Upload folder using huggingface_hub
Browse files- scheduler.pt +1 -1
- trainer_state.json +2103 -3
- training_args.bin +1 -1
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30f7e3fbfbc8f9521fe18e5679b1ea21c30fcfc40bbbf22572cdfa0818f51f9c
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1407,6 +1407,2106 @@
|
|
| 1407 |
"learning_rate": 1.9600000000000002e-05,
|
| 1408 |
"loss": 5.6614,
|
| 1409 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1410 |
}
|
| 1411 |
],
|
| 1412 |
"logging_steps": 10,
|
|
@@ -1414,7 +3514,7 @@
|
|
| 1414 |
"num_input_tokens_seen": 0,
|
| 1415 |
"num_train_epochs": 1,
|
| 1416 |
"save_steps": 1000,
|
| 1417 |
-
"total_flos":
|
| 1418 |
"train_batch_size": 4,
|
| 1419 |
"trial_name": null,
|
| 1420 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.0015618557345095149,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1407 |
"learning_rate": 1.9600000000000002e-05,
|
| 1408 |
"loss": 5.6614,
|
| 1409 |
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.000627866005272825,
|
| 1413 |
+
"grad_norm": 17.25,
|
| 1414 |
+
"learning_rate": 1.9598e-05,
|
| 1415 |
+
"loss": 5.6901,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.000630989716741844,
|
| 1420 |
+
"grad_norm": 17.375,
|
| 1421 |
+
"learning_rate": 1.9596e-05,
|
| 1422 |
+
"loss": 5.7166,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.000634113428210863,
|
| 1427 |
+
"grad_norm": 19.25,
|
| 1428 |
+
"learning_rate": 1.9594000000000002e-05,
|
| 1429 |
+
"loss": 5.7282,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.000637237139679882,
|
| 1434 |
+
"grad_norm": 18.5,
|
| 1435 |
+
"learning_rate": 1.9592e-05,
|
| 1436 |
+
"loss": 5.6836,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.0006403608511489011,
|
| 1441 |
+
"grad_norm": 19.625,
|
| 1442 |
+
"learning_rate": 1.9590000000000004e-05,
|
| 1443 |
+
"loss": 5.6814,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.0006434845626179201,
|
| 1448 |
+
"grad_norm": 17.625,
|
| 1449 |
+
"learning_rate": 1.9588000000000003e-05,
|
| 1450 |
+
"loss": 5.4854,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.0006466082740869391,
|
| 1455 |
+
"grad_norm": 13.8125,
|
| 1456 |
+
"learning_rate": 1.9586000000000002e-05,
|
| 1457 |
+
"loss": 5.7422,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.0006497319855559582,
|
| 1462 |
+
"grad_norm": 19.125,
|
| 1463 |
+
"learning_rate": 1.9584e-05,
|
| 1464 |
+
"loss": 5.2697,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.0006528556970249772,
|
| 1469 |
+
"grad_norm": 17.125,
|
| 1470 |
+
"learning_rate": 1.9582e-05,
|
| 1471 |
+
"loss": 5.4215,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.0006559794084939962,
|
| 1476 |
+
"grad_norm": 17.375,
|
| 1477 |
+
"learning_rate": 1.9580000000000002e-05,
|
| 1478 |
+
"loss": 5.4331,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.0006591031199630153,
|
| 1483 |
+
"grad_norm": 15.5625,
|
| 1484 |
+
"learning_rate": 1.9578e-05,
|
| 1485 |
+
"loss": 5.6154,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.0006622268314320342,
|
| 1490 |
+
"grad_norm": 22.625,
|
| 1491 |
+
"learning_rate": 1.9576e-05,
|
| 1492 |
+
"loss": 5.1004,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.0006653505429010533,
|
| 1497 |
+
"grad_norm": 16.75,
|
| 1498 |
+
"learning_rate": 1.9574000000000003e-05,
|
| 1499 |
+
"loss": 5.5134,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.0006684742543700724,
|
| 1504 |
+
"grad_norm": 14.875,
|
| 1505 |
+
"learning_rate": 1.9572e-05,
|
| 1506 |
+
"loss": 5.5428,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.0006715979658390913,
|
| 1511 |
+
"grad_norm": 23.0,
|
| 1512 |
+
"learning_rate": 1.957e-05,
|
| 1513 |
+
"loss": 5.4856,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.0006747216773081104,
|
| 1518 |
+
"grad_norm": 14.5,
|
| 1519 |
+
"learning_rate": 1.9568000000000003e-05,
|
| 1520 |
+
"loss": 5.7353,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.0006778453887771295,
|
| 1525 |
+
"grad_norm": 13.875,
|
| 1526 |
+
"learning_rate": 1.9566000000000002e-05,
|
| 1527 |
+
"loss": 5.4402,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.0006809691002461484,
|
| 1532 |
+
"grad_norm": 21.75,
|
| 1533 |
+
"learning_rate": 1.9564e-05,
|
| 1534 |
+
"loss": 5.842,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.0006840928117151675,
|
| 1539 |
+
"grad_norm": 15.9375,
|
| 1540 |
+
"learning_rate": 1.9562e-05,
|
| 1541 |
+
"loss": 5.4636,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.0006872165231841866,
|
| 1546 |
+
"grad_norm": 16.125,
|
| 1547 |
+
"learning_rate": 1.9560000000000002e-05,
|
| 1548 |
+
"loss": 5.3003,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.0006903402346532055,
|
| 1553 |
+
"grad_norm": 18.75,
|
| 1554 |
+
"learning_rate": 1.9558e-05,
|
| 1555 |
+
"loss": 5.6459,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.0006934639461222246,
|
| 1560 |
+
"grad_norm": 11.875,
|
| 1561 |
+
"learning_rate": 1.9556e-05,
|
| 1562 |
+
"loss": 5.5871,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.0006965876575912436,
|
| 1567 |
+
"grad_norm": 20.375,
|
| 1568 |
+
"learning_rate": 1.9554000000000003e-05,
|
| 1569 |
+
"loss": 5.5089,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.0006997113690602626,
|
| 1574 |
+
"grad_norm": 16.375,
|
| 1575 |
+
"learning_rate": 1.9552000000000002e-05,
|
| 1576 |
+
"loss": 5.7504,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.0007028350805292817,
|
| 1581 |
+
"grad_norm": 28.375,
|
| 1582 |
+
"learning_rate": 1.955e-05,
|
| 1583 |
+
"loss": 5.5179,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.0007059587919983007,
|
| 1588 |
+
"grad_norm": 13.6875,
|
| 1589 |
+
"learning_rate": 1.9548000000000003e-05,
|
| 1590 |
+
"loss": 5.4784,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.0007090825034673197,
|
| 1595 |
+
"grad_norm": 14.375,
|
| 1596 |
+
"learning_rate": 1.9546e-05,
|
| 1597 |
+
"loss": 5.6661,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.0007122062149363388,
|
| 1602 |
+
"grad_norm": 13.75,
|
| 1603 |
+
"learning_rate": 1.9544e-05,
|
| 1604 |
+
"loss": 5.4985,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.0007153299264053578,
|
| 1609 |
+
"grad_norm": 16.0,
|
| 1610 |
+
"learning_rate": 1.9542e-05,
|
| 1611 |
+
"loss": 5.8197,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.0007184536378743768,
|
| 1616 |
+
"grad_norm": 19.125,
|
| 1617 |
+
"learning_rate": 1.9540000000000003e-05,
|
| 1618 |
+
"loss": 5.3348,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.0007215773493433958,
|
| 1623 |
+
"grad_norm": 13.5,
|
| 1624 |
+
"learning_rate": 1.9538e-05,
|
| 1625 |
+
"loss": 5.6816,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.0007247010608124149,
|
| 1630 |
+
"grad_norm": 15.4375,
|
| 1631 |
+
"learning_rate": 1.9536e-05,
|
| 1632 |
+
"loss": 5.7536,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.0007278247722814339,
|
| 1637 |
+
"grad_norm": 19.625,
|
| 1638 |
+
"learning_rate": 1.9534000000000003e-05,
|
| 1639 |
+
"loss": 5.7437,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.0007309484837504529,
|
| 1644 |
+
"grad_norm": 16.875,
|
| 1645 |
+
"learning_rate": 1.9532000000000002e-05,
|
| 1646 |
+
"loss": 5.5242,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.000734072195219472,
|
| 1651 |
+
"grad_norm": 15.6875,
|
| 1652 |
+
"learning_rate": 1.953e-05,
|
| 1653 |
+
"loss": 5.7513,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.000737195906688491,
|
| 1658 |
+
"grad_norm": 12.5625,
|
| 1659 |
+
"learning_rate": 1.9528000000000003e-05,
|
| 1660 |
+
"loss": 5.7211,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.00074031961815751,
|
| 1665 |
+
"grad_norm": 16.875,
|
| 1666 |
+
"learning_rate": 1.9526e-05,
|
| 1667 |
+
"loss": 5.7414,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.0007434433296265291,
|
| 1672 |
+
"grad_norm": 23.25,
|
| 1673 |
+
"learning_rate": 1.9524e-05,
|
| 1674 |
+
"loss": 5.4911,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.000746567041095548,
|
| 1679 |
+
"grad_norm": 14.5625,
|
| 1680 |
+
"learning_rate": 1.9522e-05,
|
| 1681 |
+
"loss": 5.7355,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.0007496907525645671,
|
| 1686 |
+
"grad_norm": 20.75,
|
| 1687 |
+
"learning_rate": 1.9520000000000003e-05,
|
| 1688 |
+
"loss": 5.5423,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.0007528144640335862,
|
| 1693 |
+
"grad_norm": 16.125,
|
| 1694 |
+
"learning_rate": 1.9518000000000002e-05,
|
| 1695 |
+
"loss": 5.4726,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.0007559381755026051,
|
| 1700 |
+
"grad_norm": 19.0,
|
| 1701 |
+
"learning_rate": 1.9516e-05,
|
| 1702 |
+
"loss": 5.7674,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.0007590618869716242,
|
| 1707 |
+
"grad_norm": 13.25,
|
| 1708 |
+
"learning_rate": 1.9514000000000003e-05,
|
| 1709 |
+
"loss": 5.65,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.0007621855984406433,
|
| 1714 |
+
"grad_norm": 14.0,
|
| 1715 |
+
"learning_rate": 1.9512000000000002e-05,
|
| 1716 |
+
"loss": 5.777,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.0007653093099096622,
|
| 1721 |
+
"grad_norm": 17.125,
|
| 1722 |
+
"learning_rate": 1.951e-05,
|
| 1723 |
+
"loss": 5.2448,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.0007684330213786813,
|
| 1728 |
+
"grad_norm": 11.1875,
|
| 1729 |
+
"learning_rate": 1.9508000000000004e-05,
|
| 1730 |
+
"loss": 5.6465,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.0007715567328477003,
|
| 1735 |
+
"grad_norm": 15.3125,
|
| 1736 |
+
"learning_rate": 1.9506e-05,
|
| 1737 |
+
"loss": 5.6983,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.0007746804443167194,
|
| 1742 |
+
"grad_norm": 12.1875,
|
| 1743 |
+
"learning_rate": 1.9504e-05,
|
| 1744 |
+
"loss": 5.2056,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.0007778041557857384,
|
| 1749 |
+
"grad_norm": 22.125,
|
| 1750 |
+
"learning_rate": 1.9502e-05,
|
| 1751 |
+
"loss": 5.675,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.0007809278672547574,
|
| 1756 |
+
"grad_norm": 20.25,
|
| 1757 |
+
"learning_rate": 1.95e-05,
|
| 1758 |
+
"loss": 5.6855,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.0007840515787237765,
|
| 1763 |
+
"grad_norm": 12.25,
|
| 1764 |
+
"learning_rate": 1.9498000000000002e-05,
|
| 1765 |
+
"loss": 5.7423,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.0007871752901927955,
|
| 1770 |
+
"grad_norm": 15.5,
|
| 1771 |
+
"learning_rate": 1.9496e-05,
|
| 1772 |
+
"loss": 5.4792,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.0007902990016618145,
|
| 1777 |
+
"grad_norm": 17.375,
|
| 1778 |
+
"learning_rate": 1.9494000000000003e-05,
|
| 1779 |
+
"loss": 5.562,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.0007934227131308336,
|
| 1784 |
+
"grad_norm": 14.1875,
|
| 1785 |
+
"learning_rate": 1.9492000000000002e-05,
|
| 1786 |
+
"loss": 5.6019,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.0007965464245998525,
|
| 1791 |
+
"grad_norm": 17.625,
|
| 1792 |
+
"learning_rate": 1.949e-05,
|
| 1793 |
+
"loss": 5.6609,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.0007996701360688716,
|
| 1798 |
+
"grad_norm": 18.75,
|
| 1799 |
+
"learning_rate": 1.9488000000000004e-05,
|
| 1800 |
+
"loss": 5.374,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.0008027938475378907,
|
| 1805 |
+
"grad_norm": 22.25,
|
| 1806 |
+
"learning_rate": 1.9486000000000003e-05,
|
| 1807 |
+
"loss": 5.4583,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.0008059175590069096,
|
| 1812 |
+
"grad_norm": 13.8125,
|
| 1813 |
+
"learning_rate": 1.9484000000000002e-05,
|
| 1814 |
+
"loss": 5.2723,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.0008090412704759287,
|
| 1819 |
+
"grad_norm": 12.5625,
|
| 1820 |
+
"learning_rate": 1.9482e-05,
|
| 1821 |
+
"loss": 5.328,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.0008121649819449478,
|
| 1826 |
+
"grad_norm": 12.375,
|
| 1827 |
+
"learning_rate": 1.948e-05,
|
| 1828 |
+
"loss": 5.4062,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.0008152886934139667,
|
| 1833 |
+
"grad_norm": 13.125,
|
| 1834 |
+
"learning_rate": 1.9478000000000002e-05,
|
| 1835 |
+
"loss": 5.7013,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.0008184124048829858,
|
| 1840 |
+
"grad_norm": 17.625,
|
| 1841 |
+
"learning_rate": 1.9476e-05,
|
| 1842 |
+
"loss": 5.5071,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.0008215361163520048,
|
| 1847 |
+
"grad_norm": 14.4375,
|
| 1848 |
+
"learning_rate": 1.9474e-05,
|
| 1849 |
+
"loss": 5.6346,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.0008246598278210238,
|
| 1854 |
+
"grad_norm": 17.0,
|
| 1855 |
+
"learning_rate": 1.9472000000000003e-05,
|
| 1856 |
+
"loss": 5.1499,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.0008277835392900429,
|
| 1861 |
+
"grad_norm": 13.5,
|
| 1862 |
+
"learning_rate": 1.947e-05,
|
| 1863 |
+
"loss": 5.547,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.0008309072507590619,
|
| 1868 |
+
"grad_norm": 16.0,
|
| 1869 |
+
"learning_rate": 1.9468000000000004e-05,
|
| 1870 |
+
"loss": 5.7553,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.0008340309622280809,
|
| 1875 |
+
"grad_norm": 19.875,
|
| 1876 |
+
"learning_rate": 1.9466000000000003e-05,
|
| 1877 |
+
"loss": 5.4614,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.0008371546736971,
|
| 1882 |
+
"grad_norm": 14.5625,
|
| 1883 |
+
"learning_rate": 1.9464000000000002e-05,
|
| 1884 |
+
"loss": 5.4983,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.000840278385166119,
|
| 1889 |
+
"grad_norm": 17.75,
|
| 1890 |
+
"learning_rate": 1.9462e-05,
|
| 1891 |
+
"loss": 5.1406,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.000843402096635138,
|
| 1896 |
+
"grad_norm": 16.625,
|
| 1897 |
+
"learning_rate": 1.946e-05,
|
| 1898 |
+
"loss": 5.4609,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.000846525808104157,
|
| 1903 |
+
"grad_norm": 21.5,
|
| 1904 |
+
"learning_rate": 1.9458000000000002e-05,
|
| 1905 |
+
"loss": 5.7063,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.0008496495195731761,
|
| 1910 |
+
"grad_norm": 15.625,
|
| 1911 |
+
"learning_rate": 1.9456e-05,
|
| 1912 |
+
"loss": 5.6036,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.0008527732310421951,
|
| 1917 |
+
"grad_norm": 13.9375,
|
| 1918 |
+
"learning_rate": 1.9454e-05,
|
| 1919 |
+
"loss": 5.5167,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.0008558969425112141,
|
| 1924 |
+
"grad_norm": 10.5,
|
| 1925 |
+
"learning_rate": 1.9452000000000003e-05,
|
| 1926 |
+
"loss": 5.3084,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.0008590206539802332,
|
| 1931 |
+
"grad_norm": 12.5625,
|
| 1932 |
+
"learning_rate": 1.9450000000000002e-05,
|
| 1933 |
+
"loss": 5.2286,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.0008621443654492521,
|
| 1938 |
+
"grad_norm": 17.125,
|
| 1939 |
+
"learning_rate": 1.9448e-05,
|
| 1940 |
+
"loss": 5.4303,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.0008652680769182712,
|
| 1945 |
+
"grad_norm": 14.75,
|
| 1946 |
+
"learning_rate": 1.9446000000000003e-05,
|
| 1947 |
+
"loss": 5.476,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.0008683917883872903,
|
| 1952 |
+
"grad_norm": 16.75,
|
| 1953 |
+
"learning_rate": 1.9444000000000002e-05,
|
| 1954 |
+
"loss": 5.7369,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.0008715154998563092,
|
| 1959 |
+
"grad_norm": 12.3125,
|
| 1960 |
+
"learning_rate": 1.9442e-05,
|
| 1961 |
+
"loss": 5.7263,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.0008746392113253283,
|
| 1966 |
+
"grad_norm": 20.25,
|
| 1967 |
+
"learning_rate": 1.944e-05,
|
| 1968 |
+
"loss": 5.1745,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.0008777629227943474,
|
| 1973 |
+
"grad_norm": 11.25,
|
| 1974 |
+
"learning_rate": 1.9438000000000003e-05,
|
| 1975 |
+
"loss": 5.5658,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.0008808866342633663,
|
| 1980 |
+
"grad_norm": 16.625,
|
| 1981 |
+
"learning_rate": 1.9436e-05,
|
| 1982 |
+
"loss": 5.4065,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.0008840103457323854,
|
| 1987 |
+
"grad_norm": 15.5,
|
| 1988 |
+
"learning_rate": 1.9434e-05,
|
| 1989 |
+
"loss": 5.6371,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.0008871340572014045,
|
| 1994 |
+
"grad_norm": 13.3125,
|
| 1995 |
+
"learning_rate": 1.9432000000000003e-05,
|
| 1996 |
+
"loss": 5.771,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.0008902577686704234,
|
| 2001 |
+
"grad_norm": 16.375,
|
| 2002 |
+
"learning_rate": 1.9430000000000002e-05,
|
| 2003 |
+
"loss": 5.5594,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.0008933814801394425,
|
| 2008 |
+
"grad_norm": 14.5625,
|
| 2009 |
+
"learning_rate": 1.9428e-05,
|
| 2010 |
+
"loss": 5.5727,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.0008965051916084615,
|
| 2015 |
+
"grad_norm": 21.625,
|
| 2016 |
+
"learning_rate": 1.9426000000000003e-05,
|
| 2017 |
+
"loss": 5.6611,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.0008996289030774805,
|
| 2022 |
+
"grad_norm": 21.375,
|
| 2023 |
+
"learning_rate": 1.9424e-05,
|
| 2024 |
+
"loss": 5.5716,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.0009027526145464996,
|
| 2029 |
+
"grad_norm": 21.375,
|
| 2030 |
+
"learning_rate": 1.9422e-05,
|
| 2031 |
+
"loss": 5.4191,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.0009058763260155186,
|
| 2036 |
+
"grad_norm": 14.375,
|
| 2037 |
+
"learning_rate": 1.942e-05,
|
| 2038 |
+
"loss": 5.5003,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.0009090000374845376,
|
| 2043 |
+
"grad_norm": 18.375,
|
| 2044 |
+
"learning_rate": 1.9418000000000003e-05,
|
| 2045 |
+
"loss": 5.6882,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.0009121237489535567,
|
| 2050 |
+
"grad_norm": 16.125,
|
| 2051 |
+
"learning_rate": 1.9416000000000002e-05,
|
| 2052 |
+
"loss": 5.7119,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.0009152474604225757,
|
| 2057 |
+
"grad_norm": 15.75,
|
| 2058 |
+
"learning_rate": 1.9414e-05,
|
| 2059 |
+
"loss": 5.6209,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.0009183711718915947,
|
| 2064 |
+
"grad_norm": 15.9375,
|
| 2065 |
+
"learning_rate": 1.9412000000000003e-05,
|
| 2066 |
+
"loss": 5.7167,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.0009214948833606137,
|
| 2071 |
+
"grad_norm": 11.5625,
|
| 2072 |
+
"learning_rate": 1.9410000000000002e-05,
|
| 2073 |
+
"loss": 5.6873,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.0009246185948296328,
|
| 2078 |
+
"grad_norm": 14.4375,
|
| 2079 |
+
"learning_rate": 1.9408e-05,
|
| 2080 |
+
"loss": 5.66,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.0009277423062986518,
|
| 2085 |
+
"grad_norm": 16.375,
|
| 2086 |
+
"learning_rate": 1.9406000000000004e-05,
|
| 2087 |
+
"loss": 5.6787,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.0009308660177676708,
|
| 2092 |
+
"grad_norm": 16.0,
|
| 2093 |
+
"learning_rate": 1.9404e-05,
|
| 2094 |
+
"loss": 5.014,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.0009339897292366899,
|
| 2099 |
+
"grad_norm": 15.6875,
|
| 2100 |
+
"learning_rate": 1.9402e-05,
|
| 2101 |
+
"loss": 5.6579,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.0009371134407057088,
|
| 2106 |
+
"grad_norm": 18.125,
|
| 2107 |
+
"learning_rate": 1.94e-05,
|
| 2108 |
+
"loss": 5.4841,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.0009402371521747279,
|
| 2113 |
+
"grad_norm": 20.125,
|
| 2114 |
+
"learning_rate": 1.9398e-05,
|
| 2115 |
+
"loss": 5.7642,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.000943360863643747,
|
| 2120 |
+
"grad_norm": 14.3125,
|
| 2121 |
+
"learning_rate": 1.9396000000000002e-05,
|
| 2122 |
+
"loss": 5.7511,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.0009464845751127659,
|
| 2127 |
+
"grad_norm": 19.75,
|
| 2128 |
+
"learning_rate": 1.9394e-05,
|
| 2129 |
+
"loss": 5.6252,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.000949608286581785,
|
| 2134 |
+
"grad_norm": 14.8125,
|
| 2135 |
+
"learning_rate": 1.9392000000000003e-05,
|
| 2136 |
+
"loss": 5.6126,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.0009527319980508041,
|
| 2141 |
+
"grad_norm": 15.75,
|
| 2142 |
+
"learning_rate": 1.9390000000000002e-05,
|
| 2143 |
+
"loss": 5.2285,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.000955855709519823,
|
| 2148 |
+
"grad_norm": 21.0,
|
| 2149 |
+
"learning_rate": 1.9388e-05,
|
| 2150 |
+
"loss": 5.712,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.0009589794209888421,
|
| 2155 |
+
"grad_norm": 12.1875,
|
| 2156 |
+
"learning_rate": 1.9386000000000004e-05,
|
| 2157 |
+
"loss": 5.4932,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.0009621031324578612,
|
| 2162 |
+
"grad_norm": 16.25,
|
| 2163 |
+
"learning_rate": 1.9384e-05,
|
| 2164 |
+
"loss": 5.6017,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.0009652268439268801,
|
| 2169 |
+
"grad_norm": 14.3125,
|
| 2170 |
+
"learning_rate": 1.9382000000000002e-05,
|
| 2171 |
+
"loss": 5.6923,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.0009683505553958992,
|
| 2176 |
+
"grad_norm": 22.0,
|
| 2177 |
+
"learning_rate": 1.938e-05,
|
| 2178 |
+
"loss": 5.6061,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.0009714742668649182,
|
| 2183 |
+
"grad_norm": 12.375,
|
| 2184 |
+
"learning_rate": 1.9378e-05,
|
| 2185 |
+
"loss": 5.1761,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.0009745979783339372,
|
| 2190 |
+
"grad_norm": 10.875,
|
| 2191 |
+
"learning_rate": 1.9376000000000002e-05,
|
| 2192 |
+
"loss": 5.6632,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.0009777216898029563,
|
| 2197 |
+
"grad_norm": 18.875,
|
| 2198 |
+
"learning_rate": 1.9374e-05,
|
| 2199 |
+
"loss": 5.7265,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.0009808454012719752,
|
| 2204 |
+
"grad_norm": 13.0,
|
| 2205 |
+
"learning_rate": 1.9372000000000004e-05,
|
| 2206 |
+
"loss": 5.726,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.0009839691127409944,
|
| 2211 |
+
"grad_norm": 17.0,
|
| 2212 |
+
"learning_rate": 1.9370000000000003e-05,
|
| 2213 |
+
"loss": 5.1336,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.0009870928242100134,
|
| 2218 |
+
"grad_norm": 17.25,
|
| 2219 |
+
"learning_rate": 1.9368e-05,
|
| 2220 |
+
"loss": 5.5581,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.0009902165356790323,
|
| 2225 |
+
"grad_norm": 17.0,
|
| 2226 |
+
"learning_rate": 1.9366000000000004e-05,
|
| 2227 |
+
"loss": 5.3809,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.0009933402471480515,
|
| 2232 |
+
"grad_norm": 13.75,
|
| 2233 |
+
"learning_rate": 1.9364e-05,
|
| 2234 |
+
"loss": 5.6005,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.0009964639586170704,
|
| 2239 |
+
"grad_norm": 16.75,
|
| 2240 |
+
"learning_rate": 1.9362000000000002e-05,
|
| 2241 |
+
"loss": 5.304,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.0009995876700860894,
|
| 2246 |
+
"grad_norm": 13.125,
|
| 2247 |
+
"learning_rate": 1.936e-05,
|
| 2248 |
+
"loss": 5.4925,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.0010027113815551086,
|
| 2253 |
+
"grad_norm": 12.1875,
|
| 2254 |
+
"learning_rate": 1.9358e-05,
|
| 2255 |
+
"loss": 5.6312,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.0010058350930241275,
|
| 2260 |
+
"grad_norm": 15.625,
|
| 2261 |
+
"learning_rate": 1.9356000000000002e-05,
|
| 2262 |
+
"loss": 5.5569,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.0010089588044931465,
|
| 2267 |
+
"grad_norm": 16.5,
|
| 2268 |
+
"learning_rate": 1.9354e-05,
|
| 2269 |
+
"loss": 5.7318,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.0010120825159621657,
|
| 2274 |
+
"grad_norm": 20.75,
|
| 2275 |
+
"learning_rate": 1.9352e-05,
|
| 2276 |
+
"loss": 5.3372,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.0010152062274311846,
|
| 2281 |
+
"grad_norm": 15.8125,
|
| 2282 |
+
"learning_rate": 1.9350000000000003e-05,
|
| 2283 |
+
"loss": 5.3677,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.0010183299389002036,
|
| 2288 |
+
"grad_norm": 12.75,
|
| 2289 |
+
"learning_rate": 1.9348000000000002e-05,
|
| 2290 |
+
"loss": 5.6175,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.0010214536503692227,
|
| 2295 |
+
"grad_norm": 14.25,
|
| 2296 |
+
"learning_rate": 1.9346000000000004e-05,
|
| 2297 |
+
"loss": 5.3534,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.0010245773618382417,
|
| 2302 |
+
"grad_norm": 12.4375,
|
| 2303 |
+
"learning_rate": 1.9344e-05,
|
| 2304 |
+
"loss": 5.1925,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.0010277010733072607,
|
| 2309 |
+
"grad_norm": 13.5625,
|
| 2310 |
+
"learning_rate": 1.9342000000000002e-05,
|
| 2311 |
+
"loss": 5.8256,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.0010308247847762798,
|
| 2316 |
+
"grad_norm": 14.5625,
|
| 2317 |
+
"learning_rate": 1.934e-05,
|
| 2318 |
+
"loss": 5.7592,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.0010339484962452988,
|
| 2323 |
+
"grad_norm": 11.75,
|
| 2324 |
+
"learning_rate": 1.9338e-05,
|
| 2325 |
+
"loss": 5.6782,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.0010370722077143177,
|
| 2330 |
+
"grad_norm": 23.25,
|
| 2331 |
+
"learning_rate": 1.9336000000000003e-05,
|
| 2332 |
+
"loss": 5.4134,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.001040195919183337,
|
| 2337 |
+
"grad_norm": 14.625,
|
| 2338 |
+
"learning_rate": 1.9334e-05,
|
| 2339 |
+
"loss": 5.6013,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.0010433196306523559,
|
| 2344 |
+
"grad_norm": 17.625,
|
| 2345 |
+
"learning_rate": 1.9332e-05,
|
| 2346 |
+
"loss": 5.4116,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.0010464433421213748,
|
| 2351 |
+
"grad_norm": 10.5625,
|
| 2352 |
+
"learning_rate": 1.9330000000000003e-05,
|
| 2353 |
+
"loss": 5.5778,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.001049567053590394,
|
| 2358 |
+
"grad_norm": 16.75,
|
| 2359 |
+
"learning_rate": 1.9328000000000002e-05,
|
| 2360 |
+
"loss": 5.7124,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.001052690765059413,
|
| 2365 |
+
"grad_norm": 15.8125,
|
| 2366 |
+
"learning_rate": 1.9326e-05,
|
| 2367 |
+
"loss": 5.5873,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.001055814476528432,
|
| 2372 |
+
"grad_norm": 15.4375,
|
| 2373 |
+
"learning_rate": 1.9324e-05,
|
| 2374 |
+
"loss": 5.706,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.001058938187997451,
|
| 2379 |
+
"grad_norm": 14.375,
|
| 2380 |
+
"learning_rate": 1.9322000000000002e-05,
|
| 2381 |
+
"loss": 5.4256,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.00106206189946647,
|
| 2386 |
+
"grad_norm": 19.0,
|
| 2387 |
+
"learning_rate": 1.932e-05,
|
| 2388 |
+
"loss": 5.3917,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.001065185610935489,
|
| 2393 |
+
"grad_norm": 13.1875,
|
| 2394 |
+
"learning_rate": 1.9318e-05,
|
| 2395 |
+
"loss": 5.6195,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.0010683093224045082,
|
| 2400 |
+
"grad_norm": 14.8125,
|
| 2401 |
+
"learning_rate": 1.9316000000000003e-05,
|
| 2402 |
+
"loss": 5.3445,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.0010714330338735271,
|
| 2407 |
+
"grad_norm": 11.75,
|
| 2408 |
+
"learning_rate": 1.9314000000000002e-05,
|
| 2409 |
+
"loss": 5.155,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.001074556745342546,
|
| 2414 |
+
"grad_norm": 13.0625,
|
| 2415 |
+
"learning_rate": 1.9312e-05,
|
| 2416 |
+
"loss": 5.3025,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.0010776804568115653,
|
| 2421 |
+
"grad_norm": 14.5,
|
| 2422 |
+
"learning_rate": 1.9310000000000003e-05,
|
| 2423 |
+
"loss": 5.4074,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.0010808041682805842,
|
| 2428 |
+
"grad_norm": 16.375,
|
| 2429 |
+
"learning_rate": 1.9308000000000002e-05,
|
| 2430 |
+
"loss": 5.6609,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.0010839278797496032,
|
| 2435 |
+
"grad_norm": 19.25,
|
| 2436 |
+
"learning_rate": 1.9306e-05,
|
| 2437 |
+
"loss": 5.4313,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.0010870515912186224,
|
| 2442 |
+
"grad_norm": 15.6875,
|
| 2443 |
+
"learning_rate": 1.9304e-05,
|
| 2444 |
+
"loss": 5.5033,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.0010901753026876413,
|
| 2449 |
+
"grad_norm": 18.625,
|
| 2450 |
+
"learning_rate": 1.9302e-05,
|
| 2451 |
+
"loss": 5.7241,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.0010932990141566603,
|
| 2456 |
+
"grad_norm": 13.0625,
|
| 2457 |
+
"learning_rate": 1.93e-05,
|
| 2458 |
+
"loss": 5.6599,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.0010964227256256794,
|
| 2463 |
+
"grad_norm": 14.6875,
|
| 2464 |
+
"learning_rate": 1.9298e-05,
|
| 2465 |
+
"loss": 5.7546,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.0010995464370946984,
|
| 2470 |
+
"grad_norm": 13.1875,
|
| 2471 |
+
"learning_rate": 1.9296000000000003e-05,
|
| 2472 |
+
"loss": 5.6877,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.0011026701485637176,
|
| 2477 |
+
"grad_norm": 12.1875,
|
| 2478 |
+
"learning_rate": 1.9294000000000002e-05,
|
| 2479 |
+
"loss": 5.5043,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.0011057938600327365,
|
| 2484 |
+
"grad_norm": 23.25,
|
| 2485 |
+
"learning_rate": 1.9292e-05,
|
| 2486 |
+
"loss": 5.3722,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.0011089175715017555,
|
| 2491 |
+
"grad_norm": 16.625,
|
| 2492 |
+
"learning_rate": 1.9290000000000003e-05,
|
| 2493 |
+
"loss": 5.1707,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.0011120412829707747,
|
| 2498 |
+
"grad_norm": 19.25,
|
| 2499 |
+
"learning_rate": 1.9288000000000002e-05,
|
| 2500 |
+
"loss": 5.6722,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.0011151649944397936,
|
| 2505 |
+
"grad_norm": 12.75,
|
| 2506 |
+
"learning_rate": 1.9286e-05,
|
| 2507 |
+
"loss": 5.3163,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.0011182887059088126,
|
| 2512 |
+
"grad_norm": 17.625,
|
| 2513 |
+
"learning_rate": 1.9284e-05,
|
| 2514 |
+
"loss": 5.4662,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.0011214124173778317,
|
| 2519 |
+
"grad_norm": 15.75,
|
| 2520 |
+
"learning_rate": 1.9282e-05,
|
| 2521 |
+
"loss": 5.4683,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.0011245361288468507,
|
| 2526 |
+
"grad_norm": 12.9375,
|
| 2527 |
+
"learning_rate": 1.9280000000000002e-05,
|
| 2528 |
+
"loss": 5.8348,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.0011276598403158697,
|
| 2533 |
+
"grad_norm": 20.625,
|
| 2534 |
+
"learning_rate": 1.9278e-05,
|
| 2535 |
+
"loss": 5.5024,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.0011307835517848888,
|
| 2540 |
+
"grad_norm": 19.125,
|
| 2541 |
+
"learning_rate": 1.9276e-05,
|
| 2542 |
+
"loss": 5.8194,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.0011339072632539078,
|
| 2547 |
+
"grad_norm": 15.3125,
|
| 2548 |
+
"learning_rate": 1.9274000000000002e-05,
|
| 2549 |
+
"loss": 5.662,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.0011370309747229267,
|
| 2554 |
+
"grad_norm": 15.3125,
|
| 2555 |
+
"learning_rate": 1.9272e-05,
|
| 2556 |
+
"loss": 5.6406,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.001140154686191946,
|
| 2561 |
+
"grad_norm": 18.0,
|
| 2562 |
+
"learning_rate": 1.9270000000000004e-05,
|
| 2563 |
+
"loss": 5.5363,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.0011432783976609649,
|
| 2568 |
+
"grad_norm": 13.4375,
|
| 2569 |
+
"learning_rate": 1.9268000000000003e-05,
|
| 2570 |
+
"loss": 5.4715,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.0011464021091299838,
|
| 2575 |
+
"grad_norm": 16.25,
|
| 2576 |
+
"learning_rate": 1.9266e-05,
|
| 2577 |
+
"loss": 5.6616,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.001149525820599003,
|
| 2582 |
+
"grad_norm": 11.375,
|
| 2583 |
+
"learning_rate": 1.9264e-05,
|
| 2584 |
+
"loss": 5.3432,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.001152649532068022,
|
| 2589 |
+
"grad_norm": 20.0,
|
| 2590 |
+
"learning_rate": 1.9262e-05,
|
| 2591 |
+
"loss": 5.6658,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.001155773243537041,
|
| 2596 |
+
"grad_norm": 10.375,
|
| 2597 |
+
"learning_rate": 1.9260000000000002e-05,
|
| 2598 |
+
"loss": 5.6279,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.00115889695500606,
|
| 2603 |
+
"grad_norm": 14.4375,
|
| 2604 |
+
"learning_rate": 1.9258e-05,
|
| 2605 |
+
"loss": 5.7396,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.001162020666475079,
|
| 2610 |
+
"grad_norm": 19.0,
|
| 2611 |
+
"learning_rate": 1.9256e-05,
|
| 2612 |
+
"loss": 5.3711,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.001165144377944098,
|
| 2617 |
+
"grad_norm": 13.75,
|
| 2618 |
+
"learning_rate": 1.9254000000000002e-05,
|
| 2619 |
+
"loss": 5.0692,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.0011682680894131172,
|
| 2624 |
+
"grad_norm": 14.625,
|
| 2625 |
+
"learning_rate": 1.9252e-05,
|
| 2626 |
+
"loss": 5.6739,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.0011713918008821361,
|
| 2631 |
+
"grad_norm": 16.875,
|
| 2632 |
+
"learning_rate": 1.925e-05,
|
| 2633 |
+
"loss": 5.3631,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.001174515512351155,
|
| 2638 |
+
"grad_norm": 12.375,
|
| 2639 |
+
"learning_rate": 1.9248000000000003e-05,
|
| 2640 |
+
"loss": 5.3772,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.0011776392238201743,
|
| 2645 |
+
"grad_norm": 17.5,
|
| 2646 |
+
"learning_rate": 1.9246000000000002e-05,
|
| 2647 |
+
"loss": 5.5263,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.0011807629352891932,
|
| 2652 |
+
"grad_norm": 9.25,
|
| 2653 |
+
"learning_rate": 1.9244000000000004e-05,
|
| 2654 |
+
"loss": 5.0929,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.0011838866467582122,
|
| 2659 |
+
"grad_norm": 23.875,
|
| 2660 |
+
"learning_rate": 1.9242e-05,
|
| 2661 |
+
"loss": 5.7394,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.0011870103582272314,
|
| 2666 |
+
"grad_norm": 11.0,
|
| 2667 |
+
"learning_rate": 1.9240000000000002e-05,
|
| 2668 |
+
"loss": 5.7045,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.0011901340696962503,
|
| 2673 |
+
"grad_norm": 14.0,
|
| 2674 |
+
"learning_rate": 1.9238e-05,
|
| 2675 |
+
"loss": 5.6738,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.0011932577811652693,
|
| 2680 |
+
"grad_norm": 15.375,
|
| 2681 |
+
"learning_rate": 1.9236e-05,
|
| 2682 |
+
"loss": 5.3815,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.0011963814926342884,
|
| 2687 |
+
"grad_norm": 19.125,
|
| 2688 |
+
"learning_rate": 1.9234000000000003e-05,
|
| 2689 |
+
"loss": 5.0305,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.0011995052041033074,
|
| 2694 |
+
"grad_norm": 13.4375,
|
| 2695 |
+
"learning_rate": 1.9232e-05,
|
| 2696 |
+
"loss": 5.7874,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.0012026289155723264,
|
| 2701 |
+
"grad_norm": 17.125,
|
| 2702 |
+
"learning_rate": 1.923e-05,
|
| 2703 |
+
"loss": 5.4694,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.0012057526270413455,
|
| 2708 |
+
"grad_norm": 14.625,
|
| 2709 |
+
"learning_rate": 1.9228000000000003e-05,
|
| 2710 |
+
"loss": 5.5002,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.0012088763385103645,
|
| 2715 |
+
"grad_norm": 12.75,
|
| 2716 |
+
"learning_rate": 1.9226000000000002e-05,
|
| 2717 |
+
"loss": 5.7154,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.0012120000499793834,
|
| 2722 |
+
"grad_norm": 17.625,
|
| 2723 |
+
"learning_rate": 1.9224000000000004e-05,
|
| 2724 |
+
"loss": 5.4635,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.0012151237614484026,
|
| 2729 |
+
"grad_norm": 18.5,
|
| 2730 |
+
"learning_rate": 1.9222e-05,
|
| 2731 |
+
"loss": 5.7866,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.0012182474729174216,
|
| 2736 |
+
"grad_norm": 12.625,
|
| 2737 |
+
"learning_rate": 1.9220000000000002e-05,
|
| 2738 |
+
"loss": 5.7408,
|
| 2739 |
+
"step": 3900
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.0012213711843864405,
|
| 2743 |
+
"grad_norm": 17.25,
|
| 2744 |
+
"learning_rate": 1.9218e-05,
|
| 2745 |
+
"loss": 5.5827,
|
| 2746 |
+
"step": 3910
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.0012244948958554597,
|
| 2750 |
+
"grad_norm": 18.625,
|
| 2751 |
+
"learning_rate": 1.9216e-05,
|
| 2752 |
+
"loss": 5.5495,
|
| 2753 |
+
"step": 3920
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.0012276186073244787,
|
| 2757 |
+
"grad_norm": 23.0,
|
| 2758 |
+
"learning_rate": 1.9214000000000003e-05,
|
| 2759 |
+
"loss": 5.7164,
|
| 2760 |
+
"step": 3930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.0012307423187934976,
|
| 2764 |
+
"grad_norm": 16.625,
|
| 2765 |
+
"learning_rate": 1.9212000000000002e-05,
|
| 2766 |
+
"loss": 5.5238,
|
| 2767 |
+
"step": 3940
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.0012338660302625168,
|
| 2771 |
+
"grad_norm": 23.625,
|
| 2772 |
+
"learning_rate": 1.921e-05,
|
| 2773 |
+
"loss": 5.1463,
|
| 2774 |
+
"step": 3950
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.0012369897417315358,
|
| 2778 |
+
"grad_norm": 14.875,
|
| 2779 |
+
"learning_rate": 1.9208000000000003e-05,
|
| 2780 |
+
"loss": 5.4224,
|
| 2781 |
+
"step": 3960
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.0012401134532005547,
|
| 2785 |
+
"grad_norm": 12.3125,
|
| 2786 |
+
"learning_rate": 1.9206000000000002e-05,
|
| 2787 |
+
"loss": 5.6267,
|
| 2788 |
+
"step": 3970
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.0012432371646695739,
|
| 2792 |
+
"grad_norm": 12.9375,
|
| 2793 |
+
"learning_rate": 1.9204e-05,
|
| 2794 |
+
"loss": 5.3007,
|
| 2795 |
+
"step": 3980
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.0012463608761385928,
|
| 2799 |
+
"grad_norm": 26.75,
|
| 2800 |
+
"learning_rate": 1.9202e-05,
|
| 2801 |
+
"loss": 5.5985,
|
| 2802 |
+
"step": 3990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.0012494845876076118,
|
| 2806 |
+
"grad_norm": 13.25,
|
| 2807 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 2808 |
+
"loss": 5.8027,
|
| 2809 |
+
"step": 4000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.001252608299076631,
|
| 2813 |
+
"grad_norm": 15.5625,
|
| 2814 |
+
"learning_rate": 1.9198e-05,
|
| 2815 |
+
"loss": 5.4355,
|
| 2816 |
+
"step": 4010
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.00125573201054565,
|
| 2820 |
+
"grad_norm": 15.125,
|
| 2821 |
+
"learning_rate": 1.9196e-05,
|
| 2822 |
+
"loss": 5.2763,
|
| 2823 |
+
"step": 4020
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.0012588557220146689,
|
| 2827 |
+
"grad_norm": 11.8125,
|
| 2828 |
+
"learning_rate": 1.9194000000000003e-05,
|
| 2829 |
+
"loss": 5.3245,
|
| 2830 |
+
"step": 4030
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.001261979433483688,
|
| 2834 |
+
"grad_norm": 14.6875,
|
| 2835 |
+
"learning_rate": 1.9192000000000002e-05,
|
| 2836 |
+
"loss": 4.9815,
|
| 2837 |
+
"step": 4040
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.001265103144952707,
|
| 2841 |
+
"grad_norm": 13.875,
|
| 2842 |
+
"learning_rate": 1.919e-05,
|
| 2843 |
+
"loss": 5.5233,
|
| 2844 |
+
"step": 4050
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.001268226856421726,
|
| 2848 |
+
"grad_norm": 17.625,
|
| 2849 |
+
"learning_rate": 1.9188000000000003e-05,
|
| 2850 |
+
"loss": 5.6294,
|
| 2851 |
+
"step": 4060
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 0.0012713505678907451,
|
| 2855 |
+
"grad_norm": 20.5,
|
| 2856 |
+
"learning_rate": 1.9186000000000002e-05,
|
| 2857 |
+
"loss": 5.2967,
|
| 2858 |
+
"step": 4070
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 0.001274474279359764,
|
| 2862 |
+
"grad_norm": 14.75,
|
| 2863 |
+
"learning_rate": 1.9184e-05,
|
| 2864 |
+
"loss": 5.0256,
|
| 2865 |
+
"step": 4080
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 0.001277597990828783,
|
| 2869 |
+
"grad_norm": 20.875,
|
| 2870 |
+
"learning_rate": 1.9182e-05,
|
| 2871 |
+
"loss": 5.0885,
|
| 2872 |
+
"step": 4090
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 0.0012807217022978022,
|
| 2876 |
+
"grad_norm": 11.75,
|
| 2877 |
+
"learning_rate": 1.918e-05,
|
| 2878 |
+
"loss": 5.4286,
|
| 2879 |
+
"step": 4100
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 0.0012838454137668212,
|
| 2883 |
+
"grad_norm": 14.5625,
|
| 2884 |
+
"learning_rate": 1.9178000000000002e-05,
|
| 2885 |
+
"loss": 5.344,
|
| 2886 |
+
"step": 4110
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 0.0012869691252358401,
|
| 2890 |
+
"grad_norm": 24.125,
|
| 2891 |
+
"learning_rate": 1.9176e-05,
|
| 2892 |
+
"loss": 5.362,
|
| 2893 |
+
"step": 4120
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 0.0012900928367048593,
|
| 2897 |
+
"grad_norm": 24.125,
|
| 2898 |
+
"learning_rate": 1.9174000000000003e-05,
|
| 2899 |
+
"loss": 5.6709,
|
| 2900 |
+
"step": 4130
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 0.0012932165481738783,
|
| 2904 |
+
"grad_norm": 15.6875,
|
| 2905 |
+
"learning_rate": 1.9172000000000002e-05,
|
| 2906 |
+
"loss": 5.5542,
|
| 2907 |
+
"step": 4140
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 0.0012963402596428972,
|
| 2911 |
+
"grad_norm": 16.125,
|
| 2912 |
+
"learning_rate": 1.917e-05,
|
| 2913 |
+
"loss": 5.2315,
|
| 2914 |
+
"step": 4150
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 0.0012994639711119164,
|
| 2918 |
+
"grad_norm": 17.5,
|
| 2919 |
+
"learning_rate": 1.9168000000000004e-05,
|
| 2920 |
+
"loss": 5.538,
|
| 2921 |
+
"step": 4160
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.0013025876825809354,
|
| 2925 |
+
"grad_norm": 15.9375,
|
| 2926 |
+
"learning_rate": 1.9166000000000003e-05,
|
| 2927 |
+
"loss": 5.5096,
|
| 2928 |
+
"step": 4170
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 0.0013057113940499543,
|
| 2932 |
+
"grad_norm": 11.125,
|
| 2933 |
+
"learning_rate": 1.9164e-05,
|
| 2934 |
+
"loss": 5.7454,
|
| 2935 |
+
"step": 4180
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 0.0013088351055189735,
|
| 2939 |
+
"grad_norm": 11.6875,
|
| 2940 |
+
"learning_rate": 1.9162e-05,
|
| 2941 |
+
"loss": 5.5636,
|
| 2942 |
+
"step": 4190
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 0.0013119588169879925,
|
| 2946 |
+
"grad_norm": 12.4375,
|
| 2947 |
+
"learning_rate": 1.916e-05,
|
| 2948 |
+
"loss": 5.6127,
|
| 2949 |
+
"step": 4200
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 0.0013150825284570114,
|
| 2953 |
+
"grad_norm": 11.875,
|
| 2954 |
+
"learning_rate": 1.9158000000000002e-05,
|
| 2955 |
+
"loss": 5.6977,
|
| 2956 |
+
"step": 4210
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.0013182062399260306,
|
| 2960 |
+
"grad_norm": 16.75,
|
| 2961 |
+
"learning_rate": 1.9156e-05,
|
| 2962 |
+
"loss": 5.7171,
|
| 2963 |
+
"step": 4220
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 0.0013213299513950495,
|
| 2967 |
+
"grad_norm": 17.125,
|
| 2968 |
+
"learning_rate": 1.9154e-05,
|
| 2969 |
+
"loss": 5.4361,
|
| 2970 |
+
"step": 4230
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 0.0013244536628640685,
|
| 2974 |
+
"grad_norm": 14.625,
|
| 2975 |
+
"learning_rate": 1.9152000000000002e-05,
|
| 2976 |
+
"loss": 5.3326,
|
| 2977 |
+
"step": 4240
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 0.0013275773743330877,
|
| 2981 |
+
"grad_norm": 15.4375,
|
| 2982 |
+
"learning_rate": 1.915e-05,
|
| 2983 |
+
"loss": 5.438,
|
| 2984 |
+
"step": 4250
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 0.0013307010858021066,
|
| 2988 |
+
"grad_norm": 17.0,
|
| 2989 |
+
"learning_rate": 1.9148000000000004e-05,
|
| 2990 |
+
"loss": 5.5329,
|
| 2991 |
+
"step": 4260
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 0.0013338247972711256,
|
| 2995 |
+
"grad_norm": 21.125,
|
| 2996 |
+
"learning_rate": 1.9146000000000003e-05,
|
| 2997 |
+
"loss": 5.791,
|
| 2998 |
+
"step": 4270
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 0.0013369485087401448,
|
| 3002 |
+
"grad_norm": 20.125,
|
| 3003 |
+
"learning_rate": 1.9144000000000002e-05,
|
| 3004 |
+
"loss": 5.7484,
|
| 3005 |
+
"step": 4280
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 0.0013400722202091637,
|
| 3009 |
+
"grad_norm": 15.3125,
|
| 3010 |
+
"learning_rate": 1.9142e-05,
|
| 3011 |
+
"loss": 5.0317,
|
| 3012 |
+
"step": 4290
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 0.0013431959316781827,
|
| 3016 |
+
"grad_norm": 25.0,
|
| 3017 |
+
"learning_rate": 1.914e-05,
|
| 3018 |
+
"loss": 5.6427,
|
| 3019 |
+
"step": 4300
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 0.0013463196431472018,
|
| 3023 |
+
"grad_norm": 21.375,
|
| 3024 |
+
"learning_rate": 1.9138000000000002e-05,
|
| 3025 |
+
"loss": 5.4972,
|
| 3026 |
+
"step": 4310
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 0.0013494433546162208,
|
| 3030 |
+
"grad_norm": 11.1875,
|
| 3031 |
+
"learning_rate": 1.9136e-05,
|
| 3032 |
+
"loss": 5.5393,
|
| 3033 |
+
"step": 4320
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.0013525670660852398,
|
| 3037 |
+
"grad_norm": 27.625,
|
| 3038 |
+
"learning_rate": 1.9134e-05,
|
| 3039 |
+
"loss": 5.3816,
|
| 3040 |
+
"step": 4330
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 0.001355690777554259,
|
| 3044 |
+
"grad_norm": 24.75,
|
| 3045 |
+
"learning_rate": 1.9132000000000002e-05,
|
| 3046 |
+
"loss": 5.5932,
|
| 3047 |
+
"step": 4340
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 0.0013588144890232779,
|
| 3051 |
+
"grad_norm": 19.25,
|
| 3052 |
+
"learning_rate": 1.913e-05,
|
| 3053 |
+
"loss": 5.7437,
|
| 3054 |
+
"step": 4350
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 0.0013619382004922968,
|
| 3058 |
+
"grad_norm": 22.625,
|
| 3059 |
+
"learning_rate": 1.9128e-05,
|
| 3060 |
+
"loss": 5.453,
|
| 3061 |
+
"step": 4360
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 0.001365061911961316,
|
| 3065 |
+
"grad_norm": 13.5625,
|
| 3066 |
+
"learning_rate": 1.9126000000000003e-05,
|
| 3067 |
+
"loss": 5.4275,
|
| 3068 |
+
"step": 4370
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.001368185623430335,
|
| 3072 |
+
"grad_norm": 18.375,
|
| 3073 |
+
"learning_rate": 1.9124000000000002e-05,
|
| 3074 |
+
"loss": 5.1498,
|
| 3075 |
+
"step": 4380
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.001371309334899354,
|
| 3079 |
+
"grad_norm": 13.25,
|
| 3080 |
+
"learning_rate": 1.9122e-05,
|
| 3081 |
+
"loss": 5.72,
|
| 3082 |
+
"step": 4390
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.001374433046368373,
|
| 3086 |
+
"grad_norm": 23.0,
|
| 3087 |
+
"learning_rate": 1.912e-05,
|
| 3088 |
+
"loss": 5.2518,
|
| 3089 |
+
"step": 4400
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.001377556757837392,
|
| 3093 |
+
"grad_norm": 12.8125,
|
| 3094 |
+
"learning_rate": 1.9118000000000002e-05,
|
| 3095 |
+
"loss": 5.5847,
|
| 3096 |
+
"step": 4410
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.001380680469306411,
|
| 3100 |
+
"grad_norm": 21.375,
|
| 3101 |
+
"learning_rate": 1.9116e-05,
|
| 3102 |
+
"loss": 5.6463,
|
| 3103 |
+
"step": 4420
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.0013838041807754302,
|
| 3107 |
+
"grad_norm": 19.0,
|
| 3108 |
+
"learning_rate": 1.9114e-05,
|
| 3109 |
+
"loss": 5.5771,
|
| 3110 |
+
"step": 4430
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.0013869278922444492,
|
| 3114 |
+
"grad_norm": 19.625,
|
| 3115 |
+
"learning_rate": 1.9112000000000003e-05,
|
| 3116 |
+
"loss": 5.5444,
|
| 3117 |
+
"step": 4440
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.0013900516037134681,
|
| 3121 |
+
"grad_norm": 11.875,
|
| 3122 |
+
"learning_rate": 1.911e-05,
|
| 3123 |
+
"loss": 5.5599,
|
| 3124 |
+
"step": 4450
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.0013931753151824873,
|
| 3128 |
+
"grad_norm": 14.1875,
|
| 3129 |
+
"learning_rate": 1.9108e-05,
|
| 3130 |
+
"loss": 5.6097,
|
| 3131 |
+
"step": 4460
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.0013962990266515062,
|
| 3135 |
+
"grad_norm": 15.625,
|
| 3136 |
+
"learning_rate": 1.9106000000000003e-05,
|
| 3137 |
+
"loss": 5.4403,
|
| 3138 |
+
"step": 4470
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.0013994227381205252,
|
| 3142 |
+
"grad_norm": 15.5,
|
| 3143 |
+
"learning_rate": 1.9104000000000002e-05,
|
| 3144 |
+
"loss": 5.6646,
|
| 3145 |
+
"step": 4480
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.0014025464495895444,
|
| 3149 |
+
"grad_norm": 16.375,
|
| 3150 |
+
"learning_rate": 1.9102e-05,
|
| 3151 |
+
"loss": 5.5024,
|
| 3152 |
+
"step": 4490
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.0014056701610585633,
|
| 3156 |
+
"grad_norm": 22.0,
|
| 3157 |
+
"learning_rate": 1.91e-05,
|
| 3158 |
+
"loss": 5.5846,
|
| 3159 |
+
"step": 4500
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.0014087938725275823,
|
| 3163 |
+
"grad_norm": 12.5625,
|
| 3164 |
+
"learning_rate": 1.9098000000000002e-05,
|
| 3165 |
+
"loss": 5.7252,
|
| 3166 |
+
"step": 4510
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.0014119175839966015,
|
| 3170 |
+
"grad_norm": 16.25,
|
| 3171 |
+
"learning_rate": 1.9096e-05,
|
| 3172 |
+
"loss": 5.7331,
|
| 3173 |
+
"step": 4520
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.0014150412954656204,
|
| 3177 |
+
"grad_norm": 17.5,
|
| 3178 |
+
"learning_rate": 1.9094e-05,
|
| 3179 |
+
"loss": 5.6089,
|
| 3180 |
+
"step": 4530
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.0014181650069346394,
|
| 3184 |
+
"grad_norm": 15.625,
|
| 3185 |
+
"learning_rate": 1.9092000000000003e-05,
|
| 3186 |
+
"loss": 5.5612,
|
| 3187 |
+
"step": 4540
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.0014212887184036585,
|
| 3191 |
+
"grad_norm": 10.3125,
|
| 3192 |
+
"learning_rate": 1.9090000000000002e-05,
|
| 3193 |
+
"loss": 5.4987,
|
| 3194 |
+
"step": 4550
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.0014244124298726775,
|
| 3198 |
+
"grad_norm": 18.625,
|
| 3199 |
+
"learning_rate": 1.9088e-05,
|
| 3200 |
+
"loss": 5.4008,
|
| 3201 |
+
"step": 4560
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.0014275361413416965,
|
| 3205 |
+
"grad_norm": 15.0625,
|
| 3206 |
+
"learning_rate": 1.9086000000000003e-05,
|
| 3207 |
+
"loss": 5.8175,
|
| 3208 |
+
"step": 4570
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.0014306598528107156,
|
| 3212 |
+
"grad_norm": 17.875,
|
| 3213 |
+
"learning_rate": 1.9084000000000002e-05,
|
| 3214 |
+
"loss": 5.7524,
|
| 3215 |
+
"step": 4580
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.0014337835642797346,
|
| 3219 |
+
"grad_norm": 14.0625,
|
| 3220 |
+
"learning_rate": 1.9082e-05,
|
| 3221 |
+
"loss": 5.6834,
|
| 3222 |
+
"step": 4590
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.0014369072757487535,
|
| 3226 |
+
"grad_norm": 17.375,
|
| 3227 |
+
"learning_rate": 1.908e-05,
|
| 3228 |
+
"loss": 5.7438,
|
| 3229 |
+
"step": 4600
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.0014400309872177727,
|
| 3233 |
+
"grad_norm": 14.625,
|
| 3234 |
+
"learning_rate": 1.9078e-05,
|
| 3235 |
+
"loss": 5.6002,
|
| 3236 |
+
"step": 4610
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.0014431546986867917,
|
| 3240 |
+
"grad_norm": 17.0,
|
| 3241 |
+
"learning_rate": 1.9076e-05,
|
| 3242 |
+
"loss": 5.5971,
|
| 3243 |
+
"step": 4620
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.0014462784101558106,
|
| 3247 |
+
"grad_norm": 12.5,
|
| 3248 |
+
"learning_rate": 1.9074e-05,
|
| 3249 |
+
"loss": 5.6446,
|
| 3250 |
+
"step": 4630
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.0014494021216248298,
|
| 3254 |
+
"grad_norm": 22.75,
|
| 3255 |
+
"learning_rate": 1.9072000000000003e-05,
|
| 3256 |
+
"loss": 5.2944,
|
| 3257 |
+
"step": 4640
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.0014525258330938488,
|
| 3261 |
+
"grad_norm": 13.25,
|
| 3262 |
+
"learning_rate": 1.9070000000000002e-05,
|
| 3263 |
+
"loss": 5.6228,
|
| 3264 |
+
"step": 4650
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.0014556495445628677,
|
| 3268 |
+
"grad_norm": 13.25,
|
| 3269 |
+
"learning_rate": 1.9068e-05,
|
| 3270 |
+
"loss": 5.1766,
|
| 3271 |
+
"step": 4660
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.001458773256031887,
|
| 3275 |
+
"grad_norm": 20.875,
|
| 3276 |
+
"learning_rate": 1.9066000000000003e-05,
|
| 3277 |
+
"loss": 5.1814,
|
| 3278 |
+
"step": 4670
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.0014618969675009059,
|
| 3282 |
+
"grad_norm": 16.625,
|
| 3283 |
+
"learning_rate": 1.9064000000000002e-05,
|
| 3284 |
+
"loss": 5.2023,
|
| 3285 |
+
"step": 4680
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.0014650206789699248,
|
| 3289 |
+
"grad_norm": 18.75,
|
| 3290 |
+
"learning_rate": 1.9062e-05,
|
| 3291 |
+
"loss": 5.5188,
|
| 3292 |
+
"step": 4690
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.001468144390438944,
|
| 3296 |
+
"grad_norm": 16.875,
|
| 3297 |
+
"learning_rate": 1.906e-05,
|
| 3298 |
+
"loss": 5.5913,
|
| 3299 |
+
"step": 4700
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.001471268101907963,
|
| 3303 |
+
"grad_norm": 12.3125,
|
| 3304 |
+
"learning_rate": 1.9058e-05,
|
| 3305 |
+
"loss": 5.5999,
|
| 3306 |
+
"step": 4710
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.001474391813376982,
|
| 3310 |
+
"grad_norm": 16.0,
|
| 3311 |
+
"learning_rate": 1.9056000000000002e-05,
|
| 3312 |
+
"loss": 5.7381,
|
| 3313 |
+
"step": 4720
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.001477515524846001,
|
| 3317 |
+
"grad_norm": 12.625,
|
| 3318 |
+
"learning_rate": 1.9054e-05,
|
| 3319 |
+
"loss": 5.7001,
|
| 3320 |
+
"step": 4730
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.00148063923631502,
|
| 3324 |
+
"grad_norm": 13.875,
|
| 3325 |
+
"learning_rate": 1.9052000000000003e-05,
|
| 3326 |
+
"loss": 5.5039,
|
| 3327 |
+
"step": 4740
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.001483762947784039,
|
| 3331 |
+
"grad_norm": 16.75,
|
| 3332 |
+
"learning_rate": 1.9050000000000002e-05,
|
| 3333 |
+
"loss": 5.7352,
|
| 3334 |
+
"step": 4750
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.0014868866592530582,
|
| 3338 |
+
"grad_norm": 15.1875,
|
| 3339 |
+
"learning_rate": 1.9048e-05,
|
| 3340 |
+
"loss": 5.5764,
|
| 3341 |
+
"step": 4760
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.0014900103707220771,
|
| 3345 |
+
"grad_norm": 12.75,
|
| 3346 |
+
"learning_rate": 1.9046000000000004e-05,
|
| 3347 |
+
"loss": 5.6157,
|
| 3348 |
+
"step": 4770
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.001493134082191096,
|
| 3352 |
+
"grad_norm": 13.625,
|
| 3353 |
+
"learning_rate": 1.9044000000000003e-05,
|
| 3354 |
+
"loss": 5.5936,
|
| 3355 |
+
"step": 4780
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.0014962577936601152,
|
| 3359 |
+
"grad_norm": 16.75,
|
| 3360 |
+
"learning_rate": 1.9042e-05,
|
| 3361 |
+
"loss": 5.5932,
|
| 3362 |
+
"step": 4790
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.0014993815051291342,
|
| 3366 |
+
"grad_norm": 22.75,
|
| 3367 |
+
"learning_rate": 1.904e-05,
|
| 3368 |
+
"loss": 5.1952,
|
| 3369 |
+
"step": 4800
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.0015025052165981532,
|
| 3373 |
+
"grad_norm": 10.625,
|
| 3374 |
+
"learning_rate": 1.9038e-05,
|
| 3375 |
+
"loss": 5.2644,
|
| 3376 |
+
"step": 4810
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.0015056289280671723,
|
| 3380 |
+
"grad_norm": 12.1875,
|
| 3381 |
+
"learning_rate": 1.9036000000000002e-05,
|
| 3382 |
+
"loss": 5.6933,
|
| 3383 |
+
"step": 4820
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.0015087526395361913,
|
| 3387 |
+
"grad_norm": 14.875,
|
| 3388 |
+
"learning_rate": 1.9034e-05,
|
| 3389 |
+
"loss": 5.7847,
|
| 3390 |
+
"step": 4830
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.0015118763510052102,
|
| 3394 |
+
"grad_norm": 23.625,
|
| 3395 |
+
"learning_rate": 1.9032e-05,
|
| 3396 |
+
"loss": 5.3767,
|
| 3397 |
+
"step": 4840
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.0015150000624742294,
|
| 3401 |
+
"grad_norm": 17.875,
|
| 3402 |
+
"learning_rate": 1.9030000000000002e-05,
|
| 3403 |
+
"loss": 5.6129,
|
| 3404 |
+
"step": 4850
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.0015181237739432484,
|
| 3408 |
+
"grad_norm": 15.4375,
|
| 3409 |
+
"learning_rate": 1.9028e-05,
|
| 3410 |
+
"loss": 5.6761,
|
| 3411 |
+
"step": 4860
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.0015212474854122673,
|
| 3415 |
+
"grad_norm": 13.4375,
|
| 3416 |
+
"learning_rate": 1.9026000000000004e-05,
|
| 3417 |
+
"loss": 5.5118,
|
| 3418 |
+
"step": 4870
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.0015243711968812865,
|
| 3422 |
+
"grad_norm": 14.625,
|
| 3423 |
+
"learning_rate": 1.9024000000000003e-05,
|
| 3424 |
+
"loss": 5.5173,
|
| 3425 |
+
"step": 4880
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.0015274949083503055,
|
| 3429 |
+
"grad_norm": 15.3125,
|
| 3430 |
+
"learning_rate": 1.9022000000000002e-05,
|
| 3431 |
+
"loss": 5.5349,
|
| 3432 |
+
"step": 4890
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.0015306186198193244,
|
| 3436 |
+
"grad_norm": 17.25,
|
| 3437 |
+
"learning_rate": 1.902e-05,
|
| 3438 |
+
"loss": 5.1842,
|
| 3439 |
+
"step": 4900
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.0015337423312883436,
|
| 3443 |
+
"grad_norm": 28.5,
|
| 3444 |
+
"learning_rate": 1.9018e-05,
|
| 3445 |
+
"loss": 5.2254,
|
| 3446 |
+
"step": 4910
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.0015368660427573626,
|
| 3450 |
+
"grad_norm": 15.6875,
|
| 3451 |
+
"learning_rate": 1.9016000000000002e-05,
|
| 3452 |
+
"loss": 5.4484,
|
| 3453 |
+
"step": 4920
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.0015399897542263817,
|
| 3457 |
+
"grad_norm": 13.0,
|
| 3458 |
+
"learning_rate": 1.9014e-05,
|
| 3459 |
+
"loss": 5.1899,
|
| 3460 |
+
"step": 4930
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.0015431134656954007,
|
| 3464 |
+
"grad_norm": 15.5,
|
| 3465 |
+
"learning_rate": 1.9012e-05,
|
| 3466 |
+
"loss": 5.7682,
|
| 3467 |
+
"step": 4940
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.0015462371771644196,
|
| 3471 |
+
"grad_norm": 16.125,
|
| 3472 |
+
"learning_rate": 1.9010000000000003e-05,
|
| 3473 |
+
"loss": 5.564,
|
| 3474 |
+
"step": 4950
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.0015493608886334388,
|
| 3478 |
+
"grad_norm": 15.3125,
|
| 3479 |
+
"learning_rate": 1.9008e-05,
|
| 3480 |
+
"loss": 5.0649,
|
| 3481 |
+
"step": 4960
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.0015524846001024578,
|
| 3485 |
+
"grad_norm": 14.375,
|
| 3486 |
+
"learning_rate": 1.9006e-05,
|
| 3487 |
+
"loss": 5.1236,
|
| 3488 |
+
"step": 4970
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.0015556083115714767,
|
| 3492 |
+
"grad_norm": 13.0,
|
| 3493 |
+
"learning_rate": 1.9004000000000003e-05,
|
| 3494 |
+
"loss": 5.6358,
|
| 3495 |
+
"step": 4980
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.001558732023040496,
|
| 3499 |
+
"grad_norm": 13.0625,
|
| 3500 |
+
"learning_rate": 1.9002000000000002e-05,
|
| 3501 |
+
"loss": 5.4334,
|
| 3502 |
+
"step": 4990
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.0015618557345095149,
|
| 3506 |
+
"grad_norm": 10.8125,
|
| 3507 |
+
"learning_rate": 1.9e-05,
|
| 3508 |
+
"loss": 5.5315,
|
| 3509 |
+
"step": 5000
|
| 3510 |
}
|
| 3511 |
],
|
| 3512 |
"logging_steps": 10,
|
|
|
|
| 3514 |
"num_input_tokens_seen": 0,
|
| 3515 |
"num_train_epochs": 1,
|
| 3516 |
"save_steps": 1000,
|
| 3517 |
+
"total_flos": 1.106646172434432e+18,
|
| 3518 |
"train_batch_size": 4,
|
| 3519 |
"trial_name": null,
|
| 3520 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e98683d9284be6e0cd978529d0257d227b3826d3be4c36f5cc01d2c8b325e32
|
| 3 |
size 4984
|