Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +2622 -4
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 23641256
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a357a54ce0cda3c5d6d7dbb74cbb747788303eb44309968c70c8419797cace7f
|
| 3 |
size 23641256
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 47458874
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1adaa9b511cbff73e724b5ca110a03d7079e04ac73d245531c73df4d6540b83a
|
| 3 |
size 47458874
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cdc99c73973f503de5132b308c75e48e1433fb57f7ca3e6591034232d0c9884d
|
| 3 |
size 14244
|
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 sha256:21f918cb5d17d0f0871e1f2fbd4a3b767ff237277b6a0af30d0316caf0021eab
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 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,
|
|
@@ -917,6 +917,2624 @@
|
|
| 917 |
"learning_rate": 0.0002,
|
| 918 |
"loss": 0.6839,
|
| 919 |
"step": 130
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 920 |
}
|
| 921 |
],
|
| 922 |
"logging_steps": 1,
|
|
@@ -931,12 +3549,12 @@
|
|
| 931 |
"should_evaluate": false,
|
| 932 |
"should_log": false,
|
| 933 |
"should_save": true,
|
| 934 |
-
"should_training_stop":
|
| 935 |
},
|
| 936 |
"attributes": {}
|
| 937 |
}
|
| 938 |
},
|
| 939 |
-
"total_flos":
|
| 940 |
"train_batch_size": 1,
|
| 941 |
"trial_name": null,
|
| 942 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9977695167286247,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 504,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 917 |
"learning_rate": 0.0002,
|
| 918 |
"loss": 0.6839,
|
| 919 |
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.779182156133829,
|
| 923 |
+
"grad_norm": 1.0010913610458374,
|
| 924 |
+
"learning_rate": 0.0002,
|
| 925 |
+
"loss": 0.6247,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.7851301115241636,
|
| 930 |
+
"grad_norm": 0.8673060536384583,
|
| 931 |
+
"learning_rate": 0.0002,
|
| 932 |
+
"loss": 0.7232,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7910780669144981,
|
| 937 |
+
"grad_norm": 1.070591688156128,
|
| 938 |
+
"learning_rate": 0.0002,
|
| 939 |
+
"loss": 0.785,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.7970260223048327,
|
| 944 |
+
"grad_norm": 1.0302468538284302,
|
| 945 |
+
"learning_rate": 0.0002,
|
| 946 |
+
"loss": 0.8044,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.8029739776951673,
|
| 951 |
+
"grad_norm": 1.0886098146438599,
|
| 952 |
+
"learning_rate": 0.0002,
|
| 953 |
+
"loss": 0.6984,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.8089219330855019,
|
| 958 |
+
"grad_norm": 0.9349246025085449,
|
| 959 |
+
"learning_rate": 0.0002,
|
| 960 |
+
"loss": 0.6711,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.8148698884758364,
|
| 965 |
+
"grad_norm": 1.2482446432113647,
|
| 966 |
+
"learning_rate": 0.0002,
|
| 967 |
+
"loss": 0.6468,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.820817843866171,
|
| 972 |
+
"grad_norm": 1.184043049812317,
|
| 973 |
+
"learning_rate": 0.0002,
|
| 974 |
+
"loss": 0.6909,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.8267657992565056,
|
| 979 |
+
"grad_norm": 0.8721855878829956,
|
| 980 |
+
"learning_rate": 0.0002,
|
| 981 |
+
"loss": 0.6553,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.8327137546468402,
|
| 986 |
+
"grad_norm": 1.30323326587677,
|
| 987 |
+
"learning_rate": 0.0002,
|
| 988 |
+
"loss": 0.7109,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.8386617100371747,
|
| 993 |
+
"grad_norm": 1.0187689065933228,
|
| 994 |
+
"learning_rate": 0.0002,
|
| 995 |
+
"loss": 0.6682,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.8446096654275093,
|
| 1000 |
+
"grad_norm": 2.3475165367126465,
|
| 1001 |
+
"learning_rate": 0.0002,
|
| 1002 |
+
"loss": 0.7615,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8505576208178439,
|
| 1007 |
+
"grad_norm": 0.9803043603897095,
|
| 1008 |
+
"learning_rate": 0.0002,
|
| 1009 |
+
"loss": 0.7179,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.8565055762081785,
|
| 1014 |
+
"grad_norm": 1.2290213108062744,
|
| 1015 |
+
"learning_rate": 0.0002,
|
| 1016 |
+
"loss": 0.8696,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.862453531598513,
|
| 1021 |
+
"grad_norm": 1.1041066646575928,
|
| 1022 |
+
"learning_rate": 0.0002,
|
| 1023 |
+
"loss": 0.8196,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.8684014869888476,
|
| 1028 |
+
"grad_norm": 0.9638866186141968,
|
| 1029 |
+
"learning_rate": 0.0002,
|
| 1030 |
+
"loss": 0.655,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.8743494423791821,
|
| 1035 |
+
"grad_norm": 0.9777591824531555,
|
| 1036 |
+
"learning_rate": 0.0002,
|
| 1037 |
+
"loss": 0.8053,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.8802973977695168,
|
| 1042 |
+
"grad_norm": 0.8717353343963623,
|
| 1043 |
+
"learning_rate": 0.0002,
|
| 1044 |
+
"loss": 0.6726,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.8862453531598513,
|
| 1049 |
+
"grad_norm": 1.1772398948669434,
|
| 1050 |
+
"learning_rate": 0.0002,
|
| 1051 |
+
"loss": 0.7188,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.8921933085501859,
|
| 1056 |
+
"grad_norm": 1.1113988161087036,
|
| 1057 |
+
"learning_rate": 0.0002,
|
| 1058 |
+
"loss": 0.7177,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.8981412639405204,
|
| 1063 |
+
"grad_norm": 1.6691763401031494,
|
| 1064 |
+
"learning_rate": 0.0002,
|
| 1065 |
+
"loss": 0.5822,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.9040892193308551,
|
| 1070 |
+
"grad_norm": 1.0139896869659424,
|
| 1071 |
+
"learning_rate": 0.0002,
|
| 1072 |
+
"loss": 0.7416,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.9100371747211896,
|
| 1077 |
+
"grad_norm": 1.2538039684295654,
|
| 1078 |
+
"learning_rate": 0.0002,
|
| 1079 |
+
"loss": 0.7822,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.9159851301115242,
|
| 1084 |
+
"grad_norm": 0.833595335483551,
|
| 1085 |
+
"learning_rate": 0.0002,
|
| 1086 |
+
"loss": 0.6617,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.9219330855018587,
|
| 1091 |
+
"grad_norm": 0.869482696056366,
|
| 1092 |
+
"learning_rate": 0.0002,
|
| 1093 |
+
"loss": 0.6146,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.9278810408921933,
|
| 1098 |
+
"grad_norm": 0.973523736000061,
|
| 1099 |
+
"learning_rate": 0.0002,
|
| 1100 |
+
"loss": 0.6685,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.9338289962825279,
|
| 1105 |
+
"grad_norm": 0.982566237449646,
|
| 1106 |
+
"learning_rate": 0.0002,
|
| 1107 |
+
"loss": 0.6685,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.9397769516728625,
|
| 1112 |
+
"grad_norm": 1.0534875392913818,
|
| 1113 |
+
"learning_rate": 0.0002,
|
| 1114 |
+
"loss": 0.656,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.945724907063197,
|
| 1119 |
+
"grad_norm": 1.11860990524292,
|
| 1120 |
+
"learning_rate": 0.0002,
|
| 1121 |
+
"loss": 0.7261,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.9516728624535316,
|
| 1126 |
+
"grad_norm": 1.0286844968795776,
|
| 1127 |
+
"learning_rate": 0.0002,
|
| 1128 |
+
"loss": 0.7271,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.9576208178438662,
|
| 1133 |
+
"grad_norm": 0.8426340818405151,
|
| 1134 |
+
"learning_rate": 0.0002,
|
| 1135 |
+
"loss": 0.718,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.9635687732342008,
|
| 1140 |
+
"grad_norm": 0.990667998790741,
|
| 1141 |
+
"learning_rate": 0.0002,
|
| 1142 |
+
"loss": 0.7795,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.9695167286245353,
|
| 1147 |
+
"grad_norm": 1.1110923290252686,
|
| 1148 |
+
"learning_rate": 0.0002,
|
| 1149 |
+
"loss": 0.6946,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.9754646840148699,
|
| 1154 |
+
"grad_norm": 1.0378597974777222,
|
| 1155 |
+
"learning_rate": 0.0002,
|
| 1156 |
+
"loss": 0.8106,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.9814126394052045,
|
| 1161 |
+
"grad_norm": 0.9507467746734619,
|
| 1162 |
+
"learning_rate": 0.0002,
|
| 1163 |
+
"loss": 0.7031,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.9873605947955391,
|
| 1168 |
+
"grad_norm": 0.8636868596076965,
|
| 1169 |
+
"learning_rate": 0.0002,
|
| 1170 |
+
"loss": 0.6636,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.9933085501858736,
|
| 1175 |
+
"grad_norm": 1.0482003688812256,
|
| 1176 |
+
"learning_rate": 0.0002,
|
| 1177 |
+
"loss": 0.7204,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.9992565055762082,
|
| 1182 |
+
"grad_norm": 0.9490022659301758,
|
| 1183 |
+
"learning_rate": 0.0002,
|
| 1184 |
+
"loss": 0.5626,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.0052044609665427,
|
| 1189 |
+
"grad_norm": 0.8918917179107666,
|
| 1190 |
+
"learning_rate": 0.0002,
|
| 1191 |
+
"loss": 0.6822,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.0111524163568772,
|
| 1196 |
+
"grad_norm": 0.9100430011749268,
|
| 1197 |
+
"learning_rate": 0.0002,
|
| 1198 |
+
"loss": 0.7147,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.0171003717472118,
|
| 1203 |
+
"grad_norm": 0.9007158279418945,
|
| 1204 |
+
"learning_rate": 0.0002,
|
| 1205 |
+
"loss": 0.6015,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.0230483271375466,
|
| 1210 |
+
"grad_norm": 0.9267099499702454,
|
| 1211 |
+
"learning_rate": 0.0002,
|
| 1212 |
+
"loss": 0.6562,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.0289962825278811,
|
| 1217 |
+
"grad_norm": 1.0618972778320312,
|
| 1218 |
+
"learning_rate": 0.0002,
|
| 1219 |
+
"loss": 0.6601,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.0349442379182157,
|
| 1224 |
+
"grad_norm": 1.1782737970352173,
|
| 1225 |
+
"learning_rate": 0.0002,
|
| 1226 |
+
"loss": 0.6172,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.0408921933085502,
|
| 1231 |
+
"grad_norm": 1.141661286354065,
|
| 1232 |
+
"learning_rate": 0.0002,
|
| 1233 |
+
"loss": 0.5854,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.0468401486988848,
|
| 1238 |
+
"grad_norm": 1.1038105487823486,
|
| 1239 |
+
"learning_rate": 0.0002,
|
| 1240 |
+
"loss": 0.6655,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.0527881040892193,
|
| 1245 |
+
"grad_norm": 1.1518810987472534,
|
| 1246 |
+
"learning_rate": 0.0002,
|
| 1247 |
+
"loss": 0.493,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.0587360594795538,
|
| 1252 |
+
"grad_norm": 1.0501494407653809,
|
| 1253 |
+
"learning_rate": 0.0002,
|
| 1254 |
+
"loss": 0.6256,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.0646840148698884,
|
| 1259 |
+
"grad_norm": 0.9064037799835205,
|
| 1260 |
+
"learning_rate": 0.0002,
|
| 1261 |
+
"loss": 0.5695,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.070631970260223,
|
| 1266 |
+
"grad_norm": 1.1978446245193481,
|
| 1267 |
+
"learning_rate": 0.0002,
|
| 1268 |
+
"loss": 0.617,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.0765799256505577,
|
| 1273 |
+
"grad_norm": 0.9782500267028809,
|
| 1274 |
+
"learning_rate": 0.0002,
|
| 1275 |
+
"loss": 0.6542,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.0825278810408923,
|
| 1280 |
+
"grad_norm": 1.6090043783187866,
|
| 1281 |
+
"learning_rate": 0.0002,
|
| 1282 |
+
"loss": 0.6827,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.0884758364312268,
|
| 1287 |
+
"grad_norm": 1.2153990268707275,
|
| 1288 |
+
"learning_rate": 0.0002,
|
| 1289 |
+
"loss": 0.7469,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.0944237918215614,
|
| 1294 |
+
"grad_norm": 0.9915666580200195,
|
| 1295 |
+
"learning_rate": 0.0002,
|
| 1296 |
+
"loss": 0.6475,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.100371747211896,
|
| 1301 |
+
"grad_norm": 1.1319215297698975,
|
| 1302 |
+
"learning_rate": 0.0002,
|
| 1303 |
+
"loss": 0.7879,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.1063197026022304,
|
| 1308 |
+
"grad_norm": 1.0497454404830933,
|
| 1309 |
+
"learning_rate": 0.0002,
|
| 1310 |
+
"loss": 0.6386,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.112267657992565,
|
| 1315 |
+
"grad_norm": 1.1735246181488037,
|
| 1316 |
+
"learning_rate": 0.0002,
|
| 1317 |
+
"loss": 0.7276,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.1182156133828995,
|
| 1322 |
+
"grad_norm": 1.280543327331543,
|
| 1323 |
+
"learning_rate": 0.0002,
|
| 1324 |
+
"loss": 0.6403,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.124163568773234,
|
| 1329 |
+
"grad_norm": 1.0103743076324463,
|
| 1330 |
+
"learning_rate": 0.0002,
|
| 1331 |
+
"loss": 0.5913,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.1301115241635689,
|
| 1336 |
+
"grad_norm": 1.0629348754882812,
|
| 1337 |
+
"learning_rate": 0.0002,
|
| 1338 |
+
"loss": 0.708,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.1360594795539034,
|
| 1343 |
+
"grad_norm": 0.9152292609214783,
|
| 1344 |
+
"learning_rate": 0.0002,
|
| 1345 |
+
"loss": 0.8295,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.142007434944238,
|
| 1350 |
+
"grad_norm": 0.9847885370254517,
|
| 1351 |
+
"learning_rate": 0.0002,
|
| 1352 |
+
"loss": 0.6569,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.1479553903345725,
|
| 1357 |
+
"grad_norm": 1.5211213827133179,
|
| 1358 |
+
"learning_rate": 0.0002,
|
| 1359 |
+
"loss": 0.676,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.153903345724907,
|
| 1364 |
+
"grad_norm": 1.0376240015029907,
|
| 1365 |
+
"learning_rate": 0.0002,
|
| 1366 |
+
"loss": 0.5558,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.1598513011152416,
|
| 1371 |
+
"grad_norm": 0.9746745824813843,
|
| 1372 |
+
"learning_rate": 0.0002,
|
| 1373 |
+
"loss": 0.6051,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.1657992565055761,
|
| 1378 |
+
"grad_norm": 1.0810937881469727,
|
| 1379 |
+
"learning_rate": 0.0002,
|
| 1380 |
+
"loss": 0.5947,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.1717472118959107,
|
| 1385 |
+
"grad_norm": 0.8687509894371033,
|
| 1386 |
+
"learning_rate": 0.0002,
|
| 1387 |
+
"loss": 0.5834,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.1776951672862452,
|
| 1392 |
+
"grad_norm": 1.3437882661819458,
|
| 1393 |
+
"learning_rate": 0.0002,
|
| 1394 |
+
"loss": 0.7133,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.18364312267658,
|
| 1399 |
+
"grad_norm": 0.9247745871543884,
|
| 1400 |
+
"learning_rate": 0.0002,
|
| 1401 |
+
"loss": 0.668,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.1895910780669146,
|
| 1406 |
+
"grad_norm": 1.116870403289795,
|
| 1407 |
+
"learning_rate": 0.0002,
|
| 1408 |
+
"loss": 0.5763,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.195539033457249,
|
| 1413 |
+
"grad_norm": 1.0791046619415283,
|
| 1414 |
+
"learning_rate": 0.0002,
|
| 1415 |
+
"loss": 0.7535,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.2014869888475836,
|
| 1420 |
+
"grad_norm": 1.1156578063964844,
|
| 1421 |
+
"learning_rate": 0.0002,
|
| 1422 |
+
"loss": 0.623,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.2074349442379182,
|
| 1427 |
+
"grad_norm": 1.3306505680084229,
|
| 1428 |
+
"learning_rate": 0.0002,
|
| 1429 |
+
"loss": 0.6239,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.2133828996282527,
|
| 1434 |
+
"grad_norm": 1.1251856088638306,
|
| 1435 |
+
"learning_rate": 0.0002,
|
| 1436 |
+
"loss": 0.6122,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.2193308550185873,
|
| 1441 |
+
"grad_norm": 0.9659932255744934,
|
| 1442 |
+
"learning_rate": 0.0002,
|
| 1443 |
+
"loss": 0.7163,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.2252788104089218,
|
| 1448 |
+
"grad_norm": 1.1080381870269775,
|
| 1449 |
+
"learning_rate": 0.0002,
|
| 1450 |
+
"loss": 0.5558,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.2312267657992564,
|
| 1455 |
+
"grad_norm": 1.1085773706436157,
|
| 1456 |
+
"learning_rate": 0.0002,
|
| 1457 |
+
"loss": 0.6306,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.2371747211895912,
|
| 1462 |
+
"grad_norm": 1.5555293560028076,
|
| 1463 |
+
"learning_rate": 0.0002,
|
| 1464 |
+
"loss": 0.7266,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.2431226765799257,
|
| 1469 |
+
"grad_norm": 1.1494146585464478,
|
| 1470 |
+
"learning_rate": 0.0002,
|
| 1471 |
+
"loss": 0.6902,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.2490706319702602,
|
| 1476 |
+
"grad_norm": 1.0835429430007935,
|
| 1477 |
+
"learning_rate": 0.0002,
|
| 1478 |
+
"loss": 0.7219,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.2550185873605948,
|
| 1483 |
+
"grad_norm": 1.1306850910186768,
|
| 1484 |
+
"learning_rate": 0.0002,
|
| 1485 |
+
"loss": 0.5931,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.2609665427509293,
|
| 1490 |
+
"grad_norm": 1.148278832435608,
|
| 1491 |
+
"learning_rate": 0.0002,
|
| 1492 |
+
"loss": 0.6452,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.266914498141264,
|
| 1497 |
+
"grad_norm": 1.097596526145935,
|
| 1498 |
+
"learning_rate": 0.0002,
|
| 1499 |
+
"loss": 0.6486,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.2728624535315984,
|
| 1504 |
+
"grad_norm": 1.3407150506973267,
|
| 1505 |
+
"learning_rate": 0.0002,
|
| 1506 |
+
"loss": 0.5874,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.2788104089219332,
|
| 1511 |
+
"grad_norm": 0.8781871199607849,
|
| 1512 |
+
"learning_rate": 0.0002,
|
| 1513 |
+
"loss": 0.6079,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.2847583643122675,
|
| 1518 |
+
"grad_norm": 0.998188853263855,
|
| 1519 |
+
"learning_rate": 0.0002,
|
| 1520 |
+
"loss": 0.7292,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.2907063197026023,
|
| 1525 |
+
"grad_norm": 1.0128471851348877,
|
| 1526 |
+
"learning_rate": 0.0002,
|
| 1527 |
+
"loss": 0.6729,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.2966542750929368,
|
| 1532 |
+
"grad_norm": 1.2172327041625977,
|
| 1533 |
+
"learning_rate": 0.0002,
|
| 1534 |
+
"loss": 0.7529,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.3026022304832714,
|
| 1539 |
+
"grad_norm": 0.9904006719589233,
|
| 1540 |
+
"learning_rate": 0.0002,
|
| 1541 |
+
"loss": 0.639,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.308550185873606,
|
| 1546 |
+
"grad_norm": 0.8886059522628784,
|
| 1547 |
+
"learning_rate": 0.0002,
|
| 1548 |
+
"loss": 0.6128,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.3144981412639405,
|
| 1553 |
+
"grad_norm": 1.0350927114486694,
|
| 1554 |
+
"learning_rate": 0.0002,
|
| 1555 |
+
"loss": 0.6668,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.320446096654275,
|
| 1560 |
+
"grad_norm": 1.0321650505065918,
|
| 1561 |
+
"learning_rate": 0.0002,
|
| 1562 |
+
"loss": 0.6493,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.3263940520446096,
|
| 1567 |
+
"grad_norm": 0.8952768445014954,
|
| 1568 |
+
"learning_rate": 0.0002,
|
| 1569 |
+
"loss": 0.7005,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.3323420074349444,
|
| 1574 |
+
"grad_norm": 1.3372063636779785,
|
| 1575 |
+
"learning_rate": 0.0002,
|
| 1576 |
+
"loss": 0.6347,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.3382899628252787,
|
| 1581 |
+
"grad_norm": 0.9312218427658081,
|
| 1582 |
+
"learning_rate": 0.0002,
|
| 1583 |
+
"loss": 0.5963,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.3442379182156134,
|
| 1588 |
+
"grad_norm": 0.8845749497413635,
|
| 1589 |
+
"learning_rate": 0.0002,
|
| 1590 |
+
"loss": 0.5445,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.350185873605948,
|
| 1595 |
+
"grad_norm": 1.292598843574524,
|
| 1596 |
+
"learning_rate": 0.0002,
|
| 1597 |
+
"loss": 0.6662,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.3561338289962825,
|
| 1602 |
+
"grad_norm": 1.0537996292114258,
|
| 1603 |
+
"learning_rate": 0.0002,
|
| 1604 |
+
"loss": 0.7332,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.362081784386617,
|
| 1609 |
+
"grad_norm": 0.9492632150650024,
|
| 1610 |
+
"learning_rate": 0.0002,
|
| 1611 |
+
"loss": 0.6157,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.3680297397769516,
|
| 1616 |
+
"grad_norm": 1.0352752208709717,
|
| 1617 |
+
"learning_rate": 0.0002,
|
| 1618 |
+
"loss": 0.5248,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.3739776951672862,
|
| 1623 |
+
"grad_norm": 1.085534930229187,
|
| 1624 |
+
"learning_rate": 0.0002,
|
| 1625 |
+
"loss": 0.6358,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.3799256505576207,
|
| 1630 |
+
"grad_norm": 1.098999261856079,
|
| 1631 |
+
"learning_rate": 0.0002,
|
| 1632 |
+
"loss": 0.6181,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.3858736059479555,
|
| 1637 |
+
"grad_norm": 1.114450454711914,
|
| 1638 |
+
"learning_rate": 0.0002,
|
| 1639 |
+
"loss": 0.7067,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.3918215613382898,
|
| 1644 |
+
"grad_norm": 1.3746989965438843,
|
| 1645 |
+
"learning_rate": 0.0002,
|
| 1646 |
+
"loss": 0.6016,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.3977695167286246,
|
| 1651 |
+
"grad_norm": 0.916519820690155,
|
| 1652 |
+
"learning_rate": 0.0002,
|
| 1653 |
+
"loss": 0.6579,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.4037174721189591,
|
| 1658 |
+
"grad_norm": 1.0786117315292358,
|
| 1659 |
+
"learning_rate": 0.0002,
|
| 1660 |
+
"loss": 0.6516,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.4096654275092937,
|
| 1665 |
+
"grad_norm": 0.9264970421791077,
|
| 1666 |
+
"learning_rate": 0.0002,
|
| 1667 |
+
"loss": 0.5927,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.4156133828996282,
|
| 1672 |
+
"grad_norm": 1.0969526767730713,
|
| 1673 |
+
"learning_rate": 0.0002,
|
| 1674 |
+
"loss": 0.6747,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.4215613382899628,
|
| 1679 |
+
"grad_norm": 0.9945991635322571,
|
| 1680 |
+
"learning_rate": 0.0002,
|
| 1681 |
+
"loss": 0.6719,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.4275092936802973,
|
| 1686 |
+
"grad_norm": 1.0272929668426514,
|
| 1687 |
+
"learning_rate": 0.0002,
|
| 1688 |
+
"loss": 0.7015,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.4334572490706319,
|
| 1693 |
+
"grad_norm": 1.2321354150772095,
|
| 1694 |
+
"learning_rate": 0.0002,
|
| 1695 |
+
"loss": 0.6904,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.4394052044609666,
|
| 1700 |
+
"grad_norm": 1.1331416368484497,
|
| 1701 |
+
"learning_rate": 0.0002,
|
| 1702 |
+
"loss": 0.6444,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.4453531598513012,
|
| 1707 |
+
"grad_norm": 1.0527664422988892,
|
| 1708 |
+
"learning_rate": 0.0002,
|
| 1709 |
+
"loss": 0.6135,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.4513011152416357,
|
| 1714 |
+
"grad_norm": 1.0586967468261719,
|
| 1715 |
+
"learning_rate": 0.0002,
|
| 1716 |
+
"loss": 0.6705,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.4572490706319703,
|
| 1721 |
+
"grad_norm": 1.0302836894989014,
|
| 1722 |
+
"learning_rate": 0.0002,
|
| 1723 |
+
"loss": 0.6757,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.4631970260223048,
|
| 1728 |
+
"grad_norm": 0.9323686957359314,
|
| 1729 |
+
"learning_rate": 0.0002,
|
| 1730 |
+
"loss": 0.7458,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.4691449814126394,
|
| 1735 |
+
"grad_norm": 1.103028416633606,
|
| 1736 |
+
"learning_rate": 0.0002,
|
| 1737 |
+
"loss": 0.5832,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.475092936802974,
|
| 1742 |
+
"grad_norm": 1.1638356447219849,
|
| 1743 |
+
"learning_rate": 0.0002,
|
| 1744 |
+
"loss": 0.6286,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.4810408921933085,
|
| 1749 |
+
"grad_norm": 1.0685887336730957,
|
| 1750 |
+
"learning_rate": 0.0002,
|
| 1751 |
+
"loss": 0.6255,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.486988847583643,
|
| 1756 |
+
"grad_norm": 0.9854826927185059,
|
| 1757 |
+
"learning_rate": 0.0002,
|
| 1758 |
+
"loss": 0.7764,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.4929368029739778,
|
| 1763 |
+
"grad_norm": 1.1790441274642944,
|
| 1764 |
+
"learning_rate": 0.0002,
|
| 1765 |
+
"loss": 0.5791,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.4988847583643123,
|
| 1770 |
+
"grad_norm": 0.9097880721092224,
|
| 1771 |
+
"learning_rate": 0.0002,
|
| 1772 |
+
"loss": 0.5517,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.504832713754647,
|
| 1777 |
+
"grad_norm": 1.1351274251937866,
|
| 1778 |
+
"learning_rate": 0.0002,
|
| 1779 |
+
"loss": 0.6725,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.5107806691449814,
|
| 1784 |
+
"grad_norm": 1.2710050344467163,
|
| 1785 |
+
"learning_rate": 0.0002,
|
| 1786 |
+
"loss": 0.7106,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.516728624535316,
|
| 1791 |
+
"grad_norm": 1.2035406827926636,
|
| 1792 |
+
"learning_rate": 0.0002,
|
| 1793 |
+
"loss": 0.623,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.5226765799256505,
|
| 1798 |
+
"grad_norm": 1.100200891494751,
|
| 1799 |
+
"learning_rate": 0.0002,
|
| 1800 |
+
"loss": 0.6418,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.528624535315985,
|
| 1805 |
+
"grad_norm": 1.3976622819900513,
|
| 1806 |
+
"learning_rate": 0.0002,
|
| 1807 |
+
"loss": 0.7007,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.5345724907063198,
|
| 1812 |
+
"grad_norm": 1.2113823890686035,
|
| 1813 |
+
"learning_rate": 0.0002,
|
| 1814 |
+
"loss": 0.7271,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.5405204460966542,
|
| 1819 |
+
"grad_norm": 1.1983304023742676,
|
| 1820 |
+
"learning_rate": 0.0002,
|
| 1821 |
+
"loss": 0.6937,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.546468401486989,
|
| 1826 |
+
"grad_norm": 1.2594386339187622,
|
| 1827 |
+
"learning_rate": 0.0002,
|
| 1828 |
+
"loss": 0.708,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.5524163568773233,
|
| 1833 |
+
"grad_norm": 1.1495513916015625,
|
| 1834 |
+
"learning_rate": 0.0002,
|
| 1835 |
+
"loss": 0.621,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.558364312267658,
|
| 1840 |
+
"grad_norm": 1.0474885702133179,
|
| 1841 |
+
"learning_rate": 0.0002,
|
| 1842 |
+
"loss": 0.5433,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.5643122676579926,
|
| 1847 |
+
"grad_norm": 1.1138205528259277,
|
| 1848 |
+
"learning_rate": 0.0002,
|
| 1849 |
+
"loss": 0.6886,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.5702602230483271,
|
| 1854 |
+
"grad_norm": 0.9678700566291809,
|
| 1855 |
+
"learning_rate": 0.0002,
|
| 1856 |
+
"loss": 0.5945,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.5762081784386617,
|
| 1861 |
+
"grad_norm": 0.928419828414917,
|
| 1862 |
+
"learning_rate": 0.0002,
|
| 1863 |
+
"loss": 0.7483,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.5821561338289962,
|
| 1868 |
+
"grad_norm": 0.8806396126747131,
|
| 1869 |
+
"learning_rate": 0.0002,
|
| 1870 |
+
"loss": 0.6022,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.588104089219331,
|
| 1875 |
+
"grad_norm": 0.9389284253120422,
|
| 1876 |
+
"learning_rate": 0.0002,
|
| 1877 |
+
"loss": 0.6638,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.5940520446096653,
|
| 1882 |
+
"grad_norm": 1.0797287225723267,
|
| 1883 |
+
"learning_rate": 0.0002,
|
| 1884 |
+
"loss": 0.6298,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.6,
|
| 1889 |
+
"grad_norm": 0.9545785784721375,
|
| 1890 |
+
"learning_rate": 0.0002,
|
| 1891 |
+
"loss": 0.6335,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.6059479553903344,
|
| 1896 |
+
"grad_norm": 0.9800273776054382,
|
| 1897 |
+
"learning_rate": 0.0002,
|
| 1898 |
+
"loss": 0.5772,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.6118959107806692,
|
| 1903 |
+
"grad_norm": 1.3683196306228638,
|
| 1904 |
+
"learning_rate": 0.0002,
|
| 1905 |
+
"loss": 0.6181,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.6178438661710037,
|
| 1910 |
+
"grad_norm": 1.1559855937957764,
|
| 1911 |
+
"learning_rate": 0.0002,
|
| 1912 |
+
"loss": 0.62,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.6237918215613383,
|
| 1917 |
+
"grad_norm": 1.1240603923797607,
|
| 1918 |
+
"learning_rate": 0.0002,
|
| 1919 |
+
"loss": 0.682,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.6297397769516728,
|
| 1924 |
+
"grad_norm": 0.9673051834106445,
|
| 1925 |
+
"learning_rate": 0.0002,
|
| 1926 |
+
"loss": 0.6835,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.6356877323420074,
|
| 1931 |
+
"grad_norm": 1.1218955516815186,
|
| 1932 |
+
"learning_rate": 0.0002,
|
| 1933 |
+
"loss": 0.6591,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.6416356877323421,
|
| 1938 |
+
"grad_norm": 1.2360399961471558,
|
| 1939 |
+
"learning_rate": 0.0002,
|
| 1940 |
+
"loss": 0.7957,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.6475836431226765,
|
| 1945 |
+
"grad_norm": 1.2180172204971313,
|
| 1946 |
+
"learning_rate": 0.0002,
|
| 1947 |
+
"loss": 0.6951,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.6535315985130112,
|
| 1952 |
+
"grad_norm": 1.2104121446609497,
|
| 1953 |
+
"learning_rate": 0.0002,
|
| 1954 |
+
"loss": 0.6549,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.6594795539033456,
|
| 1959 |
+
"grad_norm": 0.9836241006851196,
|
| 1960 |
+
"learning_rate": 0.0002,
|
| 1961 |
+
"loss": 0.6077,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.6654275092936803,
|
| 1966 |
+
"grad_norm": 0.8980191349983215,
|
| 1967 |
+
"learning_rate": 0.0002,
|
| 1968 |
+
"loss": 0.7533,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.6713754646840149,
|
| 1973 |
+
"grad_norm": 1.056117296218872,
|
| 1974 |
+
"learning_rate": 0.0002,
|
| 1975 |
+
"loss": 0.6283,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.6773234200743494,
|
| 1980 |
+
"grad_norm": 1.0315310955047607,
|
| 1981 |
+
"learning_rate": 0.0002,
|
| 1982 |
+
"loss": 0.7099,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.683271375464684,
|
| 1987 |
+
"grad_norm": 1.1293710470199585,
|
| 1988 |
+
"learning_rate": 0.0002,
|
| 1989 |
+
"loss": 0.7628,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.6892193308550185,
|
| 1994 |
+
"grad_norm": 0.8841990232467651,
|
| 1995 |
+
"learning_rate": 0.0002,
|
| 1996 |
+
"loss": 0.6076,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.6951672862453533,
|
| 2001 |
+
"grad_norm": 1.0221779346466064,
|
| 2002 |
+
"learning_rate": 0.0002,
|
| 2003 |
+
"loss": 0.6003,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.7011152416356876,
|
| 2008 |
+
"grad_norm": 0.9923282861709595,
|
| 2009 |
+
"learning_rate": 0.0002,
|
| 2010 |
+
"loss": 0.5743,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.7070631970260224,
|
| 2015 |
+
"grad_norm": 1.1585432291030884,
|
| 2016 |
+
"learning_rate": 0.0002,
|
| 2017 |
+
"loss": 0.6121,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.713011152416357,
|
| 2022 |
+
"grad_norm": 0.9201356172561646,
|
| 2023 |
+
"learning_rate": 0.0002,
|
| 2024 |
+
"loss": 0.6343,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.7189591078066915,
|
| 2029 |
+
"grad_norm": 1.164581298828125,
|
| 2030 |
+
"learning_rate": 0.0002,
|
| 2031 |
+
"loss": 0.615,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.724907063197026,
|
| 2036 |
+
"grad_norm": 0.9991989135742188,
|
| 2037 |
+
"learning_rate": 0.0002,
|
| 2038 |
+
"loss": 0.62,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.7308550185873606,
|
| 2043 |
+
"grad_norm": 1.0976234674453735,
|
| 2044 |
+
"learning_rate": 0.0002,
|
| 2045 |
+
"loss": 0.687,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.7368029739776951,
|
| 2050 |
+
"grad_norm": 1.1581001281738281,
|
| 2051 |
+
"learning_rate": 0.0002,
|
| 2052 |
+
"loss": 0.6227,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.7427509293680297,
|
| 2057 |
+
"grad_norm": 1.0079922676086426,
|
| 2058 |
+
"learning_rate": 0.0002,
|
| 2059 |
+
"loss": 0.675,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.7486988847583644,
|
| 2064 |
+
"grad_norm": 1.0962276458740234,
|
| 2065 |
+
"learning_rate": 0.0002,
|
| 2066 |
+
"loss": 0.6694,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.7546468401486988,
|
| 2071 |
+
"grad_norm": 1.0988850593566895,
|
| 2072 |
+
"learning_rate": 0.0002,
|
| 2073 |
+
"loss": 0.6114,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.7605947955390335,
|
| 2078 |
+
"grad_norm": 1.4446635246276855,
|
| 2079 |
+
"learning_rate": 0.0002,
|
| 2080 |
+
"loss": 0.5885,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.766542750929368,
|
| 2085 |
+
"grad_norm": 1.2141138315200806,
|
| 2086 |
+
"learning_rate": 0.0002,
|
| 2087 |
+
"loss": 0.7856,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.7724907063197026,
|
| 2092 |
+
"grad_norm": 1.1908177137374878,
|
| 2093 |
+
"learning_rate": 0.0002,
|
| 2094 |
+
"loss": 0.7033,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.7784386617100372,
|
| 2099 |
+
"grad_norm": 1.019839882850647,
|
| 2100 |
+
"learning_rate": 0.0002,
|
| 2101 |
+
"loss": 0.6763,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.7843866171003717,
|
| 2106 |
+
"grad_norm": 1.039696216583252,
|
| 2107 |
+
"learning_rate": 0.0002,
|
| 2108 |
+
"loss": 0.6279,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.7903345724907063,
|
| 2113 |
+
"grad_norm": 0.974805474281311,
|
| 2114 |
+
"learning_rate": 0.0002,
|
| 2115 |
+
"loss": 0.5777,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.7962825278810408,
|
| 2120 |
+
"grad_norm": 1.1052793264389038,
|
| 2121 |
+
"learning_rate": 0.0002,
|
| 2122 |
+
"loss": 0.7294,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.8022304832713756,
|
| 2127 |
+
"grad_norm": 1.4657918214797974,
|
| 2128 |
+
"learning_rate": 0.0002,
|
| 2129 |
+
"loss": 0.5142,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.80817843866171,
|
| 2134 |
+
"grad_norm": 1.0391294956207275,
|
| 2135 |
+
"learning_rate": 0.0002,
|
| 2136 |
+
"loss": 0.6388,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.8141263940520447,
|
| 2141 |
+
"grad_norm": 1.0521687269210815,
|
| 2142 |
+
"learning_rate": 0.0002,
|
| 2143 |
+
"loss": 0.6944,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.8200743494423792,
|
| 2148 |
+
"grad_norm": 1.0755914449691772,
|
| 2149 |
+
"learning_rate": 0.0002,
|
| 2150 |
+
"loss": 0.6974,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.8260223048327138,
|
| 2155 |
+
"grad_norm": 1.128304123878479,
|
| 2156 |
+
"learning_rate": 0.0002,
|
| 2157 |
+
"loss": 0.7687,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.8319702602230483,
|
| 2162 |
+
"grad_norm": 1.0178970098495483,
|
| 2163 |
+
"learning_rate": 0.0002,
|
| 2164 |
+
"loss": 0.6074,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.8379182156133829,
|
| 2169 |
+
"grad_norm": 0.9115421772003174,
|
| 2170 |
+
"learning_rate": 0.0002,
|
| 2171 |
+
"loss": 0.725,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.8438661710037176,
|
| 2176 |
+
"grad_norm": 1.0200258493423462,
|
| 2177 |
+
"learning_rate": 0.0002,
|
| 2178 |
+
"loss": 0.6866,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.849814126394052,
|
| 2183 |
+
"grad_norm": 1.286431908607483,
|
| 2184 |
+
"learning_rate": 0.0002,
|
| 2185 |
+
"loss": 0.6618,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.8557620817843867,
|
| 2190 |
+
"grad_norm": 1.0576943159103394,
|
| 2191 |
+
"learning_rate": 0.0002,
|
| 2192 |
+
"loss": 0.6217,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.861710037174721,
|
| 2197 |
+
"grad_norm": 0.9450961351394653,
|
| 2198 |
+
"learning_rate": 0.0002,
|
| 2199 |
+
"loss": 0.7285,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.8676579925650558,
|
| 2204 |
+
"grad_norm": 1.2659786939620972,
|
| 2205 |
+
"learning_rate": 0.0002,
|
| 2206 |
+
"loss": 0.6004,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.8736059479553904,
|
| 2211 |
+
"grad_norm": 1.0950329303741455,
|
| 2212 |
+
"learning_rate": 0.0002,
|
| 2213 |
+
"loss": 0.6792,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.879553903345725,
|
| 2218 |
+
"grad_norm": 1.0434305667877197,
|
| 2219 |
+
"learning_rate": 0.0002,
|
| 2220 |
+
"loss": 0.6651,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.8855018587360595,
|
| 2225 |
+
"grad_norm": 2.390085458755493,
|
| 2226 |
+
"learning_rate": 0.0002,
|
| 2227 |
+
"loss": 0.6394,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.891449814126394,
|
| 2232 |
+
"grad_norm": 1.1428786516189575,
|
| 2233 |
+
"learning_rate": 0.0002,
|
| 2234 |
+
"loss": 0.6519,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.8973977695167288,
|
| 2239 |
+
"grad_norm": 1.1516354084014893,
|
| 2240 |
+
"learning_rate": 0.0002,
|
| 2241 |
+
"loss": 0.5967,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.903345724907063,
|
| 2246 |
+
"grad_norm": 0.9553952813148499,
|
| 2247 |
+
"learning_rate": 0.0002,
|
| 2248 |
+
"loss": 0.6626,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.9092936802973979,
|
| 2253 |
+
"grad_norm": 1.1295243501663208,
|
| 2254 |
+
"learning_rate": 0.0002,
|
| 2255 |
+
"loss": 0.6151,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.9152416356877322,
|
| 2260 |
+
"grad_norm": 1.1593585014343262,
|
| 2261 |
+
"learning_rate": 0.0002,
|
| 2262 |
+
"loss": 0.6412,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.921189591078067,
|
| 2267 |
+
"grad_norm": 1.830063819885254,
|
| 2268 |
+
"learning_rate": 0.0002,
|
| 2269 |
+
"loss": 0.6746,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.9271375464684015,
|
| 2274 |
+
"grad_norm": 1.3906419277191162,
|
| 2275 |
+
"learning_rate": 0.0002,
|
| 2276 |
+
"loss": 0.6351,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.933085501858736,
|
| 2281 |
+
"grad_norm": 1.4869827032089233,
|
| 2282 |
+
"learning_rate": 0.0002,
|
| 2283 |
+
"loss": 0.6806,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.9390334572490706,
|
| 2288 |
+
"grad_norm": 1.110323429107666,
|
| 2289 |
+
"learning_rate": 0.0002,
|
| 2290 |
+
"loss": 0.5748,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.9449814126394052,
|
| 2295 |
+
"grad_norm": 1.4225271940231323,
|
| 2296 |
+
"learning_rate": 0.0002,
|
| 2297 |
+
"loss": 0.6572,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.95092936802974,
|
| 2302 |
+
"grad_norm": 1.0343074798583984,
|
| 2303 |
+
"learning_rate": 0.0002,
|
| 2304 |
+
"loss": 0.5376,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.9568773234200743,
|
| 2309 |
+
"grad_norm": 0.9949336647987366,
|
| 2310 |
+
"learning_rate": 0.0002,
|
| 2311 |
+
"loss": 0.69,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.962825278810409,
|
| 2316 |
+
"grad_norm": 1.0380656719207764,
|
| 2317 |
+
"learning_rate": 0.0002,
|
| 2318 |
+
"loss": 0.63,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.9687732342007433,
|
| 2323 |
+
"grad_norm": 1.390371561050415,
|
| 2324 |
+
"learning_rate": 0.0002,
|
| 2325 |
+
"loss": 0.7011,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.9747211895910781,
|
| 2330 |
+
"grad_norm": 1.4939589500427246,
|
| 2331 |
+
"learning_rate": 0.0002,
|
| 2332 |
+
"loss": 0.5689,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.9806691449814127,
|
| 2337 |
+
"grad_norm": 1.2601418495178223,
|
| 2338 |
+
"learning_rate": 0.0002,
|
| 2339 |
+
"loss": 0.5838,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 1.9866171003717472,
|
| 2344 |
+
"grad_norm": 2.679206132888794,
|
| 2345 |
+
"learning_rate": 0.0002,
|
| 2346 |
+
"loss": 0.7981,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 1.9925650557620818,
|
| 2351 |
+
"grad_norm": 1.1042869091033936,
|
| 2352 |
+
"learning_rate": 0.0002,
|
| 2353 |
+
"loss": 0.6212,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.9985130111524163,
|
| 2358 |
+
"grad_norm": 1.4491620063781738,
|
| 2359 |
+
"learning_rate": 0.0002,
|
| 2360 |
+
"loss": 0.6194,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.004460966542751,
|
| 2365 |
+
"grad_norm": 0.9622808694839478,
|
| 2366 |
+
"learning_rate": 0.0002,
|
| 2367 |
+
"loss": 0.6566,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.0104089219330854,
|
| 2372 |
+
"grad_norm": 1.5044182538986206,
|
| 2373 |
+
"learning_rate": 0.0002,
|
| 2374 |
+
"loss": 0.6093,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.01635687732342,
|
| 2379 |
+
"grad_norm": 1.699040174484253,
|
| 2380 |
+
"learning_rate": 0.0002,
|
| 2381 |
+
"loss": 0.5237,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.0223048327137545,
|
| 2386 |
+
"grad_norm": 1.1767878532409668,
|
| 2387 |
+
"learning_rate": 0.0002,
|
| 2388 |
+
"loss": 0.6145,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.0282527881040893,
|
| 2393 |
+
"grad_norm": 1.2151747941970825,
|
| 2394 |
+
"learning_rate": 0.0002,
|
| 2395 |
+
"loss": 0.6171,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.0342007434944236,
|
| 2400 |
+
"grad_norm": 1.2429864406585693,
|
| 2401 |
+
"learning_rate": 0.0002,
|
| 2402 |
+
"loss": 0.6489,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.0401486988847584,
|
| 2407 |
+
"grad_norm": 1.164552092552185,
|
| 2408 |
+
"learning_rate": 0.0002,
|
| 2409 |
+
"loss": 0.509,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.046096654275093,
|
| 2414 |
+
"grad_norm": 1.1822024583816528,
|
| 2415 |
+
"learning_rate": 0.0002,
|
| 2416 |
+
"loss": 0.6568,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.0520446096654275,
|
| 2421 |
+
"grad_norm": 7.130686283111572,
|
| 2422 |
+
"learning_rate": 0.0002,
|
| 2423 |
+
"loss": 0.5688,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.0579925650557622,
|
| 2428 |
+
"grad_norm": 1.1000553369522095,
|
| 2429 |
+
"learning_rate": 0.0002,
|
| 2430 |
+
"loss": 0.5202,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.0639405204460965,
|
| 2435 |
+
"grad_norm": 1.0652920007705688,
|
| 2436 |
+
"learning_rate": 0.0002,
|
| 2437 |
+
"loss": 0.5797,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.0698884758364313,
|
| 2442 |
+
"grad_norm": 1.3442667722702026,
|
| 2443 |
+
"learning_rate": 0.0002,
|
| 2444 |
+
"loss": 0.6164,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.0758364312267656,
|
| 2449 |
+
"grad_norm": 1.1383881568908691,
|
| 2450 |
+
"learning_rate": 0.0002,
|
| 2451 |
+
"loss": 0.5462,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.0817843866171004,
|
| 2456 |
+
"grad_norm": 0.9077207446098328,
|
| 2457 |
+
"learning_rate": 0.0002,
|
| 2458 |
+
"loss": 0.4876,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.0877323420074347,
|
| 2463 |
+
"grad_norm": 1.0893263816833496,
|
| 2464 |
+
"learning_rate": 0.0002,
|
| 2465 |
+
"loss": 0.6335,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.0936802973977695,
|
| 2470 |
+
"grad_norm": 0.9917628169059753,
|
| 2471 |
+
"learning_rate": 0.0002,
|
| 2472 |
+
"loss": 0.5745,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.0996282527881043,
|
| 2477 |
+
"grad_norm": 1.0131194591522217,
|
| 2478 |
+
"learning_rate": 0.0002,
|
| 2479 |
+
"loss": 0.5427,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.1055762081784386,
|
| 2484 |
+
"grad_norm": 2.081542730331421,
|
| 2485 |
+
"learning_rate": 0.0002,
|
| 2486 |
+
"loss": 0.5692,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.1115241635687734,
|
| 2491 |
+
"grad_norm": 1.659941554069519,
|
| 2492 |
+
"learning_rate": 0.0002,
|
| 2493 |
+
"loss": 0.4985,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.1174721189591077,
|
| 2498 |
+
"grad_norm": 1.309465765953064,
|
| 2499 |
+
"learning_rate": 0.0002,
|
| 2500 |
+
"loss": 0.5439,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.1234200743494425,
|
| 2505 |
+
"grad_norm": 1.2212107181549072,
|
| 2506 |
+
"learning_rate": 0.0002,
|
| 2507 |
+
"loss": 0.5133,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.129368029739777,
|
| 2512 |
+
"grad_norm": 1.0576850175857544,
|
| 2513 |
+
"learning_rate": 0.0002,
|
| 2514 |
+
"loss": 0.6627,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.1353159851301116,
|
| 2519 |
+
"grad_norm": 1.2587406635284424,
|
| 2520 |
+
"learning_rate": 0.0002,
|
| 2521 |
+
"loss": 0.5235,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.141263940520446,
|
| 2526 |
+
"grad_norm": 1.05579674243927,
|
| 2527 |
+
"learning_rate": 0.0002,
|
| 2528 |
+
"loss": 0.5994,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.1472118959107807,
|
| 2533 |
+
"grad_norm": 1.2232978343963623,
|
| 2534 |
+
"learning_rate": 0.0002,
|
| 2535 |
+
"loss": 0.586,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.1531598513011154,
|
| 2540 |
+
"grad_norm": 1.0725406408309937,
|
| 2541 |
+
"learning_rate": 0.0002,
|
| 2542 |
+
"loss": 0.4801,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.1591078066914497,
|
| 2547 |
+
"grad_norm": 1.0593106746673584,
|
| 2548 |
+
"learning_rate": 0.0002,
|
| 2549 |
+
"loss": 0.4877,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.1650557620817845,
|
| 2554 |
+
"grad_norm": 2.2175445556640625,
|
| 2555 |
+
"learning_rate": 0.0002,
|
| 2556 |
+
"loss": 0.57,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.171003717472119,
|
| 2561 |
+
"grad_norm": 1.1013628244400024,
|
| 2562 |
+
"learning_rate": 0.0002,
|
| 2563 |
+
"loss": 0.5797,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 2.1769516728624536,
|
| 2568 |
+
"grad_norm": 1.186463713645935,
|
| 2569 |
+
"learning_rate": 0.0002,
|
| 2570 |
+
"loss": 0.5938,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.182899628252788,
|
| 2575 |
+
"grad_norm": 1.1608301401138306,
|
| 2576 |
+
"learning_rate": 0.0002,
|
| 2577 |
+
"loss": 0.4446,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 2.1888475836431227,
|
| 2582 |
+
"grad_norm": 1.2304465770721436,
|
| 2583 |
+
"learning_rate": 0.0002,
|
| 2584 |
+
"loss": 0.5154,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.194795539033457,
|
| 2589 |
+
"grad_norm": 1.1233623027801514,
|
| 2590 |
+
"learning_rate": 0.0002,
|
| 2591 |
+
"loss": 0.6199,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.200743494423792,
|
| 2596 |
+
"grad_norm": 1.2339355945587158,
|
| 2597 |
+
"learning_rate": 0.0002,
|
| 2598 |
+
"loss": 0.584,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 2.2066914498141266,
|
| 2603 |
+
"grad_norm": 1.6794264316558838,
|
| 2604 |
+
"learning_rate": 0.0002,
|
| 2605 |
+
"loss": 0.5609,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.212639405204461,
|
| 2610 |
+
"grad_norm": 1.1440285444259644,
|
| 2611 |
+
"learning_rate": 0.0002,
|
| 2612 |
+
"loss": 0.5306,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.2185873605947957,
|
| 2617 |
+
"grad_norm": 1.6425179243087769,
|
| 2618 |
+
"learning_rate": 0.0002,
|
| 2619 |
+
"loss": 0.6556,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 2.22453531598513,
|
| 2624 |
+
"grad_norm": 1.1068412065505981,
|
| 2625 |
+
"learning_rate": 0.0002,
|
| 2626 |
+
"loss": 0.6442,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 2.2304832713754648,
|
| 2631 |
+
"grad_norm": 1.1996163129806519,
|
| 2632 |
+
"learning_rate": 0.0002,
|
| 2633 |
+
"loss": 0.51,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.236431226765799,
|
| 2638 |
+
"grad_norm": 1.193741798400879,
|
| 2639 |
+
"learning_rate": 0.0002,
|
| 2640 |
+
"loss": 0.5323,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 2.242379182156134,
|
| 2645 |
+
"grad_norm": 1.3267923593521118,
|
| 2646 |
+
"learning_rate": 0.0002,
|
| 2647 |
+
"loss": 0.5576,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.248327137546468,
|
| 2652 |
+
"grad_norm": 1.1256170272827148,
|
| 2653 |
+
"learning_rate": 0.0002,
|
| 2654 |
+
"loss": 0.6218,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.254275092936803,
|
| 2659 |
+
"grad_norm": 1.1282093524932861,
|
| 2660 |
+
"learning_rate": 0.0002,
|
| 2661 |
+
"loss": 0.5095,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 2.2602230483271377,
|
| 2666 |
+
"grad_norm": 1.156480073928833,
|
| 2667 |
+
"learning_rate": 0.0002,
|
| 2668 |
+
"loss": 0.5423,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 2.266171003717472,
|
| 2673 |
+
"grad_norm": 1.2083227634429932,
|
| 2674 |
+
"learning_rate": 0.0002,
|
| 2675 |
+
"loss": 0.5714,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 2.272118959107807,
|
| 2680 |
+
"grad_norm": 1.5389211177825928,
|
| 2681 |
+
"learning_rate": 0.0002,
|
| 2682 |
+
"loss": 0.4797,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 2.278066914498141,
|
| 2687 |
+
"grad_norm": 1.1418561935424805,
|
| 2688 |
+
"learning_rate": 0.0002,
|
| 2689 |
+
"loss": 0.585,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.284014869888476,
|
| 2694 |
+
"grad_norm": 1.0242007970809937,
|
| 2695 |
+
"learning_rate": 0.0002,
|
| 2696 |
+
"loss": 0.562,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.2899628252788102,
|
| 2701 |
+
"grad_norm": 1.4750384092330933,
|
| 2702 |
+
"learning_rate": 0.0002,
|
| 2703 |
+
"loss": 0.558,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 2.295910780669145,
|
| 2708 |
+
"grad_norm": 1.176080346107483,
|
| 2709 |
+
"learning_rate": 0.0002,
|
| 2710 |
+
"loss": 0.4977,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.3018587360594793,
|
| 2715 |
+
"grad_norm": 1.1733489036560059,
|
| 2716 |
+
"learning_rate": 0.0002,
|
| 2717 |
+
"loss": 0.5984,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 2.307806691449814,
|
| 2722 |
+
"grad_norm": 1.0431591272354126,
|
| 2723 |
+
"learning_rate": 0.0002,
|
| 2724 |
+
"loss": 0.5805,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 2.313754646840149,
|
| 2729 |
+
"grad_norm": 1.1595654487609863,
|
| 2730 |
+
"learning_rate": 0.0002,
|
| 2731 |
+
"loss": 0.5639,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 2.319702602230483,
|
| 2736 |
+
"grad_norm": 1.2077865600585938,
|
| 2737 |
+
"learning_rate": 0.0002,
|
| 2738 |
+
"loss": 0.5021,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.325650557620818,
|
| 2743 |
+
"grad_norm": 1.4747991561889648,
|
| 2744 |
+
"learning_rate": 0.0002,
|
| 2745 |
+
"loss": 0.5809,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 2.3315985130111523,
|
| 2750 |
+
"grad_norm": 1.0486669540405273,
|
| 2751 |
+
"learning_rate": 0.0002,
|
| 2752 |
+
"loss": 0.6334,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.337546468401487,
|
| 2757 |
+
"grad_norm": 1.8145817518234253,
|
| 2758 |
+
"learning_rate": 0.0002,
|
| 2759 |
+
"loss": 0.5223,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 2.3434944237918214,
|
| 2764 |
+
"grad_norm": 1.393776535987854,
|
| 2765 |
+
"learning_rate": 0.0002,
|
| 2766 |
+
"loss": 0.4034,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 2.349442379182156,
|
| 2771 |
+
"grad_norm": 1.208957314491272,
|
| 2772 |
+
"learning_rate": 0.0002,
|
| 2773 |
+
"loss": 0.5929,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.3553903345724905,
|
| 2778 |
+
"grad_norm": 1.3021739721298218,
|
| 2779 |
+
"learning_rate": 0.0002,
|
| 2780 |
+
"loss": 0.6176,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 2.3613382899628252,
|
| 2785 |
+
"grad_norm": 1.4258112907409668,
|
| 2786 |
+
"learning_rate": 0.0002,
|
| 2787 |
+
"loss": 0.6765,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 2.36728624535316,
|
| 2792 |
+
"grad_norm": 1.2623789310455322,
|
| 2793 |
+
"learning_rate": 0.0002,
|
| 2794 |
+
"loss": 0.5973,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 2.3732342007434943,
|
| 2799 |
+
"grad_norm": 1.1510920524597168,
|
| 2800 |
+
"learning_rate": 0.0002,
|
| 2801 |
+
"loss": 0.6167,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 2.379182156133829,
|
| 2806 |
+
"grad_norm": 1.0992542505264282,
|
| 2807 |
+
"learning_rate": 0.0002,
|
| 2808 |
+
"loss": 0.4328,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 2.3851301115241634,
|
| 2813 |
+
"grad_norm": 1.3745630979537964,
|
| 2814 |
+
"learning_rate": 0.0002,
|
| 2815 |
+
"loss": 0.6485,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.391078066914498,
|
| 2820 |
+
"grad_norm": 1.2044932842254639,
|
| 2821 |
+
"learning_rate": 0.0002,
|
| 2822 |
+
"loss": 0.6345,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 2.3970260223048325,
|
| 2827 |
+
"grad_norm": 1.4290401935577393,
|
| 2828 |
+
"learning_rate": 0.0002,
|
| 2829 |
+
"loss": 0.5706,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 2.4029739776951673,
|
| 2834 |
+
"grad_norm": 1.3580994606018066,
|
| 2835 |
+
"learning_rate": 0.0002,
|
| 2836 |
+
"loss": 0.5373,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.4089219330855016,
|
| 2841 |
+
"grad_norm": 1.0747463703155518,
|
| 2842 |
+
"learning_rate": 0.0002,
|
| 2843 |
+
"loss": 0.5597,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 2.4148698884758364,
|
| 2848 |
+
"grad_norm": 1.288228154182434,
|
| 2849 |
+
"learning_rate": 0.0002,
|
| 2850 |
+
"loss": 0.6366,
|
| 2851 |
+
"step": 406
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 2.420817843866171,
|
| 2855 |
+
"grad_norm": 1.2379798889160156,
|
| 2856 |
+
"learning_rate": 0.0002,
|
| 2857 |
+
"loss": 0.474,
|
| 2858 |
+
"step": 407
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 2.4267657992565055,
|
| 2862 |
+
"grad_norm": 1.234220266342163,
|
| 2863 |
+
"learning_rate": 0.0002,
|
| 2864 |
+
"loss": 0.5363,
|
| 2865 |
+
"step": 408
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 2.4327137546468403,
|
| 2869 |
+
"grad_norm": 1.2338114976882935,
|
| 2870 |
+
"learning_rate": 0.0002,
|
| 2871 |
+
"loss": 0.4992,
|
| 2872 |
+
"step": 409
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 2.4386617100371746,
|
| 2876 |
+
"grad_norm": 1.3846346139907837,
|
| 2877 |
+
"learning_rate": 0.0002,
|
| 2878 |
+
"loss": 0.6412,
|
| 2879 |
+
"step": 410
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.4446096654275093,
|
| 2883 |
+
"grad_norm": 1.2423279285430908,
|
| 2884 |
+
"learning_rate": 0.0002,
|
| 2885 |
+
"loss": 0.528,
|
| 2886 |
+
"step": 411
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 2.4505576208178437,
|
| 2890 |
+
"grad_norm": 1.235088586807251,
|
| 2891 |
+
"learning_rate": 0.0002,
|
| 2892 |
+
"loss": 0.5763,
|
| 2893 |
+
"step": 412
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 2.4565055762081784,
|
| 2897 |
+
"grad_norm": 1.3832026720046997,
|
| 2898 |
+
"learning_rate": 0.0002,
|
| 2899 |
+
"loss": 0.6355,
|
| 2900 |
+
"step": 413
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 2.4624535315985128,
|
| 2904 |
+
"grad_norm": 1.214076280593872,
|
| 2905 |
+
"learning_rate": 0.0002,
|
| 2906 |
+
"loss": 0.6345,
|
| 2907 |
+
"step": 414
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 2.4684014869888475,
|
| 2911 |
+
"grad_norm": 1.463728904724121,
|
| 2912 |
+
"learning_rate": 0.0002,
|
| 2913 |
+
"loss": 0.6186,
|
| 2914 |
+
"step": 415
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 2.4743494423791823,
|
| 2918 |
+
"grad_norm": 1.0485203266143799,
|
| 2919 |
+
"learning_rate": 0.0002,
|
| 2920 |
+
"loss": 0.4723,
|
| 2921 |
+
"step": 416
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 2.4802973977695166,
|
| 2925 |
+
"grad_norm": 1.5590802431106567,
|
| 2926 |
+
"learning_rate": 0.0002,
|
| 2927 |
+
"loss": 0.5688,
|
| 2928 |
+
"step": 417
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 2.4862453531598514,
|
| 2932 |
+
"grad_norm": 1.1843955516815186,
|
| 2933 |
+
"learning_rate": 0.0002,
|
| 2934 |
+
"loss": 0.5043,
|
| 2935 |
+
"step": 418
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 2.4921933085501857,
|
| 2939 |
+
"grad_norm": 1.579487919807434,
|
| 2940 |
+
"learning_rate": 0.0002,
|
| 2941 |
+
"loss": 0.5257,
|
| 2942 |
+
"step": 419
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 2.4981412639405205,
|
| 2946 |
+
"grad_norm": 1.2086743116378784,
|
| 2947 |
+
"learning_rate": 0.0002,
|
| 2948 |
+
"loss": 0.6798,
|
| 2949 |
+
"step": 420
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 2.5040892193308553,
|
| 2953 |
+
"grad_norm": 1.195058822631836,
|
| 2954 |
+
"learning_rate": 0.0002,
|
| 2955 |
+
"loss": 0.6485,
|
| 2956 |
+
"step": 421
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 2.5100371747211896,
|
| 2960 |
+
"grad_norm": 1.283530354499817,
|
| 2961 |
+
"learning_rate": 0.0002,
|
| 2962 |
+
"loss": 0.5873,
|
| 2963 |
+
"step": 422
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 2.515985130111524,
|
| 2967 |
+
"grad_norm": 1.2394181489944458,
|
| 2968 |
+
"learning_rate": 0.0002,
|
| 2969 |
+
"loss": 0.6391,
|
| 2970 |
+
"step": 423
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 2.5219330855018587,
|
| 2974 |
+
"grad_norm": 1.2939765453338623,
|
| 2975 |
+
"learning_rate": 0.0002,
|
| 2976 |
+
"loss": 0.6709,
|
| 2977 |
+
"step": 424
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 2.5278810408921935,
|
| 2981 |
+
"grad_norm": 1.1421490907669067,
|
| 2982 |
+
"learning_rate": 0.0002,
|
| 2983 |
+
"loss": 0.563,
|
| 2984 |
+
"step": 425
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 2.533828996282528,
|
| 2988 |
+
"grad_norm": 1.1861711740493774,
|
| 2989 |
+
"learning_rate": 0.0002,
|
| 2990 |
+
"loss": 0.5176,
|
| 2991 |
+
"step": 426
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 2.5397769516728625,
|
| 2995 |
+
"grad_norm": 1.3675614595413208,
|
| 2996 |
+
"learning_rate": 0.0002,
|
| 2997 |
+
"loss": 0.529,
|
| 2998 |
+
"step": 427
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 2.545724907063197,
|
| 3002 |
+
"grad_norm": 1.5121257305145264,
|
| 3003 |
+
"learning_rate": 0.0002,
|
| 3004 |
+
"loss": 0.7019,
|
| 3005 |
+
"step": 428
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 2.5516728624535316,
|
| 3009 |
+
"grad_norm": 1.169859766960144,
|
| 3010 |
+
"learning_rate": 0.0002,
|
| 3011 |
+
"loss": 0.5985,
|
| 3012 |
+
"step": 429
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 2.5576208178438664,
|
| 3016 |
+
"grad_norm": 1.3540085554122925,
|
| 3017 |
+
"learning_rate": 0.0002,
|
| 3018 |
+
"loss": 0.5499,
|
| 3019 |
+
"step": 430
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 2.5635687732342007,
|
| 3023 |
+
"grad_norm": 1.0933575630187988,
|
| 3024 |
+
"learning_rate": 0.0002,
|
| 3025 |
+
"loss": 0.5208,
|
| 3026 |
+
"step": 431
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 2.569516728624535,
|
| 3030 |
+
"grad_norm": 1.1712511777877808,
|
| 3031 |
+
"learning_rate": 0.0002,
|
| 3032 |
+
"loss": 0.536,
|
| 3033 |
+
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 2.57546468401487,
|
| 3037 |
+
"grad_norm": 1.4346905946731567,
|
| 3038 |
+
"learning_rate": 0.0002,
|
| 3039 |
+
"loss": 0.6115,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 2.5814126394052046,
|
| 3044 |
+
"grad_norm": 2.1174967288970947,
|
| 3045 |
+
"learning_rate": 0.0002,
|
| 3046 |
+
"loss": 0.5341,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 2.587360594795539,
|
| 3051 |
+
"grad_norm": 1.3911654949188232,
|
| 3052 |
+
"learning_rate": 0.0002,
|
| 3053 |
+
"loss": 0.5707,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 2.5933085501858737,
|
| 3058 |
+
"grad_norm": 1.3274894952774048,
|
| 3059 |
+
"learning_rate": 0.0002,
|
| 3060 |
+
"loss": 0.6272,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 2.599256505576208,
|
| 3065 |
+
"grad_norm": 0.9820629358291626,
|
| 3066 |
+
"learning_rate": 0.0002,
|
| 3067 |
+
"loss": 0.4664,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 2.605204460966543,
|
| 3072 |
+
"grad_norm": 1.1450122594833374,
|
| 3073 |
+
"learning_rate": 0.0002,
|
| 3074 |
+
"loss": 0.4652,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 2.6111524163568776,
|
| 3079 |
+
"grad_norm": 2.7575511932373047,
|
| 3080 |
+
"learning_rate": 0.0002,
|
| 3081 |
+
"loss": 0.5652,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 2.617100371747212,
|
| 3086 |
+
"grad_norm": 1.4760148525238037,
|
| 3087 |
+
"learning_rate": 0.0002,
|
| 3088 |
+
"loss": 0.6331,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 2.623048327137546,
|
| 3093 |
+
"grad_norm": 1.2463843822479248,
|
| 3094 |
+
"learning_rate": 0.0002,
|
| 3095 |
+
"loss": 0.5721,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 2.628996282527881,
|
| 3100 |
+
"grad_norm": 1.2081701755523682,
|
| 3101 |
+
"learning_rate": 0.0002,
|
| 3102 |
+
"loss": 0.752,
|
| 3103 |
+
"step": 442
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 2.6349442379182157,
|
| 3107 |
+
"grad_norm": 1.1693692207336426,
|
| 3108 |
+
"learning_rate": 0.0002,
|
| 3109 |
+
"loss": 0.5505,
|
| 3110 |
+
"step": 443
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 2.64089219330855,
|
| 3114 |
+
"grad_norm": 1.3918544054031372,
|
| 3115 |
+
"learning_rate": 0.0002,
|
| 3116 |
+
"loss": 0.5387,
|
| 3117 |
+
"step": 444
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 2.646840148698885,
|
| 3121 |
+
"grad_norm": 1.3081449270248413,
|
| 3122 |
+
"learning_rate": 0.0002,
|
| 3123 |
+
"loss": 0.4156,
|
| 3124 |
+
"step": 445
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 2.652788104089219,
|
| 3128 |
+
"grad_norm": 1.1178191900253296,
|
| 3129 |
+
"learning_rate": 0.0002,
|
| 3130 |
+
"loss": 0.6099,
|
| 3131 |
+
"step": 446
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 2.658736059479554,
|
| 3135 |
+
"grad_norm": 1.172034740447998,
|
| 3136 |
+
"learning_rate": 0.0002,
|
| 3137 |
+
"loss": 0.6112,
|
| 3138 |
+
"step": 447
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 2.6646840148698887,
|
| 3142 |
+
"grad_norm": 1.3142459392547607,
|
| 3143 |
+
"learning_rate": 0.0002,
|
| 3144 |
+
"loss": 0.6123,
|
| 3145 |
+
"step": 448
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 2.670631970260223,
|
| 3149 |
+
"grad_norm": 1.2363723516464233,
|
| 3150 |
+
"learning_rate": 0.0002,
|
| 3151 |
+
"loss": 0.5472,
|
| 3152 |
+
"step": 449
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 2.6765799256505574,
|
| 3156 |
+
"grad_norm": 1.2881202697753906,
|
| 3157 |
+
"learning_rate": 0.0002,
|
| 3158 |
+
"loss": 0.4572,
|
| 3159 |
+
"step": 450
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 2.682527881040892,
|
| 3163 |
+
"grad_norm": 1.0761253833770752,
|
| 3164 |
+
"learning_rate": 0.0002,
|
| 3165 |
+
"loss": 0.5366,
|
| 3166 |
+
"step": 451
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 2.688475836431227,
|
| 3170 |
+
"grad_norm": 1.0405654907226562,
|
| 3171 |
+
"learning_rate": 0.0002,
|
| 3172 |
+
"loss": 0.6858,
|
| 3173 |
+
"step": 452
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 2.694423791821561,
|
| 3177 |
+
"grad_norm": 1.3384194374084473,
|
| 3178 |
+
"learning_rate": 0.0002,
|
| 3179 |
+
"loss": 0.616,
|
| 3180 |
+
"step": 453
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 2.700371747211896,
|
| 3184 |
+
"grad_norm": 8.933956146240234,
|
| 3185 |
+
"learning_rate": 0.0002,
|
| 3186 |
+
"loss": 0.6136,
|
| 3187 |
+
"step": 454
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 2.7063197026022303,
|
| 3191 |
+
"grad_norm": 1.1435190439224243,
|
| 3192 |
+
"learning_rate": 0.0002,
|
| 3193 |
+
"loss": 0.5598,
|
| 3194 |
+
"step": 455
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 2.712267657992565,
|
| 3198 |
+
"grad_norm": 1.2891956567764282,
|
| 3199 |
+
"learning_rate": 0.0002,
|
| 3200 |
+
"loss": 0.5945,
|
| 3201 |
+
"step": 456
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 2.7182156133829,
|
| 3205 |
+
"grad_norm": 1.3077706098556519,
|
| 3206 |
+
"learning_rate": 0.0002,
|
| 3207 |
+
"loss": 0.5793,
|
| 3208 |
+
"step": 457
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 2.724163568773234,
|
| 3212 |
+
"grad_norm": 1.1445353031158447,
|
| 3213 |
+
"learning_rate": 0.0002,
|
| 3214 |
+
"loss": 0.5644,
|
| 3215 |
+
"step": 458
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 2.7301115241635685,
|
| 3219 |
+
"grad_norm": 1.1466567516326904,
|
| 3220 |
+
"learning_rate": 0.0002,
|
| 3221 |
+
"loss": 0.5748,
|
| 3222 |
+
"step": 459
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 2.7360594795539033,
|
| 3226 |
+
"grad_norm": 1.0083645582199097,
|
| 3227 |
+
"learning_rate": 0.0002,
|
| 3228 |
+
"loss": 0.5676,
|
| 3229 |
+
"step": 460
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 2.742007434944238,
|
| 3233 |
+
"grad_norm": 0.9980899691581726,
|
| 3234 |
+
"learning_rate": 0.0002,
|
| 3235 |
+
"loss": 0.5975,
|
| 3236 |
+
"step": 461
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 2.7479553903345724,
|
| 3240 |
+
"grad_norm": 1.0702303647994995,
|
| 3241 |
+
"learning_rate": 0.0002,
|
| 3242 |
+
"loss": 0.5235,
|
| 3243 |
+
"step": 462
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 2.753903345724907,
|
| 3247 |
+
"grad_norm": 1.3305853605270386,
|
| 3248 |
+
"learning_rate": 0.0002,
|
| 3249 |
+
"loss": 0.6493,
|
| 3250 |
+
"step": 463
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 2.7598513011152415,
|
| 3254 |
+
"grad_norm": 1.4583408832550049,
|
| 3255 |
+
"learning_rate": 0.0002,
|
| 3256 |
+
"loss": 0.5948,
|
| 3257 |
+
"step": 464
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 2.7657992565055762,
|
| 3261 |
+
"grad_norm": 1.1704531908035278,
|
| 3262 |
+
"learning_rate": 0.0002,
|
| 3263 |
+
"loss": 0.5631,
|
| 3264 |
+
"step": 465
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 2.771747211895911,
|
| 3268 |
+
"grad_norm": 1.1165651082992554,
|
| 3269 |
+
"learning_rate": 0.0002,
|
| 3270 |
+
"loss": 0.6674,
|
| 3271 |
+
"step": 466
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 2.7776951672862453,
|
| 3275 |
+
"grad_norm": 1.2043639421463013,
|
| 3276 |
+
"learning_rate": 0.0002,
|
| 3277 |
+
"loss": 0.649,
|
| 3278 |
+
"step": 467
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 2.7836431226765797,
|
| 3282 |
+
"grad_norm": 1.0930832624435425,
|
| 3283 |
+
"learning_rate": 0.0002,
|
| 3284 |
+
"loss": 0.541,
|
| 3285 |
+
"step": 468
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 2.7895910780669144,
|
| 3289 |
+
"grad_norm": 1.1153466701507568,
|
| 3290 |
+
"learning_rate": 0.0002,
|
| 3291 |
+
"loss": 0.4666,
|
| 3292 |
+
"step": 469
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 2.795539033457249,
|
| 3296 |
+
"grad_norm": 3.27708101272583,
|
| 3297 |
+
"learning_rate": 0.0002,
|
| 3298 |
+
"loss": 0.6981,
|
| 3299 |
+
"step": 470
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 2.8014869888475835,
|
| 3303 |
+
"grad_norm": 1.200003981590271,
|
| 3304 |
+
"learning_rate": 0.0002,
|
| 3305 |
+
"loss": 0.5579,
|
| 3306 |
+
"step": 471
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 2.8074349442379183,
|
| 3310 |
+
"grad_norm": 1.2021151781082153,
|
| 3311 |
+
"learning_rate": 0.0002,
|
| 3312 |
+
"loss": 0.5383,
|
| 3313 |
+
"step": 472
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 2.8133828996282526,
|
| 3317 |
+
"grad_norm": 1.0844088792800903,
|
| 3318 |
+
"learning_rate": 0.0002,
|
| 3319 |
+
"loss": 0.6027,
|
| 3320 |
+
"step": 473
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 2.8193308550185874,
|
| 3324 |
+
"grad_norm": 1.1981035470962524,
|
| 3325 |
+
"learning_rate": 0.0002,
|
| 3326 |
+
"loss": 0.5589,
|
| 3327 |
+
"step": 474
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 2.825278810408922,
|
| 3331 |
+
"grad_norm": 1.23332679271698,
|
| 3332 |
+
"learning_rate": 0.0002,
|
| 3333 |
+
"loss": 0.6733,
|
| 3334 |
+
"step": 475
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 2.8312267657992565,
|
| 3338 |
+
"grad_norm": 1.2242364883422852,
|
| 3339 |
+
"learning_rate": 0.0002,
|
| 3340 |
+
"loss": 0.6392,
|
| 3341 |
+
"step": 476
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 2.8371747211895912,
|
| 3345 |
+
"grad_norm": 1.2482764720916748,
|
| 3346 |
+
"learning_rate": 0.0002,
|
| 3347 |
+
"loss": 0.6185,
|
| 3348 |
+
"step": 477
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 2.8431226765799256,
|
| 3352 |
+
"grad_norm": 1.3755487203598022,
|
| 3353 |
+
"learning_rate": 0.0002,
|
| 3354 |
+
"loss": 0.5761,
|
| 3355 |
+
"step": 478
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 2.8490706319702603,
|
| 3359 |
+
"grad_norm": 1.2065231800079346,
|
| 3360 |
+
"learning_rate": 0.0002,
|
| 3361 |
+
"loss": 0.5185,
|
| 3362 |
+
"step": 479
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 2.8550185873605947,
|
| 3366 |
+
"grad_norm": 1.1161603927612305,
|
| 3367 |
+
"learning_rate": 0.0002,
|
| 3368 |
+
"loss": 0.5697,
|
| 3369 |
+
"step": 480
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 2.8609665427509294,
|
| 3373 |
+
"grad_norm": 1.7466390132904053,
|
| 3374 |
+
"learning_rate": 0.0002,
|
| 3375 |
+
"loss": 0.5835,
|
| 3376 |
+
"step": 481
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 2.8669144981412638,
|
| 3380 |
+
"grad_norm": 1.371319055557251,
|
| 3381 |
+
"learning_rate": 0.0002,
|
| 3382 |
+
"loss": 0.6031,
|
| 3383 |
+
"step": 482
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 2.8728624535315985,
|
| 3387 |
+
"grad_norm": 1.4363592863082886,
|
| 3388 |
+
"learning_rate": 0.0002,
|
| 3389 |
+
"loss": 0.6028,
|
| 3390 |
+
"step": 483
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 2.8788104089219333,
|
| 3394 |
+
"grad_norm": 1.1025314331054688,
|
| 3395 |
+
"learning_rate": 0.0002,
|
| 3396 |
+
"loss": 0.492,
|
| 3397 |
+
"step": 484
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 2.8847583643122676,
|
| 3401 |
+
"grad_norm": 1.04302978515625,
|
| 3402 |
+
"learning_rate": 0.0002,
|
| 3403 |
+
"loss": 0.6032,
|
| 3404 |
+
"step": 485
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 2.8907063197026024,
|
| 3408 |
+
"grad_norm": 1.5093481540679932,
|
| 3409 |
+
"learning_rate": 0.0002,
|
| 3410 |
+
"loss": 0.6832,
|
| 3411 |
+
"step": 486
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 2.8966542750929367,
|
| 3415 |
+
"grad_norm": 1.068484902381897,
|
| 3416 |
+
"learning_rate": 0.0002,
|
| 3417 |
+
"loss": 0.4942,
|
| 3418 |
+
"step": 487
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 2.9026022304832715,
|
| 3422 |
+
"grad_norm": 2.1092681884765625,
|
| 3423 |
+
"learning_rate": 0.0002,
|
| 3424 |
+
"loss": 0.4909,
|
| 3425 |
+
"step": 488
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 2.908550185873606,
|
| 3429 |
+
"grad_norm": 1.22842276096344,
|
| 3430 |
+
"learning_rate": 0.0002,
|
| 3431 |
+
"loss": 0.6226,
|
| 3432 |
+
"step": 489
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 2.9144981412639406,
|
| 3436 |
+
"grad_norm": 1.1664717197418213,
|
| 3437 |
+
"learning_rate": 0.0002,
|
| 3438 |
+
"loss": 0.6093,
|
| 3439 |
+
"step": 490
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 2.920446096654275,
|
| 3443 |
+
"grad_norm": 1.2886883020401,
|
| 3444 |
+
"learning_rate": 0.0002,
|
| 3445 |
+
"loss": 0.5866,
|
| 3446 |
+
"step": 491
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 2.9263940520446097,
|
| 3450 |
+
"grad_norm": 1.1186504364013672,
|
| 3451 |
+
"learning_rate": 0.0002,
|
| 3452 |
+
"loss": 0.5942,
|
| 3453 |
+
"step": 492
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 2.9323420074349444,
|
| 3457 |
+
"grad_norm": 1.2734028100967407,
|
| 3458 |
+
"learning_rate": 0.0002,
|
| 3459 |
+
"loss": 0.5871,
|
| 3460 |
+
"step": 493
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 2.9382899628252788,
|
| 3464 |
+
"grad_norm": 1.1976778507232666,
|
| 3465 |
+
"learning_rate": 0.0002,
|
| 3466 |
+
"loss": 0.6766,
|
| 3467 |
+
"step": 494
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 2.9442379182156135,
|
| 3471 |
+
"grad_norm": 1.524681806564331,
|
| 3472 |
+
"learning_rate": 0.0002,
|
| 3473 |
+
"loss": 0.602,
|
| 3474 |
+
"step": 495
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 2.950185873605948,
|
| 3478 |
+
"grad_norm": 1.4174754619598389,
|
| 3479 |
+
"learning_rate": 0.0002,
|
| 3480 |
+
"loss": 0.6079,
|
| 3481 |
+
"step": 496
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 2.9561338289962826,
|
| 3485 |
+
"grad_norm": 1.1006587743759155,
|
| 3486 |
+
"learning_rate": 0.0002,
|
| 3487 |
+
"loss": 0.6393,
|
| 3488 |
+
"step": 497
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 2.962081784386617,
|
| 3492 |
+
"grad_norm": 1.3037843704223633,
|
| 3493 |
+
"learning_rate": 0.0002,
|
| 3494 |
+
"loss": 0.6065,
|
| 3495 |
+
"step": 498
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 2.9680297397769517,
|
| 3499 |
+
"grad_norm": 1.5767035484313965,
|
| 3500 |
+
"learning_rate": 0.0002,
|
| 3501 |
+
"loss": 0.6444,
|
| 3502 |
+
"step": 499
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 2.973977695167286,
|
| 3506 |
+
"grad_norm": 1.2918823957443237,
|
| 3507 |
+
"learning_rate": 0.0002,
|
| 3508 |
+
"loss": 0.5221,
|
| 3509 |
+
"step": 500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 2.979925650557621,
|
| 3513 |
+
"grad_norm": 1.0898538827896118,
|
| 3514 |
+
"learning_rate": 0.0002,
|
| 3515 |
+
"loss": 0.687,
|
| 3516 |
+
"step": 501
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 2.9858736059479556,
|
| 3520 |
+
"grad_norm": 1.6398361921310425,
|
| 3521 |
+
"learning_rate": 0.0002,
|
| 3522 |
+
"loss": 0.5781,
|
| 3523 |
+
"step": 502
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 2.99182156133829,
|
| 3527 |
+
"grad_norm": 1.2504217624664307,
|
| 3528 |
+
"learning_rate": 0.0002,
|
| 3529 |
+
"loss": 0.5916,
|
| 3530 |
+
"step": 503
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 2.9977695167286247,
|
| 3534 |
+
"grad_norm": 1.8298507928848267,
|
| 3535 |
+
"learning_rate": 0.0002,
|
| 3536 |
+
"loss": 0.5777,
|
| 3537 |
+
"step": 504
|
| 3538 |
}
|
| 3539 |
],
|
| 3540 |
"logging_steps": 1,
|
|
|
|
| 3549 |
"should_evaluate": false,
|
| 3550 |
"should_log": false,
|
| 3551 |
"should_save": true,
|
| 3552 |
+
"should_training_stop": true
|
| 3553 |
},
|
| 3554 |
"attributes": {}
|
| 3555 |
}
|
| 3556 |
},
|
| 3557 |
+
"total_flos": 3.641281849472852e+16,
|
| 3558 |
"train_batch_size": 1,
|
| 3559 |
"trial_name": null,
|
| 3560 |
"trial_params": null
|