Update to checkpoint 2000 (20% trained)
Browse files- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- trainer_state.json +1234 -0
- training_args.bin +3 -0
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4990120184
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21a0739be3bb3af3aef54bbd801ad9b72e447a36097840ff5da6ac12e9b5fbab
|
| 3 |
size 4990120184
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4823190320
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:577c53d3fdcad82c9e55554ee82db790a9715a97e97aaf32501b1d5d026a82e3
|
| 3 |
size 4823190320
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1234 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"grad_norm": 1.8000761270523071,
|
| 14 |
+
"learning_rate": 1.8e-06,
|
| 15 |
+
"loss": 1.1194,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"grad_norm": 0.8661630153656006,
|
| 20 |
+
"learning_rate": 3.8e-06,
|
| 21 |
+
"loss": 1.1135,
|
| 22 |
+
"step": 20
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"grad_norm": 0.7289697527885437,
|
| 26 |
+
"learning_rate": 5.8e-06,
|
| 27 |
+
"loss": 1.0689,
|
| 28 |
+
"step": 30
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"grad_norm": 0.4166327118873596,
|
| 32 |
+
"learning_rate": 7.8e-06,
|
| 33 |
+
"loss": 1.0905,
|
| 34 |
+
"step": 40
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"grad_norm": 0.526554524898529,
|
| 38 |
+
"learning_rate": 9.800000000000001e-06,
|
| 39 |
+
"loss": 1.0538,
|
| 40 |
+
"step": 50
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"grad_norm": 0.4327709972858429,
|
| 44 |
+
"learning_rate": 1.18e-05,
|
| 45 |
+
"loss": 1.0508,
|
| 46 |
+
"step": 60
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"grad_norm": 0.538928210735321,
|
| 50 |
+
"learning_rate": 1.3800000000000002e-05,
|
| 51 |
+
"loss": 1.0425,
|
| 52 |
+
"step": 70
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"grad_norm": 0.5299176573753357,
|
| 56 |
+
"learning_rate": 1.58e-05,
|
| 57 |
+
"loss": 1.0375,
|
| 58 |
+
"step": 80
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"grad_norm": 0.5128000974655151,
|
| 62 |
+
"learning_rate": 1.78e-05,
|
| 63 |
+
"loss": 1.0294,
|
| 64 |
+
"step": 90
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"grad_norm": 0.5661309361457825,
|
| 68 |
+
"learning_rate": 1.9800000000000004e-05,
|
| 69 |
+
"loss": 1.0436,
|
| 70 |
+
"step": 100
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"grad_norm": 0.4856761395931244,
|
| 74 |
+
"learning_rate": 2.18e-05,
|
| 75 |
+
"loss": 1.0431,
|
| 76 |
+
"step": 110
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"grad_norm": 0.4439884126186371,
|
| 80 |
+
"learning_rate": 2.38e-05,
|
| 81 |
+
"loss": 1.0226,
|
| 82 |
+
"step": 120
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"grad_norm": 0.6393558382987976,
|
| 86 |
+
"learning_rate": 2.58e-05,
|
| 87 |
+
"loss": 1.0128,
|
| 88 |
+
"step": 130
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"grad_norm": 1.510035753250122,
|
| 92 |
+
"learning_rate": 2.7800000000000005e-05,
|
| 93 |
+
"loss": 0.9712,
|
| 94 |
+
"step": 140
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"grad_norm": 0.8571694493293762,
|
| 98 |
+
"learning_rate": 2.98e-05,
|
| 99 |
+
"loss": 0.9299,
|
| 100 |
+
"step": 150
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"grad_norm": 1.6079281568527222,
|
| 104 |
+
"learning_rate": 3.18e-05,
|
| 105 |
+
"loss": 0.9007,
|
| 106 |
+
"step": 160
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"grad_norm": 1.5867490768432617,
|
| 110 |
+
"learning_rate": 3.38e-05,
|
| 111 |
+
"loss": 0.8255,
|
| 112 |
+
"step": 170
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"grad_norm": 2.30320405960083,
|
| 116 |
+
"learning_rate": 3.58e-05,
|
| 117 |
+
"loss": 0.751,
|
| 118 |
+
"step": 180
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"grad_norm": 1.6800912618637085,
|
| 122 |
+
"learning_rate": 3.7800000000000004e-05,
|
| 123 |
+
"loss": 0.6726,
|
| 124 |
+
"step": 190
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"grad_norm": 2.149949789047241,
|
| 128 |
+
"learning_rate": 3.9800000000000005e-05,
|
| 129 |
+
"loss": 0.6235,
|
| 130 |
+
"step": 200
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"grad_norm": 1.355879545211792,
|
| 134 |
+
"learning_rate": 4.18e-05,
|
| 135 |
+
"loss": 0.5208,
|
| 136 |
+
"step": 210
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"grad_norm": 2.6391069889068604,
|
| 140 |
+
"learning_rate": 4.38e-05,
|
| 141 |
+
"loss": 0.4078,
|
| 142 |
+
"step": 220
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"grad_norm": 2.550145387649536,
|
| 146 |
+
"learning_rate": 4.58e-05,
|
| 147 |
+
"loss": 0.2918,
|
| 148 |
+
"step": 230
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"grad_norm": 1.527336835861206,
|
| 152 |
+
"learning_rate": 4.78e-05,
|
| 153 |
+
"loss": 0.2215,
|
| 154 |
+
"step": 240
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"grad_norm": 1.369310975074768,
|
| 158 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 159 |
+
"loss": 0.1697,
|
| 160 |
+
"step": 250
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"grad_norm": 1.200178861618042,
|
| 164 |
+
"learning_rate": 5.1800000000000005e-05,
|
| 165 |
+
"loss": 0.1496,
|
| 166 |
+
"step": 260
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"grad_norm": 1.654726505279541,
|
| 170 |
+
"learning_rate": 5.380000000000001e-05,
|
| 171 |
+
"loss": 0.1612,
|
| 172 |
+
"step": 270
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"grad_norm": 1.003008484840393,
|
| 176 |
+
"learning_rate": 5.580000000000001e-05,
|
| 177 |
+
"loss": 0.1355,
|
| 178 |
+
"step": 280
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"grad_norm": 1.2864620685577393,
|
| 182 |
+
"learning_rate": 5.7799999999999995e-05,
|
| 183 |
+
"loss": 0.1646,
|
| 184 |
+
"step": 290
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"grad_norm": 1.5691300630569458,
|
| 188 |
+
"learning_rate": 5.9800000000000003e-05,
|
| 189 |
+
"loss": 0.14,
|
| 190 |
+
"step": 300
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"grad_norm": 1.1779141426086426,
|
| 194 |
+
"learning_rate": 6.18e-05,
|
| 195 |
+
"loss": 0.1405,
|
| 196 |
+
"step": 310
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"grad_norm": 1.0536258220672607,
|
| 200 |
+
"learning_rate": 6.38e-05,
|
| 201 |
+
"loss": 0.1245,
|
| 202 |
+
"step": 320
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"grad_norm": 0.939816951751709,
|
| 206 |
+
"learning_rate": 6.58e-05,
|
| 207 |
+
"loss": 0.1273,
|
| 208 |
+
"step": 330
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"grad_norm": 1.1611407995224,
|
| 212 |
+
"learning_rate": 6.780000000000001e-05,
|
| 213 |
+
"loss": 0.1054,
|
| 214 |
+
"step": 340
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"grad_norm": 1.0939171314239502,
|
| 218 |
+
"learning_rate": 6.98e-05,
|
| 219 |
+
"loss": 0.0992,
|
| 220 |
+
"step": 350
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"grad_norm": 1.1488924026489258,
|
| 224 |
+
"learning_rate": 7.18e-05,
|
| 225 |
+
"loss": 0.1146,
|
| 226 |
+
"step": 360
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"grad_norm": 0.8890052437782288,
|
| 230 |
+
"learning_rate": 7.38e-05,
|
| 231 |
+
"loss": 0.1211,
|
| 232 |
+
"step": 370
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"grad_norm": 0.7940606474876404,
|
| 236 |
+
"learning_rate": 7.58e-05,
|
| 237 |
+
"loss": 0.0995,
|
| 238 |
+
"step": 380
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"grad_norm": 0.8308754563331604,
|
| 242 |
+
"learning_rate": 7.780000000000001e-05,
|
| 243 |
+
"loss": 0.1037,
|
| 244 |
+
"step": 390
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"grad_norm": 0.8870435953140259,
|
| 248 |
+
"learning_rate": 7.98e-05,
|
| 249 |
+
"loss": 0.097,
|
| 250 |
+
"step": 400
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"grad_norm": 0.6374945640563965,
|
| 254 |
+
"learning_rate": 8.18e-05,
|
| 255 |
+
"loss": 0.1025,
|
| 256 |
+
"step": 410
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"grad_norm": 1.0475709438323975,
|
| 260 |
+
"learning_rate": 8.38e-05,
|
| 261 |
+
"loss": 0.097,
|
| 262 |
+
"step": 420
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"grad_norm": 0.8428962230682373,
|
| 266 |
+
"learning_rate": 8.58e-05,
|
| 267 |
+
"loss": 0.1012,
|
| 268 |
+
"step": 430
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"grad_norm": 0.7878981828689575,
|
| 272 |
+
"learning_rate": 8.78e-05,
|
| 273 |
+
"loss": 0.095,
|
| 274 |
+
"step": 440
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"grad_norm": 0.8133223652839661,
|
| 278 |
+
"learning_rate": 8.98e-05,
|
| 279 |
+
"loss": 0.0933,
|
| 280 |
+
"step": 450
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"grad_norm": 0.5507503747940063,
|
| 284 |
+
"learning_rate": 9.180000000000001e-05,
|
| 285 |
+
"loss": 0.1006,
|
| 286 |
+
"step": 460
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"grad_norm": 0.7642120122909546,
|
| 290 |
+
"learning_rate": 9.38e-05,
|
| 291 |
+
"loss": 0.1012,
|
| 292 |
+
"step": 470
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"grad_norm": 0.8028876185417175,
|
| 296 |
+
"learning_rate": 9.58e-05,
|
| 297 |
+
"loss": 0.093,
|
| 298 |
+
"step": 480
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"grad_norm": 0.6212483048439026,
|
| 302 |
+
"learning_rate": 9.78e-05,
|
| 303 |
+
"loss": 0.0958,
|
| 304 |
+
"step": 490
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"grad_norm": 0.6925771236419678,
|
| 308 |
+
"learning_rate": 9.98e-05,
|
| 309 |
+
"loss": 0.0934,
|
| 310 |
+
"step": 500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"grad_norm": 0.7765862345695496,
|
| 314 |
+
"learning_rate": 9.9999778549206e-05,
|
| 315 |
+
"loss": 0.0968,
|
| 316 |
+
"step": 510
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"grad_norm": 0.5598022937774658,
|
| 320 |
+
"learning_rate": 9.999901304280685e-05,
|
| 321 |
+
"loss": 0.0951,
|
| 322 |
+
"step": 520
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"grad_norm": 0.6049850583076477,
|
| 326 |
+
"learning_rate": 9.999770075521164e-05,
|
| 327 |
+
"loss": 0.1006,
|
| 328 |
+
"step": 530
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"grad_norm": 0.8623297810554504,
|
| 332 |
+
"learning_rate": 9.99958417007713e-05,
|
| 333 |
+
"loss": 0.0938,
|
| 334 |
+
"step": 540
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"grad_norm": 0.5017800331115723,
|
| 338 |
+
"learning_rate": 9.999343589981615e-05,
|
| 339 |
+
"loss": 0.0961,
|
| 340 |
+
"step": 550
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"grad_norm": 0.586384117603302,
|
| 344 |
+
"learning_rate": 9.999048337865568e-05,
|
| 345 |
+
"loss": 0.0807,
|
| 346 |
+
"step": 560
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"grad_norm": 0.5691239237785339,
|
| 350 |
+
"learning_rate": 9.998698416957815e-05,
|
| 351 |
+
"loss": 0.0761,
|
| 352 |
+
"step": 570
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"grad_norm": 0.5205549597740173,
|
| 356 |
+
"learning_rate": 9.998293831085037e-05,
|
| 357 |
+
"loss": 0.079,
|
| 358 |
+
"step": 580
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"grad_norm": 0.7439537048339844,
|
| 362 |
+
"learning_rate": 9.997834584671719e-05,
|
| 363 |
+
"loss": 0.0848,
|
| 364 |
+
"step": 590
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"grad_norm": 0.5913628935813904,
|
| 368 |
+
"learning_rate": 9.997320682740107e-05,
|
| 369 |
+
"loss": 0.0821,
|
| 370 |
+
"step": 600
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"grad_norm": 0.6202144026756287,
|
| 374 |
+
"learning_rate": 9.996752130910149e-05,
|
| 375 |
+
"loss": 0.0909,
|
| 376 |
+
"step": 610
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"grad_norm": 0.5822686553001404,
|
| 380 |
+
"learning_rate": 9.99612893539944e-05,
|
| 381 |
+
"loss": 0.0775,
|
| 382 |
+
"step": 620
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"grad_norm": 0.6942011713981628,
|
| 386 |
+
"learning_rate": 9.995451103023144e-05,
|
| 387 |
+
"loss": 0.0762,
|
| 388 |
+
"step": 630
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"grad_norm": 0.5743650794029236,
|
| 392 |
+
"learning_rate": 9.994718641193928e-05,
|
| 393 |
+
"loss": 0.0721,
|
| 394 |
+
"step": 640
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"grad_norm": 0.6001242399215698,
|
| 398 |
+
"learning_rate": 9.993931557921874e-05,
|
| 399 |
+
"loss": 0.0786,
|
| 400 |
+
"step": 650
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"grad_norm": 0.4573017954826355,
|
| 404 |
+
"learning_rate": 9.993089861814402e-05,
|
| 405 |
+
"loss": 0.0742,
|
| 406 |
+
"step": 660
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"grad_norm": 0.6226702332496643,
|
| 410 |
+
"learning_rate": 9.992193562076166e-05,
|
| 411 |
+
"loss": 0.0874,
|
| 412 |
+
"step": 670
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"grad_norm": 0.6129893660545349,
|
| 416 |
+
"learning_rate": 9.991242668508954e-05,
|
| 417 |
+
"loss": 0.0742,
|
| 418 |
+
"step": 680
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"grad_norm": 0.4674377739429474,
|
| 422 |
+
"learning_rate": 9.990237191511587e-05,
|
| 423 |
+
"loss": 0.0725,
|
| 424 |
+
"step": 690
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"grad_norm": 0.4674297869205475,
|
| 428 |
+
"learning_rate": 9.989177142079802e-05,
|
| 429 |
+
"loss": 0.0738,
|
| 430 |
+
"step": 700
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"grad_norm": 0.5865145325660706,
|
| 434 |
+
"learning_rate": 9.988062531806126e-05,
|
| 435 |
+
"loss": 0.0887,
|
| 436 |
+
"step": 710
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"grad_norm": 0.7082783579826355,
|
| 440 |
+
"learning_rate": 9.986893372879762e-05,
|
| 441 |
+
"loss": 0.081,
|
| 442 |
+
"step": 720
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"grad_norm": 0.3989603817462921,
|
| 446 |
+
"learning_rate": 9.985669678086443e-05,
|
| 447 |
+
"loss": 0.0712,
|
| 448 |
+
"step": 730
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"grad_norm": 0.6835122108459473,
|
| 452 |
+
"learning_rate": 9.984391460808298e-05,
|
| 453 |
+
"loss": 0.0634,
|
| 454 |
+
"step": 740
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"grad_norm": 0.5364878177642822,
|
| 458 |
+
"learning_rate": 9.983058735023709e-05,
|
| 459 |
+
"loss": 0.0732,
|
| 460 |
+
"step": 750
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"grad_norm": 0.35529661178588867,
|
| 464 |
+
"learning_rate": 9.98167151530715e-05,
|
| 465 |
+
"loss": 0.0634,
|
| 466 |
+
"step": 760
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"grad_norm": 0.40558916330337524,
|
| 470 |
+
"learning_rate": 9.980229816829034e-05,
|
| 471 |
+
"loss": 0.0711,
|
| 472 |
+
"step": 770
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"grad_norm": 0.5186157822608948,
|
| 476 |
+
"learning_rate": 9.978733655355544e-05,
|
| 477 |
+
"loss": 0.0716,
|
| 478 |
+
"step": 780
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"grad_norm": 0.5230529308319092,
|
| 482 |
+
"learning_rate": 9.977183047248464e-05,
|
| 483 |
+
"loss": 0.0665,
|
| 484 |
+
"step": 790
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"grad_norm": 0.55573970079422,
|
| 488 |
+
"learning_rate": 9.975578009464992e-05,
|
| 489 |
+
"loss": 0.0731,
|
| 490 |
+
"step": 800
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"grad_norm": 0.4610231816768646,
|
| 494 |
+
"learning_rate": 9.97391855955757e-05,
|
| 495 |
+
"loss": 0.0774,
|
| 496 |
+
"step": 810
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"grad_norm": 0.42752256989479065,
|
| 500 |
+
"learning_rate": 9.972204715673669e-05,
|
| 501 |
+
"loss": 0.0684,
|
| 502 |
+
"step": 820
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"grad_norm": 0.45933523774147034,
|
| 506 |
+
"learning_rate": 9.970436496555617e-05,
|
| 507 |
+
"loss": 0.0655,
|
| 508 |
+
"step": 830
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"grad_norm": 0.42024242877960205,
|
| 512 |
+
"learning_rate": 9.968613921540373e-05,
|
| 513 |
+
"loss": 0.0622,
|
| 514 |
+
"step": 840
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"grad_norm": 0.5874196290969849,
|
| 518 |
+
"learning_rate": 9.966737010559326e-05,
|
| 519 |
+
"loss": 0.0727,
|
| 520 |
+
"step": 850
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"grad_norm": 0.5695536732673645,
|
| 524 |
+
"learning_rate": 9.964805784138072e-05,
|
| 525 |
+
"loss": 0.0692,
|
| 526 |
+
"step": 860
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"grad_norm": 0.385803759098053,
|
| 530 |
+
"learning_rate": 9.962820263396195e-05,
|
| 531 |
+
"loss": 0.0633,
|
| 532 |
+
"step": 870
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"grad_norm": 0.43449321389198303,
|
| 536 |
+
"learning_rate": 9.960780470047033e-05,
|
| 537 |
+
"loss": 0.0564,
|
| 538 |
+
"step": 880
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"grad_norm": 0.3207187354564667,
|
| 542 |
+
"learning_rate": 9.958686426397437e-05,
|
| 543 |
+
"loss": 0.0621,
|
| 544 |
+
"step": 890
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"grad_norm": 0.3849020302295685,
|
| 548 |
+
"learning_rate": 9.956538155347534e-05,
|
| 549 |
+
"loss": 0.0732,
|
| 550 |
+
"step": 900
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"grad_norm": 0.3251587748527527,
|
| 554 |
+
"learning_rate": 9.95433568039047e-05,
|
| 555 |
+
"loss": 0.0578,
|
| 556 |
+
"step": 910
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"grad_norm": 0.35808417201042175,
|
| 560 |
+
"learning_rate": 9.952079025612162e-05,
|
| 561 |
+
"loss": 0.0601,
|
| 562 |
+
"step": 920
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"grad_norm": 0.2598479390144348,
|
| 566 |
+
"learning_rate": 9.949768215691022e-05,
|
| 567 |
+
"loss": 0.0587,
|
| 568 |
+
"step": 930
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"grad_norm": 0.44453999400138855,
|
| 572 |
+
"learning_rate": 9.9474032758977e-05,
|
| 573 |
+
"loss": 0.0613,
|
| 574 |
+
"step": 940
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"grad_norm": 0.5126887559890747,
|
| 578 |
+
"learning_rate": 9.944984232094794e-05,
|
| 579 |
+
"loss": 0.0697,
|
| 580 |
+
"step": 950
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"grad_norm": 0.34455230832099915,
|
| 584 |
+
"learning_rate": 9.942511110736584e-05,
|
| 585 |
+
"loss": 0.0602,
|
| 586 |
+
"step": 960
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"grad_norm": 0.5941815376281738,
|
| 590 |
+
"learning_rate": 9.939983938868726e-05,
|
| 591 |
+
"loss": 0.0611,
|
| 592 |
+
"step": 970
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"grad_norm": 0.4752832055091858,
|
| 596 |
+
"learning_rate": 9.93740274412797e-05,
|
| 597 |
+
"loss": 0.0741,
|
| 598 |
+
"step": 980
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"grad_norm": 0.46665894985198975,
|
| 602 |
+
"learning_rate": 9.934767554741846e-05,
|
| 603 |
+
"loss": 0.0674,
|
| 604 |
+
"step": 990
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"grad_norm": 0.5232179164886475,
|
| 608 |
+
"learning_rate": 9.932078399528361e-05,
|
| 609 |
+
"loss": 0.0624,
|
| 610 |
+
"step": 1000
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"grad_norm": 0.28243017196655273,
|
| 614 |
+
"learning_rate": 9.929335307895689e-05,
|
| 615 |
+
"loss": 0.0571,
|
| 616 |
+
"step": 1010
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"grad_norm": 0.6062475442886353,
|
| 620 |
+
"learning_rate": 9.926538309841839e-05,
|
| 621 |
+
"loss": 0.0665,
|
| 622 |
+
"step": 1020
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"grad_norm": 0.40058252215385437,
|
| 626 |
+
"learning_rate": 9.923687435954334e-05,
|
| 627 |
+
"loss": 0.0642,
|
| 628 |
+
"step": 1030
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"grad_norm": 0.5141247510910034,
|
| 632 |
+
"learning_rate": 9.920782717409873e-05,
|
| 633 |
+
"loss": 0.0672,
|
| 634 |
+
"step": 1040
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"grad_norm": 0.4424152076244354,
|
| 638 |
+
"learning_rate": 9.917824185973994e-05,
|
| 639 |
+
"loss": 0.0585,
|
| 640 |
+
"step": 1050
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"grad_norm": 0.5799531936645508,
|
| 644 |
+
"learning_rate": 9.914811874000723e-05,
|
| 645 |
+
"loss": 0.0671,
|
| 646 |
+
"step": 1060
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"grad_norm": 0.2700813412666321,
|
| 650 |
+
"learning_rate": 9.911745814432218e-05,
|
| 651 |
+
"loss": 0.0597,
|
| 652 |
+
"step": 1070
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"grad_norm": 0.5105420351028442,
|
| 656 |
+
"learning_rate": 9.90862604079842e-05,
|
| 657 |
+
"loss": 0.0659,
|
| 658 |
+
"step": 1080
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"grad_norm": 0.28891175985336304,
|
| 662 |
+
"learning_rate": 9.90545258721667e-05,
|
| 663 |
+
"loss": 0.0537,
|
| 664 |
+
"step": 1090
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"grad_norm": 0.3393675982952118,
|
| 668 |
+
"learning_rate": 9.90222548839135e-05,
|
| 669 |
+
"loss": 0.0588,
|
| 670 |
+
"step": 1100
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"grad_norm": 0.5993463397026062,
|
| 674 |
+
"learning_rate": 9.898944779613495e-05,
|
| 675 |
+
"loss": 0.0597,
|
| 676 |
+
"step": 1110
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"grad_norm": 0.36953073740005493,
|
| 680 |
+
"learning_rate": 9.89561049676041e-05,
|
| 681 |
+
"loss": 0.0663,
|
| 682 |
+
"step": 1120
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"grad_norm": 0.4737951159477234,
|
| 686 |
+
"learning_rate": 9.89222267629528e-05,
|
| 687 |
+
"loss": 0.0723,
|
| 688 |
+
"step": 1130
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"grad_norm": 0.4001386761665344,
|
| 692 |
+
"learning_rate": 9.888781355266763e-05,
|
| 693 |
+
"loss": 0.0677,
|
| 694 |
+
"step": 1140
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"grad_norm": 0.3812955617904663,
|
| 698 |
+
"learning_rate": 9.885286571308598e-05,
|
| 699 |
+
"loss": 0.0611,
|
| 700 |
+
"step": 1150
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"grad_norm": 0.4578457176685333,
|
| 704 |
+
"learning_rate": 9.881738362639182e-05,
|
| 705 |
+
"loss": 0.0661,
|
| 706 |
+
"step": 1160
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"grad_norm": 0.4562247693538666,
|
| 710 |
+
"learning_rate": 9.878136768061154e-05,
|
| 711 |
+
"loss": 0.0608,
|
| 712 |
+
"step": 1170
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"grad_norm": 0.3326525390148163,
|
| 716 |
+
"learning_rate": 9.874481826960979e-05,
|
| 717 |
+
"loss": 0.0542,
|
| 718 |
+
"step": 1180
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"grad_norm": 0.4416123330593109,
|
| 722 |
+
"learning_rate": 9.870773579308503e-05,
|
| 723 |
+
"loss": 0.0549,
|
| 724 |
+
"step": 1190
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"grad_norm": 0.3314540684223175,
|
| 728 |
+
"learning_rate": 9.867012065656533e-05,
|
| 729 |
+
"loss": 0.0617,
|
| 730 |
+
"step": 1200
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"grad_norm": 0.45206043124198914,
|
| 734 |
+
"learning_rate": 9.863197327140376e-05,
|
| 735 |
+
"loss": 0.0616,
|
| 736 |
+
"step": 1210
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"grad_norm": 0.23829543590545654,
|
| 740 |
+
"learning_rate": 9.859329405477403e-05,
|
| 741 |
+
"loss": 0.0538,
|
| 742 |
+
"step": 1220
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"grad_norm": 0.40245768427848816,
|
| 746 |
+
"learning_rate": 9.855408342966585e-05,
|
| 747 |
+
"loss": 0.0599,
|
| 748 |
+
"step": 1230
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"grad_norm": 0.36202260851860046,
|
| 752 |
+
"learning_rate": 9.851434182488033e-05,
|
| 753 |
+
"loss": 0.0563,
|
| 754 |
+
"step": 1240
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"grad_norm": 0.3963845372200012,
|
| 758 |
+
"learning_rate": 9.84740696750253e-05,
|
| 759 |
+
"loss": 0.0631,
|
| 760 |
+
"step": 1250
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"grad_norm": 0.5504999756813049,
|
| 764 |
+
"learning_rate": 9.843326742051055e-05,
|
| 765 |
+
"loss": 0.0547,
|
| 766 |
+
"step": 1260
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"grad_norm": 0.34831398725509644,
|
| 770 |
+
"learning_rate": 9.839193550754297e-05,
|
| 771 |
+
"loss": 0.0624,
|
| 772 |
+
"step": 1270
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"grad_norm": 0.3962138295173645,
|
| 776 |
+
"learning_rate": 9.835007438812177e-05,
|
| 777 |
+
"loss": 0.065,
|
| 778 |
+
"step": 1280
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"grad_norm": 0.5264518857002258,
|
| 782 |
+
"learning_rate": 9.830768452003341e-05,
|
| 783 |
+
"loss": 0.0629,
|
| 784 |
+
"step": 1290
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"grad_norm": 0.4762389659881592,
|
| 788 |
+
"learning_rate": 9.826476636684671e-05,
|
| 789 |
+
"loss": 0.061,
|
| 790 |
+
"step": 1300
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"grad_norm": 0.4322468638420105,
|
| 794 |
+
"learning_rate": 9.822132039790773e-05,
|
| 795 |
+
"loss": 0.0587,
|
| 796 |
+
"step": 1310
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"grad_norm": 0.3566872477531433,
|
| 800 |
+
"learning_rate": 9.817734708833461e-05,
|
| 801 |
+
"loss": 0.0591,
|
| 802 |
+
"step": 1320
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"grad_norm": 0.35532960295677185,
|
| 806 |
+
"learning_rate": 9.813284691901243e-05,
|
| 807 |
+
"loss": 0.0562,
|
| 808 |
+
"step": 1330
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"grad_norm": 0.46946966648101807,
|
| 812 |
+
"learning_rate": 9.808782037658792e-05,
|
| 813 |
+
"loss": 0.0571,
|
| 814 |
+
"step": 1340
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"grad_norm": 0.332731693983078,
|
| 818 |
+
"learning_rate": 9.804226795346411e-05,
|
| 819 |
+
"loss": 0.054,
|
| 820 |
+
"step": 1350
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"grad_norm": 0.28058159351348877,
|
| 824 |
+
"learning_rate": 9.799619014779503e-05,
|
| 825 |
+
"loss": 0.0594,
|
| 826 |
+
"step": 1360
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"grad_norm": 0.4327380955219269,
|
| 830 |
+
"learning_rate": 9.794958746348013e-05,
|
| 831 |
+
"loss": 0.049,
|
| 832 |
+
"step": 1370
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"grad_norm": 0.2480960488319397,
|
| 836 |
+
"learning_rate": 9.790246041015896e-05,
|
| 837 |
+
"loss": 0.056,
|
| 838 |
+
"step": 1380
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"grad_norm": 0.3483447730541229,
|
| 842 |
+
"learning_rate": 9.785480950320538e-05,
|
| 843 |
+
"loss": 0.0495,
|
| 844 |
+
"step": 1390
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"grad_norm": 0.41755494475364685,
|
| 848 |
+
"learning_rate": 9.78066352637221e-05,
|
| 849 |
+
"loss": 0.0601,
|
| 850 |
+
"step": 1400
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"grad_norm": 0.31297025084495544,
|
| 854 |
+
"learning_rate": 9.775793821853488e-05,
|
| 855 |
+
"loss": 0.0524,
|
| 856 |
+
"step": 1410
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"grad_norm": 0.3274155259132385,
|
| 860 |
+
"learning_rate": 9.77087189001868e-05,
|
| 861 |
+
"loss": 0.0583,
|
| 862 |
+
"step": 1420
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"grad_norm": 0.3719927966594696,
|
| 866 |
+
"learning_rate": 9.765897784693243e-05,
|
| 867 |
+
"loss": 0.0552,
|
| 868 |
+
"step": 1430
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"grad_norm": 0.23558516800403595,
|
| 872 |
+
"learning_rate": 9.760871560273197e-05,
|
| 873 |
+
"loss": 0.0541,
|
| 874 |
+
"step": 1440
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"grad_norm": 0.3183964490890503,
|
| 878 |
+
"learning_rate": 9.755793271724526e-05,
|
| 879 |
+
"loss": 0.0518,
|
| 880 |
+
"step": 1450
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"grad_norm": 0.5059468150138855,
|
| 884 |
+
"learning_rate": 9.750662974582584e-05,
|
| 885 |
+
"loss": 0.0557,
|
| 886 |
+
"step": 1460
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"grad_norm": 0.3334318995475769,
|
| 890 |
+
"learning_rate": 9.745480724951473e-05,
|
| 891 |
+
"loss": 0.0549,
|
| 892 |
+
"step": 1470
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"grad_norm": 0.3796740770339966,
|
| 896 |
+
"learning_rate": 9.740246579503447e-05,
|
| 897 |
+
"loss": 0.0621,
|
| 898 |
+
"step": 1480
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"grad_norm": 0.3182413578033447,
|
| 902 |
+
"learning_rate": 9.734960595478284e-05,
|
| 903 |
+
"loss": 0.0494,
|
| 904 |
+
"step": 1490
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"grad_norm": 0.361194372177124,
|
| 908 |
+
"learning_rate": 9.729622830682657e-05,
|
| 909 |
+
"loss": 0.0633,
|
| 910 |
+
"step": 1500
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"grad_norm": 0.3800562620162964,
|
| 914 |
+
"learning_rate": 9.724233343489504e-05,
|
| 915 |
+
"loss": 0.0522,
|
| 916 |
+
"step": 1510
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"grad_norm": 0.3530648946762085,
|
| 920 |
+
"learning_rate": 9.718792192837396e-05,
|
| 921 |
+
"loss": 0.0534,
|
| 922 |
+
"step": 1520
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"grad_norm": 0.2889000177383423,
|
| 926 |
+
"learning_rate": 9.713299438229886e-05,
|
| 927 |
+
"loss": 0.0545,
|
| 928 |
+
"step": 1530
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"grad_norm": 0.49150562286376953,
|
| 932 |
+
"learning_rate": 9.707755139734855e-05,
|
| 933 |
+
"loss": 0.0646,
|
| 934 |
+
"step": 1540
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"grad_norm": 0.2493639588356018,
|
| 938 |
+
"learning_rate": 9.702159357983866e-05,
|
| 939 |
+
"loss": 0.0561,
|
| 940 |
+
"step": 1550
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"grad_norm": 0.264142245054245,
|
| 944 |
+
"learning_rate": 9.696512154171492e-05,
|
| 945 |
+
"loss": 0.0541,
|
| 946 |
+
"step": 1560
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"grad_norm": 0.366263747215271,
|
| 950 |
+
"learning_rate": 9.690813590054645e-05,
|
| 951 |
+
"loss": 0.048,
|
| 952 |
+
"step": 1570
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"grad_norm": 0.3640550374984741,
|
| 956 |
+
"learning_rate": 9.685063727951914e-05,
|
| 957 |
+
"loss": 0.0501,
|
| 958 |
+
"step": 1580
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"grad_norm": 0.3413008451461792,
|
| 962 |
+
"learning_rate": 9.679262630742865e-05,
|
| 963 |
+
"loss": 0.0544,
|
| 964 |
+
"step": 1590
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"grad_norm": 0.37865203619003296,
|
| 968 |
+
"learning_rate": 9.673410361867373e-05,
|
| 969 |
+
"loss": 0.053,
|
| 970 |
+
"step": 1600
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"grad_norm": 0.5088744759559631,
|
| 974 |
+
"learning_rate": 9.667506985324909e-05,
|
| 975 |
+
"loss": 0.0547,
|
| 976 |
+
"step": 1610
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"grad_norm": 0.25608953833580017,
|
| 980 |
+
"learning_rate": 9.661552565673855e-05,
|
| 981 |
+
"loss": 0.0541,
|
| 982 |
+
"step": 1620
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"grad_norm": 0.41425254940986633,
|
| 986 |
+
"learning_rate": 9.655547168030789e-05,
|
| 987 |
+
"loss": 0.0491,
|
| 988 |
+
"step": 1630
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"grad_norm": 0.5099269151687622,
|
| 992 |
+
"learning_rate": 9.649490858069777e-05,
|
| 993 |
+
"loss": 0.0574,
|
| 994 |
+
"step": 1640
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"grad_norm": 0.42447859048843384,
|
| 998 |
+
"learning_rate": 9.643383702021658e-05,
|
| 999 |
+
"loss": 0.063,
|
| 1000 |
+
"step": 1650
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"grad_norm": 0.2408001720905304,
|
| 1004 |
+
"learning_rate": 9.637225766673307e-05,
|
| 1005 |
+
"loss": 0.0502,
|
| 1006 |
+
"step": 1660
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"grad_norm": 0.3616074025630951,
|
| 1010 |
+
"learning_rate": 9.631017119366922e-05,
|
| 1011 |
+
"loss": 0.0623,
|
| 1012 |
+
"step": 1670
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"grad_norm": 0.3710896968841553,
|
| 1016 |
+
"learning_rate": 9.624757827999273e-05,
|
| 1017 |
+
"loss": 0.0566,
|
| 1018 |
+
"step": 1680
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"grad_norm": 0.2255239635705948,
|
| 1022 |
+
"learning_rate": 9.618447961020971e-05,
|
| 1023 |
+
"loss": 0.0556,
|
| 1024 |
+
"step": 1690
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"grad_norm": 0.4593660831451416,
|
| 1028 |
+
"learning_rate": 9.612087587435707e-05,
|
| 1029 |
+
"loss": 0.0522,
|
| 1030 |
+
"step": 1700
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"grad_norm": 0.3208771347999573,
|
| 1034 |
+
"learning_rate": 9.605676776799508e-05,
|
| 1035 |
+
"loss": 0.0489,
|
| 1036 |
+
"step": 1710
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"grad_norm": 0.2732645869255066,
|
| 1040 |
+
"learning_rate": 9.599215599219973e-05,
|
| 1041 |
+
"loss": 0.0522,
|
| 1042 |
+
"step": 1720
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"grad_norm": 0.3936018943786621,
|
| 1046 |
+
"learning_rate": 9.592704125355505e-05,
|
| 1047 |
+
"loss": 0.0546,
|
| 1048 |
+
"step": 1730
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"grad_norm": 0.31143778562545776,
|
| 1052 |
+
"learning_rate": 9.586142426414538e-05,
|
| 1053 |
+
"loss": 0.0493,
|
| 1054 |
+
"step": 1740
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"grad_norm": 0.3694917857646942,
|
| 1058 |
+
"learning_rate": 9.57953057415476e-05,
|
| 1059 |
+
"loss": 0.0502,
|
| 1060 |
+
"step": 1750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"grad_norm": 0.33363762497901917,
|
| 1064 |
+
"learning_rate": 9.572868640882328e-05,
|
| 1065 |
+
"loss": 0.0518,
|
| 1066 |
+
"step": 1760
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"grad_norm": 0.36443617939949036,
|
| 1070 |
+
"learning_rate": 9.56615669945108e-05,
|
| 1071 |
+
"loss": 0.0519,
|
| 1072 |
+
"step": 1770
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"grad_norm": 0.2904825806617737,
|
| 1076 |
+
"learning_rate": 9.55939482326173e-05,
|
| 1077 |
+
"loss": 0.0501,
|
| 1078 |
+
"step": 1780
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"grad_norm": 0.37458282709121704,
|
| 1082 |
+
"learning_rate": 9.552583086261069e-05,
|
| 1083 |
+
"loss": 0.0535,
|
| 1084 |
+
"step": 1790
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"grad_norm": 0.43567535281181335,
|
| 1088 |
+
"learning_rate": 9.545721562941168e-05,
|
| 1089 |
+
"loss": 0.0529,
|
| 1090 |
+
"step": 1800
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"grad_norm": 0.3915671110153198,
|
| 1094 |
+
"learning_rate": 9.538810328338543e-05,
|
| 1095 |
+
"loss": 0.0515,
|
| 1096 |
+
"step": 1810
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"grad_norm": 0.27103352546691895,
|
| 1100 |
+
"learning_rate": 9.531849458033349e-05,
|
| 1101 |
+
"loss": 0.0464,
|
| 1102 |
+
"step": 1820
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"grad_norm": 0.25404107570648193,
|
| 1106 |
+
"learning_rate": 9.524839028148547e-05,
|
| 1107 |
+
"loss": 0.0498,
|
| 1108 |
+
"step": 1830
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"grad_norm": 0.4265500605106354,
|
| 1112 |
+
"learning_rate": 9.517779115349077e-05,
|
| 1113 |
+
"loss": 0.0487,
|
| 1114 |
+
"step": 1840
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"grad_norm": 0.32784515619277954,
|
| 1118 |
+
"learning_rate": 9.510669796841014e-05,
|
| 1119 |
+
"loss": 0.0427,
|
| 1120 |
+
"step": 1850
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"grad_norm": 0.38013437390327454,
|
| 1124 |
+
"learning_rate": 9.503511150370727e-05,
|
| 1125 |
+
"loss": 0.0511,
|
| 1126 |
+
"step": 1860
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"grad_norm": 0.3897137939929962,
|
| 1130 |
+
"learning_rate": 9.496303254224024e-05,
|
| 1131 |
+
"loss": 0.0519,
|
| 1132 |
+
"step": 1870
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"grad_norm": 0.3544720411300659,
|
| 1136 |
+
"learning_rate": 9.489046187225306e-05,
|
| 1137 |
+
"loss": 0.0407,
|
| 1138 |
+
"step": 1880
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"grad_norm": 0.33186250925064087,
|
| 1142 |
+
"learning_rate": 9.481740028736692e-05,
|
| 1143 |
+
"loss": 0.0438,
|
| 1144 |
+
"step": 1890
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"grad_norm": 0.40961360931396484,
|
| 1148 |
+
"learning_rate": 9.474384858657164e-05,
|
| 1149 |
+
"loss": 0.0516,
|
| 1150 |
+
"step": 1900
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"grad_norm": 0.37814396619796753,
|
| 1154 |
+
"learning_rate": 9.466980757421679e-05,
|
| 1155 |
+
"loss": 0.0502,
|
| 1156 |
+
"step": 1910
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"grad_norm": 0.3576529026031494,
|
| 1160 |
+
"learning_rate": 9.459527806000305e-05,
|
| 1161 |
+
"loss": 0.0473,
|
| 1162 |
+
"step": 1920
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"grad_norm": 0.4338109493255615,
|
| 1166 |
+
"learning_rate": 9.452026085897325e-05,
|
| 1167 |
+
"loss": 0.0547,
|
| 1168 |
+
"step": 1930
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"grad_norm": 0.4323793053627014,
|
| 1172 |
+
"learning_rate": 9.444475679150348e-05,
|
| 1173 |
+
"loss": 0.0512,
|
| 1174 |
+
"step": 1940
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"grad_norm": 0.4074627459049225,
|
| 1178 |
+
"learning_rate": 9.436876668329411e-05,
|
| 1179 |
+
"loss": 0.0469,
|
| 1180 |
+
"step": 1950
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"grad_norm": 0.2311258316040039,
|
| 1184 |
+
"learning_rate": 9.429229136536079e-05,
|
| 1185 |
+
"loss": 0.0431,
|
| 1186 |
+
"step": 1960
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"grad_norm": 0.4605922996997833,
|
| 1190 |
+
"learning_rate": 9.421533167402534e-05,
|
| 1191 |
+
"loss": 0.0561,
|
| 1192 |
+
"step": 1970
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"grad_norm": 0.3769131898880005,
|
| 1196 |
+
"learning_rate": 9.413788845090666e-05,
|
| 1197 |
+
"loss": 0.0582,
|
| 1198 |
+
"step": 1980
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"grad_norm": 0.4106776714324951,
|
| 1202 |
+
"learning_rate": 9.405996254291136e-05,
|
| 1203 |
+
"loss": 0.0512,
|
| 1204 |
+
"step": 1990
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"grad_norm": 0.33406883478164673,
|
| 1208 |
+
"learning_rate": 9.398155480222474e-05,
|
| 1209 |
+
"loss": 0.0486,
|
| 1210 |
+
"step": 2000
|
| 1211 |
+
}
|
| 1212 |
+
],
|
| 1213 |
+
"logging_steps": 10,
|
| 1214 |
+
"max_steps": 10000,
|
| 1215 |
+
"num_input_tokens_seen": 0,
|
| 1216 |
+
"num_train_epochs": 9223372036854775807,
|
| 1217 |
+
"save_steps": 1000,
|
| 1218 |
+
"stateful_callbacks": {
|
| 1219 |
+
"TrainerControl": {
|
| 1220 |
+
"args": {
|
| 1221 |
+
"should_epoch_stop": false,
|
| 1222 |
+
"should_evaluate": false,
|
| 1223 |
+
"should_log": false,
|
| 1224 |
+
"should_save": true,
|
| 1225 |
+
"should_training_stop": false
|
| 1226 |
+
},
|
| 1227 |
+
"attributes": {}
|
| 1228 |
+
}
|
| 1229 |
+
},
|
| 1230 |
+
"total_flos": 0.0,
|
| 1231 |
+
"train_batch_size": 32,
|
| 1232 |
+
"trial_name": null,
|
| 1233 |
+
"trial_params": null
|
| 1234 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5681d3ba5ecda11b72a2cd5a91279a8ec4ee9abf1f93cfc8581149b81eabf7d6
|
| 3 |
+
size 5777
|