 676 |
"eval_steps_per_second": 2.803,
|
| 677 |
"step": 93
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 678 |
}
|
| 679 |
],
|
| 680 |
"logging_steps": 1,
|
|
@@ -694,7 +2676,7 @@
|
|
| 694 |
"attributes": {}
|
| 695 |
}
|
| 696 |
},
|
| 697 |
-
"total_flos":
|
| 698 |
"train_batch_size": 60,
|
| 699 |
"trial_name": null,
|
| 700 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.21623180049012541,
|
| 6 |
"eval_steps": 1734,
|
| 7 |
+
"global_step": 375,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 675 |
"eval_samples_per_second": 130.347,
|
| 676 |
"eval_steps_per_second": 2.803,
|
| 677 |
"step": 93
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.054202104656191435,
|
| 681 |
+
"grad_norm": 0.6640625,
|
| 682 |
+
"learning_rate": 9.995587006107607e-05,
|
| 683 |
+
"loss": 0.9885,
|
| 684 |
+
"step": 94
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.05477872279083177,
|
| 688 |
+
"grad_norm": 0.69921875,
|
| 689 |
+
"learning_rate": 9.995480051351141e-05,
|
| 690 |
+
"loss": 0.9397,
|
| 691 |
+
"step": 95
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.055355340925472105,
|
| 695 |
+
"grad_norm": 0.62109375,
|
| 696 |
+
"learning_rate": 9.99537181680915e-05,
|
| 697 |
+
"loss": 0.9706,
|
| 698 |
+
"step": 96
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.05593195906011244,
|
| 702 |
+
"grad_norm": 0.671875,
|
| 703 |
+
"learning_rate": 9.995262302521265e-05,
|
| 704 |
+
"loss": 0.9613,
|
| 705 |
+
"step": 97
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.056508577194752774,
|
| 709 |
+
"grad_norm": 0.67578125,
|
| 710 |
+
"learning_rate": 9.995151508527583e-05,
|
| 711 |
+
"loss": 0.9783,
|
| 712 |
+
"step": 98
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 0.05708519532939311,
|
| 716 |
+
"grad_norm": 0.64453125,
|
| 717 |
+
"learning_rate": 9.995039434868667e-05,
|
| 718 |
+
"loss": 0.9635,
|
| 719 |
+
"step": 99
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.05766181346403344,
|
| 723 |
+
"grad_norm": 0.62109375,
|
| 724 |
+
"learning_rate": 9.994926081585551e-05,
|
| 725 |
+
"loss": 0.95,
|
| 726 |
+
"step": 100
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.05823843159867378,
|
| 730 |
+
"grad_norm": 0.59375,
|
| 731 |
+
"learning_rate": 9.994811448719736e-05,
|
| 732 |
+
"loss": 0.9274,
|
| 733 |
+
"step": 101
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.05881504973331411,
|
| 737 |
+
"grad_norm": 0.64453125,
|
| 738 |
+
"learning_rate": 9.994695536313194e-05,
|
| 739 |
+
"loss": 0.964,
|
| 740 |
+
"step": 102
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.05939166786795445,
|
| 744 |
+
"grad_norm": 0.64453125,
|
| 745 |
+
"learning_rate": 9.994578344408361e-05,
|
| 746 |
+
"loss": 0.9469,
|
| 747 |
+
"step": 103
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.05996828600259478,
|
| 751 |
+
"grad_norm": 0.640625,
|
| 752 |
+
"learning_rate": 9.994459873048146e-05,
|
| 753 |
+
"loss": 0.979,
|
| 754 |
+
"step": 104
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.060544904137235116,
|
| 758 |
+
"grad_norm": 0.640625,
|
| 759 |
+
"learning_rate": 9.99434012227592e-05,
|
| 760 |
+
"loss": 0.9555,
|
| 761 |
+
"step": 105
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.06112152227187545,
|
| 765 |
+
"grad_norm": 0.6796875,
|
| 766 |
+
"learning_rate": 9.994219092135533e-05,
|
| 767 |
+
"loss": 0.9661,
|
| 768 |
+
"step": 106
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.061698140406515785,
|
| 772 |
+
"grad_norm": 0.640625,
|
| 773 |
+
"learning_rate": 9.994096782671293e-05,
|
| 774 |
+
"loss": 0.9808,
|
| 775 |
+
"step": 107
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.06227475854115612,
|
| 779 |
+
"grad_norm": 0.625,
|
| 780 |
+
"learning_rate": 9.993973193927984e-05,
|
| 781 |
+
"loss": 0.9517,
|
| 782 |
+
"step": 108
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.06285137667579645,
|
| 786 |
+
"grad_norm": 0.625,
|
| 787 |
+
"learning_rate": 9.993848325950852e-05,
|
| 788 |
+
"loss": 0.9701,
|
| 789 |
+
"step": 109
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.06342799481043679,
|
| 793 |
+
"grad_norm": 0.6484375,
|
| 794 |
+
"learning_rate": 9.993722178785616e-05,
|
| 795 |
+
"loss": 1.0023,
|
| 796 |
+
"step": 110
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.06400461294507712,
|
| 800 |
+
"grad_norm": 0.65234375,
|
| 801 |
+
"learning_rate": 9.993594752478461e-05,
|
| 802 |
+
"loss": 1.0208,
|
| 803 |
+
"step": 111
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.06458123107971746,
|
| 807 |
+
"grad_norm": 0.61328125,
|
| 808 |
+
"learning_rate": 9.993466047076041e-05,
|
| 809 |
+
"loss": 0.9391,
|
| 810 |
+
"step": 112
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.06515784921435779,
|
| 814 |
+
"grad_norm": 0.66015625,
|
| 815 |
+
"learning_rate": 9.99333606262548e-05,
|
| 816 |
+
"loss": 1.0017,
|
| 817 |
+
"step": 113
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.06573446734899813,
|
| 821 |
+
"grad_norm": 0.60546875,
|
| 822 |
+
"learning_rate": 9.993204799174367e-05,
|
| 823 |
+
"loss": 0.9599,
|
| 824 |
+
"step": 114
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.06631108548363845,
|
| 828 |
+
"grad_norm": 0.6640625,
|
| 829 |
+
"learning_rate": 9.993072256770759e-05,
|
| 830 |
+
"loss": 0.973,
|
| 831 |
+
"step": 115
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.0668877036182788,
|
| 835 |
+
"grad_norm": 0.609375,
|
| 836 |
+
"learning_rate": 9.992938435463189e-05,
|
| 837 |
+
"loss": 0.9892,
|
| 838 |
+
"step": 116
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.06746432175291912,
|
| 842 |
+
"grad_norm": 0.6484375,
|
| 843 |
+
"learning_rate": 9.992803335300647e-05,
|
| 844 |
+
"loss": 0.9819,
|
| 845 |
+
"step": 117
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.06804093988755947,
|
| 849 |
+
"grad_norm": 0.62109375,
|
| 850 |
+
"learning_rate": 9.992666956332599e-05,
|
| 851 |
+
"loss": 0.9517,
|
| 852 |
+
"step": 118
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.06861755802219979,
|
| 856 |
+
"grad_norm": 0.625,
|
| 857 |
+
"learning_rate": 9.992529298608977e-05,
|
| 858 |
+
"loss": 0.9475,
|
| 859 |
+
"step": 119
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.06919417615684013,
|
| 863 |
+
"grad_norm": 0.609375,
|
| 864 |
+
"learning_rate": 9.99239036218018e-05,
|
| 865 |
+
"loss": 0.9438,
|
| 866 |
+
"step": 120
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.06977079429148046,
|
| 870 |
+
"grad_norm": 0.61328125,
|
| 871 |
+
"learning_rate": 9.992250147097076e-05,
|
| 872 |
+
"loss": 0.9475,
|
| 873 |
+
"step": 121
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.0703474124261208,
|
| 877 |
+
"grad_norm": 0.625,
|
| 878 |
+
"learning_rate": 9.992108653411e-05,
|
| 879 |
+
"loss": 0.9157,
|
| 880 |
+
"step": 122
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.07092403056076113,
|
| 884 |
+
"grad_norm": 0.6171875,
|
| 885 |
+
"learning_rate": 9.991965881173764e-05,
|
| 886 |
+
"loss": 0.9425,
|
| 887 |
+
"step": 123
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.07150064869540147,
|
| 891 |
+
"grad_norm": 0.62890625,
|
| 892 |
+
"learning_rate": 9.991821830437631e-05,
|
| 893 |
+
"loss": 0.9465,
|
| 894 |
+
"step": 124
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.0720772668300418,
|
| 898 |
+
"grad_norm": 0.66015625,
|
| 899 |
+
"learning_rate": 9.991676501255346e-05,
|
| 900 |
+
"loss": 0.9809,
|
| 901 |
+
"step": 125
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.07265388496468214,
|
| 905 |
+
"grad_norm": 0.63671875,
|
| 906 |
+
"learning_rate": 9.991529893680121e-05,
|
| 907 |
+
"loss": 0.952,
|
| 908 |
+
"step": 126
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.07323050309932247,
|
| 912 |
+
"grad_norm": 0.640625,
|
| 913 |
+
"learning_rate": 9.991382007765626e-05,
|
| 914 |
+
"loss": 0.9734,
|
| 915 |
+
"step": 127
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.07380712123396281,
|
| 919 |
+
"grad_norm": 0.63671875,
|
| 920 |
+
"learning_rate": 9.991232843566011e-05,
|
| 921 |
+
"loss": 0.9436,
|
| 922 |
+
"step": 128
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.07438373936860314,
|
| 926 |
+
"grad_norm": 0.66015625,
|
| 927 |
+
"learning_rate": 9.991082401135887e-05,
|
| 928 |
+
"loss": 0.9497,
|
| 929 |
+
"step": 129
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.07496035750324348,
|
| 933 |
+
"grad_norm": 0.640625,
|
| 934 |
+
"learning_rate": 9.990930680530335e-05,
|
| 935 |
+
"loss": 0.973,
|
| 936 |
+
"step": 130
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.07553697563788381,
|
| 940 |
+
"grad_norm": 0.640625,
|
| 941 |
+
"learning_rate": 9.990777681804903e-05,
|
| 942 |
+
"loss": 0.9116,
|
| 943 |
+
"step": 131
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.07611359377252415,
|
| 947 |
+
"grad_norm": 0.640625,
|
| 948 |
+
"learning_rate": 9.990623405015612e-05,
|
| 949 |
+
"loss": 0.9332,
|
| 950 |
+
"step": 132
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.07669021190716448,
|
| 954 |
+
"grad_norm": 0.62109375,
|
| 955 |
+
"learning_rate": 9.990467850218938e-05,
|
| 956 |
+
"loss": 0.9892,
|
| 957 |
+
"step": 133
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.07726683004180482,
|
| 961 |
+
"grad_norm": 0.64453125,
|
| 962 |
+
"learning_rate": 9.990311017471842e-05,
|
| 963 |
+
"loss": 0.9529,
|
| 964 |
+
"step": 134
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.07784344817644515,
|
| 968 |
+
"grad_norm": 0.640625,
|
| 969 |
+
"learning_rate": 9.990152906831743e-05,
|
| 970 |
+
"loss": 0.9713,
|
| 971 |
+
"step": 135
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.07842006631108549,
|
| 975 |
+
"grad_norm": 0.62890625,
|
| 976 |
+
"learning_rate": 9.989993518356526e-05,
|
| 977 |
+
"loss": 0.9428,
|
| 978 |
+
"step": 136
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.07899668444572581,
|
| 982 |
+
"grad_norm": 0.6640625,
|
| 983 |
+
"learning_rate": 9.98983285210455e-05,
|
| 984 |
+
"loss": 0.9495,
|
| 985 |
+
"step": 137
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.07957330258036616,
|
| 989 |
+
"grad_norm": 0.6171875,
|
| 990 |
+
"learning_rate": 9.989670908134637e-05,
|
| 991 |
+
"loss": 0.9647,
|
| 992 |
+
"step": 138
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.08014992071500648,
|
| 996 |
+
"grad_norm": 0.671875,
|
| 997 |
+
"learning_rate": 9.989507686506082e-05,
|
| 998 |
+
"loss": 0.9841,
|
| 999 |
+
"step": 139
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.08072653884964683,
|
| 1003 |
+
"grad_norm": 0.62109375,
|
| 1004 |
+
"learning_rate": 9.98934318727864e-05,
|
| 1005 |
+
"loss": 0.9468,
|
| 1006 |
+
"step": 140
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.08130315698428715,
|
| 1010 |
+
"grad_norm": 0.625,
|
| 1011 |
+
"learning_rate": 9.989177410512543e-05,
|
| 1012 |
+
"loss": 0.9491,
|
| 1013 |
+
"step": 141
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.0818797751189275,
|
| 1017 |
+
"grad_norm": 0.6015625,
|
| 1018 |
+
"learning_rate": 9.989010356268484e-05,
|
| 1019 |
+
"loss": 0.8968,
|
| 1020 |
+
"step": 142
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.08245639325356782,
|
| 1024 |
+
"grad_norm": 0.63671875,
|
| 1025 |
+
"learning_rate": 9.988842024607625e-05,
|
| 1026 |
+
"loss": 0.9615,
|
| 1027 |
+
"step": 143
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.08303301138820816,
|
| 1031 |
+
"grad_norm": 0.6484375,
|
| 1032 |
+
"learning_rate": 9.9886724155916e-05,
|
| 1033 |
+
"loss": 0.9638,
|
| 1034 |
+
"step": 144
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.08360962952284849,
|
| 1038 |
+
"grad_norm": 0.63671875,
|
| 1039 |
+
"learning_rate": 9.988501529282504e-05,
|
| 1040 |
+
"loss": 0.9358,
|
| 1041 |
+
"step": 145
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.08418624765748883,
|
| 1045 |
+
"grad_norm": 0.6875,
|
| 1046 |
+
"learning_rate": 9.988329365742903e-05,
|
| 1047 |
+
"loss": 0.9658,
|
| 1048 |
+
"step": 146
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.08476286579212916,
|
| 1052 |
+
"grad_norm": 0.625,
|
| 1053 |
+
"learning_rate": 9.988155925035832e-05,
|
| 1054 |
+
"loss": 0.9569,
|
| 1055 |
+
"step": 147
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.0853394839267695,
|
| 1059 |
+
"grad_norm": 0.6171875,
|
| 1060 |
+
"learning_rate": 9.987981207224793e-05,
|
| 1061 |
+
"loss": 0.9371,
|
| 1062 |
+
"step": 148
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.08591610206140983,
|
| 1066 |
+
"grad_norm": 0.65234375,
|
| 1067 |
+
"learning_rate": 9.98780521237375e-05,
|
| 1068 |
+
"loss": 0.9796,
|
| 1069 |
+
"step": 149
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.08649272019605017,
|
| 1073 |
+
"grad_norm": 0.6015625,
|
| 1074 |
+
"learning_rate": 9.987627940547145e-05,
|
| 1075 |
+
"loss": 0.9195,
|
| 1076 |
+
"step": 150
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.0870693383306905,
|
| 1080 |
+
"grad_norm": 0.64453125,
|
| 1081 |
+
"learning_rate": 9.987449391809878e-05,
|
| 1082 |
+
"loss": 0.9468,
|
| 1083 |
+
"step": 151
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.08764595646533084,
|
| 1087 |
+
"grad_norm": 0.66015625,
|
| 1088 |
+
"learning_rate": 9.987269566227322e-05,
|
| 1089 |
+
"loss": 0.9438,
|
| 1090 |
+
"step": 152
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.08822257459997117,
|
| 1094 |
+
"grad_norm": 0.61328125,
|
| 1095 |
+
"learning_rate": 9.987088463865317e-05,
|
| 1096 |
+
"loss": 0.9673,
|
| 1097 |
+
"step": 153
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.08879919273461151,
|
| 1101 |
+
"grad_norm": 0.6484375,
|
| 1102 |
+
"learning_rate": 9.986906084790164e-05,
|
| 1103 |
+
"loss": 0.9103,
|
| 1104 |
+
"step": 154
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.08937581086925184,
|
| 1108 |
+
"grad_norm": 0.58203125,
|
| 1109 |
+
"learning_rate": 9.986722429068644e-05,
|
| 1110 |
+
"loss": 0.918,
|
| 1111 |
+
"step": 155
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.08995242900389218,
|
| 1115 |
+
"grad_norm": 0.62890625,
|
| 1116 |
+
"learning_rate": 9.986537496767993e-05,
|
| 1117 |
+
"loss": 0.9399,
|
| 1118 |
+
"step": 156
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.0905290471385325,
|
| 1122 |
+
"grad_norm": 0.6328125,
|
| 1123 |
+
"learning_rate": 9.986351287955922e-05,
|
| 1124 |
+
"loss": 0.9584,
|
| 1125 |
+
"step": 157
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.09110566527317285,
|
| 1129 |
+
"grad_norm": 0.6484375,
|
| 1130 |
+
"learning_rate": 9.986163802700604e-05,
|
| 1131 |
+
"loss": 0.9801,
|
| 1132 |
+
"step": 158
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.09168228340781318,
|
| 1136 |
+
"grad_norm": 0.6171875,
|
| 1137 |
+
"learning_rate": 9.985975041070683e-05,
|
| 1138 |
+
"loss": 0.9435,
|
| 1139 |
+
"step": 159
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.09225890154245352,
|
| 1143 |
+
"grad_norm": 0.625,
|
| 1144 |
+
"learning_rate": 9.985785003135272e-05,
|
| 1145 |
+
"loss": 0.9362,
|
| 1146 |
+
"step": 160
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.09283551967709384,
|
| 1150 |
+
"grad_norm": 0.6328125,
|
| 1151 |
+
"learning_rate": 9.985593688963948e-05,
|
| 1152 |
+
"loss": 1.0124,
|
| 1153 |
+
"step": 161
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.09341213781173417,
|
| 1157 |
+
"grad_norm": 0.6015625,
|
| 1158 |
+
"learning_rate": 9.985401098626754e-05,
|
| 1159 |
+
"loss": 0.9245,
|
| 1160 |
+
"step": 162
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.09398875594637451,
|
| 1164 |
+
"grad_norm": 0.59765625,
|
| 1165 |
+
"learning_rate": 9.985207232194205e-05,
|
| 1166 |
+
"loss": 0.9019,
|
| 1167 |
+
"step": 163
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.09456537408101484,
|
| 1171 |
+
"grad_norm": 0.62890625,
|
| 1172 |
+
"learning_rate": 9.985012089737278e-05,
|
| 1173 |
+
"loss": 0.9756,
|
| 1174 |
+
"step": 164
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.09514199221565518,
|
| 1178 |
+
"grad_norm": 0.60546875,
|
| 1179 |
+
"learning_rate": 9.98481567132742e-05,
|
| 1180 |
+
"loss": 0.9209,
|
| 1181 |
+
"step": 165
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.09571861035029551,
|
| 1185 |
+
"grad_norm": 0.609375,
|
| 1186 |
+
"learning_rate": 9.98461797703655e-05,
|
| 1187 |
+
"loss": 0.9595,
|
| 1188 |
+
"step": 166
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.09629522848493585,
|
| 1192 |
+
"grad_norm": 0.6328125,
|
| 1193 |
+
"learning_rate": 9.98441900693704e-05,
|
| 1194 |
+
"loss": 0.9633,
|
| 1195 |
+
"step": 167
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.09687184661957618,
|
| 1199 |
+
"grad_norm": 0.65234375,
|
| 1200 |
+
"learning_rate": 9.984218761101744e-05,
|
| 1201 |
+
"loss": 0.9551,
|
| 1202 |
+
"step": 168
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.09744846475421652,
|
| 1206 |
+
"grad_norm": 0.61328125,
|
| 1207 |
+
"learning_rate": 9.984017239603978e-05,
|
| 1208 |
+
"loss": 0.927,
|
| 1209 |
+
"step": 169
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.09802508288885685,
|
| 1213 |
+
"grad_norm": 0.59375,
|
| 1214 |
+
"learning_rate": 9.98381444251752e-05,
|
| 1215 |
+
"loss": 0.9283,
|
| 1216 |
+
"step": 170
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.09860170102349719,
|
| 1220 |
+
"grad_norm": 0.65625,
|
| 1221 |
+
"learning_rate": 9.983610369916621e-05,
|
| 1222 |
+
"loss": 0.9794,
|
| 1223 |
+
"step": 171
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.09917831915813752,
|
| 1227 |
+
"grad_norm": 0.64453125,
|
| 1228 |
+
"learning_rate": 9.983405021875998e-05,
|
| 1229 |
+
"loss": 0.9629,
|
| 1230 |
+
"step": 172
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.09975493729277786,
|
| 1234 |
+
"grad_norm": 0.6171875,
|
| 1235 |
+
"learning_rate": 9.983198398470834e-05,
|
| 1236 |
+
"loss": 0.9343,
|
| 1237 |
+
"step": 173
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.10033155542741819,
|
| 1241 |
+
"grad_norm": 0.62109375,
|
| 1242 |
+
"learning_rate": 9.982990499776782e-05,
|
| 1243 |
+
"loss": 0.9536,
|
| 1244 |
+
"step": 174
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.10090817356205853,
|
| 1248 |
+
"grad_norm": 0.625,
|
| 1249 |
+
"learning_rate": 9.982781325869952e-05,
|
| 1250 |
+
"loss": 0.9283,
|
| 1251 |
+
"step": 175
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.10148479169669886,
|
| 1255 |
+
"grad_norm": 0.609375,
|
| 1256 |
+
"learning_rate": 9.982570876826934e-05,
|
| 1257 |
+
"loss": 0.9128,
|
| 1258 |
+
"step": 176
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.1020614098313392,
|
| 1262 |
+
"grad_norm": 0.6328125,
|
| 1263 |
+
"learning_rate": 9.982359152724777e-05,
|
| 1264 |
+
"loss": 0.9672,
|
| 1265 |
+
"step": 177
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.10263802796597953,
|
| 1269 |
+
"grad_norm": 0.6171875,
|
| 1270 |
+
"learning_rate": 9.982146153640997e-05,
|
| 1271 |
+
"loss": 0.9524,
|
| 1272 |
+
"step": 178
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.10321464610061987,
|
| 1276 |
+
"grad_norm": 0.60546875,
|
| 1277 |
+
"learning_rate": 9.981931879653582e-05,
|
| 1278 |
+
"loss": 0.9426,
|
| 1279 |
+
"step": 179
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.1037912642352602,
|
| 1283 |
+
"grad_norm": 0.63671875,
|
| 1284 |
+
"learning_rate": 9.981716330840977e-05,
|
| 1285 |
+
"loss": 0.9469,
|
| 1286 |
+
"step": 180
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.10436788236990054,
|
| 1290 |
+
"grad_norm": 0.625,
|
| 1291 |
+
"learning_rate": 9.981499507282109e-05,
|
| 1292 |
+
"loss": 0.8872,
|
| 1293 |
+
"step": 181
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.10494450050454086,
|
| 1297 |
+
"grad_norm": 0.60546875,
|
| 1298 |
+
"learning_rate": 9.981281409056358e-05,
|
| 1299 |
+
"loss": 0.9412,
|
| 1300 |
+
"step": 182
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.1055211186391812,
|
| 1304 |
+
"grad_norm": 0.60546875,
|
| 1305 |
+
"learning_rate": 9.981062036243573e-05,
|
| 1306 |
+
"loss": 0.9138,
|
| 1307 |
+
"step": 183
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.10609773677382153,
|
| 1311 |
+
"grad_norm": 0.6484375,
|
| 1312 |
+
"learning_rate": 9.980841388924076e-05,
|
| 1313 |
+
"loss": 0.9755,
|
| 1314 |
+
"step": 184
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.10667435490846187,
|
| 1318 |
+
"grad_norm": 0.59765625,
|
| 1319 |
+
"learning_rate": 9.980619467178646e-05,
|
| 1320 |
+
"loss": 0.9032,
|
| 1321 |
+
"step": 185
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.1072509730431022,
|
| 1325 |
+
"grad_norm": 0.625,
|
| 1326 |
+
"learning_rate": 9.980396271088544e-05,
|
| 1327 |
+
"loss": 0.9294,
|
| 1328 |
+
"step": 186
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.10782759117774254,
|
| 1332 |
+
"grad_norm": 0.609375,
|
| 1333 |
+
"learning_rate": 9.98017180073548e-05,
|
| 1334 |
+
"loss": 0.9219,
|
| 1335 |
+
"step": 187
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.10840420931238287,
|
| 1339 |
+
"grad_norm": 0.625,
|
| 1340 |
+
"learning_rate": 9.97994605620164e-05,
|
| 1341 |
+
"loss": 0.9394,
|
| 1342 |
+
"step": 188
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.10898082744702321,
|
| 1346 |
+
"grad_norm": 0.58984375,
|
| 1347 |
+
"learning_rate": 9.979719037569677e-05,
|
| 1348 |
+
"loss": 0.9266,
|
| 1349 |
+
"step": 189
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.10955744558166354,
|
| 1353 |
+
"grad_norm": 0.6015625,
|
| 1354 |
+
"learning_rate": 9.979490744922706e-05,
|
| 1355 |
+
"loss": 0.9293,
|
| 1356 |
+
"step": 190
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.11013406371630388,
|
| 1360 |
+
"grad_norm": 0.6171875,
|
| 1361 |
+
"learning_rate": 9.979261178344313e-05,
|
| 1362 |
+
"loss": 0.9627,
|
| 1363 |
+
"step": 191
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.11071068185094421,
|
| 1367 |
+
"grad_norm": 0.609375,
|
| 1368 |
+
"learning_rate": 9.979030337918544e-05,
|
| 1369 |
+
"loss": 0.9455,
|
| 1370 |
+
"step": 192
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.11128729998558455,
|
| 1374 |
+
"grad_norm": 0.62890625,
|
| 1375 |
+
"learning_rate": 9.978798223729921e-05,
|
| 1376 |
+
"loss": 0.9105,
|
| 1377 |
+
"step": 193
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.11186391812022488,
|
| 1381 |
+
"grad_norm": 0.625,
|
| 1382 |
+
"learning_rate": 9.978564835863424e-05,
|
| 1383 |
+
"loss": 0.9269,
|
| 1384 |
+
"step": 194
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.11244053625486522,
|
| 1388 |
+
"grad_norm": 0.5859375,
|
| 1389 |
+
"learning_rate": 9.978330174404504e-05,
|
| 1390 |
+
"loss": 0.8918,
|
| 1391 |
+
"step": 195
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.11301715438950555,
|
| 1395 |
+
"grad_norm": 0.64453125,
|
| 1396 |
+
"learning_rate": 9.978094239439072e-05,
|
| 1397 |
+
"loss": 0.9321,
|
| 1398 |
+
"step": 196
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.11359377252414589,
|
| 1402 |
+
"grad_norm": 0.59765625,
|
| 1403 |
+
"learning_rate": 9.977857031053517e-05,
|
| 1404 |
+
"loss": 0.881,
|
| 1405 |
+
"step": 197
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.11417039065878622,
|
| 1409 |
+
"grad_norm": 0.609375,
|
| 1410 |
+
"learning_rate": 9.977618549334684e-05,
|
| 1411 |
+
"loss": 0.9129,
|
| 1412 |
+
"step": 198
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.11474700879342656,
|
| 1416 |
+
"grad_norm": 0.609375,
|
| 1417 |
+
"learning_rate": 9.977378794369885e-05,
|
| 1418 |
+
"loss": 0.9191,
|
| 1419 |
+
"step": 199
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.11532362692806689,
|
| 1423 |
+
"grad_norm": 0.609375,
|
| 1424 |
+
"learning_rate": 9.977137766246905e-05,
|
| 1425 |
+
"loss": 0.9051,
|
| 1426 |
+
"step": 200
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.11590024506270723,
|
| 1430 |
+
"grad_norm": 0.6015625,
|
| 1431 |
+
"learning_rate": 9.976895465053986e-05,
|
| 1432 |
+
"loss": 0.9171,
|
| 1433 |
+
"step": 201
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.11647686319734755,
|
| 1437 |
+
"grad_norm": 0.62109375,
|
| 1438 |
+
"learning_rate": 9.976651890879842e-05,
|
| 1439 |
+
"loss": 0.9307,
|
| 1440 |
+
"step": 202
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.1170534813319879,
|
| 1444 |
+
"grad_norm": 0.640625,
|
| 1445 |
+
"learning_rate": 9.976407043813654e-05,
|
| 1446 |
+
"loss": 1.0037,
|
| 1447 |
+
"step": 203
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.11763009946662822,
|
| 1451 |
+
"grad_norm": 0.63671875,
|
| 1452 |
+
"learning_rate": 9.976160923945063e-05,
|
| 1453 |
+
"loss": 0.8872,
|
| 1454 |
+
"step": 204
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.11820671760126857,
|
| 1458 |
+
"grad_norm": 0.58203125,
|
| 1459 |
+
"learning_rate": 9.975913531364185e-05,
|
| 1460 |
+
"loss": 0.9196,
|
| 1461 |
+
"step": 205
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.1187833357359089,
|
| 1465 |
+
"grad_norm": 0.609375,
|
| 1466 |
+
"learning_rate": 9.975664866161594e-05,
|
| 1467 |
+
"loss": 0.9431,
|
| 1468 |
+
"step": 206
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 0.11935995387054923,
|
| 1472 |
+
"grad_norm": 0.62890625,
|
| 1473 |
+
"learning_rate": 9.975414928428331e-05,
|
| 1474 |
+
"loss": 0.9603,
|
| 1475 |
+
"step": 207
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.11993657200518956,
|
| 1479 |
+
"grad_norm": 0.62890625,
|
| 1480 |
+
"learning_rate": 9.975163718255906e-05,
|
| 1481 |
+
"loss": 0.9841,
|
| 1482 |
+
"step": 208
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.1205131901398299,
|
| 1486 |
+
"grad_norm": 0.6171875,
|
| 1487 |
+
"learning_rate": 9.974911235736295e-05,
|
| 1488 |
+
"loss": 0.9446,
|
| 1489 |
+
"step": 209
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.12108980827447023,
|
| 1493 |
+
"grad_norm": 0.6171875,
|
| 1494 |
+
"learning_rate": 9.974657480961938e-05,
|
| 1495 |
+
"loss": 0.9353,
|
| 1496 |
+
"step": 210
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.12166642640911057,
|
| 1500 |
+
"grad_norm": 0.6171875,
|
| 1501 |
+
"learning_rate": 9.97440245402574e-05,
|
| 1502 |
+
"loss": 0.914,
|
| 1503 |
+
"step": 211
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.1222430445437509,
|
| 1507 |
+
"grad_norm": 0.6484375,
|
| 1508 |
+
"learning_rate": 9.974146155021074e-05,
|
| 1509 |
+
"loss": 0.9604,
|
| 1510 |
+
"step": 212
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.12281966267839124,
|
| 1514 |
+
"grad_norm": 0.609375,
|
| 1515 |
+
"learning_rate": 9.973888584041776e-05,
|
| 1516 |
+
"loss": 0.9225,
|
| 1517 |
+
"step": 213
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.12339628081303157,
|
| 1521 |
+
"grad_norm": 0.5859375,
|
| 1522 |
+
"learning_rate": 9.973629741182151e-05,
|
| 1523 |
+
"loss": 0.8992,
|
| 1524 |
+
"step": 214
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.12397289894767191,
|
| 1528 |
+
"grad_norm": 0.60546875,
|
| 1529 |
+
"learning_rate": 9.973369626536968e-05,
|
| 1530 |
+
"loss": 0.9262,
|
| 1531 |
+
"step": 215
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.12454951708231224,
|
| 1535 |
+
"grad_norm": 0.59375,
|
| 1536 |
+
"learning_rate": 9.973108240201461e-05,
|
| 1537 |
+
"loss": 0.9602,
|
| 1538 |
+
"step": 216
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.12512613521695257,
|
| 1542 |
+
"grad_norm": 0.62109375,
|
| 1543 |
+
"learning_rate": 9.97284558227133e-05,
|
| 1544 |
+
"loss": 0.9615,
|
| 1545 |
+
"step": 217
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.1257027533515929,
|
| 1549 |
+
"grad_norm": 0.59375,
|
| 1550 |
+
"learning_rate": 9.972581652842743e-05,
|
| 1551 |
+
"loss": 0.9679,
|
| 1552 |
+
"step": 218
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.12627937148623325,
|
| 1556 |
+
"grad_norm": 0.59375,
|
| 1557 |
+
"learning_rate": 9.972316452012327e-05,
|
| 1558 |
+
"loss": 0.9701,
|
| 1559 |
+
"step": 219
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.12685598962087358,
|
| 1563 |
+
"grad_norm": 0.61328125,
|
| 1564 |
+
"learning_rate": 9.972049979877183e-05,
|
| 1565 |
+
"loss": 0.92,
|
| 1566 |
+
"step": 220
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.1274326077555139,
|
| 1570 |
+
"grad_norm": 0.59375,
|
| 1571 |
+
"learning_rate": 9.971782236534872e-05,
|
| 1572 |
+
"loss": 0.9352,
|
| 1573 |
+
"step": 221
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.12800922589015423,
|
| 1577 |
+
"grad_norm": 0.59765625,
|
| 1578 |
+
"learning_rate": 9.971513222083423e-05,
|
| 1579 |
+
"loss": 0.8721,
|
| 1580 |
+
"step": 222
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.1285858440247946,
|
| 1584 |
+
"grad_norm": 0.62109375,
|
| 1585 |
+
"learning_rate": 9.971242936621328e-05,
|
| 1586 |
+
"loss": 0.9416,
|
| 1587 |
+
"step": 223
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.12916246215943492,
|
| 1591 |
+
"grad_norm": 0.6171875,
|
| 1592 |
+
"learning_rate": 9.970971380247545e-05,
|
| 1593 |
+
"loss": 0.9501,
|
| 1594 |
+
"step": 224
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.12973908029407524,
|
| 1598 |
+
"grad_norm": 0.59765625,
|
| 1599 |
+
"learning_rate": 9.970698553061497e-05,
|
| 1600 |
+
"loss": 0.947,
|
| 1601 |
+
"step": 225
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.13031569842871557,
|
| 1605 |
+
"grad_norm": 0.67578125,
|
| 1606 |
+
"learning_rate": 9.970424455163074e-05,
|
| 1607 |
+
"loss": 0.8953,
|
| 1608 |
+
"step": 226
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.13089231656335593,
|
| 1612 |
+
"grad_norm": 0.6015625,
|
| 1613 |
+
"learning_rate": 9.970149086652631e-05,
|
| 1614 |
+
"loss": 0.9052,
|
| 1615 |
+
"step": 227
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.13146893469799625,
|
| 1619 |
+
"grad_norm": 0.60546875,
|
| 1620 |
+
"learning_rate": 9.969872447630988e-05,
|
| 1621 |
+
"loss": 0.9875,
|
| 1622 |
+
"step": 228
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.13204555283263658,
|
| 1626 |
+
"grad_norm": 0.59375,
|
| 1627 |
+
"learning_rate": 9.969594538199429e-05,
|
| 1628 |
+
"loss": 0.8976,
|
| 1629 |
+
"step": 229
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.1326221709672769,
|
| 1633 |
+
"grad_norm": 0.60546875,
|
| 1634 |
+
"learning_rate": 9.969315358459704e-05,
|
| 1635 |
+
"loss": 0.9022,
|
| 1636 |
+
"step": 230
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.13319878910191726,
|
| 1640 |
+
"grad_norm": 0.62109375,
|
| 1641 |
+
"learning_rate": 9.969034908514026e-05,
|
| 1642 |
+
"loss": 0.8922,
|
| 1643 |
+
"step": 231
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.1337754072365576,
|
| 1647 |
+
"grad_norm": 0.60546875,
|
| 1648 |
+
"learning_rate": 9.968753188465077e-05,
|
| 1649 |
+
"loss": 0.9523,
|
| 1650 |
+
"step": 232
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.13435202537119792,
|
| 1654 |
+
"grad_norm": 0.59375,
|
| 1655 |
+
"learning_rate": 9.968470198416e-05,
|
| 1656 |
+
"loss": 0.9256,
|
| 1657 |
+
"step": 233
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.13492864350583825,
|
| 1661 |
+
"grad_norm": 0.59765625,
|
| 1662 |
+
"learning_rate": 9.968185938470409e-05,
|
| 1663 |
+
"loss": 0.9132,
|
| 1664 |
+
"step": 234
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.1355052616404786,
|
| 1668 |
+
"grad_norm": 0.58984375,
|
| 1669 |
+
"learning_rate": 9.967900408732373e-05,
|
| 1670 |
+
"loss": 0.8912,
|
| 1671 |
+
"step": 235
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.13608187977511893,
|
| 1675 |
+
"grad_norm": 0.6328125,
|
| 1676 |
+
"learning_rate": 9.967613609306439e-05,
|
| 1677 |
+
"loss": 0.9169,
|
| 1678 |
+
"step": 236
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 0.13665849790975926,
|
| 1682 |
+
"grad_norm": 0.61328125,
|
| 1683 |
+
"learning_rate": 9.967325540297608e-05,
|
| 1684 |
+
"loss": 0.9064,
|
| 1685 |
+
"step": 237
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.13723511604439959,
|
| 1689 |
+
"grad_norm": 0.58984375,
|
| 1690 |
+
"learning_rate": 9.967036201811346e-05,
|
| 1691 |
+
"loss": 0.9257,
|
| 1692 |
+
"step": 238
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 0.13781173417903994,
|
| 1696 |
+
"grad_norm": 0.6328125,
|
| 1697 |
+
"learning_rate": 9.966745593953593e-05,
|
| 1698 |
+
"loss": 0.9532,
|
| 1699 |
+
"step": 239
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 0.13838835231368027,
|
| 1703 |
+
"grad_norm": 0.63671875,
|
| 1704 |
+
"learning_rate": 9.966453716830743e-05,
|
| 1705 |
+
"loss": 0.9413,
|
| 1706 |
+
"step": 240
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.1389649704483206,
|
| 1710 |
+
"grad_norm": 0.62109375,
|
| 1711 |
+
"learning_rate": 9.966160570549666e-05,
|
| 1712 |
+
"loss": 0.9554,
|
| 1713 |
+
"step": 241
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.13954158858296092,
|
| 1717 |
+
"grad_norm": 0.62109375,
|
| 1718 |
+
"learning_rate": 9.965866155217685e-05,
|
| 1719 |
+
"loss": 0.938,
|
| 1720 |
+
"step": 242
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.14011820671760128,
|
| 1724 |
+
"grad_norm": 0.625,
|
| 1725 |
+
"learning_rate": 9.965570470942594e-05,
|
| 1726 |
+
"loss": 0.8946,
|
| 1727 |
+
"step": 243
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.1406948248522416,
|
| 1731 |
+
"grad_norm": 0.6171875,
|
| 1732 |
+
"learning_rate": 9.965273517832652e-05,
|
| 1733 |
+
"loss": 0.9377,
|
| 1734 |
+
"step": 244
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 0.14127144298688193,
|
| 1738 |
+
"grad_norm": 0.625,
|
| 1739 |
+
"learning_rate": 9.964975295996582e-05,
|
| 1740 |
+
"loss": 0.9558,
|
| 1741 |
+
"step": 245
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.14184806112152226,
|
| 1745 |
+
"grad_norm": 0.6484375,
|
| 1746 |
+
"learning_rate": 9.964675805543569e-05,
|
| 1747 |
+
"loss": 0.9083,
|
| 1748 |
+
"step": 246
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.14242467925616262,
|
| 1752 |
+
"grad_norm": 0.62890625,
|
| 1753 |
+
"learning_rate": 9.964375046583265e-05,
|
| 1754 |
+
"loss": 0.9289,
|
| 1755 |
+
"step": 247
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.14300129739080294,
|
| 1759 |
+
"grad_norm": 0.609375,
|
| 1760 |
+
"learning_rate": 9.964073019225784e-05,
|
| 1761 |
+
"loss": 0.9139,
|
| 1762 |
+
"step": 248
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.14357791552544327,
|
| 1766 |
+
"grad_norm": 0.66796875,
|
| 1767 |
+
"learning_rate": 9.963769723581709e-05,
|
| 1768 |
+
"loss": 0.9616,
|
| 1769 |
+
"step": 249
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.1441545336600836,
|
| 1773 |
+
"grad_norm": 0.609375,
|
| 1774 |
+
"learning_rate": 9.963465159762082e-05,
|
| 1775 |
+
"loss": 0.9082,
|
| 1776 |
+
"step": 250
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.14473115179472396,
|
| 1780 |
+
"grad_norm": 0.625,
|
| 1781 |
+
"learning_rate": 9.963159327878411e-05,
|
| 1782 |
+
"loss": 0.8906,
|
| 1783 |
+
"step": 251
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.14530776992936428,
|
| 1787 |
+
"grad_norm": 0.63671875,
|
| 1788 |
+
"learning_rate": 9.962852228042671e-05,
|
| 1789 |
+
"loss": 0.9086,
|
| 1790 |
+
"step": 252
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.1458843880640046,
|
| 1794 |
+
"grad_norm": 0.6171875,
|
| 1795 |
+
"learning_rate": 9.962543860367299e-05,
|
| 1796 |
+
"loss": 0.914,
|
| 1797 |
+
"step": 253
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.14646100619864494,
|
| 1801 |
+
"grad_norm": 0.609375,
|
| 1802 |
+
"learning_rate": 9.962234224965196e-05,
|
| 1803 |
+
"loss": 0.955,
|
| 1804 |
+
"step": 254
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.1470376243332853,
|
| 1808 |
+
"grad_norm": 0.59765625,
|
| 1809 |
+
"learning_rate": 9.961923321949729e-05,
|
| 1810 |
+
"loss": 0.9088,
|
| 1811 |
+
"step": 255
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.14761424246792562,
|
| 1815 |
+
"grad_norm": 0.59765625,
|
| 1816 |
+
"learning_rate": 9.961611151434722e-05,
|
| 1817 |
+
"loss": 0.9276,
|
| 1818 |
+
"step": 256
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.14819086060256595,
|
| 1822 |
+
"grad_norm": 0.60546875,
|
| 1823 |
+
"learning_rate": 9.961297713534476e-05,
|
| 1824 |
+
"loss": 0.9369,
|
| 1825 |
+
"step": 257
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.14876747873720628,
|
| 1829 |
+
"grad_norm": 0.59765625,
|
| 1830 |
+
"learning_rate": 9.960983008363745e-05,
|
| 1831 |
+
"loss": 0.9465,
|
| 1832 |
+
"step": 258
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.14934409687184663,
|
| 1836 |
+
"grad_norm": 0.58203125,
|
| 1837 |
+
"learning_rate": 9.96066703603775e-05,
|
| 1838 |
+
"loss": 0.9,
|
| 1839 |
+
"step": 259
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.14992071500648696,
|
| 1843 |
+
"grad_norm": 0.609375,
|
| 1844 |
+
"learning_rate": 9.960349796672177e-05,
|
| 1845 |
+
"loss": 0.9486,
|
| 1846 |
+
"step": 260
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.1504973331411273,
|
| 1850 |
+
"grad_norm": 0.57421875,
|
| 1851 |
+
"learning_rate": 9.960031290383179e-05,
|
| 1852 |
+
"loss": 0.9293,
|
| 1853 |
+
"step": 261
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.15107395127576762,
|
| 1857 |
+
"grad_norm": 0.5859375,
|
| 1858 |
+
"learning_rate": 9.959711517287364e-05,
|
| 1859 |
+
"loss": 0.8732,
|
| 1860 |
+
"step": 262
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.15165056941040797,
|
| 1864 |
+
"grad_norm": 0.609375,
|
| 1865 |
+
"learning_rate": 9.959390477501814e-05,
|
| 1866 |
+
"loss": 0.9095,
|
| 1867 |
+
"step": 263
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.1522271875450483,
|
| 1871 |
+
"grad_norm": 0.59375,
|
| 1872 |
+
"learning_rate": 9.959068171144063e-05,
|
| 1873 |
+
"loss": 0.9583,
|
| 1874 |
+
"step": 264
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.15280380567968863,
|
| 1878 |
+
"grad_norm": 0.56640625,
|
| 1879 |
+
"learning_rate": 9.958744598332126e-05,
|
| 1880 |
+
"loss": 0.9175,
|
| 1881 |
+
"step": 265
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.15338042381432895,
|
| 1885 |
+
"grad_norm": 0.6328125,
|
| 1886 |
+
"learning_rate": 9.958419759184463e-05,
|
| 1887 |
+
"loss": 0.9034,
|
| 1888 |
+
"step": 266
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.1539570419489693,
|
| 1892 |
+
"grad_norm": 0.57421875,
|
| 1893 |
+
"learning_rate": 9.95809365382001e-05,
|
| 1894 |
+
"loss": 0.8857,
|
| 1895 |
+
"step": 267
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.15453366008360964,
|
| 1899 |
+
"grad_norm": 0.59765625,
|
| 1900 |
+
"learning_rate": 9.957766282358161e-05,
|
| 1901 |
+
"loss": 0.9408,
|
| 1902 |
+
"step": 268
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.15511027821824996,
|
| 1906 |
+
"grad_norm": 0.609375,
|
| 1907 |
+
"learning_rate": 9.957437644918775e-05,
|
| 1908 |
+
"loss": 0.9403,
|
| 1909 |
+
"step": 269
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.1556868963528903,
|
| 1913 |
+
"grad_norm": 0.640625,
|
| 1914 |
+
"learning_rate": 9.957107741622176e-05,
|
| 1915 |
+
"loss": 0.8624,
|
| 1916 |
+
"step": 270
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.15626351448753062,
|
| 1920 |
+
"grad_norm": 0.57421875,
|
| 1921 |
+
"learning_rate": 9.956776572589148e-05,
|
| 1922 |
+
"loss": 0.9076,
|
| 1923 |
+
"step": 271
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.15684013262217097,
|
| 1927 |
+
"grad_norm": 0.59765625,
|
| 1928 |
+
"learning_rate": 9.956444137940943e-05,
|
| 1929 |
+
"loss": 0.9275,
|
| 1930 |
+
"step": 272
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.1574167507568113,
|
| 1934 |
+
"grad_norm": 0.5859375,
|
| 1935 |
+
"learning_rate": 9.956110437799273e-05,
|
| 1936 |
+
"loss": 0.8866,
|
| 1937 |
+
"step": 273
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.15799336889145163,
|
| 1941 |
+
"grad_norm": 0.58984375,
|
| 1942 |
+
"learning_rate": 9.955775472286315e-05,
|
| 1943 |
+
"loss": 0.9608,
|
| 1944 |
+
"step": 274
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.15856998702609196,
|
| 1948 |
+
"grad_norm": 0.65234375,
|
| 1949 |
+
"learning_rate": 9.955439241524707e-05,
|
| 1950 |
+
"loss": 0.9419,
|
| 1951 |
+
"step": 275
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.1591466051607323,
|
| 1955 |
+
"grad_norm": 0.61328125,
|
| 1956 |
+
"learning_rate": 9.955101745637552e-05,
|
| 1957 |
+
"loss": 0.951,
|
| 1958 |
+
"step": 276
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.15972322329537264,
|
| 1962 |
+
"grad_norm": 0.6328125,
|
| 1963 |
+
"learning_rate": 9.954762984748413e-05,
|
| 1964 |
+
"loss": 0.9377,
|
| 1965 |
+
"step": 277
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.16029984143001297,
|
| 1969 |
+
"grad_norm": 0.6171875,
|
| 1970 |
+
"learning_rate": 9.954422958981326e-05,
|
| 1971 |
+
"loss": 0.8692,
|
| 1972 |
+
"step": 278
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.1608764595646533,
|
| 1976 |
+
"grad_norm": 0.58203125,
|
| 1977 |
+
"learning_rate": 9.95408166846078e-05,
|
| 1978 |
+
"loss": 0.9131,
|
| 1979 |
+
"step": 279
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.16145307769929365,
|
| 1983 |
+
"grad_norm": 0.58984375,
|
| 1984 |
+
"learning_rate": 9.953739113311726e-05,
|
| 1985 |
+
"loss": 0.9578,
|
| 1986 |
+
"step": 280
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 0.16202969583393398,
|
| 1990 |
+
"grad_norm": 0.6328125,
|
| 1991 |
+
"learning_rate": 9.953395293659591e-05,
|
| 1992 |
+
"loss": 0.9629,
|
| 1993 |
+
"step": 281
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.1626063139685743,
|
| 1997 |
+
"grad_norm": 0.61328125,
|
| 1998 |
+
"learning_rate": 9.953050209630249e-05,
|
| 1999 |
+
"loss": 0.9274,
|
| 2000 |
+
"step": 282
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.16318293210321463,
|
| 2004 |
+
"grad_norm": 0.61328125,
|
| 2005 |
+
"learning_rate": 9.952703861350046e-05,
|
| 2006 |
+
"loss": 0.8947,
|
| 2007 |
+
"step": 283
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.163759550237855,
|
| 2011 |
+
"grad_norm": 0.58203125,
|
| 2012 |
+
"learning_rate": 9.952356248945791e-05,
|
| 2013 |
+
"loss": 0.9305,
|
| 2014 |
+
"step": 284
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.16433616837249532,
|
| 2018 |
+
"grad_norm": 0.5625,
|
| 2019 |
+
"learning_rate": 9.952007372544751e-05,
|
| 2020 |
+
"loss": 0.898,
|
| 2021 |
+
"step": 285
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.16491278650713564,
|
| 2025 |
+
"grad_norm": 0.5859375,
|
| 2026 |
+
"learning_rate": 9.951657232274662e-05,
|
| 2027 |
+
"loss": 0.9065,
|
| 2028 |
+
"step": 286
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.16548940464177597,
|
| 2032 |
+
"grad_norm": 0.57421875,
|
| 2033 |
+
"learning_rate": 9.951305828263715e-05,
|
| 2034 |
+
"loss": 0.9443,
|
| 2035 |
+
"step": 287
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.16606602277641633,
|
| 2039 |
+
"grad_norm": 0.5859375,
|
| 2040 |
+
"learning_rate": 9.950953160640573e-05,
|
| 2041 |
+
"loss": 0.8977,
|
| 2042 |
+
"step": 288
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.16664264091105666,
|
| 2046 |
+
"grad_norm": 0.59765625,
|
| 2047 |
+
"learning_rate": 9.950599229534354e-05,
|
| 2048 |
+
"loss": 0.9149,
|
| 2049 |
+
"step": 289
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.16721925904569698,
|
| 2053 |
+
"grad_norm": 0.59375,
|
| 2054 |
+
"learning_rate": 9.950244035074641e-05,
|
| 2055 |
+
"loss": 0.8839,
|
| 2056 |
+
"step": 290
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.1677958771803373,
|
| 2060 |
+
"grad_norm": 0.58984375,
|
| 2061 |
+
"learning_rate": 9.949887577391482e-05,
|
| 2062 |
+
"loss": 0.9395,
|
| 2063 |
+
"step": 291
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.16837249531497767,
|
| 2067 |
+
"grad_norm": 0.61328125,
|
| 2068 |
+
"learning_rate": 9.949529856615382e-05,
|
| 2069 |
+
"loss": 0.9681,
|
| 2070 |
+
"step": 292
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.168949113449618,
|
| 2074 |
+
"grad_norm": 0.58203125,
|
| 2075 |
+
"learning_rate": 9.949170872877314e-05,
|
| 2076 |
+
"loss": 0.9334,
|
| 2077 |
+
"step": 293
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.16952573158425832,
|
| 2081 |
+
"grad_norm": 0.59765625,
|
| 2082 |
+
"learning_rate": 9.948810626308711e-05,
|
| 2083 |
+
"loss": 0.8906,
|
| 2084 |
+
"step": 294
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.17010234971889865,
|
| 2088 |
+
"grad_norm": 0.59765625,
|
| 2089 |
+
"learning_rate": 9.94844911704147e-05,
|
| 2090 |
+
"loss": 0.936,
|
| 2091 |
+
"step": 295
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.170678967853539,
|
| 2095 |
+
"grad_norm": 0.59765625,
|
| 2096 |
+
"learning_rate": 9.948086345207945e-05,
|
| 2097 |
+
"loss": 0.9193,
|
| 2098 |
+
"step": 296
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.17125558598817933,
|
| 2102 |
+
"grad_norm": 0.59765625,
|
| 2103 |
+
"learning_rate": 9.94772231094096e-05,
|
| 2104 |
+
"loss": 0.924,
|
| 2105 |
+
"step": 297
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.17183220412281966,
|
| 2109 |
+
"grad_norm": 0.59765625,
|
| 2110 |
+
"learning_rate": 9.947357014373797e-05,
|
| 2111 |
+
"loss": 0.921,
|
| 2112 |
+
"step": 298
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.17240882225746,
|
| 2116 |
+
"grad_norm": 0.60546875,
|
| 2117 |
+
"learning_rate": 9.946990455640197e-05,
|
| 2118 |
+
"loss": 0.9262,
|
| 2119 |
+
"step": 299
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.17298544039210034,
|
| 2123 |
+
"grad_norm": 0.578125,
|
| 2124 |
+
"learning_rate": 9.94662263487437e-05,
|
| 2125 |
+
"loss": 0.9545,
|
| 2126 |
+
"step": 300
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.17356205852674067,
|
| 2130 |
+
"grad_norm": 0.60546875,
|
| 2131 |
+
"learning_rate": 9.946253552210984e-05,
|
| 2132 |
+
"loss": 0.9526,
|
| 2133 |
+
"step": 301
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.174138676661381,
|
| 2137 |
+
"grad_norm": 0.578125,
|
| 2138 |
+
"learning_rate": 9.94588320778517e-05,
|
| 2139 |
+
"loss": 0.8684,
|
| 2140 |
+
"step": 302
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.17471529479602133,
|
| 2144 |
+
"grad_norm": 0.60546875,
|
| 2145 |
+
"learning_rate": 9.945511601732518e-05,
|
| 2146 |
+
"loss": 0.933,
|
| 2147 |
+
"step": 303
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.17529191293066168,
|
| 2151 |
+
"grad_norm": 0.58984375,
|
| 2152 |
+
"learning_rate": 9.945138734189088e-05,
|
| 2153 |
+
"loss": 0.9365,
|
| 2154 |
+
"step": 304
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.175868531065302,
|
| 2158 |
+
"grad_norm": 0.59765625,
|
| 2159 |
+
"learning_rate": 9.94476460529139e-05,
|
| 2160 |
+
"loss": 0.9153,
|
| 2161 |
+
"step": 305
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.17644514919994234,
|
| 2165 |
+
"grad_norm": 0.58203125,
|
| 2166 |
+
"learning_rate": 9.944389215176406e-05,
|
| 2167 |
+
"loss": 0.9356,
|
| 2168 |
+
"step": 306
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.17702176733458266,
|
| 2172 |
+
"grad_norm": 0.59765625,
|
| 2173 |
+
"learning_rate": 9.944012563981575e-05,
|
| 2174 |
+
"loss": 0.9362,
|
| 2175 |
+
"step": 307
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.17759838546922302,
|
| 2179 |
+
"grad_norm": 0.54296875,
|
| 2180 |
+
"learning_rate": 9.9436346518448e-05,
|
| 2181 |
+
"loss": 0.9098,
|
| 2182 |
+
"step": 308
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.17817500360386335,
|
| 2186 |
+
"grad_norm": 0.58984375,
|
| 2187 |
+
"learning_rate": 9.943255478904444e-05,
|
| 2188 |
+
"loss": 0.9465,
|
| 2189 |
+
"step": 309
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.17875162173850367,
|
| 2193 |
+
"grad_norm": 0.58984375,
|
| 2194 |
+
"learning_rate": 9.942875045299331e-05,
|
| 2195 |
+
"loss": 0.8589,
|
| 2196 |
+
"step": 310
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.179328239873144,
|
| 2200 |
+
"grad_norm": 0.5859375,
|
| 2201 |
+
"learning_rate": 9.942493351168747e-05,
|
| 2202 |
+
"loss": 0.9228,
|
| 2203 |
+
"step": 311
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.17990485800778436,
|
| 2207 |
+
"grad_norm": 0.57421875,
|
| 2208 |
+
"learning_rate": 9.942110396652442e-05,
|
| 2209 |
+
"loss": 0.925,
|
| 2210 |
+
"step": 312
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.18048147614242468,
|
| 2214 |
+
"grad_norm": 0.61328125,
|
| 2215 |
+
"learning_rate": 9.941726181890625e-05,
|
| 2216 |
+
"loss": 0.9151,
|
| 2217 |
+
"step": 313
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.181058094277065,
|
| 2221 |
+
"grad_norm": 0.58203125,
|
| 2222 |
+
"learning_rate": 9.941340707023965e-05,
|
| 2223 |
+
"loss": 0.9018,
|
| 2224 |
+
"step": 314
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.18163471241170534,
|
| 2228 |
+
"grad_norm": 0.5859375,
|
| 2229 |
+
"learning_rate": 9.940953972193596e-05,
|
| 2230 |
+
"loss": 0.9353,
|
| 2231 |
+
"step": 315
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.1822113305463457,
|
| 2235 |
+
"grad_norm": 0.66015625,
|
| 2236 |
+
"learning_rate": 9.940565977541112e-05,
|
| 2237 |
+
"loss": 0.9979,
|
| 2238 |
+
"step": 316
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.18278794868098602,
|
| 2242 |
+
"grad_norm": 0.60546875,
|
| 2243 |
+
"learning_rate": 9.940176723208569e-05,
|
| 2244 |
+
"loss": 0.9077,
|
| 2245 |
+
"step": 317
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.18336456681562635,
|
| 2249 |
+
"grad_norm": 0.60546875,
|
| 2250 |
+
"learning_rate": 9.93978620933848e-05,
|
| 2251 |
+
"loss": 0.9142,
|
| 2252 |
+
"step": 318
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.18394118495026668,
|
| 2256 |
+
"grad_norm": 0.6015625,
|
| 2257 |
+
"learning_rate": 9.939394436073824e-05,
|
| 2258 |
+
"loss": 0.9162,
|
| 2259 |
+
"step": 319
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.18451780308490703,
|
| 2263 |
+
"grad_norm": 0.609375,
|
| 2264 |
+
"learning_rate": 9.939001403558036e-05,
|
| 2265 |
+
"loss": 0.9548,
|
| 2266 |
+
"step": 320
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.18509442121954736,
|
| 2270 |
+
"grad_norm": 0.59765625,
|
| 2271 |
+
"learning_rate": 9.938607111935024e-05,
|
| 2272 |
+
"loss": 0.9308,
|
| 2273 |
+
"step": 321
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.1856710393541877,
|
| 2277 |
+
"grad_norm": 0.6328125,
|
| 2278 |
+
"learning_rate": 9.938211561349137e-05,
|
| 2279 |
+
"loss": 0.9203,
|
| 2280 |
+
"step": 322
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.18624765748882802,
|
| 2284 |
+
"grad_norm": 0.58203125,
|
| 2285 |
+
"learning_rate": 9.937814751945207e-05,
|
| 2286 |
+
"loss": 0.8803,
|
| 2287 |
+
"step": 323
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.18682427562346834,
|
| 2291 |
+
"grad_norm": 0.60546875,
|
| 2292 |
+
"learning_rate": 9.937416683868508e-05,
|
| 2293 |
+
"loss": 0.8631,
|
| 2294 |
+
"step": 324
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.1874008937581087,
|
| 2298 |
+
"grad_norm": 0.609375,
|
| 2299 |
+
"learning_rate": 9.937017357264786e-05,
|
| 2300 |
+
"loss": 0.9173,
|
| 2301 |
+
"step": 325
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.18797751189274903,
|
| 2305 |
+
"grad_norm": 0.60546875,
|
| 2306 |
+
"learning_rate": 9.936616772280247e-05,
|
| 2307 |
+
"loss": 0.9349,
|
| 2308 |
+
"step": 326
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.18855413002738936,
|
| 2312 |
+
"grad_norm": 0.578125,
|
| 2313 |
+
"learning_rate": 9.936214929061552e-05,
|
| 2314 |
+
"loss": 0.9278,
|
| 2315 |
+
"step": 327
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.18913074816202968,
|
| 2319 |
+
"grad_norm": 0.5859375,
|
| 2320 |
+
"learning_rate": 9.935811827755827e-05,
|
| 2321 |
+
"loss": 0.9365,
|
| 2322 |
+
"step": 328
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.18970736629667004,
|
| 2326 |
+
"grad_norm": 0.6015625,
|
| 2327 |
+
"learning_rate": 9.935407468510658e-05,
|
| 2328 |
+
"loss": 0.8856,
|
| 2329 |
+
"step": 329
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.19028398443131037,
|
| 2333 |
+
"grad_norm": 0.59765625,
|
| 2334 |
+
"learning_rate": 9.935001851474093e-05,
|
| 2335 |
+
"loss": 0.9021,
|
| 2336 |
+
"step": 330
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.1908606025659507,
|
| 2340 |
+
"grad_norm": 0.6015625,
|
| 2341 |
+
"learning_rate": 9.934594976794638e-05,
|
| 2342 |
+
"loss": 0.9066,
|
| 2343 |
+
"step": 331
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.19143722070059102,
|
| 2347 |
+
"grad_norm": 0.60546875,
|
| 2348 |
+
"learning_rate": 9.93418684462126e-05,
|
| 2349 |
+
"loss": 0.9428,
|
| 2350 |
+
"step": 332
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.19201383883523138,
|
| 2354 |
+
"grad_norm": 0.58203125,
|
| 2355 |
+
"learning_rate": 9.933777455103385e-05,
|
| 2356 |
+
"loss": 0.9251,
|
| 2357 |
+
"step": 333
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.1925904569698717,
|
| 2361 |
+
"grad_norm": 0.59765625,
|
| 2362 |
+
"learning_rate": 9.933366808390904e-05,
|
| 2363 |
+
"loss": 0.9136,
|
| 2364 |
+
"step": 334
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.19316707510451203,
|
| 2368 |
+
"grad_norm": 0.60546875,
|
| 2369 |
+
"learning_rate": 9.932954904634165e-05,
|
| 2370 |
+
"loss": 0.9046,
|
| 2371 |
+
"step": 335
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.19374369323915236,
|
| 2375 |
+
"grad_norm": 0.59375,
|
| 2376 |
+
"learning_rate": 9.932541743983974e-05,
|
| 2377 |
+
"loss": 0.9416,
|
| 2378 |
+
"step": 336
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.19432031137379271,
|
| 2382 |
+
"grad_norm": 0.59765625,
|
| 2383 |
+
"learning_rate": 9.9321273265916e-05,
|
| 2384 |
+
"loss": 0.8579,
|
| 2385 |
+
"step": 337
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.19489692950843304,
|
| 2389 |
+
"grad_norm": 0.58984375,
|
| 2390 |
+
"learning_rate": 9.931711652608777e-05,
|
| 2391 |
+
"loss": 0.9288,
|
| 2392 |
+
"step": 338
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.19547354764307337,
|
| 2396 |
+
"grad_norm": 0.5625,
|
| 2397 |
+
"learning_rate": 9.931294722187689e-05,
|
| 2398 |
+
"loss": 0.8769,
|
| 2399 |
+
"step": 339
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.1960501657777137,
|
| 2403 |
+
"grad_norm": 0.609375,
|
| 2404 |
+
"learning_rate": 9.930876535480986e-05,
|
| 2405 |
+
"loss": 0.9835,
|
| 2406 |
+
"step": 340
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.19662678391235405,
|
| 2410 |
+
"grad_norm": 0.6015625,
|
| 2411 |
+
"learning_rate": 9.930457092641778e-05,
|
| 2412 |
+
"loss": 0.9814,
|
| 2413 |
+
"step": 341
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.19720340204699438,
|
| 2417 |
+
"grad_norm": 0.609375,
|
| 2418 |
+
"learning_rate": 9.930036393823632e-05,
|
| 2419 |
+
"loss": 0.9236,
|
| 2420 |
+
"step": 342
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.1977800201816347,
|
| 2424 |
+
"grad_norm": 0.578125,
|
| 2425 |
+
"learning_rate": 9.92961443918058e-05,
|
| 2426 |
+
"loss": 0.8814,
|
| 2427 |
+
"step": 343
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.19835663831627504,
|
| 2431 |
+
"grad_norm": 0.578125,
|
| 2432 |
+
"learning_rate": 9.929191228867105e-05,
|
| 2433 |
+
"loss": 0.901,
|
| 2434 |
+
"step": 344
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.1989332564509154,
|
| 2438 |
+
"grad_norm": 0.5859375,
|
| 2439 |
+
"learning_rate": 9.928766763038162e-05,
|
| 2440 |
+
"loss": 0.897,
|
| 2441 |
+
"step": 345
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.19950987458555572,
|
| 2445 |
+
"grad_norm": 0.578125,
|
| 2446 |
+
"learning_rate": 9.928341041849154e-05,
|
| 2447 |
+
"loss": 0.915,
|
| 2448 |
+
"step": 346
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.20008649272019605,
|
| 2452 |
+
"grad_norm": 0.6015625,
|
| 2453 |
+
"learning_rate": 9.927914065455952e-05,
|
| 2454 |
+
"loss": 0.9253,
|
| 2455 |
+
"step": 347
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.20066311085483637,
|
| 2459 |
+
"grad_norm": 0.58984375,
|
| 2460 |
+
"learning_rate": 9.927485834014878e-05,
|
| 2461 |
+
"loss": 0.9277,
|
| 2462 |
+
"step": 348
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.20123972898947673,
|
| 2466 |
+
"grad_norm": 0.60546875,
|
| 2467 |
+
"learning_rate": 9.927056347682724e-05,
|
| 2468 |
+
"loss": 0.8951,
|
| 2469 |
+
"step": 349
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.20181634712411706,
|
| 2473 |
+
"grad_norm": 0.609375,
|
| 2474 |
+
"learning_rate": 9.92662560661673e-05,
|
| 2475 |
+
"loss": 0.9207,
|
| 2476 |
+
"step": 350
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.20239296525875738,
|
| 2480 |
+
"grad_norm": 0.63671875,
|
| 2481 |
+
"learning_rate": 9.92619361097461e-05,
|
| 2482 |
+
"loss": 0.9634,
|
| 2483 |
+
"step": 351
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.2029695833933977,
|
| 2487 |
+
"grad_norm": 0.56640625,
|
| 2488 |
+
"learning_rate": 9.925760360914523e-05,
|
| 2489 |
+
"loss": 0.919,
|
| 2490 |
+
"step": 352
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.20354620152803807,
|
| 2494 |
+
"grad_norm": 0.57421875,
|
| 2495 |
+
"learning_rate": 9.925325856595091e-05,
|
| 2496 |
+
"loss": 0.8943,
|
| 2497 |
+
"step": 353
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.2041228196626784,
|
| 2501 |
+
"grad_norm": 0.6171875,
|
| 2502 |
+
"learning_rate": 9.924890098175403e-05,
|
| 2503 |
+
"loss": 0.8739,
|
| 2504 |
+
"step": 354
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.20469943779731872,
|
| 2508 |
+
"grad_norm": 0.58984375,
|
| 2509 |
+
"learning_rate": 9.924453085814997e-05,
|
| 2510 |
+
"loss": 0.946,
|
| 2511 |
+
"step": 355
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.20527605593195905,
|
| 2515 |
+
"grad_norm": 0.59765625,
|
| 2516 |
+
"learning_rate": 9.924014819673877e-05,
|
| 2517 |
+
"loss": 0.8982,
|
| 2518 |
+
"step": 356
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.2058526740665994,
|
| 2522 |
+
"grad_norm": 0.64453125,
|
| 2523 |
+
"learning_rate": 9.923575299912504e-05,
|
| 2524 |
+
"loss": 0.9856,
|
| 2525 |
+
"step": 357
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.20642929220123973,
|
| 2529 |
+
"grad_norm": 0.62109375,
|
| 2530 |
+
"learning_rate": 9.923134526691795e-05,
|
| 2531 |
+
"loss": 0.8953,
|
| 2532 |
+
"step": 358
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.20700591033588006,
|
| 2536 |
+
"grad_norm": 0.60546875,
|
| 2537 |
+
"learning_rate": 9.92269250017313e-05,
|
| 2538 |
+
"loss": 0.9021,
|
| 2539 |
+
"step": 359
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.2075825284705204,
|
| 2543 |
+
"grad_norm": 0.5859375,
|
| 2544 |
+
"learning_rate": 9.922249220518345e-05,
|
| 2545 |
+
"loss": 0.9057,
|
| 2546 |
+
"step": 360
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.20815914660516074,
|
| 2550 |
+
"grad_norm": 0.59375,
|
| 2551 |
+
"learning_rate": 9.921804687889738e-05,
|
| 2552 |
+
"loss": 0.9107,
|
| 2553 |
+
"step": 361
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.20873576473980107,
|
| 2557 |
+
"grad_norm": 0.59765625,
|
| 2558 |
+
"learning_rate": 9.921358902450062e-05,
|
| 2559 |
+
"loss": 0.9374,
|
| 2560 |
+
"step": 362
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.2093123828744414,
|
| 2564 |
+
"grad_norm": 0.60546875,
|
| 2565 |
+
"learning_rate": 9.920911864362534e-05,
|
| 2566 |
+
"loss": 0.9108,
|
| 2567 |
+
"step": 363
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.20988900100908173,
|
| 2571 |
+
"grad_norm": 0.5859375,
|
| 2572 |
+
"learning_rate": 9.920463573790822e-05,
|
| 2573 |
+
"loss": 0.8881,
|
| 2574 |
+
"step": 364
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.21046561914372208,
|
| 2578 |
+
"grad_norm": 0.5859375,
|
| 2579 |
+
"learning_rate": 9.920014030899058e-05,
|
| 2580 |
+
"loss": 0.9157,
|
| 2581 |
+
"step": 365
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.2110422372783624,
|
| 2585 |
+
"grad_norm": 0.5859375,
|
| 2586 |
+
"learning_rate": 9.919563235851835e-05,
|
| 2587 |
+
"loss": 0.8968,
|
| 2588 |
+
"step": 366
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.21161885541300274,
|
| 2592 |
+
"grad_norm": 0.63671875,
|
| 2593 |
+
"learning_rate": 9.919111188814192e-05,
|
| 2594 |
+
"loss": 0.9315,
|
| 2595 |
+
"step": 367
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.21219547354764307,
|
| 2599 |
+
"grad_norm": 0.59765625,
|
| 2600 |
+
"learning_rate": 9.918657889951645e-05,
|
| 2601 |
+
"loss": 0.9069,
|
| 2602 |
+
"step": 368
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.21277209168228342,
|
| 2606 |
+
"grad_norm": 0.578125,
|
| 2607 |
+
"learning_rate": 9.918203339430152e-05,
|
| 2608 |
+
"loss": 0.8905,
|
| 2609 |
+
"step": 369
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.21334870981692375,
|
| 2613 |
+
"grad_norm": 0.60546875,
|
| 2614 |
+
"learning_rate": 9.917747537416138e-05,
|
| 2615 |
+
"loss": 0.914,
|
| 2616 |
+
"step": 370
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.21392532795156408,
|
| 2620 |
+
"grad_norm": 0.578125,
|
| 2621 |
+
"learning_rate": 9.917290484076485e-05,
|
| 2622 |
+
"loss": 0.8891,
|
| 2623 |
+
"step": 371
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.2145019460862044,
|
| 2627 |
+
"grad_norm": 0.58203125,
|
| 2628 |
+
"learning_rate": 9.91683217957853e-05,
|
| 2629 |
+
"loss": 0.898,
|
| 2630 |
+
"step": 372
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.21507856422084476,
|
| 2634 |
+
"grad_norm": 0.58203125,
|
| 2635 |
+
"learning_rate": 9.916372624090071e-05,
|
| 2636 |
+
"loss": 0.9283,
|
| 2637 |
+
"step": 373
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.2156551823554851,
|
| 2641 |
+
"grad_norm": 0.58203125,
|
| 2642 |
+
"learning_rate": 9.915911817779362e-05,
|
| 2643 |
+
"loss": 0.932,
|
| 2644 |
+
"step": 374
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.21623180049012541,
|
| 2648 |
+
"grad_norm": 0.58984375,
|
| 2649 |
+
"learning_rate": 9.915449760815117e-05,
|
| 2650 |
+
"loss": 0.9047,
|
| 2651 |
+
"step": 375
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 0.21623180049012541,
|
| 2655 |
+
"eval_loss": 0.9402498006820679,
|
| 2656 |
+
"eval_runtime": 1.4105,
|
| 2657 |
+
"eval_samples_per_second": 131.872,
|
| 2658 |
+
"eval_steps_per_second": 2.836,
|
| 2659 |
+
"step": 375
|
| 2660 |
}
|
| 2661 |
],
|
| 2662 |
"logging_steps": 1,
|
|
|
|
| 2676 |
"attributes": {}
|
| 2677 |
}
|
| 2678 |
},
|
| 2679 |
+
"total_flos": 5.717261656522752e+17,
|
| 2680 |
"train_batch_size": 60,
|
| 2681 |
"trial_name": null,
|
| 2682 |
"trial_params": null
|