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 +1753 -3
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3158328
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1b7f2f278e302865c60c8d05e6357ee7af1a0ad27a440e023bc9791b3008605
|
| 3 |
size 3158328
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1687034
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a8e21e40536ee857c1b466c73591efdd0af3fbc6924011c28cb079111a4e1ce
|
| 3 |
size 1687034
|
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:c44cd996c8ed27623113c0cd6c151d153a43988c9f84544695b9113185a594c6
|
| 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 sha256:3345390e24b1316ef013c9196db09afdb816048bd66321ccb9bd8a5ae5fa8c0f
|
| 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,
|
|
@@ -1757,6 +1757,1756 @@
|
|
| 1757 |
"learning_rate": 0.0001678688524590164,
|
| 1758 |
"loss": 0.2152,
|
| 1759 |
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1760 |
}
|
| 1761 |
],
|
| 1762 |
"logging_steps": 20,
|
|
@@ -1776,7 +3526,7 @@
|
|
| 1776 |
"attributes": {}
|
| 1777 |
}
|
| 1778 |
},
|
| 1779 |
-
"total_flos":
|
| 1780 |
"train_batch_size": 1,
|
| 1781 |
"trial_name": null,
|
| 1782 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 48.810250152532035,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1757 |
"learning_rate": 0.0001678688524590164,
|
| 1758 |
"loss": 0.2152,
|
| 1759 |
"step": 5000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 24.50274557657108,
|
| 1763 |
+
"grad_norm": 3.6262013912200928,
|
| 1764 |
+
"learning_rate": 0.0001677377049180328,
|
| 1765 |
+
"loss": 0.2023,
|
| 1766 |
+
"step": 5020
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 24.600366076876146,
|
| 1770 |
+
"grad_norm": 3.2124197483062744,
|
| 1771 |
+
"learning_rate": 0.00016760655737704919,
|
| 1772 |
+
"loss": 0.231,
|
| 1773 |
+
"step": 5040
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 24.697986577181208,
|
| 1777 |
+
"grad_norm": 3.997265100479126,
|
| 1778 |
+
"learning_rate": 0.00016747540983606558,
|
| 1779 |
+
"loss": 0.2249,
|
| 1780 |
+
"step": 5060
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 24.795607077486274,
|
| 1784 |
+
"grad_norm": 3.5564634799957275,
|
| 1785 |
+
"learning_rate": 0.00016734426229508196,
|
| 1786 |
+
"loss": 0.2337,
|
| 1787 |
+
"step": 5080
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 24.893227577791336,
|
| 1791 |
+
"grad_norm": 3.023794651031494,
|
| 1792 |
+
"learning_rate": 0.00016721311475409838,
|
| 1793 |
+
"loss": 0.2523,
|
| 1794 |
+
"step": 5100
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 24.990848078096402,
|
| 1798 |
+
"grad_norm": 3.3614046573638916,
|
| 1799 |
+
"learning_rate": 0.00016708196721311474,
|
| 1800 |
+
"loss": 0.2524,
|
| 1801 |
+
"step": 5120
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 25.088468578401464,
|
| 1805 |
+
"grad_norm": 2.63189435005188,
|
| 1806 |
+
"learning_rate": 0.00016695081967213116,
|
| 1807 |
+
"loss": 0.1942,
|
| 1808 |
+
"step": 5140
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 25.18608907870653,
|
| 1812 |
+
"grad_norm": 3.4099018573760986,
|
| 1813 |
+
"learning_rate": 0.00016681967213114755,
|
| 1814 |
+
"loss": 0.1751,
|
| 1815 |
+
"step": 5160
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 25.283709579011592,
|
| 1819 |
+
"grad_norm": 3.533768653869629,
|
| 1820 |
+
"learning_rate": 0.00016668852459016397,
|
| 1821 |
+
"loss": 0.1953,
|
| 1822 |
+
"step": 5180
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 25.381330079316655,
|
| 1826 |
+
"grad_norm": 3.330538511276245,
|
| 1827 |
+
"learning_rate": 0.00016655737704918033,
|
| 1828 |
+
"loss": 0.2124,
|
| 1829 |
+
"step": 5200
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 25.47895057962172,
|
| 1833 |
+
"grad_norm": 3.8602449893951416,
|
| 1834 |
+
"learning_rate": 0.00016642622950819672,
|
| 1835 |
+
"loss": 0.209,
|
| 1836 |
+
"step": 5220
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 25.576571079926783,
|
| 1840 |
+
"grad_norm": 3.290445566177368,
|
| 1841 |
+
"learning_rate": 0.00016629508196721313,
|
| 1842 |
+
"loss": 0.2002,
|
| 1843 |
+
"step": 5240
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 25.67419158023185,
|
| 1847 |
+
"grad_norm": 3.143129825592041,
|
| 1848 |
+
"learning_rate": 0.0001661639344262295,
|
| 1849 |
+
"loss": 0.2376,
|
| 1850 |
+
"step": 5260
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 25.77181208053691,
|
| 1854 |
+
"grad_norm": 4.425613880157471,
|
| 1855 |
+
"learning_rate": 0.0001660327868852459,
|
| 1856 |
+
"loss": 0.2255,
|
| 1857 |
+
"step": 5280
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 25.869432580841977,
|
| 1861 |
+
"grad_norm": 5.267921447753906,
|
| 1862 |
+
"learning_rate": 0.0001659016393442623,
|
| 1863 |
+
"loss": 0.2334,
|
| 1864 |
+
"step": 5300
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 25.96705308114704,
|
| 1868 |
+
"grad_norm": 3.738940715789795,
|
| 1869 |
+
"learning_rate": 0.0001657704918032787,
|
| 1870 |
+
"loss": 0.2354,
|
| 1871 |
+
"step": 5320
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 26.064673581452105,
|
| 1875 |
+
"grad_norm": 2.8138153553009033,
|
| 1876 |
+
"learning_rate": 0.00016563934426229508,
|
| 1877 |
+
"loss": 0.1706,
|
| 1878 |
+
"step": 5340
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 26.162294081757167,
|
| 1882 |
+
"grad_norm": 3.58404278755188,
|
| 1883 |
+
"learning_rate": 0.00016550819672131147,
|
| 1884 |
+
"loss": 0.199,
|
| 1885 |
+
"step": 5360
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 26.259914582062233,
|
| 1889 |
+
"grad_norm": 3.536153554916382,
|
| 1890 |
+
"learning_rate": 0.00016537704918032789,
|
| 1891 |
+
"loss": 0.1991,
|
| 1892 |
+
"step": 5380
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 26.357535082367296,
|
| 1896 |
+
"grad_norm": 2.7656655311584473,
|
| 1897 |
+
"learning_rate": 0.00016524590163934428,
|
| 1898 |
+
"loss": 0.179,
|
| 1899 |
+
"step": 5400
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 26.45515558267236,
|
| 1903 |
+
"grad_norm": 3.2351813316345215,
|
| 1904 |
+
"learning_rate": 0.00016511475409836067,
|
| 1905 |
+
"loss": 0.2142,
|
| 1906 |
+
"step": 5420
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 26.552776082977424,
|
| 1910 |
+
"grad_norm": 4.154785633087158,
|
| 1911 |
+
"learning_rate": 0.00016498360655737705,
|
| 1912 |
+
"loss": 0.2007,
|
| 1913 |
+
"step": 5440
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 26.65039658328249,
|
| 1917 |
+
"grad_norm": 3.327596664428711,
|
| 1918 |
+
"learning_rate": 0.00016485245901639344,
|
| 1919 |
+
"loss": 0.1934,
|
| 1920 |
+
"step": 5460
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 26.748017083587552,
|
| 1924 |
+
"grad_norm": 3.7125210762023926,
|
| 1925 |
+
"learning_rate": 0.00016472131147540983,
|
| 1926 |
+
"loss": 0.2013,
|
| 1927 |
+
"step": 5480
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 26.845637583892618,
|
| 1931 |
+
"grad_norm": 3.177676200866699,
|
| 1932 |
+
"learning_rate": 0.00016459016393442622,
|
| 1933 |
+
"loss": 0.2156,
|
| 1934 |
+
"step": 5500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 26.94325808419768,
|
| 1938 |
+
"grad_norm": 5.258444786071777,
|
| 1939 |
+
"learning_rate": 0.00016445901639344264,
|
| 1940 |
+
"loss": 0.2068,
|
| 1941 |
+
"step": 5520
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 27.040878584502746,
|
| 1945 |
+
"grad_norm": 2.750441074371338,
|
| 1946 |
+
"learning_rate": 0.00016432786885245903,
|
| 1947 |
+
"loss": 0.2059,
|
| 1948 |
+
"step": 5540
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 27.13849908480781,
|
| 1952 |
+
"grad_norm": 2.5060107707977295,
|
| 1953 |
+
"learning_rate": 0.00016419672131147542,
|
| 1954 |
+
"loss": 0.1696,
|
| 1955 |
+
"step": 5560
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 27.236119585112874,
|
| 1959 |
+
"grad_norm": 3.026069164276123,
|
| 1960 |
+
"learning_rate": 0.0001640655737704918,
|
| 1961 |
+
"loss": 0.1874,
|
| 1962 |
+
"step": 5580
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 27.333740085417936,
|
| 1966 |
+
"grad_norm": 3.3759682178497314,
|
| 1967 |
+
"learning_rate": 0.0001639344262295082,
|
| 1968 |
+
"loss": 0.1936,
|
| 1969 |
+
"step": 5600
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 27.431360585723002,
|
| 1973 |
+
"grad_norm": 4.374515533447266,
|
| 1974 |
+
"learning_rate": 0.0001638032786885246,
|
| 1975 |
+
"loss": 0.1839,
|
| 1976 |
+
"step": 5620
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 27.528981086028065,
|
| 1980 |
+
"grad_norm": 3.855091094970703,
|
| 1981 |
+
"learning_rate": 0.00016367213114754098,
|
| 1982 |
+
"loss": 0.1998,
|
| 1983 |
+
"step": 5640
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 27.62660158633313,
|
| 1987 |
+
"grad_norm": 3.3890366554260254,
|
| 1988 |
+
"learning_rate": 0.0001635409836065574,
|
| 1989 |
+
"loss": 0.1956,
|
| 1990 |
+
"step": 5660
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 27.724222086638193,
|
| 1994 |
+
"grad_norm": 3.516590118408203,
|
| 1995 |
+
"learning_rate": 0.00016340983606557378,
|
| 1996 |
+
"loss": 0.1784,
|
| 1997 |
+
"step": 5680
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 27.82184258694326,
|
| 2001 |
+
"grad_norm": 3.740670680999756,
|
| 2002 |
+
"learning_rate": 0.00016327868852459017,
|
| 2003 |
+
"loss": 0.1957,
|
| 2004 |
+
"step": 5700
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 27.91946308724832,
|
| 2008 |
+
"grad_norm": 5.360738754272461,
|
| 2009 |
+
"learning_rate": 0.00016314754098360656,
|
| 2010 |
+
"loss": 0.219,
|
| 2011 |
+
"step": 5720
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 28.017083587553387,
|
| 2015 |
+
"grad_norm": 2.9260716438293457,
|
| 2016 |
+
"learning_rate": 0.00016301639344262295,
|
| 2017 |
+
"loss": 0.1956,
|
| 2018 |
+
"step": 5740
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 28.11470408785845,
|
| 2022 |
+
"grad_norm": 2.962049722671509,
|
| 2023 |
+
"learning_rate": 0.00016288524590163937,
|
| 2024 |
+
"loss": 0.1514,
|
| 2025 |
+
"step": 5760
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 28.212324588163515,
|
| 2029 |
+
"grad_norm": 3.02143931388855,
|
| 2030 |
+
"learning_rate": 0.00016275409836065573,
|
| 2031 |
+
"loss": 0.1586,
|
| 2032 |
+
"step": 5780
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 28.309945088468577,
|
| 2036 |
+
"grad_norm": 3.138089895248413,
|
| 2037 |
+
"learning_rate": 0.00016262295081967214,
|
| 2038 |
+
"loss": 0.1706,
|
| 2039 |
+
"step": 5800
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 28.407565588773643,
|
| 2043 |
+
"grad_norm": 3.2356488704681396,
|
| 2044 |
+
"learning_rate": 0.00016249180327868853,
|
| 2045 |
+
"loss": 0.1794,
|
| 2046 |
+
"step": 5820
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 28.505186089078705,
|
| 2050 |
+
"grad_norm": 4.481720447540283,
|
| 2051 |
+
"learning_rate": 0.00016236065573770492,
|
| 2052 |
+
"loss": 0.1871,
|
| 2053 |
+
"step": 5840
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 28.60280658938377,
|
| 2057 |
+
"grad_norm": 3.6736230850219727,
|
| 2058 |
+
"learning_rate": 0.0001622295081967213,
|
| 2059 |
+
"loss": 0.1649,
|
| 2060 |
+
"step": 5860
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 28.700427089688834,
|
| 2064 |
+
"grad_norm": 4.480273723602295,
|
| 2065 |
+
"learning_rate": 0.0001620983606557377,
|
| 2066 |
+
"loss": 0.1837,
|
| 2067 |
+
"step": 5880
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 28.7980475899939,
|
| 2071 |
+
"grad_norm": 3.7047078609466553,
|
| 2072 |
+
"learning_rate": 0.00016196721311475412,
|
| 2073 |
+
"loss": 0.2099,
|
| 2074 |
+
"step": 5900
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 28.89566809029896,
|
| 2078 |
+
"grad_norm": 3.0603742599487305,
|
| 2079 |
+
"learning_rate": 0.00016183606557377048,
|
| 2080 |
+
"loss": 0.2015,
|
| 2081 |
+
"step": 5920
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 28.993288590604028,
|
| 2085 |
+
"grad_norm": 3.1472814083099365,
|
| 2086 |
+
"learning_rate": 0.0001617049180327869,
|
| 2087 |
+
"loss": 0.2016,
|
| 2088 |
+
"step": 5940
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 29.09090909090909,
|
| 2092 |
+
"grad_norm": 2.6735455989837646,
|
| 2093 |
+
"learning_rate": 0.0001615737704918033,
|
| 2094 |
+
"loss": 0.1586,
|
| 2095 |
+
"step": 5960
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 29.188529591214156,
|
| 2099 |
+
"grad_norm": 3.0757932662963867,
|
| 2100 |
+
"learning_rate": 0.00016144262295081968,
|
| 2101 |
+
"loss": 0.1466,
|
| 2102 |
+
"step": 5980
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 29.286150091519218,
|
| 2106 |
+
"grad_norm": 3.696460008621216,
|
| 2107 |
+
"learning_rate": 0.00016131147540983607,
|
| 2108 |
+
"loss": 0.1411,
|
| 2109 |
+
"step": 6000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 29.383770591824284,
|
| 2113 |
+
"grad_norm": 3.981374979019165,
|
| 2114 |
+
"learning_rate": 0.00016118032786885245,
|
| 2115 |
+
"loss": 0.1703,
|
| 2116 |
+
"step": 6020
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 29.481391092129346,
|
| 2120 |
+
"grad_norm": 3.1289327144622803,
|
| 2121 |
+
"learning_rate": 0.00016104918032786887,
|
| 2122 |
+
"loss": 0.1708,
|
| 2123 |
+
"step": 6040
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 29.579011592434412,
|
| 2127 |
+
"grad_norm": 2.825981855392456,
|
| 2128 |
+
"learning_rate": 0.00016091803278688526,
|
| 2129 |
+
"loss": 0.1731,
|
| 2130 |
+
"step": 6060
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 29.676632092739474,
|
| 2134 |
+
"grad_norm": 3.0257232189178467,
|
| 2135 |
+
"learning_rate": 0.00016078688524590165,
|
| 2136 |
+
"loss": 0.1799,
|
| 2137 |
+
"step": 6080
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 29.77425259304454,
|
| 2141 |
+
"grad_norm": 3.0449230670928955,
|
| 2142 |
+
"learning_rate": 0.00016065573770491804,
|
| 2143 |
+
"loss": 0.1776,
|
| 2144 |
+
"step": 6100
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 29.871873093349603,
|
| 2148 |
+
"grad_norm": 3.108060121536255,
|
| 2149 |
+
"learning_rate": 0.00016052459016393443,
|
| 2150 |
+
"loss": 0.1839,
|
| 2151 |
+
"step": 6120
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 29.96949359365467,
|
| 2155 |
+
"grad_norm": 3.475794792175293,
|
| 2156 |
+
"learning_rate": 0.00016039344262295082,
|
| 2157 |
+
"loss": 0.2044,
|
| 2158 |
+
"step": 6140
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 30.06711409395973,
|
| 2162 |
+
"grad_norm": 2.7312111854553223,
|
| 2163 |
+
"learning_rate": 0.0001602622950819672,
|
| 2164 |
+
"loss": 0.1509,
|
| 2165 |
+
"step": 6160
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 30.164734594264797,
|
| 2169 |
+
"grad_norm": 2.8779571056365967,
|
| 2170 |
+
"learning_rate": 0.00016013114754098362,
|
| 2171 |
+
"loss": 0.1387,
|
| 2172 |
+
"step": 6180
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 30.26235509456986,
|
| 2176 |
+
"grad_norm": 4.596887111663818,
|
| 2177 |
+
"learning_rate": 0.00016,
|
| 2178 |
+
"loss": 0.1625,
|
| 2179 |
+
"step": 6200
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 30.359975594874925,
|
| 2183 |
+
"grad_norm": 2.2877578735351562,
|
| 2184 |
+
"learning_rate": 0.0001598688524590164,
|
| 2185 |
+
"loss": 0.1611,
|
| 2186 |
+
"step": 6220
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 30.457596095179987,
|
| 2190 |
+
"grad_norm": 2.9951751232147217,
|
| 2191 |
+
"learning_rate": 0.0001597377049180328,
|
| 2192 |
+
"loss": 0.1579,
|
| 2193 |
+
"step": 6240
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 30.555216595485053,
|
| 2197 |
+
"grad_norm": 2.9729604721069336,
|
| 2198 |
+
"learning_rate": 0.00015960655737704918,
|
| 2199 |
+
"loss": 0.1522,
|
| 2200 |
+
"step": 6260
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 30.652837095790115,
|
| 2204 |
+
"grad_norm": 3.4268481731414795,
|
| 2205 |
+
"learning_rate": 0.0001594754098360656,
|
| 2206 |
+
"loss": 0.1673,
|
| 2207 |
+
"step": 6280
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 30.75045759609518,
|
| 2211 |
+
"grad_norm": 3.1322598457336426,
|
| 2212 |
+
"learning_rate": 0.00015934426229508196,
|
| 2213 |
+
"loss": 0.1769,
|
| 2214 |
+
"step": 6300
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 30.848078096400243,
|
| 2218 |
+
"grad_norm": 3.1607165336608887,
|
| 2219 |
+
"learning_rate": 0.00015921311475409838,
|
| 2220 |
+
"loss": 0.1795,
|
| 2221 |
+
"step": 6320
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 30.94569859670531,
|
| 2225 |
+
"grad_norm": 3.9670820236206055,
|
| 2226 |
+
"learning_rate": 0.00015908196721311477,
|
| 2227 |
+
"loss": 0.1714,
|
| 2228 |
+
"step": 6340
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 31.04331909701037,
|
| 2232 |
+
"grad_norm": 3.0311009883880615,
|
| 2233 |
+
"learning_rate": 0.00015895081967213116,
|
| 2234 |
+
"loss": 0.1546,
|
| 2235 |
+
"step": 6360
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 31.140939597315437,
|
| 2239 |
+
"grad_norm": 2.023759126663208,
|
| 2240 |
+
"learning_rate": 0.00015881967213114754,
|
| 2241 |
+
"loss": 0.1351,
|
| 2242 |
+
"step": 6380
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 31.2385600976205,
|
| 2246 |
+
"grad_norm": 2.746744155883789,
|
| 2247 |
+
"learning_rate": 0.00015868852459016393,
|
| 2248 |
+
"loss": 0.1446,
|
| 2249 |
+
"step": 6400
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 31.336180597925566,
|
| 2253 |
+
"grad_norm": 2.9396300315856934,
|
| 2254 |
+
"learning_rate": 0.00015855737704918035,
|
| 2255 |
+
"loss": 0.1505,
|
| 2256 |
+
"step": 6420
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 31.433801098230628,
|
| 2260 |
+
"grad_norm": 4.066432476043701,
|
| 2261 |
+
"learning_rate": 0.0001584262295081967,
|
| 2262 |
+
"loss": 0.1576,
|
| 2263 |
+
"step": 6440
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 31.531421598535694,
|
| 2267 |
+
"grad_norm": 3.7003040313720703,
|
| 2268 |
+
"learning_rate": 0.00015829508196721313,
|
| 2269 |
+
"loss": 0.1629,
|
| 2270 |
+
"step": 6460
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 31.629042098840756,
|
| 2274 |
+
"grad_norm": 3.1914453506469727,
|
| 2275 |
+
"learning_rate": 0.00015816393442622952,
|
| 2276 |
+
"loss": 0.1497,
|
| 2277 |
+
"step": 6480
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 31.726662599145822,
|
| 2281 |
+
"grad_norm": 3.582771062850952,
|
| 2282 |
+
"learning_rate": 0.0001580327868852459,
|
| 2283 |
+
"loss": 0.1535,
|
| 2284 |
+
"step": 6500
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 31.824283099450884,
|
| 2288 |
+
"grad_norm": 2.601606845855713,
|
| 2289 |
+
"learning_rate": 0.0001579016393442623,
|
| 2290 |
+
"loss": 0.1692,
|
| 2291 |
+
"step": 6520
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 31.92190359975595,
|
| 2295 |
+
"grad_norm": 3.206125020980835,
|
| 2296 |
+
"learning_rate": 0.0001577704918032787,
|
| 2297 |
+
"loss": 0.1529,
|
| 2298 |
+
"step": 6540
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 32.01952410006101,
|
| 2302 |
+
"grad_norm": 2.8698012828826904,
|
| 2303 |
+
"learning_rate": 0.0001576393442622951,
|
| 2304 |
+
"loss": 0.1804,
|
| 2305 |
+
"step": 6560
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 32.117144600366075,
|
| 2309 |
+
"grad_norm": 2.53930401802063,
|
| 2310 |
+
"learning_rate": 0.00015750819672131147,
|
| 2311 |
+
"loss": 0.1227,
|
| 2312 |
+
"step": 6580
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 32.214765100671144,
|
| 2316 |
+
"grad_norm": 2.9354329109191895,
|
| 2317 |
+
"learning_rate": 0.00015737704918032788,
|
| 2318 |
+
"loss": 0.1299,
|
| 2319 |
+
"step": 6600
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 32.31238560097621,
|
| 2323 |
+
"grad_norm": 3.0723047256469727,
|
| 2324 |
+
"learning_rate": 0.00015724590163934427,
|
| 2325 |
+
"loss": 0.1467,
|
| 2326 |
+
"step": 6620
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 32.41000610128127,
|
| 2330 |
+
"grad_norm": 2.8421099185943604,
|
| 2331 |
+
"learning_rate": 0.00015711475409836066,
|
| 2332 |
+
"loss": 0.1468,
|
| 2333 |
+
"step": 6640
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 32.50762660158633,
|
| 2337 |
+
"grad_norm": 3.185232400894165,
|
| 2338 |
+
"learning_rate": 0.00015698360655737705,
|
| 2339 |
+
"loss": 0.1601,
|
| 2340 |
+
"step": 6660
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 32.6052471018914,
|
| 2344 |
+
"grad_norm": 3.1354596614837646,
|
| 2345 |
+
"learning_rate": 0.00015685245901639344,
|
| 2346 |
+
"loss": 0.1466,
|
| 2347 |
+
"step": 6680
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 32.70286760219646,
|
| 2351 |
+
"grad_norm": 3.695125102996826,
|
| 2352 |
+
"learning_rate": 0.00015672131147540986,
|
| 2353 |
+
"loss": 0.1535,
|
| 2354 |
+
"step": 6700
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 32.800488102501525,
|
| 2358 |
+
"grad_norm": 3.092844009399414,
|
| 2359 |
+
"learning_rate": 0.00015659016393442622,
|
| 2360 |
+
"loss": 0.1625,
|
| 2361 |
+
"step": 6720
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 32.89810860280659,
|
| 2365 |
+
"grad_norm": 3.8023808002471924,
|
| 2366 |
+
"learning_rate": 0.00015645901639344263,
|
| 2367 |
+
"loss": 0.1623,
|
| 2368 |
+
"step": 6740
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 32.99572910311166,
|
| 2372 |
+
"grad_norm": 2.934077739715576,
|
| 2373 |
+
"learning_rate": 0.00015632786885245902,
|
| 2374 |
+
"loss": 0.1629,
|
| 2375 |
+
"step": 6760
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 33.09334960341672,
|
| 2379 |
+
"grad_norm": 2.633199453353882,
|
| 2380 |
+
"learning_rate": 0.0001561967213114754,
|
| 2381 |
+
"loss": 0.1267,
|
| 2382 |
+
"step": 6780
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 33.19097010372178,
|
| 2386 |
+
"grad_norm": 2.3645660877227783,
|
| 2387 |
+
"learning_rate": 0.0001560655737704918,
|
| 2388 |
+
"loss": 0.124,
|
| 2389 |
+
"step": 6800
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 33.288590604026844,
|
| 2393 |
+
"grad_norm": 3.1801674365997314,
|
| 2394 |
+
"learning_rate": 0.0001559344262295082,
|
| 2395 |
+
"loss": 0.1207,
|
| 2396 |
+
"step": 6820
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 33.38621110433191,
|
| 2400 |
+
"grad_norm": 3.998974323272705,
|
| 2401 |
+
"learning_rate": 0.0001558032786885246,
|
| 2402 |
+
"loss": 0.1397,
|
| 2403 |
+
"step": 6840
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 33.483831604636975,
|
| 2407 |
+
"grad_norm": 2.8653316497802734,
|
| 2408 |
+
"learning_rate": 0.000155672131147541,
|
| 2409 |
+
"loss": 0.1366,
|
| 2410 |
+
"step": 6860
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 33.58145210494204,
|
| 2414 |
+
"grad_norm": 3.3324031829833984,
|
| 2415 |
+
"learning_rate": 0.0001555409836065574,
|
| 2416 |
+
"loss": 0.1418,
|
| 2417 |
+
"step": 6880
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 33.6790726052471,
|
| 2421 |
+
"grad_norm": 2.5941219329833984,
|
| 2422 |
+
"learning_rate": 0.00015540983606557378,
|
| 2423 |
+
"loss": 0.1431,
|
| 2424 |
+
"step": 6900
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 33.77669310555217,
|
| 2428 |
+
"grad_norm": 3.388183116912842,
|
| 2429 |
+
"learning_rate": 0.00015527868852459017,
|
| 2430 |
+
"loss": 0.15,
|
| 2431 |
+
"step": 6920
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 33.87431360585723,
|
| 2435 |
+
"grad_norm": 2.9687812328338623,
|
| 2436 |
+
"learning_rate": 0.00015514754098360656,
|
| 2437 |
+
"loss": 0.1471,
|
| 2438 |
+
"step": 6940
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 33.971934106162294,
|
| 2442 |
+
"grad_norm": 3.783984899520874,
|
| 2443 |
+
"learning_rate": 0.00015501639344262294,
|
| 2444 |
+
"loss": 0.1822,
|
| 2445 |
+
"step": 6960
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 34.06955460646736,
|
| 2449 |
+
"grad_norm": 2.6898393630981445,
|
| 2450 |
+
"learning_rate": 0.00015488524590163936,
|
| 2451 |
+
"loss": 0.1217,
|
| 2452 |
+
"step": 6980
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 34.16717510677242,
|
| 2456 |
+
"grad_norm": 2.337120532989502,
|
| 2457 |
+
"learning_rate": 0.00015475409836065575,
|
| 2458 |
+
"loss": 0.1261,
|
| 2459 |
+
"step": 7000
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 34.26479560707749,
|
| 2463 |
+
"grad_norm": 2.8725621700286865,
|
| 2464 |
+
"learning_rate": 0.00015462295081967214,
|
| 2465 |
+
"loss": 0.1376,
|
| 2466 |
+
"step": 7020
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 34.36241610738255,
|
| 2470 |
+
"grad_norm": 2.8058671951293945,
|
| 2471 |
+
"learning_rate": 0.00015449180327868853,
|
| 2472 |
+
"loss": 0.1307,
|
| 2473 |
+
"step": 7040
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 34.46003660768761,
|
| 2477 |
+
"grad_norm": 2.90006685256958,
|
| 2478 |
+
"learning_rate": 0.00015436065573770492,
|
| 2479 |
+
"loss": 0.1364,
|
| 2480 |
+
"step": 7060
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 34.557657107992675,
|
| 2484 |
+
"grad_norm": 4.001739978790283,
|
| 2485 |
+
"learning_rate": 0.00015422950819672133,
|
| 2486 |
+
"loss": 0.1416,
|
| 2487 |
+
"step": 7080
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 34.655277608297745,
|
| 2491 |
+
"grad_norm": 3.0421645641326904,
|
| 2492 |
+
"learning_rate": 0.0001540983606557377,
|
| 2493 |
+
"loss": 0.137,
|
| 2494 |
+
"step": 7100
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 34.75289810860281,
|
| 2498 |
+
"grad_norm": 3.429624080657959,
|
| 2499 |
+
"learning_rate": 0.00015396721311475411,
|
| 2500 |
+
"loss": 0.1524,
|
| 2501 |
+
"step": 7120
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 34.85051860890787,
|
| 2505 |
+
"grad_norm": 2.7907767295837402,
|
| 2506 |
+
"learning_rate": 0.0001538360655737705,
|
| 2507 |
+
"loss": 0.1344,
|
| 2508 |
+
"step": 7140
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 34.94813910921293,
|
| 2512 |
+
"grad_norm": 2.5428404808044434,
|
| 2513 |
+
"learning_rate": 0.0001537049180327869,
|
| 2514 |
+
"loss": 0.1329,
|
| 2515 |
+
"step": 7160
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 35.045759609518,
|
| 2519 |
+
"grad_norm": 3.0310592651367188,
|
| 2520 |
+
"learning_rate": 0.00015357377049180328,
|
| 2521 |
+
"loss": 0.1289,
|
| 2522 |
+
"step": 7180
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 35.14338010982306,
|
| 2526 |
+
"grad_norm": 2.757962226867676,
|
| 2527 |
+
"learning_rate": 0.00015344262295081967,
|
| 2528 |
+
"loss": 0.1148,
|
| 2529 |
+
"step": 7200
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 35.241000610128125,
|
| 2533 |
+
"grad_norm": 2.47044038772583,
|
| 2534 |
+
"learning_rate": 0.0001533114754098361,
|
| 2535 |
+
"loss": 0.1234,
|
| 2536 |
+
"step": 7220
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 35.33862111043319,
|
| 2540 |
+
"grad_norm": 2.766308546066284,
|
| 2541 |
+
"learning_rate": 0.00015318032786885245,
|
| 2542 |
+
"loss": 0.1154,
|
| 2543 |
+
"step": 7240
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 35.43624161073826,
|
| 2547 |
+
"grad_norm": 3.6302130222320557,
|
| 2548 |
+
"learning_rate": 0.00015304918032786887,
|
| 2549 |
+
"loss": 0.1323,
|
| 2550 |
+
"step": 7260
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 35.53386211104332,
|
| 2554 |
+
"grad_norm": 3.4185805320739746,
|
| 2555 |
+
"learning_rate": 0.00015291803278688526,
|
| 2556 |
+
"loss": 0.1409,
|
| 2557 |
+
"step": 7280
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 35.63148261134838,
|
| 2561 |
+
"grad_norm": 2.238168954849243,
|
| 2562 |
+
"learning_rate": 0.00015278688524590165,
|
| 2563 |
+
"loss": 0.1216,
|
| 2564 |
+
"step": 7300
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 35.729103111653444,
|
| 2568 |
+
"grad_norm": 3.636718988418579,
|
| 2569 |
+
"learning_rate": 0.00015265573770491803,
|
| 2570 |
+
"loss": 0.1431,
|
| 2571 |
+
"step": 7320
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 35.82672361195851,
|
| 2575 |
+
"grad_norm": 3.1300854682922363,
|
| 2576 |
+
"learning_rate": 0.00015252459016393442,
|
| 2577 |
+
"loss": 0.1472,
|
| 2578 |
+
"step": 7340
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 35.924344112263576,
|
| 2582 |
+
"grad_norm": 3.463369846343994,
|
| 2583 |
+
"learning_rate": 0.00015239344262295084,
|
| 2584 |
+
"loss": 0.1461,
|
| 2585 |
+
"step": 7360
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 36.02196461256864,
|
| 2589 |
+
"grad_norm": 4.020668983459473,
|
| 2590 |
+
"learning_rate": 0.0001522622950819672,
|
| 2591 |
+
"loss": 0.1276,
|
| 2592 |
+
"step": 7380
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 36.1195851128737,
|
| 2596 |
+
"grad_norm": 3.0597307682037354,
|
| 2597 |
+
"learning_rate": 0.00015213114754098362,
|
| 2598 |
+
"loss": 0.1137,
|
| 2599 |
+
"step": 7400
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 36.21720561317877,
|
| 2603 |
+
"grad_norm": 2.3329412937164307,
|
| 2604 |
+
"learning_rate": 0.000152,
|
| 2605 |
+
"loss": 0.1271,
|
| 2606 |
+
"step": 7420
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 36.31482611348383,
|
| 2610 |
+
"grad_norm": 3.5883936882019043,
|
| 2611 |
+
"learning_rate": 0.0001518688524590164,
|
| 2612 |
+
"loss": 0.1112,
|
| 2613 |
+
"step": 7440
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 36.412446613788894,
|
| 2617 |
+
"grad_norm": 2.7755954265594482,
|
| 2618 |
+
"learning_rate": 0.0001517377049180328,
|
| 2619 |
+
"loss": 0.1123,
|
| 2620 |
+
"step": 7460
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 36.51006711409396,
|
| 2624 |
+
"grad_norm": 3.594489336013794,
|
| 2625 |
+
"learning_rate": 0.00015160655737704918,
|
| 2626 |
+
"loss": 0.1338,
|
| 2627 |
+
"step": 7480
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 36.607687614399026,
|
| 2631 |
+
"grad_norm": 2.8571972846984863,
|
| 2632 |
+
"learning_rate": 0.0001514754098360656,
|
| 2633 |
+
"loss": 0.1256,
|
| 2634 |
+
"step": 7500
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 36.70530811470409,
|
| 2638 |
+
"grad_norm": 3.561603307723999,
|
| 2639 |
+
"learning_rate": 0.00015134426229508198,
|
| 2640 |
+
"loss": 0.1239,
|
| 2641 |
+
"step": 7520
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 36.80292861500915,
|
| 2645 |
+
"grad_norm": 3.265953302383423,
|
| 2646 |
+
"learning_rate": 0.00015121311475409837,
|
| 2647 |
+
"loss": 0.1348,
|
| 2648 |
+
"step": 7540
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 36.90054911531421,
|
| 2652 |
+
"grad_norm": 3.5381383895874023,
|
| 2653 |
+
"learning_rate": 0.00015108196721311476,
|
| 2654 |
+
"loss": 0.1402,
|
| 2655 |
+
"step": 7560
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 36.99816961561928,
|
| 2659 |
+
"grad_norm": 3.6995930671691895,
|
| 2660 |
+
"learning_rate": 0.00015095081967213115,
|
| 2661 |
+
"loss": 0.1323,
|
| 2662 |
+
"step": 7580
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 37.095790115924345,
|
| 2666 |
+
"grad_norm": 2.1127350330352783,
|
| 2667 |
+
"learning_rate": 0.00015081967213114754,
|
| 2668 |
+
"loss": 0.1045,
|
| 2669 |
+
"step": 7600
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 37.19341061622941,
|
| 2673 |
+
"grad_norm": 2.4780166149139404,
|
| 2674 |
+
"learning_rate": 0.00015068852459016393,
|
| 2675 |
+
"loss": 0.1173,
|
| 2676 |
+
"step": 7620
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 37.29103111653447,
|
| 2680 |
+
"grad_norm": 2.4735465049743652,
|
| 2681 |
+
"learning_rate": 0.00015055737704918035,
|
| 2682 |
+
"loss": 0.1158,
|
| 2683 |
+
"step": 7640
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 37.38865161683954,
|
| 2687 |
+
"grad_norm": 4.070529937744141,
|
| 2688 |
+
"learning_rate": 0.00015042622950819673,
|
| 2689 |
+
"loss": 0.1222,
|
| 2690 |
+
"step": 7660
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 37.4862721171446,
|
| 2694 |
+
"grad_norm": 2.87009596824646,
|
| 2695 |
+
"learning_rate": 0.00015029508196721312,
|
| 2696 |
+
"loss": 0.1237,
|
| 2697 |
+
"step": 7680
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 37.58389261744966,
|
| 2701 |
+
"grad_norm": 1.9312514066696167,
|
| 2702 |
+
"learning_rate": 0.00015016393442622951,
|
| 2703 |
+
"loss": 0.1277,
|
| 2704 |
+
"step": 7700
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 37.681513117754726,
|
| 2708 |
+
"grad_norm": 2.3069088459014893,
|
| 2709 |
+
"learning_rate": 0.0001500327868852459,
|
| 2710 |
+
"loss": 0.1204,
|
| 2711 |
+
"step": 7720
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 37.779133618059795,
|
| 2715 |
+
"grad_norm": 3.6415116786956787,
|
| 2716 |
+
"learning_rate": 0.00014990163934426232,
|
| 2717 |
+
"loss": 0.1251,
|
| 2718 |
+
"step": 7740
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 37.87675411836486,
|
| 2722 |
+
"grad_norm": 2.912848472595215,
|
| 2723 |
+
"learning_rate": 0.00014977049180327868,
|
| 2724 |
+
"loss": 0.1277,
|
| 2725 |
+
"step": 7760
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 37.97437461866992,
|
| 2729 |
+
"grad_norm": 3.3791749477386475,
|
| 2730 |
+
"learning_rate": 0.0001496393442622951,
|
| 2731 |
+
"loss": 0.1215,
|
| 2732 |
+
"step": 7780
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 38.07199511897498,
|
| 2736 |
+
"grad_norm": 3.6682169437408447,
|
| 2737 |
+
"learning_rate": 0.0001495081967213115,
|
| 2738 |
+
"loss": 0.1131,
|
| 2739 |
+
"step": 7800
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 38.16961561928005,
|
| 2743 |
+
"grad_norm": 2.891191005706787,
|
| 2744 |
+
"learning_rate": 0.00014937704918032788,
|
| 2745 |
+
"loss": 0.1105,
|
| 2746 |
+
"step": 7820
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 38.267236119585114,
|
| 2750 |
+
"grad_norm": 2.662431478500366,
|
| 2751 |
+
"learning_rate": 0.00014924590163934427,
|
| 2752 |
+
"loss": 0.1144,
|
| 2753 |
+
"step": 7840
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 38.364856619890176,
|
| 2757 |
+
"grad_norm": 3.193325996398926,
|
| 2758 |
+
"learning_rate": 0.00014911475409836066,
|
| 2759 |
+
"loss": 0.1102,
|
| 2760 |
+
"step": 7860
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 38.46247712019524,
|
| 2764 |
+
"grad_norm": 2.497185230255127,
|
| 2765 |
+
"learning_rate": 0.00014898360655737707,
|
| 2766 |
+
"loss": 0.101,
|
| 2767 |
+
"step": 7880
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 38.56009762050031,
|
| 2771 |
+
"grad_norm": 2.6551053524017334,
|
| 2772 |
+
"learning_rate": 0.00014885245901639343,
|
| 2773 |
+
"loss": 0.1178,
|
| 2774 |
+
"step": 7900
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 38.65771812080537,
|
| 2778 |
+
"grad_norm": 3.329808473587036,
|
| 2779 |
+
"learning_rate": 0.00014872131147540985,
|
| 2780 |
+
"loss": 0.11,
|
| 2781 |
+
"step": 7920
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 38.75533862111043,
|
| 2785 |
+
"grad_norm": 2.581897258758545,
|
| 2786 |
+
"learning_rate": 0.00014859016393442624,
|
| 2787 |
+
"loss": 0.1228,
|
| 2788 |
+
"step": 7940
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 38.852959121415495,
|
| 2792 |
+
"grad_norm": 2.805368423461914,
|
| 2793 |
+
"learning_rate": 0.00014845901639344263,
|
| 2794 |
+
"loss": 0.1249,
|
| 2795 |
+
"step": 7960
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 38.950579621720564,
|
| 2799 |
+
"grad_norm": 2.7007319927215576,
|
| 2800 |
+
"learning_rate": 0.00014832786885245902,
|
| 2801 |
+
"loss": 0.1319,
|
| 2802 |
+
"step": 7980
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 39.04820012202563,
|
| 2806 |
+
"grad_norm": 2.029555082321167,
|
| 2807 |
+
"learning_rate": 0.0001481967213114754,
|
| 2808 |
+
"loss": 0.1132,
|
| 2809 |
+
"step": 8000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 39.14582062233069,
|
| 2813 |
+
"grad_norm": 1.6910994052886963,
|
| 2814 |
+
"learning_rate": 0.00014806557377049182,
|
| 2815 |
+
"loss": 0.0968,
|
| 2816 |
+
"step": 8020
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 39.24344112263575,
|
| 2820 |
+
"grad_norm": 2.706334114074707,
|
| 2821 |
+
"learning_rate": 0.0001479344262295082,
|
| 2822 |
+
"loss": 0.1014,
|
| 2823 |
+
"step": 8040
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 39.34106162294082,
|
| 2827 |
+
"grad_norm": 2.848043441772461,
|
| 2828 |
+
"learning_rate": 0.0001478032786885246,
|
| 2829 |
+
"loss": 0.114,
|
| 2830 |
+
"step": 8060
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 39.43868212324588,
|
| 2834 |
+
"grad_norm": 2.6262383460998535,
|
| 2835 |
+
"learning_rate": 0.000147672131147541,
|
| 2836 |
+
"loss": 0.1144,
|
| 2837 |
+
"step": 8080
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 39.536302623550945,
|
| 2841 |
+
"grad_norm": 2.2859811782836914,
|
| 2842 |
+
"learning_rate": 0.00014754098360655738,
|
| 2843 |
+
"loss": 0.1151,
|
| 2844 |
+
"step": 8100
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 39.63392312385601,
|
| 2848 |
+
"grad_norm": 2.8703114986419678,
|
| 2849 |
+
"learning_rate": 0.00014740983606557377,
|
| 2850 |
+
"loss": 0.1084,
|
| 2851 |
+
"step": 8120
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 39.73154362416108,
|
| 2855 |
+
"grad_norm": 2.1996986865997314,
|
| 2856 |
+
"learning_rate": 0.00014727868852459016,
|
| 2857 |
+
"loss": 0.1268,
|
| 2858 |
+
"step": 8140
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 39.82916412446614,
|
| 2862 |
+
"grad_norm": 2.9079976081848145,
|
| 2863 |
+
"learning_rate": 0.00014714754098360658,
|
| 2864 |
+
"loss": 0.1184,
|
| 2865 |
+
"step": 8160
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 39.9267846247712,
|
| 2869 |
+
"grad_norm": 3.584907293319702,
|
| 2870 |
+
"learning_rate": 0.00014701639344262297,
|
| 2871 |
+
"loss": 0.131,
|
| 2872 |
+
"step": 8180
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 40.024405125076264,
|
| 2876 |
+
"grad_norm": 2.2893424034118652,
|
| 2877 |
+
"learning_rate": 0.00014688524590163936,
|
| 2878 |
+
"loss": 0.1111,
|
| 2879 |
+
"step": 8200
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 40.12202562538133,
|
| 2883 |
+
"grad_norm": 3.5768001079559326,
|
| 2884 |
+
"learning_rate": 0.00014675409836065575,
|
| 2885 |
+
"loss": 0.0993,
|
| 2886 |
+
"step": 8220
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 40.219646125686396,
|
| 2890 |
+
"grad_norm": 4.004862308502197,
|
| 2891 |
+
"learning_rate": 0.00014662295081967214,
|
| 2892 |
+
"loss": 0.1048,
|
| 2893 |
+
"step": 8240
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 40.31726662599146,
|
| 2897 |
+
"grad_norm": 2.0014472007751465,
|
| 2898 |
+
"learning_rate": 0.00014649180327868852,
|
| 2899 |
+
"loss": 0.1003,
|
| 2900 |
+
"step": 8260
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 40.41488712629652,
|
| 2904 |
+
"grad_norm": 2.492799997329712,
|
| 2905 |
+
"learning_rate": 0.00014636065573770491,
|
| 2906 |
+
"loss": 0.102,
|
| 2907 |
+
"step": 8280
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 40.51250762660159,
|
| 2911 |
+
"grad_norm": 2.779287099838257,
|
| 2912 |
+
"learning_rate": 0.00014622950819672133,
|
| 2913 |
+
"loss": 0.1035,
|
| 2914 |
+
"step": 8300
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 40.61012812690665,
|
| 2918 |
+
"grad_norm": 3.0162174701690674,
|
| 2919 |
+
"learning_rate": 0.00014609836065573772,
|
| 2920 |
+
"loss": 0.1112,
|
| 2921 |
+
"step": 8320
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 40.707748627211714,
|
| 2925 |
+
"grad_norm": 3.1944477558135986,
|
| 2926 |
+
"learning_rate": 0.0001459672131147541,
|
| 2927 |
+
"loss": 0.1167,
|
| 2928 |
+
"step": 8340
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 40.80536912751678,
|
| 2932 |
+
"grad_norm": 2.849069833755493,
|
| 2933 |
+
"learning_rate": 0.0001458360655737705,
|
| 2934 |
+
"loss": 0.1111,
|
| 2935 |
+
"step": 8360
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 40.902989627821846,
|
| 2939 |
+
"grad_norm": 3.3523714542388916,
|
| 2940 |
+
"learning_rate": 0.0001457049180327869,
|
| 2941 |
+
"loss": 0.1157,
|
| 2942 |
+
"step": 8380
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 41.00061012812691,
|
| 2946 |
+
"grad_norm": 3.1365504264831543,
|
| 2947 |
+
"learning_rate": 0.0001455737704918033,
|
| 2948 |
+
"loss": 0.1228,
|
| 2949 |
+
"step": 8400
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 41.09823062843197,
|
| 2953 |
+
"grad_norm": 2.1375391483306885,
|
| 2954 |
+
"learning_rate": 0.00014544262295081967,
|
| 2955 |
+
"loss": 0.0856,
|
| 2956 |
+
"step": 8420
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 41.19585112873703,
|
| 2960 |
+
"grad_norm": 2.6286661624908447,
|
| 2961 |
+
"learning_rate": 0.00014531147540983608,
|
| 2962 |
+
"loss": 0.0944,
|
| 2963 |
+
"step": 8440
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 41.2934716290421,
|
| 2967 |
+
"grad_norm": 1.7328729629516602,
|
| 2968 |
+
"learning_rate": 0.00014518032786885247,
|
| 2969 |
+
"loss": 0.0933,
|
| 2970 |
+
"step": 8460
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 41.391092129347165,
|
| 2974 |
+
"grad_norm": 2.313880205154419,
|
| 2975 |
+
"learning_rate": 0.00014504918032786886,
|
| 2976 |
+
"loss": 0.1063,
|
| 2977 |
+
"step": 8480
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 41.48871262965223,
|
| 2981 |
+
"grad_norm": 2.3604514598846436,
|
| 2982 |
+
"learning_rate": 0.00014491803278688525,
|
| 2983 |
+
"loss": 0.0963,
|
| 2984 |
+
"step": 8500
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 41.58633312995729,
|
| 2988 |
+
"grad_norm": 2.1344621181488037,
|
| 2989 |
+
"learning_rate": 0.00014478688524590164,
|
| 2990 |
+
"loss": 0.1084,
|
| 2991 |
+
"step": 8520
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 41.68395363026236,
|
| 2995 |
+
"grad_norm": 3.828547239303589,
|
| 2996 |
+
"learning_rate": 0.00014465573770491806,
|
| 2997 |
+
"loss": 0.112,
|
| 2998 |
+
"step": 8540
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 41.78157413056742,
|
| 3002 |
+
"grad_norm": 2.653571367263794,
|
| 3003 |
+
"learning_rate": 0.00014452459016393442,
|
| 3004 |
+
"loss": 0.1241,
|
| 3005 |
+
"step": 8560
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 41.87919463087248,
|
| 3009 |
+
"grad_norm": 2.482323408126831,
|
| 3010 |
+
"learning_rate": 0.00014439344262295084,
|
| 3011 |
+
"loss": 0.1096,
|
| 3012 |
+
"step": 8580
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 41.976815131177545,
|
| 3016 |
+
"grad_norm": 2.309413433074951,
|
| 3017 |
+
"learning_rate": 0.00014426229508196722,
|
| 3018 |
+
"loss": 0.1191,
|
| 3019 |
+
"step": 8600
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 42.074435631482615,
|
| 3023 |
+
"grad_norm": 2.5040862560272217,
|
| 3024 |
+
"learning_rate": 0.00014413114754098361,
|
| 3025 |
+
"loss": 0.0932,
|
| 3026 |
+
"step": 8620
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 42.17205613178768,
|
| 3030 |
+
"grad_norm": 2.5942790508270264,
|
| 3031 |
+
"learning_rate": 0.000144,
|
| 3032 |
+
"loss": 0.0934,
|
| 3033 |
+
"step": 8640
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 42.26967663209274,
|
| 3037 |
+
"grad_norm": 2.0345799922943115,
|
| 3038 |
+
"learning_rate": 0.0001438688524590164,
|
| 3039 |
+
"loss": 0.0913,
|
| 3040 |
+
"step": 8660
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 42.3672971323978,
|
| 3044 |
+
"grad_norm": 2.216151714324951,
|
| 3045 |
+
"learning_rate": 0.0001437377049180328,
|
| 3046 |
+
"loss": 0.0966,
|
| 3047 |
+
"step": 8680
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 42.464917632702864,
|
| 3051 |
+
"grad_norm": 2.8628721237182617,
|
| 3052 |
+
"learning_rate": 0.00014360655737704917,
|
| 3053 |
+
"loss": 0.102,
|
| 3054 |
+
"step": 8700
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 42.56253813300793,
|
| 3058 |
+
"grad_norm": 3.280900239944458,
|
| 3059 |
+
"learning_rate": 0.0001434754098360656,
|
| 3060 |
+
"loss": 0.0922,
|
| 3061 |
+
"step": 8720
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 42.660158633312996,
|
| 3065 |
+
"grad_norm": 2.90677809715271,
|
| 3066 |
+
"learning_rate": 0.00014334426229508198,
|
| 3067 |
+
"loss": 0.1027,
|
| 3068 |
+
"step": 8740
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 42.75777913361806,
|
| 3072 |
+
"grad_norm": 3.169517755508423,
|
| 3073 |
+
"learning_rate": 0.00014321311475409837,
|
| 3074 |
+
"loss": 0.112,
|
| 3075 |
+
"step": 8760
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 42.85539963392312,
|
| 3079 |
+
"grad_norm": 2.617112159729004,
|
| 3080 |
+
"learning_rate": 0.00014308196721311476,
|
| 3081 |
+
"loss": 0.1063,
|
| 3082 |
+
"step": 8780
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 42.95302013422819,
|
| 3086 |
+
"grad_norm": 2.3260226249694824,
|
| 3087 |
+
"learning_rate": 0.00014295081967213115,
|
| 3088 |
+
"loss": 0.1129,
|
| 3089 |
+
"step": 8800
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 43.05064063453325,
|
| 3093 |
+
"grad_norm": 2.255563735961914,
|
| 3094 |
+
"learning_rate": 0.00014281967213114756,
|
| 3095 |
+
"loss": 0.0943,
|
| 3096 |
+
"step": 8820
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 43.148261134838314,
|
| 3100 |
+
"grad_norm": 2.7513527870178223,
|
| 3101 |
+
"learning_rate": 0.00014268852459016395,
|
| 3102 |
+
"loss": 0.0929,
|
| 3103 |
+
"step": 8840
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 43.24588163514338,
|
| 3107 |
+
"grad_norm": 1.92955482006073,
|
| 3108 |
+
"learning_rate": 0.00014255737704918034,
|
| 3109 |
+
"loss": 0.0871,
|
| 3110 |
+
"step": 8860
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 43.343502135448446,
|
| 3114 |
+
"grad_norm": 2.5168724060058594,
|
| 3115 |
+
"learning_rate": 0.00014242622950819673,
|
| 3116 |
+
"loss": 0.0915,
|
| 3117 |
+
"step": 8880
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 43.44112263575351,
|
| 3121 |
+
"grad_norm": 2.497450351715088,
|
| 3122 |
+
"learning_rate": 0.00014229508196721312,
|
| 3123 |
+
"loss": 0.094,
|
| 3124 |
+
"step": 8900
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 43.53874313605857,
|
| 3128 |
+
"grad_norm": 2.645716667175293,
|
| 3129 |
+
"learning_rate": 0.0001421639344262295,
|
| 3130 |
+
"loss": 0.1039,
|
| 3131 |
+
"step": 8920
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 43.63636363636363,
|
| 3135 |
+
"grad_norm": 2.7844698429107666,
|
| 3136 |
+
"learning_rate": 0.0001420327868852459,
|
| 3137 |
+
"loss": 0.0979,
|
| 3138 |
+
"step": 8940
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 43.7339841366687,
|
| 3142 |
+
"grad_norm": 2.3445496559143066,
|
| 3143 |
+
"learning_rate": 0.00014190163934426231,
|
| 3144 |
+
"loss": 0.0936,
|
| 3145 |
+
"step": 8960
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 43.831604636973765,
|
| 3149 |
+
"grad_norm": 3.3879520893096924,
|
| 3150 |
+
"learning_rate": 0.0001417704918032787,
|
| 3151 |
+
"loss": 0.1147,
|
| 3152 |
+
"step": 8980
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 43.92922513727883,
|
| 3156 |
+
"grad_norm": 2.5144670009613037,
|
| 3157 |
+
"learning_rate": 0.0001416393442622951,
|
| 3158 |
+
"loss": 0.1064,
|
| 3159 |
+
"step": 9000
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 44.02684563758389,
|
| 3163 |
+
"grad_norm": 1.741097092628479,
|
| 3164 |
+
"learning_rate": 0.00014150819672131148,
|
| 3165 |
+
"loss": 0.0966,
|
| 3166 |
+
"step": 9020
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 44.12446613788896,
|
| 3170 |
+
"grad_norm": 2.3388848304748535,
|
| 3171 |
+
"learning_rate": 0.00014137704918032787,
|
| 3172 |
+
"loss": 0.0877,
|
| 3173 |
+
"step": 9040
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 44.22208663819402,
|
| 3177 |
+
"grad_norm": 1.8241106271743774,
|
| 3178 |
+
"learning_rate": 0.00014124590163934426,
|
| 3179 |
+
"loss": 0.0937,
|
| 3180 |
+
"step": 9060
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 44.31970713849908,
|
| 3184 |
+
"grad_norm": 2.50146222114563,
|
| 3185 |
+
"learning_rate": 0.00014111475409836065,
|
| 3186 |
+
"loss": 0.0877,
|
| 3187 |
+
"step": 9080
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 44.417327638804146,
|
| 3191 |
+
"grad_norm": 2.029155969619751,
|
| 3192 |
+
"learning_rate": 0.00014098360655737707,
|
| 3193 |
+
"loss": 0.0841,
|
| 3194 |
+
"step": 9100
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 44.514948139109215,
|
| 3198 |
+
"grad_norm": 3.042222738265991,
|
| 3199 |
+
"learning_rate": 0.00014085245901639346,
|
| 3200 |
+
"loss": 0.0956,
|
| 3201 |
+
"step": 9120
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 44.61256863941428,
|
| 3205 |
+
"grad_norm": 3.1099298000335693,
|
| 3206 |
+
"learning_rate": 0.00014072131147540985,
|
| 3207 |
+
"loss": 0.1026,
|
| 3208 |
+
"step": 9140
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 44.71018913971934,
|
| 3212 |
+
"grad_norm": 2.7761776447296143,
|
| 3213 |
+
"learning_rate": 0.00014059016393442624,
|
| 3214 |
+
"loss": 0.0994,
|
| 3215 |
+
"step": 9160
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 44.8078096400244,
|
| 3219 |
+
"grad_norm": 4.281795501708984,
|
| 3220 |
+
"learning_rate": 0.00014045901639344262,
|
| 3221 |
+
"loss": 0.0995,
|
| 3222 |
+
"step": 9180
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 44.90543014032947,
|
| 3226 |
+
"grad_norm": 1.9752144813537598,
|
| 3227 |
+
"learning_rate": 0.00014032786885245904,
|
| 3228 |
+
"loss": 0.1037,
|
| 3229 |
+
"step": 9200
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 45.003050640634534,
|
| 3233 |
+
"grad_norm": 2.5847785472869873,
|
| 3234 |
+
"learning_rate": 0.0001401967213114754,
|
| 3235 |
+
"loss": 0.1025,
|
| 3236 |
+
"step": 9220
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 45.100671140939596,
|
| 3240 |
+
"grad_norm": 3.0032763481140137,
|
| 3241 |
+
"learning_rate": 0.00014006557377049182,
|
| 3242 |
+
"loss": 0.0794,
|
| 3243 |
+
"step": 9240
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 45.19829164124466,
|
| 3247 |
+
"grad_norm": 3.35851788520813,
|
| 3248 |
+
"learning_rate": 0.0001399344262295082,
|
| 3249 |
+
"loss": 0.091,
|
| 3250 |
+
"step": 9260
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 45.29591214154973,
|
| 3254 |
+
"grad_norm": 2.5405397415161133,
|
| 3255 |
+
"learning_rate": 0.0001398032786885246,
|
| 3256 |
+
"loss": 0.0839,
|
| 3257 |
+
"step": 9280
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 45.39353264185479,
|
| 3261 |
+
"grad_norm": 3.7618603706359863,
|
| 3262 |
+
"learning_rate": 0.000139672131147541,
|
| 3263 |
+
"loss": 0.0909,
|
| 3264 |
+
"step": 9300
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 45.49115314215985,
|
| 3268 |
+
"grad_norm": 2.1094727516174316,
|
| 3269 |
+
"learning_rate": 0.00013954098360655738,
|
| 3270 |
+
"loss": 0.0884,
|
| 3271 |
+
"step": 9320
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 45.588773642464915,
|
| 3275 |
+
"grad_norm": 2.060983657836914,
|
| 3276 |
+
"learning_rate": 0.0001394098360655738,
|
| 3277 |
+
"loss": 0.0918,
|
| 3278 |
+
"step": 9340
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 45.686394142769984,
|
| 3282 |
+
"grad_norm": 2.426604747772217,
|
| 3283 |
+
"learning_rate": 0.00013927868852459016,
|
| 3284 |
+
"loss": 0.0943,
|
| 3285 |
+
"step": 9360
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 45.78401464307505,
|
| 3289 |
+
"grad_norm": 2.716404438018799,
|
| 3290 |
+
"learning_rate": 0.00013914754098360657,
|
| 3291 |
+
"loss": 0.0985,
|
| 3292 |
+
"step": 9380
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 45.88163514338011,
|
| 3296 |
+
"grad_norm": 2.1184298992156982,
|
| 3297 |
+
"learning_rate": 0.00013901639344262296,
|
| 3298 |
+
"loss": 0.0941,
|
| 3299 |
+
"step": 9400
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 45.97925564368517,
|
| 3303 |
+
"grad_norm": 2.4884073734283447,
|
| 3304 |
+
"learning_rate": 0.00013888524590163935,
|
| 3305 |
+
"loss": 0.1097,
|
| 3306 |
+
"step": 9420
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 46.07687614399024,
|
| 3310 |
+
"grad_norm": 2.272751808166504,
|
| 3311 |
+
"learning_rate": 0.00013875409836065574,
|
| 3312 |
+
"loss": 0.088,
|
| 3313 |
+
"step": 9440
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 46.1744966442953,
|
| 3317 |
+
"grad_norm": 2.296656847000122,
|
| 3318 |
+
"learning_rate": 0.00013862295081967213,
|
| 3319 |
+
"loss": 0.0813,
|
| 3320 |
+
"step": 9460
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 46.272117144600365,
|
| 3324 |
+
"grad_norm": 2.6703407764434814,
|
| 3325 |
+
"learning_rate": 0.00013849180327868855,
|
| 3326 |
+
"loss": 0.0837,
|
| 3327 |
+
"step": 9480
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 46.36973764490543,
|
| 3331 |
+
"grad_norm": 3.1443161964416504,
|
| 3332 |
+
"learning_rate": 0.0001383606557377049,
|
| 3333 |
+
"loss": 0.0876,
|
| 3334 |
+
"step": 9500
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 46.4673581452105,
|
| 3338 |
+
"grad_norm": 2.6503586769104004,
|
| 3339 |
+
"learning_rate": 0.00013822950819672133,
|
| 3340 |
+
"loss": 0.0835,
|
| 3341 |
+
"step": 9520
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 46.56497864551556,
|
| 3345 |
+
"grad_norm": 2.3241240978240967,
|
| 3346 |
+
"learning_rate": 0.00013809836065573771,
|
| 3347 |
+
"loss": 0.0996,
|
| 3348 |
+
"step": 9540
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 46.66259914582062,
|
| 3352 |
+
"grad_norm": 2.8960094451904297,
|
| 3353 |
+
"learning_rate": 0.0001379672131147541,
|
| 3354 |
+
"loss": 0.0904,
|
| 3355 |
+
"step": 9560
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 46.760219646125684,
|
| 3359 |
+
"grad_norm": 3.4559171199798584,
|
| 3360 |
+
"learning_rate": 0.0001378360655737705,
|
| 3361 |
+
"loss": 0.0906,
|
| 3362 |
+
"step": 9580
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 46.85784014643075,
|
| 3366 |
+
"grad_norm": 2.325892925262451,
|
| 3367 |
+
"learning_rate": 0.00013770491803278688,
|
| 3368 |
+
"loss": 0.1012,
|
| 3369 |
+
"step": 9600
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 46.955460646735816,
|
| 3373 |
+
"grad_norm": 2.833704948425293,
|
| 3374 |
+
"learning_rate": 0.0001375737704918033,
|
| 3375 |
+
"loss": 0.0947,
|
| 3376 |
+
"step": 9620
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 47.05308114704088,
|
| 3380 |
+
"grad_norm": 2.060310125350952,
|
| 3381 |
+
"learning_rate": 0.0001374426229508197,
|
| 3382 |
+
"loss": 0.0861,
|
| 3383 |
+
"step": 9640
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 47.15070164734594,
|
| 3387 |
+
"grad_norm": 2.1415886878967285,
|
| 3388 |
+
"learning_rate": 0.00013731147540983608,
|
| 3389 |
+
"loss": 0.0775,
|
| 3390 |
+
"step": 9660
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 47.24832214765101,
|
| 3394 |
+
"grad_norm": 1.9478198289871216,
|
| 3395 |
+
"learning_rate": 0.00013718032786885247,
|
| 3396 |
+
"loss": 0.082,
|
| 3397 |
+
"step": 9680
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 47.34594264795607,
|
| 3401 |
+
"grad_norm": 4.712246417999268,
|
| 3402 |
+
"learning_rate": 0.00013704918032786886,
|
| 3403 |
+
"loss": 0.0753,
|
| 3404 |
+
"step": 9700
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 47.443563148261134,
|
| 3408 |
+
"grad_norm": 2.2505266666412354,
|
| 3409 |
+
"learning_rate": 0.00013691803278688525,
|
| 3410 |
+
"loss": 0.0809,
|
| 3411 |
+
"step": 9720
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 47.5411836485662,
|
| 3415 |
+
"grad_norm": 2.4680440425872803,
|
| 3416 |
+
"learning_rate": 0.00013678688524590164,
|
| 3417 |
+
"loss": 0.0923,
|
| 3418 |
+
"step": 9740
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 47.638804148871266,
|
| 3422 |
+
"grad_norm": 2.5806798934936523,
|
| 3423 |
+
"learning_rate": 0.00013665573770491805,
|
| 3424 |
+
"loss": 0.0938,
|
| 3425 |
+
"step": 9760
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 47.73642464917633,
|
| 3429 |
+
"grad_norm": 2.5440475940704346,
|
| 3430 |
+
"learning_rate": 0.00013652459016393444,
|
| 3431 |
+
"loss": 0.0911,
|
| 3432 |
+
"step": 9780
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 47.83404514948139,
|
| 3436 |
+
"grad_norm": 2.316375732421875,
|
| 3437 |
+
"learning_rate": 0.00013639344262295083,
|
| 3438 |
+
"loss": 0.0958,
|
| 3439 |
+
"step": 9800
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 47.93166564978645,
|
| 3443 |
+
"grad_norm": 2.1451222896575928,
|
| 3444 |
+
"learning_rate": 0.00013626229508196722,
|
| 3445 |
+
"loss": 0.0971,
|
| 3446 |
+
"step": 9820
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 48.02928615009152,
|
| 3450 |
+
"grad_norm": 2.381439685821533,
|
| 3451 |
+
"learning_rate": 0.0001361311475409836,
|
| 3452 |
+
"loss": 0.0944,
|
| 3453 |
+
"step": 9840
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 48.126906650396585,
|
| 3457 |
+
"grad_norm": 2.9612882137298584,
|
| 3458 |
+
"learning_rate": 0.00013600000000000003,
|
| 3459 |
+
"loss": 0.0676,
|
| 3460 |
+
"step": 9860
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 48.22452715070165,
|
| 3464 |
+
"grad_norm": 1.7020941972732544,
|
| 3465 |
+
"learning_rate": 0.0001358688524590164,
|
| 3466 |
+
"loss": 0.0835,
|
| 3467 |
+
"step": 9880
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 48.32214765100671,
|
| 3471 |
+
"grad_norm": 1.820823311805725,
|
| 3472 |
+
"learning_rate": 0.0001357377049180328,
|
| 3473 |
+
"loss": 0.0773,
|
| 3474 |
+
"step": 9900
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 48.41976815131178,
|
| 3478 |
+
"grad_norm": 1.7456953525543213,
|
| 3479 |
+
"learning_rate": 0.0001356065573770492,
|
| 3480 |
+
"loss": 0.0847,
|
| 3481 |
+
"step": 9920
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 48.51738865161684,
|
| 3485 |
+
"grad_norm": 2.277359962463379,
|
| 3486 |
+
"learning_rate": 0.00013547540983606556,
|
| 3487 |
+
"loss": 0.0879,
|
| 3488 |
+
"step": 9940
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 48.6150091519219,
|
| 3492 |
+
"grad_norm": 1.9201923608779907,
|
| 3493 |
+
"learning_rate": 0.00013534426229508197,
|
| 3494 |
+
"loss": 0.0886,
|
| 3495 |
+
"step": 9960
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 48.712629652226966,
|
| 3499 |
+
"grad_norm": 2.5674726963043213,
|
| 3500 |
+
"learning_rate": 0.00013521311475409836,
|
| 3501 |
+
"loss": 0.0828,
|
| 3502 |
+
"step": 9980
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 48.810250152532035,
|
| 3506 |
+
"grad_norm": 2.4883131980895996,
|
| 3507 |
+
"learning_rate": 0.00013508196721311478,
|
| 3508 |
+
"loss": 0.0949,
|
| 3509 |
+
"step": 10000
|
| 3510 |
}
|
| 3511 |
],
|
| 3512 |
"logging_steps": 20,
|
|
|
|
| 3526 |
"attributes": {}
|
| 3527 |
}
|
| 3528 |
},
|
| 3529 |
+
"total_flos": 1.0956762891676877e+17,
|
| 3530 |
"train_batch_size": 1,
|
| 3531 |
"trial_name": null,
|
| 3532 |
"trial_params": null
|