Uploaded checkpoint-22500
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1795 -5
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 119975656
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e72ed832ca82b47560da4b70a43bfb85762b41d54a4b1df89cee4b8816cb6fc
|
| 3 |
size 119975656
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 240145026
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19756b617f2bf91f55b9e8c9b87ec2279b8dca12dd91f8f9e92a075a7d6745b9
|
| 3 |
size 240145026
|
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:da5d678c8111a8bb6e7f07c6d826c3d293cb2dc841c1a7d8cdada1cef59bd3c9
|
| 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 sha256:2ebf87c3777a880efd4523ce05af816d67a6a12edb3e1d54f156890382c1db41
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -14327,6 +14327,1796 @@
|
|
| 14327 |
"eval_samples_per_second": 15.124,
|
| 14328 |
"eval_steps_per_second": 15.124,
|
| 14329 |
"step": 20000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14330 |
}
|
| 14331 |
],
|
| 14332 |
"logging_steps": 10,
|
|
@@ -14334,7 +16124,7 @@
|
|
| 14334 |
"num_input_tokens_seen": 0,
|
| 14335 |
"num_train_epochs": 1,
|
| 14336 |
"save_steps": 2500,
|
| 14337 |
-
"total_flos": 3.
|
| 14338 |
"train_batch_size": 1,
|
| 14339 |
"trial_name": null,
|
| 14340 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.3439137935638428,
|
| 3 |
+
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-22500",
|
| 4 |
+
"epoch": 0.5625,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 22500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 14327 |
"eval_samples_per_second": 15.124,
|
| 14328 |
"eval_steps_per_second": 15.124,
|
| 14329 |
"step": 20000
|
| 14330 |
+
},
|
| 14331 |
+
{
|
| 14332 |
+
"epoch": 0.5,
|
| 14333 |
+
"grad_norm": 2.159421920776367,
|
| 14334 |
+
"learning_rate": 6.772881355932204e-06,
|
| 14335 |
+
"loss": 1.4193,
|
| 14336 |
+
"step": 20010
|
| 14337 |
+
},
|
| 14338 |
+
{
|
| 14339 |
+
"epoch": 0.5,
|
| 14340 |
+
"grad_norm": 7.60541296005249,
|
| 14341 |
+
"learning_rate": 6.766101694915255e-06,
|
| 14342 |
+
"loss": 1.4747,
|
| 14343 |
+
"step": 20020
|
| 14344 |
+
},
|
| 14345 |
+
{
|
| 14346 |
+
"epoch": 0.5,
|
| 14347 |
+
"grad_norm": 2.0991897583007812,
|
| 14348 |
+
"learning_rate": 6.759322033898306e-06,
|
| 14349 |
+
"loss": 1.2775,
|
| 14350 |
+
"step": 20030
|
| 14351 |
+
},
|
| 14352 |
+
{
|
| 14353 |
+
"epoch": 0.5,
|
| 14354 |
+
"grad_norm": 7.756378173828125,
|
| 14355 |
+
"learning_rate": 6.7525423728813565e-06,
|
| 14356 |
+
"loss": 1.343,
|
| 14357 |
+
"step": 20040
|
| 14358 |
+
},
|
| 14359 |
+
{
|
| 14360 |
+
"epoch": 0.5,
|
| 14361 |
+
"grad_norm": 7.062932014465332,
|
| 14362 |
+
"learning_rate": 6.745762711864408e-06,
|
| 14363 |
+
"loss": 1.5168,
|
| 14364 |
+
"step": 20050
|
| 14365 |
+
},
|
| 14366 |
+
{
|
| 14367 |
+
"epoch": 0.5,
|
| 14368 |
+
"grad_norm": 6.689007759094238,
|
| 14369 |
+
"learning_rate": 6.7389830508474585e-06,
|
| 14370 |
+
"loss": 1.2894,
|
| 14371 |
+
"step": 20060
|
| 14372 |
+
},
|
| 14373 |
+
{
|
| 14374 |
+
"epoch": 0.5,
|
| 14375 |
+
"grad_norm": 5.31237268447876,
|
| 14376 |
+
"learning_rate": 6.73220338983051e-06,
|
| 14377 |
+
"loss": 1.3382,
|
| 14378 |
+
"step": 20070
|
| 14379 |
+
},
|
| 14380 |
+
{
|
| 14381 |
+
"epoch": 0.5,
|
| 14382 |
+
"grad_norm": 3.6598782539367676,
|
| 14383 |
+
"learning_rate": 6.7254237288135604e-06,
|
| 14384 |
+
"loss": 1.3839,
|
| 14385 |
+
"step": 20080
|
| 14386 |
+
},
|
| 14387 |
+
{
|
| 14388 |
+
"epoch": 0.5,
|
| 14389 |
+
"grad_norm": 3.3091630935668945,
|
| 14390 |
+
"learning_rate": 6.71864406779661e-06,
|
| 14391 |
+
"loss": 1.3654,
|
| 14392 |
+
"step": 20090
|
| 14393 |
+
},
|
| 14394 |
+
{
|
| 14395 |
+
"epoch": 0.5,
|
| 14396 |
+
"grad_norm": 2.1399312019348145,
|
| 14397 |
+
"learning_rate": 6.7118644067796615e-06,
|
| 14398 |
+
"loss": 1.5429,
|
| 14399 |
+
"step": 20100
|
| 14400 |
+
},
|
| 14401 |
+
{
|
| 14402 |
+
"epoch": 0.5,
|
| 14403 |
+
"grad_norm": 4.380695819854736,
|
| 14404 |
+
"learning_rate": 6.705084745762712e-06,
|
| 14405 |
+
"loss": 1.4629,
|
| 14406 |
+
"step": 20110
|
| 14407 |
+
},
|
| 14408 |
+
{
|
| 14409 |
+
"epoch": 0.5,
|
| 14410 |
+
"grad_norm": 9.330241203308105,
|
| 14411 |
+
"learning_rate": 6.6983050847457635e-06,
|
| 14412 |
+
"loss": 1.3547,
|
| 14413 |
+
"step": 20120
|
| 14414 |
+
},
|
| 14415 |
+
{
|
| 14416 |
+
"epoch": 0.5,
|
| 14417 |
+
"grad_norm": 5.803046226501465,
|
| 14418 |
+
"learning_rate": 6.691525423728814e-06,
|
| 14419 |
+
"loss": 1.3534,
|
| 14420 |
+
"step": 20130
|
| 14421 |
+
},
|
| 14422 |
+
{
|
| 14423 |
+
"epoch": 0.5,
|
| 14424 |
+
"grad_norm": 4.967430591583252,
|
| 14425 |
+
"learning_rate": 6.6847457627118655e-06,
|
| 14426 |
+
"loss": 1.4965,
|
| 14427 |
+
"step": 20140
|
| 14428 |
+
},
|
| 14429 |
+
{
|
| 14430 |
+
"epoch": 0.5,
|
| 14431 |
+
"grad_norm": 8.131784439086914,
|
| 14432 |
+
"learning_rate": 6.677966101694916e-06,
|
| 14433 |
+
"loss": 1.4012,
|
| 14434 |
+
"step": 20150
|
| 14435 |
+
},
|
| 14436 |
+
{
|
| 14437 |
+
"epoch": 0.5,
|
| 14438 |
+
"grad_norm": 9.220160484313965,
|
| 14439 |
+
"learning_rate": 6.6711864406779666e-06,
|
| 14440 |
+
"loss": 1.3637,
|
| 14441 |
+
"step": 20160
|
| 14442 |
+
},
|
| 14443 |
+
{
|
| 14444 |
+
"epoch": 0.5,
|
| 14445 |
+
"grad_norm": 6.833899974822998,
|
| 14446 |
+
"learning_rate": 6.664406779661018e-06,
|
| 14447 |
+
"loss": 1.3748,
|
| 14448 |
+
"step": 20170
|
| 14449 |
+
},
|
| 14450 |
+
{
|
| 14451 |
+
"epoch": 0.5,
|
| 14452 |
+
"grad_norm": 13.869956970214844,
|
| 14453 |
+
"learning_rate": 6.6576271186440685e-06,
|
| 14454 |
+
"loss": 1.4474,
|
| 14455 |
+
"step": 20180
|
| 14456 |
+
},
|
| 14457 |
+
{
|
| 14458 |
+
"epoch": 0.5,
|
| 14459 |
+
"grad_norm": 9.462739944458008,
|
| 14460 |
+
"learning_rate": 6.650847457627119e-06,
|
| 14461 |
+
"loss": 1.3893,
|
| 14462 |
+
"step": 20190
|
| 14463 |
+
},
|
| 14464 |
+
{
|
| 14465 |
+
"epoch": 0.51,
|
| 14466 |
+
"grad_norm": 2.321714162826538,
|
| 14467 |
+
"learning_rate": 6.64406779661017e-06,
|
| 14468 |
+
"loss": 1.3836,
|
| 14469 |
+
"step": 20200
|
| 14470 |
+
},
|
| 14471 |
+
{
|
| 14472 |
+
"epoch": 0.51,
|
| 14473 |
+
"grad_norm": 11.65925407409668,
|
| 14474 |
+
"learning_rate": 6.637288135593221e-06,
|
| 14475 |
+
"loss": 1.493,
|
| 14476 |
+
"step": 20210
|
| 14477 |
+
},
|
| 14478 |
+
{
|
| 14479 |
+
"epoch": 0.51,
|
| 14480 |
+
"grad_norm": 6.808838367462158,
|
| 14481 |
+
"learning_rate": 6.6305084745762716e-06,
|
| 14482 |
+
"loss": 1.3632,
|
| 14483 |
+
"step": 20220
|
| 14484 |
+
},
|
| 14485 |
+
{
|
| 14486 |
+
"epoch": 0.51,
|
| 14487 |
+
"grad_norm": 4.534874439239502,
|
| 14488 |
+
"learning_rate": 6.623728813559322e-06,
|
| 14489 |
+
"loss": 1.4303,
|
| 14490 |
+
"step": 20230
|
| 14491 |
+
},
|
| 14492 |
+
{
|
| 14493 |
+
"epoch": 0.51,
|
| 14494 |
+
"grad_norm": 1.9837802648544312,
|
| 14495 |
+
"learning_rate": 6.6169491525423735e-06,
|
| 14496 |
+
"loss": 1.446,
|
| 14497 |
+
"step": 20240
|
| 14498 |
+
},
|
| 14499 |
+
{
|
| 14500 |
+
"epoch": 0.51,
|
| 14501 |
+
"grad_norm": 7.779874324798584,
|
| 14502 |
+
"learning_rate": 6.610169491525424e-06,
|
| 14503 |
+
"loss": 1.4566,
|
| 14504 |
+
"step": 20250
|
| 14505 |
+
},
|
| 14506 |
+
{
|
| 14507 |
+
"epoch": 0.51,
|
| 14508 |
+
"grad_norm": 5.729377746582031,
|
| 14509 |
+
"learning_rate": 6.6033898305084755e-06,
|
| 14510 |
+
"loss": 1.401,
|
| 14511 |
+
"step": 20260
|
| 14512 |
+
},
|
| 14513 |
+
{
|
| 14514 |
+
"epoch": 0.51,
|
| 14515 |
+
"grad_norm": 12.633749008178711,
|
| 14516 |
+
"learning_rate": 6.596610169491526e-06,
|
| 14517 |
+
"loss": 1.2969,
|
| 14518 |
+
"step": 20270
|
| 14519 |
+
},
|
| 14520 |
+
{
|
| 14521 |
+
"epoch": 0.51,
|
| 14522 |
+
"grad_norm": 2.7961552143096924,
|
| 14523 |
+
"learning_rate": 6.5898305084745774e-06,
|
| 14524 |
+
"loss": 1.4151,
|
| 14525 |
+
"step": 20280
|
| 14526 |
+
},
|
| 14527 |
+
{
|
| 14528 |
+
"epoch": 0.51,
|
| 14529 |
+
"grad_norm": 8.461606979370117,
|
| 14530 |
+
"learning_rate": 6.583050847457627e-06,
|
| 14531 |
+
"loss": 1.2665,
|
| 14532 |
+
"step": 20290
|
| 14533 |
+
},
|
| 14534 |
+
{
|
| 14535 |
+
"epoch": 0.51,
|
| 14536 |
+
"grad_norm": 10.325592041015625,
|
| 14537 |
+
"learning_rate": 6.576271186440678e-06,
|
| 14538 |
+
"loss": 1.3666,
|
| 14539 |
+
"step": 20300
|
| 14540 |
+
},
|
| 14541 |
+
{
|
| 14542 |
+
"epoch": 0.51,
|
| 14543 |
+
"grad_norm": 3.0818538665771484,
|
| 14544 |
+
"learning_rate": 6.569491525423729e-06,
|
| 14545 |
+
"loss": 1.4307,
|
| 14546 |
+
"step": 20310
|
| 14547 |
+
},
|
| 14548 |
+
{
|
| 14549 |
+
"epoch": 0.51,
|
| 14550 |
+
"grad_norm": 7.449318885803223,
|
| 14551 |
+
"learning_rate": 6.56271186440678e-06,
|
| 14552 |
+
"loss": 1.4585,
|
| 14553 |
+
"step": 20320
|
| 14554 |
+
},
|
| 14555 |
+
{
|
| 14556 |
+
"epoch": 0.51,
|
| 14557 |
+
"grad_norm": 4.6587042808532715,
|
| 14558 |
+
"learning_rate": 6.555932203389831e-06,
|
| 14559 |
+
"loss": 1.4053,
|
| 14560 |
+
"step": 20330
|
| 14561 |
+
},
|
| 14562 |
+
{
|
| 14563 |
+
"epoch": 0.51,
|
| 14564 |
+
"grad_norm": 5.837299346923828,
|
| 14565 |
+
"learning_rate": 6.549152542372882e-06,
|
| 14566 |
+
"loss": 1.3456,
|
| 14567 |
+
"step": 20340
|
| 14568 |
+
},
|
| 14569 |
+
{
|
| 14570 |
+
"epoch": 0.51,
|
| 14571 |
+
"grad_norm": 7.345305442810059,
|
| 14572 |
+
"learning_rate": 6.542372881355933e-06,
|
| 14573 |
+
"loss": 1.3618,
|
| 14574 |
+
"step": 20350
|
| 14575 |
+
},
|
| 14576 |
+
{
|
| 14577 |
+
"epoch": 0.51,
|
| 14578 |
+
"grad_norm": 5.679592609405518,
|
| 14579 |
+
"learning_rate": 6.5355932203389836e-06,
|
| 14580 |
+
"loss": 1.3875,
|
| 14581 |
+
"step": 20360
|
| 14582 |
+
},
|
| 14583 |
+
{
|
| 14584 |
+
"epoch": 0.51,
|
| 14585 |
+
"grad_norm": 1.062429428100586,
|
| 14586 |
+
"learning_rate": 6.528813559322035e-06,
|
| 14587 |
+
"loss": 1.3083,
|
| 14588 |
+
"step": 20370
|
| 14589 |
+
},
|
| 14590 |
+
{
|
| 14591 |
+
"epoch": 0.51,
|
| 14592 |
+
"grad_norm": 5.2157769203186035,
|
| 14593 |
+
"learning_rate": 6.5220338983050855e-06,
|
| 14594 |
+
"loss": 1.3582,
|
| 14595 |
+
"step": 20380
|
| 14596 |
+
},
|
| 14597 |
+
{
|
| 14598 |
+
"epoch": 0.51,
|
| 14599 |
+
"grad_norm": 4.890625476837158,
|
| 14600 |
+
"learning_rate": 6.515254237288137e-06,
|
| 14601 |
+
"loss": 1.3184,
|
| 14602 |
+
"step": 20390
|
| 14603 |
+
},
|
| 14604 |
+
{
|
| 14605 |
+
"epoch": 0.51,
|
| 14606 |
+
"grad_norm": 1.587312936782837,
|
| 14607 |
+
"learning_rate": 6.508474576271187e-06,
|
| 14608 |
+
"loss": 1.2191,
|
| 14609 |
+
"step": 20400
|
| 14610 |
+
},
|
| 14611 |
+
{
|
| 14612 |
+
"epoch": 0.51,
|
| 14613 |
+
"grad_norm": 4.013046741485596,
|
| 14614 |
+
"learning_rate": 6.501694915254237e-06,
|
| 14615 |
+
"loss": 1.3331,
|
| 14616 |
+
"step": 20410
|
| 14617 |
+
},
|
| 14618 |
+
{
|
| 14619 |
+
"epoch": 0.51,
|
| 14620 |
+
"grad_norm": 4.076744556427002,
|
| 14621 |
+
"learning_rate": 6.4949152542372886e-06,
|
| 14622 |
+
"loss": 1.4351,
|
| 14623 |
+
"step": 20420
|
| 14624 |
+
},
|
| 14625 |
+
{
|
| 14626 |
+
"epoch": 0.51,
|
| 14627 |
+
"grad_norm": 6.493119239807129,
|
| 14628 |
+
"learning_rate": 6.488135593220339e-06,
|
| 14629 |
+
"loss": 1.3519,
|
| 14630 |
+
"step": 20430
|
| 14631 |
+
},
|
| 14632 |
+
{
|
| 14633 |
+
"epoch": 0.51,
|
| 14634 |
+
"grad_norm": 6.127922058105469,
|
| 14635 |
+
"learning_rate": 6.4813559322033905e-06,
|
| 14636 |
+
"loss": 1.4255,
|
| 14637 |
+
"step": 20440
|
| 14638 |
+
},
|
| 14639 |
+
{
|
| 14640 |
+
"epoch": 0.51,
|
| 14641 |
+
"grad_norm": 4.424916744232178,
|
| 14642 |
+
"learning_rate": 6.474576271186441e-06,
|
| 14643 |
+
"loss": 1.44,
|
| 14644 |
+
"step": 20450
|
| 14645 |
+
},
|
| 14646 |
+
{
|
| 14647 |
+
"epoch": 0.51,
|
| 14648 |
+
"grad_norm": 9.894845008850098,
|
| 14649 |
+
"learning_rate": 6.4677966101694925e-06,
|
| 14650 |
+
"loss": 1.2922,
|
| 14651 |
+
"step": 20460
|
| 14652 |
+
},
|
| 14653 |
+
{
|
| 14654 |
+
"epoch": 0.51,
|
| 14655 |
+
"grad_norm": 5.174190521240234,
|
| 14656 |
+
"learning_rate": 6.461016949152543e-06,
|
| 14657 |
+
"loss": 1.439,
|
| 14658 |
+
"step": 20470
|
| 14659 |
+
},
|
| 14660 |
+
{
|
| 14661 |
+
"epoch": 0.51,
|
| 14662 |
+
"grad_norm": 6.313706398010254,
|
| 14663 |
+
"learning_rate": 6.4542372881355944e-06,
|
| 14664 |
+
"loss": 1.3229,
|
| 14665 |
+
"step": 20480
|
| 14666 |
+
},
|
| 14667 |
+
{
|
| 14668 |
+
"epoch": 0.51,
|
| 14669 |
+
"grad_norm": 3.240434169769287,
|
| 14670 |
+
"learning_rate": 6.447457627118645e-06,
|
| 14671 |
+
"loss": 1.3911,
|
| 14672 |
+
"step": 20490
|
| 14673 |
+
},
|
| 14674 |
+
{
|
| 14675 |
+
"epoch": 0.51,
|
| 14676 |
+
"grad_norm": 7.466948986053467,
|
| 14677 |
+
"learning_rate": 6.440677966101695e-06,
|
| 14678 |
+
"loss": 1.4467,
|
| 14679 |
+
"step": 20500
|
| 14680 |
+
},
|
| 14681 |
+
{
|
| 14682 |
+
"epoch": 0.51,
|
| 14683 |
+
"eval_loss": 1.3339221477508545,
|
| 14684 |
+
"eval_runtime": 66.152,
|
| 14685 |
+
"eval_samples_per_second": 15.117,
|
| 14686 |
+
"eval_steps_per_second": 15.117,
|
| 14687 |
+
"step": 20500
|
| 14688 |
+
},
|
| 14689 |
+
{
|
| 14690 |
+
"epoch": 0.51,
|
| 14691 |
+
"grad_norm": 32.95411682128906,
|
| 14692 |
+
"learning_rate": 6.433898305084746e-06,
|
| 14693 |
+
"loss": 1.2821,
|
| 14694 |
+
"step": 20510
|
| 14695 |
+
},
|
| 14696 |
+
{
|
| 14697 |
+
"epoch": 0.51,
|
| 14698 |
+
"grad_norm": 11.001514434814453,
|
| 14699 |
+
"learning_rate": 6.427118644067797e-06,
|
| 14700 |
+
"loss": 1.3523,
|
| 14701 |
+
"step": 20520
|
| 14702 |
+
},
|
| 14703 |
+
{
|
| 14704 |
+
"epoch": 0.51,
|
| 14705 |
+
"grad_norm": 1.6923043727874756,
|
| 14706 |
+
"learning_rate": 6.420338983050848e-06,
|
| 14707 |
+
"loss": 1.3424,
|
| 14708 |
+
"step": 20530
|
| 14709 |
+
},
|
| 14710 |
+
{
|
| 14711 |
+
"epoch": 0.51,
|
| 14712 |
+
"grad_norm": 6.267253398895264,
|
| 14713 |
+
"learning_rate": 6.413559322033899e-06,
|
| 14714 |
+
"loss": 1.2163,
|
| 14715 |
+
"step": 20540
|
| 14716 |
+
},
|
| 14717 |
+
{
|
| 14718 |
+
"epoch": 0.51,
|
| 14719 |
+
"grad_norm": 15.797646522521973,
|
| 14720 |
+
"learning_rate": 6.40677966101695e-06,
|
| 14721 |
+
"loss": 1.2104,
|
| 14722 |
+
"step": 20550
|
| 14723 |
+
},
|
| 14724 |
+
{
|
| 14725 |
+
"epoch": 0.51,
|
| 14726 |
+
"grad_norm": 3.1398627758026123,
|
| 14727 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 14728 |
+
"loss": 1.2264,
|
| 14729 |
+
"step": 20560
|
| 14730 |
+
},
|
| 14731 |
+
{
|
| 14732 |
+
"epoch": 0.51,
|
| 14733 |
+
"grad_norm": 1.7001848220825195,
|
| 14734 |
+
"learning_rate": 6.393220338983052e-06,
|
| 14735 |
+
"loss": 1.345,
|
| 14736 |
+
"step": 20570
|
| 14737 |
+
},
|
| 14738 |
+
{
|
| 14739 |
+
"epoch": 0.51,
|
| 14740 |
+
"grad_norm": 6.551231861114502,
|
| 14741 |
+
"learning_rate": 6.3864406779661025e-06,
|
| 14742 |
+
"loss": 1.295,
|
| 14743 |
+
"step": 20580
|
| 14744 |
+
},
|
| 14745 |
+
{
|
| 14746 |
+
"epoch": 0.51,
|
| 14747 |
+
"grad_norm": 2.501774787902832,
|
| 14748 |
+
"learning_rate": 6.379661016949154e-06,
|
| 14749 |
+
"loss": 1.2727,
|
| 14750 |
+
"step": 20590
|
| 14751 |
+
},
|
| 14752 |
+
{
|
| 14753 |
+
"epoch": 0.52,
|
| 14754 |
+
"grad_norm": 5.2971906661987305,
|
| 14755 |
+
"learning_rate": 6.372881355932204e-06,
|
| 14756 |
+
"loss": 1.1617,
|
| 14757 |
+
"step": 20600
|
| 14758 |
+
},
|
| 14759 |
+
{
|
| 14760 |
+
"epoch": 0.52,
|
| 14761 |
+
"grad_norm": 8.737128257751465,
|
| 14762 |
+
"learning_rate": 6.366101694915254e-06,
|
| 14763 |
+
"loss": 1.3792,
|
| 14764 |
+
"step": 20610
|
| 14765 |
+
},
|
| 14766 |
+
{
|
| 14767 |
+
"epoch": 0.52,
|
| 14768 |
+
"grad_norm": 1.8445500135421753,
|
| 14769 |
+
"learning_rate": 6.3593220338983056e-06,
|
| 14770 |
+
"loss": 1.2818,
|
| 14771 |
+
"step": 20620
|
| 14772 |
+
},
|
| 14773 |
+
{
|
| 14774 |
+
"epoch": 0.52,
|
| 14775 |
+
"grad_norm": 6.586302280426025,
|
| 14776 |
+
"learning_rate": 6.352542372881356e-06,
|
| 14777 |
+
"loss": 1.471,
|
| 14778 |
+
"step": 20630
|
| 14779 |
+
},
|
| 14780 |
+
{
|
| 14781 |
+
"epoch": 0.52,
|
| 14782 |
+
"grad_norm": 2.3581125736236572,
|
| 14783 |
+
"learning_rate": 6.3457627118644075e-06,
|
| 14784 |
+
"loss": 1.26,
|
| 14785 |
+
"step": 20640
|
| 14786 |
+
},
|
| 14787 |
+
{
|
| 14788 |
+
"epoch": 0.52,
|
| 14789 |
+
"grad_norm": 3.006883382797241,
|
| 14790 |
+
"learning_rate": 6.338983050847458e-06,
|
| 14791 |
+
"loss": 1.4312,
|
| 14792 |
+
"step": 20650
|
| 14793 |
+
},
|
| 14794 |
+
{
|
| 14795 |
+
"epoch": 0.52,
|
| 14796 |
+
"grad_norm": 1.8936930894851685,
|
| 14797 |
+
"learning_rate": 6.3322033898305095e-06,
|
| 14798 |
+
"loss": 1.3705,
|
| 14799 |
+
"step": 20660
|
| 14800 |
+
},
|
| 14801 |
+
{
|
| 14802 |
+
"epoch": 0.52,
|
| 14803 |
+
"grad_norm": 5.22953462600708,
|
| 14804 |
+
"learning_rate": 6.32542372881356e-06,
|
| 14805 |
+
"loss": 1.3252,
|
| 14806 |
+
"step": 20670
|
| 14807 |
+
},
|
| 14808 |
+
{
|
| 14809 |
+
"epoch": 0.52,
|
| 14810 |
+
"grad_norm": 7.669064044952393,
|
| 14811 |
+
"learning_rate": 6.318644067796611e-06,
|
| 14812 |
+
"loss": 1.4673,
|
| 14813 |
+
"step": 20680
|
| 14814 |
+
},
|
| 14815 |
+
{
|
| 14816 |
+
"epoch": 0.52,
|
| 14817 |
+
"grad_norm": 3.7508702278137207,
|
| 14818 |
+
"learning_rate": 6.311864406779662e-06,
|
| 14819 |
+
"loss": 1.3465,
|
| 14820 |
+
"step": 20690
|
| 14821 |
+
},
|
| 14822 |
+
{
|
| 14823 |
+
"epoch": 0.52,
|
| 14824 |
+
"grad_norm": 7.959824085235596,
|
| 14825 |
+
"learning_rate": 6.3050847457627125e-06,
|
| 14826 |
+
"loss": 1.4148,
|
| 14827 |
+
"step": 20700
|
| 14828 |
+
},
|
| 14829 |
+
{
|
| 14830 |
+
"epoch": 0.52,
|
| 14831 |
+
"grad_norm": 8.888174057006836,
|
| 14832 |
+
"learning_rate": 6.298305084745763e-06,
|
| 14833 |
+
"loss": 1.4874,
|
| 14834 |
+
"step": 20710
|
| 14835 |
+
},
|
| 14836 |
+
{
|
| 14837 |
+
"epoch": 0.52,
|
| 14838 |
+
"grad_norm": 11.385110855102539,
|
| 14839 |
+
"learning_rate": 6.291525423728814e-06,
|
| 14840 |
+
"loss": 1.3519,
|
| 14841 |
+
"step": 20720
|
| 14842 |
+
},
|
| 14843 |
+
{
|
| 14844 |
+
"epoch": 0.52,
|
| 14845 |
+
"grad_norm": 11.636566162109375,
|
| 14846 |
+
"learning_rate": 6.284745762711865e-06,
|
| 14847 |
+
"loss": 1.3237,
|
| 14848 |
+
"step": 20730
|
| 14849 |
+
},
|
| 14850 |
+
{
|
| 14851 |
+
"epoch": 0.52,
|
| 14852 |
+
"grad_norm": 2.6630842685699463,
|
| 14853 |
+
"learning_rate": 6.277966101694916e-06,
|
| 14854 |
+
"loss": 1.5615,
|
| 14855 |
+
"step": 20740
|
| 14856 |
+
},
|
| 14857 |
+
{
|
| 14858 |
+
"epoch": 0.52,
|
| 14859 |
+
"grad_norm": 15.339405059814453,
|
| 14860 |
+
"learning_rate": 6.271186440677966e-06,
|
| 14861 |
+
"loss": 1.0971,
|
| 14862 |
+
"step": 20750
|
| 14863 |
+
},
|
| 14864 |
+
{
|
| 14865 |
+
"epoch": 0.52,
|
| 14866 |
+
"grad_norm": 5.97685432434082,
|
| 14867 |
+
"learning_rate": 6.2644067796610176e-06,
|
| 14868 |
+
"loss": 1.2582,
|
| 14869 |
+
"step": 20760
|
| 14870 |
+
},
|
| 14871 |
+
{
|
| 14872 |
+
"epoch": 0.52,
|
| 14873 |
+
"grad_norm": 5.218130111694336,
|
| 14874 |
+
"learning_rate": 6.257627118644068e-06,
|
| 14875 |
+
"loss": 1.3272,
|
| 14876 |
+
"step": 20770
|
| 14877 |
+
},
|
| 14878 |
+
{
|
| 14879 |
+
"epoch": 0.52,
|
| 14880 |
+
"grad_norm": 8.943811416625977,
|
| 14881 |
+
"learning_rate": 6.2508474576271195e-06,
|
| 14882 |
+
"loss": 1.4001,
|
| 14883 |
+
"step": 20780
|
| 14884 |
+
},
|
| 14885 |
+
{
|
| 14886 |
+
"epoch": 0.52,
|
| 14887 |
+
"grad_norm": 3.288783550262451,
|
| 14888 |
+
"learning_rate": 6.24406779661017e-06,
|
| 14889 |
+
"loss": 1.3729,
|
| 14890 |
+
"step": 20790
|
| 14891 |
+
},
|
| 14892 |
+
{
|
| 14893 |
+
"epoch": 0.52,
|
| 14894 |
+
"grad_norm": 2.7321832180023193,
|
| 14895 |
+
"learning_rate": 6.2372881355932215e-06,
|
| 14896 |
+
"loss": 1.4365,
|
| 14897 |
+
"step": 20800
|
| 14898 |
+
},
|
| 14899 |
+
{
|
| 14900 |
+
"epoch": 0.52,
|
| 14901 |
+
"grad_norm": 7.656303405761719,
|
| 14902 |
+
"learning_rate": 6.230508474576271e-06,
|
| 14903 |
+
"loss": 1.3646,
|
| 14904 |
+
"step": 20810
|
| 14905 |
+
},
|
| 14906 |
+
{
|
| 14907 |
+
"epoch": 0.52,
|
| 14908 |
+
"grad_norm": 3.2849278450012207,
|
| 14909 |
+
"learning_rate": 6.223728813559322e-06,
|
| 14910 |
+
"loss": 1.0755,
|
| 14911 |
+
"step": 20820
|
| 14912 |
+
},
|
| 14913 |
+
{
|
| 14914 |
+
"epoch": 0.52,
|
| 14915 |
+
"grad_norm": 3.6145408153533936,
|
| 14916 |
+
"learning_rate": 6.216949152542373e-06,
|
| 14917 |
+
"loss": 1.2836,
|
| 14918 |
+
"step": 20830
|
| 14919 |
+
},
|
| 14920 |
+
{
|
| 14921 |
+
"epoch": 0.52,
|
| 14922 |
+
"grad_norm": 6.668073654174805,
|
| 14923 |
+
"learning_rate": 6.210169491525424e-06,
|
| 14924 |
+
"loss": 1.3706,
|
| 14925 |
+
"step": 20840
|
| 14926 |
+
},
|
| 14927 |
+
{
|
| 14928 |
+
"epoch": 0.52,
|
| 14929 |
+
"grad_norm": 10.408007621765137,
|
| 14930 |
+
"learning_rate": 6.203389830508475e-06,
|
| 14931 |
+
"loss": 1.5154,
|
| 14932 |
+
"step": 20850
|
| 14933 |
+
},
|
| 14934 |
+
{
|
| 14935 |
+
"epoch": 0.52,
|
| 14936 |
+
"grad_norm": 2.708711862564087,
|
| 14937 |
+
"learning_rate": 6.196610169491526e-06,
|
| 14938 |
+
"loss": 1.3841,
|
| 14939 |
+
"step": 20860
|
| 14940 |
+
},
|
| 14941 |
+
{
|
| 14942 |
+
"epoch": 0.52,
|
| 14943 |
+
"grad_norm": 1.7991631031036377,
|
| 14944 |
+
"learning_rate": 6.189830508474577e-06,
|
| 14945 |
+
"loss": 1.261,
|
| 14946 |
+
"step": 20870
|
| 14947 |
+
},
|
| 14948 |
+
{
|
| 14949 |
+
"epoch": 0.52,
|
| 14950 |
+
"grad_norm": 6.492557048797607,
|
| 14951 |
+
"learning_rate": 6.183050847457628e-06,
|
| 14952 |
+
"loss": 1.469,
|
| 14953 |
+
"step": 20880
|
| 14954 |
+
},
|
| 14955 |
+
{
|
| 14956 |
+
"epoch": 0.52,
|
| 14957 |
+
"grad_norm": 3.5670902729034424,
|
| 14958 |
+
"learning_rate": 6.176271186440679e-06,
|
| 14959 |
+
"loss": 1.4484,
|
| 14960 |
+
"step": 20890
|
| 14961 |
+
},
|
| 14962 |
+
{
|
| 14963 |
+
"epoch": 0.52,
|
| 14964 |
+
"grad_norm": 3.4511330127716064,
|
| 14965 |
+
"learning_rate": 6.1694915254237295e-06,
|
| 14966 |
+
"loss": 1.3566,
|
| 14967 |
+
"step": 20900
|
| 14968 |
+
},
|
| 14969 |
+
{
|
| 14970 |
+
"epoch": 0.52,
|
| 14971 |
+
"grad_norm": 7.958609580993652,
|
| 14972 |
+
"learning_rate": 6.162711864406781e-06,
|
| 14973 |
+
"loss": 1.3272,
|
| 14974 |
+
"step": 20910
|
| 14975 |
+
},
|
| 14976 |
+
{
|
| 14977 |
+
"epoch": 0.52,
|
| 14978 |
+
"grad_norm": 6.594333648681641,
|
| 14979 |
+
"learning_rate": 6.155932203389831e-06,
|
| 14980 |
+
"loss": 1.3926,
|
| 14981 |
+
"step": 20920
|
| 14982 |
+
},
|
| 14983 |
+
{
|
| 14984 |
+
"epoch": 0.52,
|
| 14985 |
+
"grad_norm": 7.329288005828857,
|
| 14986 |
+
"learning_rate": 6.149152542372881e-06,
|
| 14987 |
+
"loss": 1.4948,
|
| 14988 |
+
"step": 20930
|
| 14989 |
+
},
|
| 14990 |
+
{
|
| 14991 |
+
"epoch": 0.52,
|
| 14992 |
+
"grad_norm": 10.120712280273438,
|
| 14993 |
+
"learning_rate": 6.142372881355933e-06,
|
| 14994 |
+
"loss": 1.5394,
|
| 14995 |
+
"step": 20940
|
| 14996 |
+
},
|
| 14997 |
+
{
|
| 14998 |
+
"epoch": 0.52,
|
| 14999 |
+
"grad_norm": 4.3365983963012695,
|
| 15000 |
+
"learning_rate": 6.135593220338983e-06,
|
| 15001 |
+
"loss": 1.2156,
|
| 15002 |
+
"step": 20950
|
| 15003 |
+
},
|
| 15004 |
+
{
|
| 15005 |
+
"epoch": 0.52,
|
| 15006 |
+
"grad_norm": 2.0522942543029785,
|
| 15007 |
+
"learning_rate": 6.1288135593220346e-06,
|
| 15008 |
+
"loss": 1.3206,
|
| 15009 |
+
"step": 20960
|
| 15010 |
+
},
|
| 15011 |
+
{
|
| 15012 |
+
"epoch": 0.52,
|
| 15013 |
+
"grad_norm": 5.730597496032715,
|
| 15014 |
+
"learning_rate": 6.122033898305085e-06,
|
| 15015 |
+
"loss": 1.4396,
|
| 15016 |
+
"step": 20970
|
| 15017 |
+
},
|
| 15018 |
+
{
|
| 15019 |
+
"epoch": 0.52,
|
| 15020 |
+
"grad_norm": 13.125938415527344,
|
| 15021 |
+
"learning_rate": 6.1152542372881365e-06,
|
| 15022 |
+
"loss": 1.1969,
|
| 15023 |
+
"step": 20980
|
| 15024 |
+
},
|
| 15025 |
+
{
|
| 15026 |
+
"epoch": 0.52,
|
| 15027 |
+
"grad_norm": 3.8410260677337646,
|
| 15028 |
+
"learning_rate": 6.108474576271187e-06,
|
| 15029 |
+
"loss": 1.3842,
|
| 15030 |
+
"step": 20990
|
| 15031 |
+
},
|
| 15032 |
+
{
|
| 15033 |
+
"epoch": 0.53,
|
| 15034 |
+
"grad_norm": 3.415696144104004,
|
| 15035 |
+
"learning_rate": 6.1016949152542385e-06,
|
| 15036 |
+
"loss": 1.3296,
|
| 15037 |
+
"step": 21000
|
| 15038 |
+
},
|
| 15039 |
+
{
|
| 15040 |
+
"epoch": 0.53,
|
| 15041 |
+
"eval_loss": 1.344160556793213,
|
| 15042 |
+
"eval_runtime": 66.1512,
|
| 15043 |
+
"eval_samples_per_second": 15.117,
|
| 15044 |
+
"eval_steps_per_second": 15.117,
|
| 15045 |
+
"step": 21000
|
| 15046 |
+
},
|
| 15047 |
+
{
|
| 15048 |
+
"epoch": 0.53,
|
| 15049 |
+
"grad_norm": 9.95438003540039,
|
| 15050 |
+
"learning_rate": 6.094915254237289e-06,
|
| 15051 |
+
"loss": 1.3206,
|
| 15052 |
+
"step": 21010
|
| 15053 |
+
},
|
| 15054 |
+
{
|
| 15055 |
+
"epoch": 0.53,
|
| 15056 |
+
"grad_norm": 10.410721778869629,
|
| 15057 |
+
"learning_rate": 6.088135593220339e-06,
|
| 15058 |
+
"loss": 1.4014,
|
| 15059 |
+
"step": 21020
|
| 15060 |
+
},
|
| 15061 |
+
{
|
| 15062 |
+
"epoch": 0.53,
|
| 15063 |
+
"grad_norm": 9.159972190856934,
|
| 15064 |
+
"learning_rate": 6.08135593220339e-06,
|
| 15065 |
+
"loss": 1.3827,
|
| 15066 |
+
"step": 21030
|
| 15067 |
+
},
|
| 15068 |
+
{
|
| 15069 |
+
"epoch": 0.53,
|
| 15070 |
+
"grad_norm": 4.856491565704346,
|
| 15071 |
+
"learning_rate": 6.074576271186441e-06,
|
| 15072 |
+
"loss": 1.4561,
|
| 15073 |
+
"step": 21040
|
| 15074 |
+
},
|
| 15075 |
+
{
|
| 15076 |
+
"epoch": 0.53,
|
| 15077 |
+
"grad_norm": 5.863302707672119,
|
| 15078 |
+
"learning_rate": 6.067796610169492e-06,
|
| 15079 |
+
"loss": 1.3225,
|
| 15080 |
+
"step": 21050
|
| 15081 |
+
},
|
| 15082 |
+
{
|
| 15083 |
+
"epoch": 0.53,
|
| 15084 |
+
"grad_norm": 2.968809127807617,
|
| 15085 |
+
"learning_rate": 6.061016949152543e-06,
|
| 15086 |
+
"loss": 1.3713,
|
| 15087 |
+
"step": 21060
|
| 15088 |
+
},
|
| 15089 |
+
{
|
| 15090 |
+
"epoch": 0.53,
|
| 15091 |
+
"grad_norm": 4.19352388381958,
|
| 15092 |
+
"learning_rate": 6.054237288135594e-06,
|
| 15093 |
+
"loss": 1.3894,
|
| 15094 |
+
"step": 21070
|
| 15095 |
+
},
|
| 15096 |
+
{
|
| 15097 |
+
"epoch": 0.53,
|
| 15098 |
+
"grad_norm": 4.841989040374756,
|
| 15099 |
+
"learning_rate": 6.047457627118645e-06,
|
| 15100 |
+
"loss": 1.3609,
|
| 15101 |
+
"step": 21080
|
| 15102 |
+
},
|
| 15103 |
+
{
|
| 15104 |
+
"epoch": 0.53,
|
| 15105 |
+
"grad_norm": 8.693398475646973,
|
| 15106 |
+
"learning_rate": 6.040677966101696e-06,
|
| 15107 |
+
"loss": 1.4118,
|
| 15108 |
+
"step": 21090
|
| 15109 |
+
},
|
| 15110 |
+
{
|
| 15111 |
+
"epoch": 0.53,
|
| 15112 |
+
"grad_norm": 10.902780532836914,
|
| 15113 |
+
"learning_rate": 6.0338983050847465e-06,
|
| 15114 |
+
"loss": 1.4529,
|
| 15115 |
+
"step": 21100
|
| 15116 |
+
},
|
| 15117 |
+
{
|
| 15118 |
+
"epoch": 0.53,
|
| 15119 |
+
"grad_norm": 5.1729607582092285,
|
| 15120 |
+
"learning_rate": 6.027118644067798e-06,
|
| 15121 |
+
"loss": 1.3302,
|
| 15122 |
+
"step": 21110
|
| 15123 |
+
},
|
| 15124 |
+
{
|
| 15125 |
+
"epoch": 0.53,
|
| 15126 |
+
"grad_norm": 1.9906094074249268,
|
| 15127 |
+
"learning_rate": 6.020338983050848e-06,
|
| 15128 |
+
"loss": 1.2294,
|
| 15129 |
+
"step": 21120
|
| 15130 |
+
},
|
| 15131 |
+
{
|
| 15132 |
+
"epoch": 0.53,
|
| 15133 |
+
"grad_norm": 3.285928964614868,
|
| 15134 |
+
"learning_rate": 6.013559322033898e-06,
|
| 15135 |
+
"loss": 1.5006,
|
| 15136 |
+
"step": 21130
|
| 15137 |
+
},
|
| 15138 |
+
{
|
| 15139 |
+
"epoch": 0.53,
|
| 15140 |
+
"grad_norm": 13.098603248596191,
|
| 15141 |
+
"learning_rate": 6.00677966101695e-06,
|
| 15142 |
+
"loss": 1.425,
|
| 15143 |
+
"step": 21140
|
| 15144 |
+
},
|
| 15145 |
+
{
|
| 15146 |
+
"epoch": 0.53,
|
| 15147 |
+
"grad_norm": 4.391754150390625,
|
| 15148 |
+
"learning_rate": 6e-06,
|
| 15149 |
+
"loss": 1.2668,
|
| 15150 |
+
"step": 21150
|
| 15151 |
+
},
|
| 15152 |
+
{
|
| 15153 |
+
"epoch": 0.53,
|
| 15154 |
+
"grad_norm": 8.409893035888672,
|
| 15155 |
+
"learning_rate": 5.9932203389830516e-06,
|
| 15156 |
+
"loss": 1.2061,
|
| 15157 |
+
"step": 21160
|
| 15158 |
+
},
|
| 15159 |
+
{
|
| 15160 |
+
"epoch": 0.53,
|
| 15161 |
+
"grad_norm": 10.568397521972656,
|
| 15162 |
+
"learning_rate": 5.986440677966102e-06,
|
| 15163 |
+
"loss": 1.4191,
|
| 15164 |
+
"step": 21170
|
| 15165 |
+
},
|
| 15166 |
+
{
|
| 15167 |
+
"epoch": 0.53,
|
| 15168 |
+
"grad_norm": 7.371358394622803,
|
| 15169 |
+
"learning_rate": 5.9796610169491535e-06,
|
| 15170 |
+
"loss": 1.2942,
|
| 15171 |
+
"step": 21180
|
| 15172 |
+
},
|
| 15173 |
+
{
|
| 15174 |
+
"epoch": 0.53,
|
| 15175 |
+
"grad_norm": 2.6429450511932373,
|
| 15176 |
+
"learning_rate": 5.972881355932204e-06,
|
| 15177 |
+
"loss": 1.4216,
|
| 15178 |
+
"step": 21190
|
| 15179 |
+
},
|
| 15180 |
+
{
|
| 15181 |
+
"epoch": 0.53,
|
| 15182 |
+
"grad_norm": 5.85234260559082,
|
| 15183 |
+
"learning_rate": 5.9661016949152555e-06,
|
| 15184 |
+
"loss": 1.3541,
|
| 15185 |
+
"step": 21200
|
| 15186 |
+
},
|
| 15187 |
+
{
|
| 15188 |
+
"epoch": 0.53,
|
| 15189 |
+
"grad_norm": 3.2903425693511963,
|
| 15190 |
+
"learning_rate": 5.959322033898306e-06,
|
| 15191 |
+
"loss": 1.2969,
|
| 15192 |
+
"step": 21210
|
| 15193 |
+
},
|
| 15194 |
+
{
|
| 15195 |
+
"epoch": 0.53,
|
| 15196 |
+
"grad_norm": 3.9652106761932373,
|
| 15197 |
+
"learning_rate": 5.9525423728813566e-06,
|
| 15198 |
+
"loss": 1.2853,
|
| 15199 |
+
"step": 21220
|
| 15200 |
+
},
|
| 15201 |
+
{
|
| 15202 |
+
"epoch": 0.53,
|
| 15203 |
+
"grad_norm": 3.8703153133392334,
|
| 15204 |
+
"learning_rate": 5.945762711864407e-06,
|
| 15205 |
+
"loss": 1.3094,
|
| 15206 |
+
"step": 21230
|
| 15207 |
+
},
|
| 15208 |
+
{
|
| 15209 |
+
"epoch": 0.53,
|
| 15210 |
+
"grad_norm": 8.018983840942383,
|
| 15211 |
+
"learning_rate": 5.938983050847458e-06,
|
| 15212 |
+
"loss": 1.4798,
|
| 15213 |
+
"step": 21240
|
| 15214 |
+
},
|
| 15215 |
+
{
|
| 15216 |
+
"epoch": 0.53,
|
| 15217 |
+
"grad_norm": 2.752399206161499,
|
| 15218 |
+
"learning_rate": 5.932203389830509e-06,
|
| 15219 |
+
"loss": 1.4141,
|
| 15220 |
+
"step": 21250
|
| 15221 |
+
},
|
| 15222 |
+
{
|
| 15223 |
+
"epoch": 0.53,
|
| 15224 |
+
"grad_norm": 5.574487686157227,
|
| 15225 |
+
"learning_rate": 5.92542372881356e-06,
|
| 15226 |
+
"loss": 1.2121,
|
| 15227 |
+
"step": 21260
|
| 15228 |
+
},
|
| 15229 |
+
{
|
| 15230 |
+
"epoch": 0.53,
|
| 15231 |
+
"grad_norm": 10.352456092834473,
|
| 15232 |
+
"learning_rate": 5.91864406779661e-06,
|
| 15233 |
+
"loss": 1.3264,
|
| 15234 |
+
"step": 21270
|
| 15235 |
+
},
|
| 15236 |
+
{
|
| 15237 |
+
"epoch": 0.53,
|
| 15238 |
+
"grad_norm": 4.182732582092285,
|
| 15239 |
+
"learning_rate": 5.911864406779662e-06,
|
| 15240 |
+
"loss": 1.4334,
|
| 15241 |
+
"step": 21280
|
| 15242 |
+
},
|
| 15243 |
+
{
|
| 15244 |
+
"epoch": 0.53,
|
| 15245 |
+
"grad_norm": 5.310421943664551,
|
| 15246 |
+
"learning_rate": 5.905084745762712e-06,
|
| 15247 |
+
"loss": 1.5641,
|
| 15248 |
+
"step": 21290
|
| 15249 |
+
},
|
| 15250 |
+
{
|
| 15251 |
+
"epoch": 0.53,
|
| 15252 |
+
"grad_norm": 5.489622592926025,
|
| 15253 |
+
"learning_rate": 5.8983050847457635e-06,
|
| 15254 |
+
"loss": 1.4699,
|
| 15255 |
+
"step": 21300
|
| 15256 |
+
},
|
| 15257 |
+
{
|
| 15258 |
+
"epoch": 0.53,
|
| 15259 |
+
"grad_norm": 6.644534111022949,
|
| 15260 |
+
"learning_rate": 5.891525423728814e-06,
|
| 15261 |
+
"loss": 1.3273,
|
| 15262 |
+
"step": 21310
|
| 15263 |
+
},
|
| 15264 |
+
{
|
| 15265 |
+
"epoch": 0.53,
|
| 15266 |
+
"grad_norm": 13.480459213256836,
|
| 15267 |
+
"learning_rate": 5.8847457627118655e-06,
|
| 15268 |
+
"loss": 1.2957,
|
| 15269 |
+
"step": 21320
|
| 15270 |
+
},
|
| 15271 |
+
{
|
| 15272 |
+
"epoch": 0.53,
|
| 15273 |
+
"grad_norm": 9.34183120727539,
|
| 15274 |
+
"learning_rate": 5.877966101694915e-06,
|
| 15275 |
+
"loss": 1.4145,
|
| 15276 |
+
"step": 21330
|
| 15277 |
+
},
|
| 15278 |
+
{
|
| 15279 |
+
"epoch": 0.53,
|
| 15280 |
+
"grad_norm": 4.792973518371582,
|
| 15281 |
+
"learning_rate": 5.871186440677966e-06,
|
| 15282 |
+
"loss": 1.4912,
|
| 15283 |
+
"step": 21340
|
| 15284 |
+
},
|
| 15285 |
+
{
|
| 15286 |
+
"epoch": 0.53,
|
| 15287 |
+
"grad_norm": 4.098564147949219,
|
| 15288 |
+
"learning_rate": 5.864406779661017e-06,
|
| 15289 |
+
"loss": 1.1895,
|
| 15290 |
+
"step": 21350
|
| 15291 |
+
},
|
| 15292 |
+
{
|
| 15293 |
+
"epoch": 0.53,
|
| 15294 |
+
"grad_norm": 7.23917293548584,
|
| 15295 |
+
"learning_rate": 5.857627118644068e-06,
|
| 15296 |
+
"loss": 1.3875,
|
| 15297 |
+
"step": 21360
|
| 15298 |
+
},
|
| 15299 |
+
{
|
| 15300 |
+
"epoch": 0.53,
|
| 15301 |
+
"grad_norm": 4.86613130569458,
|
| 15302 |
+
"learning_rate": 5.850847457627119e-06,
|
| 15303 |
+
"loss": 1.4827,
|
| 15304 |
+
"step": 21370
|
| 15305 |
+
},
|
| 15306 |
+
{
|
| 15307 |
+
"epoch": 0.53,
|
| 15308 |
+
"grad_norm": 7.629755973815918,
|
| 15309 |
+
"learning_rate": 5.84406779661017e-06,
|
| 15310 |
+
"loss": 1.2189,
|
| 15311 |
+
"step": 21380
|
| 15312 |
+
},
|
| 15313 |
+
{
|
| 15314 |
+
"epoch": 0.53,
|
| 15315 |
+
"grad_norm": 3.80531907081604,
|
| 15316 |
+
"learning_rate": 5.837288135593221e-06,
|
| 15317 |
+
"loss": 1.4064,
|
| 15318 |
+
"step": 21390
|
| 15319 |
+
},
|
| 15320 |
+
{
|
| 15321 |
+
"epoch": 0.54,
|
| 15322 |
+
"grad_norm": 3.432089328765869,
|
| 15323 |
+
"learning_rate": 5.830508474576272e-06,
|
| 15324 |
+
"loss": 1.1929,
|
| 15325 |
+
"step": 21400
|
| 15326 |
+
},
|
| 15327 |
+
{
|
| 15328 |
+
"epoch": 0.54,
|
| 15329 |
+
"grad_norm": 9.766077995300293,
|
| 15330 |
+
"learning_rate": 5.823728813559323e-06,
|
| 15331 |
+
"loss": 1.4525,
|
| 15332 |
+
"step": 21410
|
| 15333 |
+
},
|
| 15334 |
+
{
|
| 15335 |
+
"epoch": 0.54,
|
| 15336 |
+
"grad_norm": 4.745760440826416,
|
| 15337 |
+
"learning_rate": 5.8169491525423736e-06,
|
| 15338 |
+
"loss": 1.4924,
|
| 15339 |
+
"step": 21420
|
| 15340 |
+
},
|
| 15341 |
+
{
|
| 15342 |
+
"epoch": 0.54,
|
| 15343 |
+
"grad_norm": 5.188168525695801,
|
| 15344 |
+
"learning_rate": 5.810169491525425e-06,
|
| 15345 |
+
"loss": 1.27,
|
| 15346 |
+
"step": 21430
|
| 15347 |
+
},
|
| 15348 |
+
{
|
| 15349 |
+
"epoch": 0.54,
|
| 15350 |
+
"grad_norm": 4.576213359832764,
|
| 15351 |
+
"learning_rate": 5.803389830508475e-06,
|
| 15352 |
+
"loss": 1.3165,
|
| 15353 |
+
"step": 21440
|
| 15354 |
+
},
|
| 15355 |
+
{
|
| 15356 |
+
"epoch": 0.54,
|
| 15357 |
+
"grad_norm": 10.540860176086426,
|
| 15358 |
+
"learning_rate": 5.796610169491525e-06,
|
| 15359 |
+
"loss": 1.2652,
|
| 15360 |
+
"step": 21450
|
| 15361 |
+
},
|
| 15362 |
+
{
|
| 15363 |
+
"epoch": 0.54,
|
| 15364 |
+
"grad_norm": 4.210390567779541,
|
| 15365 |
+
"learning_rate": 5.789830508474577e-06,
|
| 15366 |
+
"loss": 1.4372,
|
| 15367 |
+
"step": 21460
|
| 15368 |
+
},
|
| 15369 |
+
{
|
| 15370 |
+
"epoch": 0.54,
|
| 15371 |
+
"grad_norm": 8.733638763427734,
|
| 15372 |
+
"learning_rate": 5.783050847457627e-06,
|
| 15373 |
+
"loss": 1.2691,
|
| 15374 |
+
"step": 21470
|
| 15375 |
+
},
|
| 15376 |
+
{
|
| 15377 |
+
"epoch": 0.54,
|
| 15378 |
+
"grad_norm": 4.997326374053955,
|
| 15379 |
+
"learning_rate": 5.776271186440679e-06,
|
| 15380 |
+
"loss": 1.1454,
|
| 15381 |
+
"step": 21480
|
| 15382 |
+
},
|
| 15383 |
+
{
|
| 15384 |
+
"epoch": 0.54,
|
| 15385 |
+
"grad_norm": 10.108692169189453,
|
| 15386 |
+
"learning_rate": 5.769491525423729e-06,
|
| 15387 |
+
"loss": 1.2982,
|
| 15388 |
+
"step": 21490
|
| 15389 |
+
},
|
| 15390 |
+
{
|
| 15391 |
+
"epoch": 0.54,
|
| 15392 |
+
"grad_norm": 13.393025398254395,
|
| 15393 |
+
"learning_rate": 5.7627118644067805e-06,
|
| 15394 |
+
"loss": 1.4995,
|
| 15395 |
+
"step": 21500
|
| 15396 |
+
},
|
| 15397 |
+
{
|
| 15398 |
+
"epoch": 0.54,
|
| 15399 |
+
"eval_loss": 1.3565526008605957,
|
| 15400 |
+
"eval_runtime": 66.1537,
|
| 15401 |
+
"eval_samples_per_second": 15.116,
|
| 15402 |
+
"eval_steps_per_second": 15.116,
|
| 15403 |
+
"step": 21500
|
| 15404 |
+
},
|
| 15405 |
+
{
|
| 15406 |
+
"epoch": 0.54,
|
| 15407 |
+
"grad_norm": 5.675364017486572,
|
| 15408 |
+
"learning_rate": 5.755932203389831e-06,
|
| 15409 |
+
"loss": 1.2256,
|
| 15410 |
+
"step": 21510
|
| 15411 |
+
},
|
| 15412 |
+
{
|
| 15413 |
+
"epoch": 0.54,
|
| 15414 |
+
"grad_norm": 2.069751262664795,
|
| 15415 |
+
"learning_rate": 5.7491525423728825e-06,
|
| 15416 |
+
"loss": 1.2355,
|
| 15417 |
+
"step": 21520
|
| 15418 |
+
},
|
| 15419 |
+
{
|
| 15420 |
+
"epoch": 0.54,
|
| 15421 |
+
"grad_norm": 4.536093711853027,
|
| 15422 |
+
"learning_rate": 5.742372881355933e-06,
|
| 15423 |
+
"loss": 1.1957,
|
| 15424 |
+
"step": 21530
|
| 15425 |
+
},
|
| 15426 |
+
{
|
| 15427 |
+
"epoch": 0.54,
|
| 15428 |
+
"grad_norm": 2.5274765491485596,
|
| 15429 |
+
"learning_rate": 5.735593220338983e-06,
|
| 15430 |
+
"loss": 1.1843,
|
| 15431 |
+
"step": 21540
|
| 15432 |
+
},
|
| 15433 |
+
{
|
| 15434 |
+
"epoch": 0.54,
|
| 15435 |
+
"grad_norm": 4.418458461761475,
|
| 15436 |
+
"learning_rate": 5.728813559322034e-06,
|
| 15437 |
+
"loss": 1.1994,
|
| 15438 |
+
"step": 21550
|
| 15439 |
+
},
|
| 15440 |
+
{
|
| 15441 |
+
"epoch": 0.54,
|
| 15442 |
+
"grad_norm": 13.488496780395508,
|
| 15443 |
+
"learning_rate": 5.722033898305085e-06,
|
| 15444 |
+
"loss": 1.4173,
|
| 15445 |
+
"step": 21560
|
| 15446 |
+
},
|
| 15447 |
+
{
|
| 15448 |
+
"epoch": 0.54,
|
| 15449 |
+
"grad_norm": 5.223592758178711,
|
| 15450 |
+
"learning_rate": 5.715254237288136e-06,
|
| 15451 |
+
"loss": 1.3032,
|
| 15452 |
+
"step": 21570
|
| 15453 |
+
},
|
| 15454 |
+
{
|
| 15455 |
+
"epoch": 0.54,
|
| 15456 |
+
"grad_norm": 5.894464492797852,
|
| 15457 |
+
"learning_rate": 5.708474576271187e-06,
|
| 15458 |
+
"loss": 1.3027,
|
| 15459 |
+
"step": 21580
|
| 15460 |
+
},
|
| 15461 |
+
{
|
| 15462 |
+
"epoch": 0.54,
|
| 15463 |
+
"grad_norm": 6.945793151855469,
|
| 15464 |
+
"learning_rate": 5.701694915254238e-06,
|
| 15465 |
+
"loss": 1.3102,
|
| 15466 |
+
"step": 21590
|
| 15467 |
+
},
|
| 15468 |
+
{
|
| 15469 |
+
"epoch": 0.54,
|
| 15470 |
+
"grad_norm": 7.386875629425049,
|
| 15471 |
+
"learning_rate": 5.694915254237289e-06,
|
| 15472 |
+
"loss": 1.3643,
|
| 15473 |
+
"step": 21600
|
| 15474 |
+
},
|
| 15475 |
+
{
|
| 15476 |
+
"epoch": 0.54,
|
| 15477 |
+
"grad_norm": 3.5999162197113037,
|
| 15478 |
+
"learning_rate": 5.68813559322034e-06,
|
| 15479 |
+
"loss": 1.2456,
|
| 15480 |
+
"step": 21610
|
| 15481 |
+
},
|
| 15482 |
+
{
|
| 15483 |
+
"epoch": 0.54,
|
| 15484 |
+
"grad_norm": 5.8258490562438965,
|
| 15485 |
+
"learning_rate": 5.6813559322033906e-06,
|
| 15486 |
+
"loss": 1.1922,
|
| 15487 |
+
"step": 21620
|
| 15488 |
+
},
|
| 15489 |
+
{
|
| 15490 |
+
"epoch": 0.54,
|
| 15491 |
+
"grad_norm": 10.920169830322266,
|
| 15492 |
+
"learning_rate": 5.674576271186442e-06,
|
| 15493 |
+
"loss": 1.5312,
|
| 15494 |
+
"step": 21630
|
| 15495 |
+
},
|
| 15496 |
+
{
|
| 15497 |
+
"epoch": 0.54,
|
| 15498 |
+
"grad_norm": 3.898834705352783,
|
| 15499 |
+
"learning_rate": 5.667796610169492e-06,
|
| 15500 |
+
"loss": 1.5012,
|
| 15501 |
+
"step": 21640
|
| 15502 |
+
},
|
| 15503 |
+
{
|
| 15504 |
+
"epoch": 0.54,
|
| 15505 |
+
"grad_norm": 6.2130866050720215,
|
| 15506 |
+
"learning_rate": 5.661016949152542e-06,
|
| 15507 |
+
"loss": 1.543,
|
| 15508 |
+
"step": 21650
|
| 15509 |
+
},
|
| 15510 |
+
{
|
| 15511 |
+
"epoch": 0.54,
|
| 15512 |
+
"grad_norm": 3.604144811630249,
|
| 15513 |
+
"learning_rate": 5.654237288135594e-06,
|
| 15514 |
+
"loss": 1.4586,
|
| 15515 |
+
"step": 21660
|
| 15516 |
+
},
|
| 15517 |
+
{
|
| 15518 |
+
"epoch": 0.54,
|
| 15519 |
+
"grad_norm": 4.859696388244629,
|
| 15520 |
+
"learning_rate": 5.647457627118644e-06,
|
| 15521 |
+
"loss": 1.3056,
|
| 15522 |
+
"step": 21670
|
| 15523 |
+
},
|
| 15524 |
+
{
|
| 15525 |
+
"epoch": 0.54,
|
| 15526 |
+
"grad_norm": 2.9596614837646484,
|
| 15527 |
+
"learning_rate": 5.640677966101696e-06,
|
| 15528 |
+
"loss": 1.3888,
|
| 15529 |
+
"step": 21680
|
| 15530 |
+
},
|
| 15531 |
+
{
|
| 15532 |
+
"epoch": 0.54,
|
| 15533 |
+
"grad_norm": 5.489665985107422,
|
| 15534 |
+
"learning_rate": 5.633898305084746e-06,
|
| 15535 |
+
"loss": 1.2491,
|
| 15536 |
+
"step": 21690
|
| 15537 |
+
},
|
| 15538 |
+
{
|
| 15539 |
+
"epoch": 0.54,
|
| 15540 |
+
"grad_norm": 5.618114948272705,
|
| 15541 |
+
"learning_rate": 5.6271186440677975e-06,
|
| 15542 |
+
"loss": 1.2746,
|
| 15543 |
+
"step": 21700
|
| 15544 |
+
},
|
| 15545 |
+
{
|
| 15546 |
+
"epoch": 0.54,
|
| 15547 |
+
"grad_norm": 2.145024061203003,
|
| 15548 |
+
"learning_rate": 5.620338983050848e-06,
|
| 15549 |
+
"loss": 1.4036,
|
| 15550 |
+
"step": 21710
|
| 15551 |
+
},
|
| 15552 |
+
{
|
| 15553 |
+
"epoch": 0.54,
|
| 15554 |
+
"grad_norm": 2.4870400428771973,
|
| 15555 |
+
"learning_rate": 5.6135593220338995e-06,
|
| 15556 |
+
"loss": 1.4355,
|
| 15557 |
+
"step": 21720
|
| 15558 |
+
},
|
| 15559 |
+
{
|
| 15560 |
+
"epoch": 0.54,
|
| 15561 |
+
"grad_norm": 2.576144218444824,
|
| 15562 |
+
"learning_rate": 5.60677966101695e-06,
|
| 15563 |
+
"loss": 1.2988,
|
| 15564 |
+
"step": 21730
|
| 15565 |
+
},
|
| 15566 |
+
{
|
| 15567 |
+
"epoch": 0.54,
|
| 15568 |
+
"grad_norm": 5.971595764160156,
|
| 15569 |
+
"learning_rate": 5.600000000000001e-06,
|
| 15570 |
+
"loss": 1.3201,
|
| 15571 |
+
"step": 21740
|
| 15572 |
+
},
|
| 15573 |
+
{
|
| 15574 |
+
"epoch": 0.54,
|
| 15575 |
+
"grad_norm": 7.581085205078125,
|
| 15576 |
+
"learning_rate": 5.593220338983051e-06,
|
| 15577 |
+
"loss": 1.3358,
|
| 15578 |
+
"step": 21750
|
| 15579 |
+
},
|
| 15580 |
+
{
|
| 15581 |
+
"epoch": 0.54,
|
| 15582 |
+
"grad_norm": 4.148537635803223,
|
| 15583 |
+
"learning_rate": 5.586440677966102e-06,
|
| 15584 |
+
"loss": 1.2464,
|
| 15585 |
+
"step": 21760
|
| 15586 |
+
},
|
| 15587 |
+
{
|
| 15588 |
+
"epoch": 0.54,
|
| 15589 |
+
"grad_norm": 6.613537788391113,
|
| 15590 |
+
"learning_rate": 5.579661016949153e-06,
|
| 15591 |
+
"loss": 1.3156,
|
| 15592 |
+
"step": 21770
|
| 15593 |
+
},
|
| 15594 |
+
{
|
| 15595 |
+
"epoch": 0.54,
|
| 15596 |
+
"grad_norm": 10.526129722595215,
|
| 15597 |
+
"learning_rate": 5.572881355932204e-06,
|
| 15598 |
+
"loss": 1.2483,
|
| 15599 |
+
"step": 21780
|
| 15600 |
+
},
|
| 15601 |
+
{
|
| 15602 |
+
"epoch": 0.54,
|
| 15603 |
+
"grad_norm": 7.221047401428223,
|
| 15604 |
+
"learning_rate": 5.566101694915255e-06,
|
| 15605 |
+
"loss": 1.5878,
|
| 15606 |
+
"step": 21790
|
| 15607 |
+
},
|
| 15608 |
+
{
|
| 15609 |
+
"epoch": 0.55,
|
| 15610 |
+
"grad_norm": 6.365529537200928,
|
| 15611 |
+
"learning_rate": 5.559322033898306e-06,
|
| 15612 |
+
"loss": 1.4117,
|
| 15613 |
+
"step": 21800
|
| 15614 |
+
},
|
| 15615 |
+
{
|
| 15616 |
+
"epoch": 0.55,
|
| 15617 |
+
"grad_norm": 3.8305916786193848,
|
| 15618 |
+
"learning_rate": 5.552542372881356e-06,
|
| 15619 |
+
"loss": 1.1894,
|
| 15620 |
+
"step": 21810
|
| 15621 |
+
},
|
| 15622 |
+
{
|
| 15623 |
+
"epoch": 0.55,
|
| 15624 |
+
"grad_norm": 3.672477960586548,
|
| 15625 |
+
"learning_rate": 5.5457627118644076e-06,
|
| 15626 |
+
"loss": 1.2199,
|
| 15627 |
+
"step": 21820
|
| 15628 |
+
},
|
| 15629 |
+
{
|
| 15630 |
+
"epoch": 0.55,
|
| 15631 |
+
"grad_norm": 2.586512565612793,
|
| 15632 |
+
"learning_rate": 5.538983050847458e-06,
|
| 15633 |
+
"loss": 1.392,
|
| 15634 |
+
"step": 21830
|
| 15635 |
+
},
|
| 15636 |
+
{
|
| 15637 |
+
"epoch": 0.55,
|
| 15638 |
+
"grad_norm": 4.2184624671936035,
|
| 15639 |
+
"learning_rate": 5.5322033898305095e-06,
|
| 15640 |
+
"loss": 1.3008,
|
| 15641 |
+
"step": 21840
|
| 15642 |
+
},
|
| 15643 |
+
{
|
| 15644 |
+
"epoch": 0.55,
|
| 15645 |
+
"grad_norm": 7.834671974182129,
|
| 15646 |
+
"learning_rate": 5.525423728813559e-06,
|
| 15647 |
+
"loss": 1.1794,
|
| 15648 |
+
"step": 21850
|
| 15649 |
+
},
|
| 15650 |
+
{
|
| 15651 |
+
"epoch": 0.55,
|
| 15652 |
+
"grad_norm": 3.5877692699432373,
|
| 15653 |
+
"learning_rate": 5.518644067796611e-06,
|
| 15654 |
+
"loss": 1.4339,
|
| 15655 |
+
"step": 21860
|
| 15656 |
+
},
|
| 15657 |
+
{
|
| 15658 |
+
"epoch": 0.55,
|
| 15659 |
+
"grad_norm": 3.0174179077148438,
|
| 15660 |
+
"learning_rate": 5.511864406779661e-06,
|
| 15661 |
+
"loss": 1.2704,
|
| 15662 |
+
"step": 21870
|
| 15663 |
+
},
|
| 15664 |
+
{
|
| 15665 |
+
"epoch": 0.55,
|
| 15666 |
+
"grad_norm": 2.9889588356018066,
|
| 15667 |
+
"learning_rate": 5.505084745762712e-06,
|
| 15668 |
+
"loss": 1.3588,
|
| 15669 |
+
"step": 21880
|
| 15670 |
+
},
|
| 15671 |
+
{
|
| 15672 |
+
"epoch": 0.55,
|
| 15673 |
+
"grad_norm": 10.810959815979004,
|
| 15674 |
+
"learning_rate": 5.498305084745763e-06,
|
| 15675 |
+
"loss": 1.3539,
|
| 15676 |
+
"step": 21890
|
| 15677 |
+
},
|
| 15678 |
+
{
|
| 15679 |
+
"epoch": 0.55,
|
| 15680 |
+
"grad_norm": 5.771850109100342,
|
| 15681 |
+
"learning_rate": 5.491525423728814e-06,
|
| 15682 |
+
"loss": 1.4278,
|
| 15683 |
+
"step": 21900
|
| 15684 |
+
},
|
| 15685 |
+
{
|
| 15686 |
+
"epoch": 0.55,
|
| 15687 |
+
"grad_norm": 4.13969612121582,
|
| 15688 |
+
"learning_rate": 5.484745762711865e-06,
|
| 15689 |
+
"loss": 1.2798,
|
| 15690 |
+
"step": 21910
|
| 15691 |
+
},
|
| 15692 |
+
{
|
| 15693 |
+
"epoch": 0.55,
|
| 15694 |
+
"grad_norm": 15.295929908752441,
|
| 15695 |
+
"learning_rate": 5.477966101694916e-06,
|
| 15696 |
+
"loss": 1.5213,
|
| 15697 |
+
"step": 21920
|
| 15698 |
+
},
|
| 15699 |
+
{
|
| 15700 |
+
"epoch": 0.55,
|
| 15701 |
+
"grad_norm": 6.445948600769043,
|
| 15702 |
+
"learning_rate": 5.471186440677967e-06,
|
| 15703 |
+
"loss": 1.3498,
|
| 15704 |
+
"step": 21930
|
| 15705 |
+
},
|
| 15706 |
+
{
|
| 15707 |
+
"epoch": 0.55,
|
| 15708 |
+
"grad_norm": 9.28097152709961,
|
| 15709 |
+
"learning_rate": 5.464406779661018e-06,
|
| 15710 |
+
"loss": 1.3046,
|
| 15711 |
+
"step": 21940
|
| 15712 |
+
},
|
| 15713 |
+
{
|
| 15714 |
+
"epoch": 0.55,
|
| 15715 |
+
"grad_norm": 6.094447135925293,
|
| 15716 |
+
"learning_rate": 5.457627118644067e-06,
|
| 15717 |
+
"loss": 1.4336,
|
| 15718 |
+
"step": 21950
|
| 15719 |
+
},
|
| 15720 |
+
{
|
| 15721 |
+
"epoch": 0.55,
|
| 15722 |
+
"grad_norm": 9.818504333496094,
|
| 15723 |
+
"learning_rate": 5.450847457627119e-06,
|
| 15724 |
+
"loss": 1.516,
|
| 15725 |
+
"step": 21960
|
| 15726 |
+
},
|
| 15727 |
+
{
|
| 15728 |
+
"epoch": 0.55,
|
| 15729 |
+
"grad_norm": 11.956009864807129,
|
| 15730 |
+
"learning_rate": 5.444067796610169e-06,
|
| 15731 |
+
"loss": 1.3435,
|
| 15732 |
+
"step": 21970
|
| 15733 |
+
},
|
| 15734 |
+
{
|
| 15735 |
+
"epoch": 0.55,
|
| 15736 |
+
"grad_norm": 7.544681072235107,
|
| 15737 |
+
"learning_rate": 5.437288135593221e-06,
|
| 15738 |
+
"loss": 1.4625,
|
| 15739 |
+
"step": 21980
|
| 15740 |
+
},
|
| 15741 |
+
{
|
| 15742 |
+
"epoch": 0.55,
|
| 15743 |
+
"grad_norm": 3.394897222518921,
|
| 15744 |
+
"learning_rate": 5.430508474576271e-06,
|
| 15745 |
+
"loss": 1.4565,
|
| 15746 |
+
"step": 21990
|
| 15747 |
+
},
|
| 15748 |
+
{
|
| 15749 |
+
"epoch": 0.55,
|
| 15750 |
+
"grad_norm": 5.722468852996826,
|
| 15751 |
+
"learning_rate": 5.423728813559323e-06,
|
| 15752 |
+
"loss": 1.2133,
|
| 15753 |
+
"step": 22000
|
| 15754 |
+
},
|
| 15755 |
+
{
|
| 15756 |
+
"epoch": 0.55,
|
| 15757 |
+
"eval_loss": 1.3403723239898682,
|
| 15758 |
+
"eval_runtime": 66.1471,
|
| 15759 |
+
"eval_samples_per_second": 15.118,
|
| 15760 |
+
"eval_steps_per_second": 15.118,
|
| 15761 |
+
"step": 22000
|
| 15762 |
+
},
|
| 15763 |
+
{
|
| 15764 |
+
"epoch": 0.55,
|
| 15765 |
+
"grad_norm": 7.046730995178223,
|
| 15766 |
+
"learning_rate": 5.416949152542373e-06,
|
| 15767 |
+
"loss": 1.1353,
|
| 15768 |
+
"step": 22010
|
| 15769 |
+
},
|
| 15770 |
+
{
|
| 15771 |
+
"epoch": 0.55,
|
| 15772 |
+
"grad_norm": 7.013365268707275,
|
| 15773 |
+
"learning_rate": 5.4101694915254246e-06,
|
| 15774 |
+
"loss": 1.4333,
|
| 15775 |
+
"step": 22020
|
| 15776 |
+
},
|
| 15777 |
+
{
|
| 15778 |
+
"epoch": 0.55,
|
| 15779 |
+
"grad_norm": 3.9469892978668213,
|
| 15780 |
+
"learning_rate": 5.403389830508475e-06,
|
| 15781 |
+
"loss": 1.3601,
|
| 15782 |
+
"step": 22030
|
| 15783 |
+
},
|
| 15784 |
+
{
|
| 15785 |
+
"epoch": 0.55,
|
| 15786 |
+
"grad_norm": 3.740983486175537,
|
| 15787 |
+
"learning_rate": 5.3966101694915265e-06,
|
| 15788 |
+
"loss": 1.0232,
|
| 15789 |
+
"step": 22040
|
| 15790 |
+
},
|
| 15791 |
+
{
|
| 15792 |
+
"epoch": 0.55,
|
| 15793 |
+
"grad_norm": 4.604060649871826,
|
| 15794 |
+
"learning_rate": 5.389830508474577e-06,
|
| 15795 |
+
"loss": 1.4011,
|
| 15796 |
+
"step": 22050
|
| 15797 |
+
},
|
| 15798 |
+
{
|
| 15799 |
+
"epoch": 0.55,
|
| 15800 |
+
"grad_norm": 0.5601249933242798,
|
| 15801 |
+
"learning_rate": 5.383050847457627e-06,
|
| 15802 |
+
"loss": 1.2068,
|
| 15803 |
+
"step": 22060
|
| 15804 |
+
},
|
| 15805 |
+
{
|
| 15806 |
+
"epoch": 0.55,
|
| 15807 |
+
"grad_norm": 2.7781484127044678,
|
| 15808 |
+
"learning_rate": 5.376271186440678e-06,
|
| 15809 |
+
"loss": 1.3976,
|
| 15810 |
+
"step": 22070
|
| 15811 |
+
},
|
| 15812 |
+
{
|
| 15813 |
+
"epoch": 0.55,
|
| 15814 |
+
"grad_norm": 10.417901992797852,
|
| 15815 |
+
"learning_rate": 5.369491525423729e-06,
|
| 15816 |
+
"loss": 1.2444,
|
| 15817 |
+
"step": 22080
|
| 15818 |
+
},
|
| 15819 |
+
{
|
| 15820 |
+
"epoch": 0.55,
|
| 15821 |
+
"grad_norm": 8.263280868530273,
|
| 15822 |
+
"learning_rate": 5.36271186440678e-06,
|
| 15823 |
+
"loss": 1.0937,
|
| 15824 |
+
"step": 22090
|
| 15825 |
+
},
|
| 15826 |
+
{
|
| 15827 |
+
"epoch": 0.55,
|
| 15828 |
+
"grad_norm": 6.128343105316162,
|
| 15829 |
+
"learning_rate": 5.355932203389831e-06,
|
| 15830 |
+
"loss": 1.2891,
|
| 15831 |
+
"step": 22100
|
| 15832 |
+
},
|
| 15833 |
+
{
|
| 15834 |
+
"epoch": 0.55,
|
| 15835 |
+
"grad_norm": 4.099800109863281,
|
| 15836 |
+
"learning_rate": 5.349152542372882e-06,
|
| 15837 |
+
"loss": 1.117,
|
| 15838 |
+
"step": 22110
|
| 15839 |
+
},
|
| 15840 |
+
{
|
| 15841 |
+
"epoch": 0.55,
|
| 15842 |
+
"grad_norm": 7.756937026977539,
|
| 15843 |
+
"learning_rate": 5.342372881355933e-06,
|
| 15844 |
+
"loss": 1.3112,
|
| 15845 |
+
"step": 22120
|
| 15846 |
+
},
|
| 15847 |
+
{
|
| 15848 |
+
"epoch": 0.55,
|
| 15849 |
+
"grad_norm": 5.372160911560059,
|
| 15850 |
+
"learning_rate": 5.335593220338984e-06,
|
| 15851 |
+
"loss": 1.5276,
|
| 15852 |
+
"step": 22130
|
| 15853 |
+
},
|
| 15854 |
+
{
|
| 15855 |
+
"epoch": 0.55,
|
| 15856 |
+
"grad_norm": 5.017634391784668,
|
| 15857 |
+
"learning_rate": 5.328813559322035e-06,
|
| 15858 |
+
"loss": 1.2825,
|
| 15859 |
+
"step": 22140
|
| 15860 |
+
},
|
| 15861 |
+
{
|
| 15862 |
+
"epoch": 0.55,
|
| 15863 |
+
"grad_norm": 5.629271030426025,
|
| 15864 |
+
"learning_rate": 5.322033898305086e-06,
|
| 15865 |
+
"loss": 1.2428,
|
| 15866 |
+
"step": 22150
|
| 15867 |
+
},
|
| 15868 |
+
{
|
| 15869 |
+
"epoch": 0.55,
|
| 15870 |
+
"grad_norm": 6.938544273376465,
|
| 15871 |
+
"learning_rate": 5.315254237288136e-06,
|
| 15872 |
+
"loss": 1.4563,
|
| 15873 |
+
"step": 22160
|
| 15874 |
+
},
|
| 15875 |
+
{
|
| 15876 |
+
"epoch": 0.55,
|
| 15877 |
+
"grad_norm": 13.804441452026367,
|
| 15878 |
+
"learning_rate": 5.308474576271186e-06,
|
| 15879 |
+
"loss": 1.332,
|
| 15880 |
+
"step": 22170
|
| 15881 |
+
},
|
| 15882 |
+
{
|
| 15883 |
+
"epoch": 0.55,
|
| 15884 |
+
"grad_norm": 10.347596168518066,
|
| 15885 |
+
"learning_rate": 5.301694915254238e-06,
|
| 15886 |
+
"loss": 1.3879,
|
| 15887 |
+
"step": 22180
|
| 15888 |
+
},
|
| 15889 |
+
{
|
| 15890 |
+
"epoch": 0.55,
|
| 15891 |
+
"grad_norm": 2.613632917404175,
|
| 15892 |
+
"learning_rate": 5.294915254237288e-06,
|
| 15893 |
+
"loss": 1.3298,
|
| 15894 |
+
"step": 22190
|
| 15895 |
+
},
|
| 15896 |
+
{
|
| 15897 |
+
"epoch": 0.56,
|
| 15898 |
+
"grad_norm": 14.637787818908691,
|
| 15899 |
+
"learning_rate": 5.28813559322034e-06,
|
| 15900 |
+
"loss": 1.3301,
|
| 15901 |
+
"step": 22200
|
| 15902 |
+
},
|
| 15903 |
+
{
|
| 15904 |
+
"epoch": 0.56,
|
| 15905 |
+
"grad_norm": 2.5796003341674805,
|
| 15906 |
+
"learning_rate": 5.28135593220339e-06,
|
| 15907 |
+
"loss": 1.41,
|
| 15908 |
+
"step": 22210
|
| 15909 |
+
},
|
| 15910 |
+
{
|
| 15911 |
+
"epoch": 0.56,
|
| 15912 |
+
"grad_norm": 5.326439380645752,
|
| 15913 |
+
"learning_rate": 5.2745762711864416e-06,
|
| 15914 |
+
"loss": 1.2483,
|
| 15915 |
+
"step": 22220
|
| 15916 |
+
},
|
| 15917 |
+
{
|
| 15918 |
+
"epoch": 0.56,
|
| 15919 |
+
"grad_norm": 8.928110122680664,
|
| 15920 |
+
"learning_rate": 5.267796610169492e-06,
|
| 15921 |
+
"loss": 1.2566,
|
| 15922 |
+
"step": 22230
|
| 15923 |
+
},
|
| 15924 |
+
{
|
| 15925 |
+
"epoch": 0.56,
|
| 15926 |
+
"grad_norm": 9.285192489624023,
|
| 15927 |
+
"learning_rate": 5.2610169491525435e-06,
|
| 15928 |
+
"loss": 1.2982,
|
| 15929 |
+
"step": 22240
|
| 15930 |
+
},
|
| 15931 |
+
{
|
| 15932 |
+
"epoch": 0.56,
|
| 15933 |
+
"grad_norm": 2.101649045944214,
|
| 15934 |
+
"learning_rate": 5.254237288135594e-06,
|
| 15935 |
+
"loss": 1.3181,
|
| 15936 |
+
"step": 22250
|
| 15937 |
+
},
|
| 15938 |
+
{
|
| 15939 |
+
"epoch": 0.56,
|
| 15940 |
+
"grad_norm": 5.57994270324707,
|
| 15941 |
+
"learning_rate": 5.247457627118645e-06,
|
| 15942 |
+
"loss": 1.3745,
|
| 15943 |
+
"step": 22260
|
| 15944 |
+
},
|
| 15945 |
+
{
|
| 15946 |
+
"epoch": 0.56,
|
| 15947 |
+
"grad_norm": 12.0460844039917,
|
| 15948 |
+
"learning_rate": 5.240677966101695e-06,
|
| 15949 |
+
"loss": 1.202,
|
| 15950 |
+
"step": 22270
|
| 15951 |
+
},
|
| 15952 |
+
{
|
| 15953 |
+
"epoch": 0.56,
|
| 15954 |
+
"grad_norm": 5.367649078369141,
|
| 15955 |
+
"learning_rate": 5.233898305084746e-06,
|
| 15956 |
+
"loss": 1.2969,
|
| 15957 |
+
"step": 22280
|
| 15958 |
+
},
|
| 15959 |
+
{
|
| 15960 |
+
"epoch": 0.56,
|
| 15961 |
+
"grad_norm": 4.784932613372803,
|
| 15962 |
+
"learning_rate": 5.227118644067797e-06,
|
| 15963 |
+
"loss": 1.363,
|
| 15964 |
+
"step": 22290
|
| 15965 |
+
},
|
| 15966 |
+
{
|
| 15967 |
+
"epoch": 0.56,
|
| 15968 |
+
"grad_norm": 8.103744506835938,
|
| 15969 |
+
"learning_rate": 5.220338983050848e-06,
|
| 15970 |
+
"loss": 1.4144,
|
| 15971 |
+
"step": 22300
|
| 15972 |
+
},
|
| 15973 |
+
{
|
| 15974 |
+
"epoch": 0.56,
|
| 15975 |
+
"grad_norm": 5.561450958251953,
|
| 15976 |
+
"learning_rate": 5.213559322033899e-06,
|
| 15977 |
+
"loss": 1.3018,
|
| 15978 |
+
"step": 22310
|
| 15979 |
+
},
|
| 15980 |
+
{
|
| 15981 |
+
"epoch": 0.56,
|
| 15982 |
+
"grad_norm": 8.294450759887695,
|
| 15983 |
+
"learning_rate": 5.20677966101695e-06,
|
| 15984 |
+
"loss": 1.3509,
|
| 15985 |
+
"step": 22320
|
| 15986 |
+
},
|
| 15987 |
+
{
|
| 15988 |
+
"epoch": 0.56,
|
| 15989 |
+
"grad_norm": 8.498016357421875,
|
| 15990 |
+
"learning_rate": 5.2e-06,
|
| 15991 |
+
"loss": 1.4932,
|
| 15992 |
+
"step": 22330
|
| 15993 |
+
},
|
| 15994 |
+
{
|
| 15995 |
+
"epoch": 0.56,
|
| 15996 |
+
"grad_norm": 3.3640925884246826,
|
| 15997 |
+
"learning_rate": 5.193220338983052e-06,
|
| 15998 |
+
"loss": 1.4413,
|
| 15999 |
+
"step": 22340
|
| 16000 |
+
},
|
| 16001 |
+
{
|
| 16002 |
+
"epoch": 0.56,
|
| 16003 |
+
"grad_norm": 3.236924886703491,
|
| 16004 |
+
"learning_rate": 5.186440677966102e-06,
|
| 16005 |
+
"loss": 1.2535,
|
| 16006 |
+
"step": 22350
|
| 16007 |
+
},
|
| 16008 |
+
{
|
| 16009 |
+
"epoch": 0.56,
|
| 16010 |
+
"grad_norm": 6.234226703643799,
|
| 16011 |
+
"learning_rate": 5.1796610169491535e-06,
|
| 16012 |
+
"loss": 1.1816,
|
| 16013 |
+
"step": 22360
|
| 16014 |
+
},
|
| 16015 |
+
{
|
| 16016 |
+
"epoch": 0.56,
|
| 16017 |
+
"grad_norm": 6.9503703117370605,
|
| 16018 |
+
"learning_rate": 5.172881355932203e-06,
|
| 16019 |
+
"loss": 1.2022,
|
| 16020 |
+
"step": 22370
|
| 16021 |
+
},
|
| 16022 |
+
{
|
| 16023 |
+
"epoch": 0.56,
|
| 16024 |
+
"grad_norm": 11.745012283325195,
|
| 16025 |
+
"learning_rate": 5.166101694915255e-06,
|
| 16026 |
+
"loss": 1.543,
|
| 16027 |
+
"step": 22380
|
| 16028 |
+
},
|
| 16029 |
+
{
|
| 16030 |
+
"epoch": 0.56,
|
| 16031 |
+
"grad_norm": 8.977341651916504,
|
| 16032 |
+
"learning_rate": 5.159322033898305e-06,
|
| 16033 |
+
"loss": 1.4185,
|
| 16034 |
+
"step": 22390
|
| 16035 |
+
},
|
| 16036 |
+
{
|
| 16037 |
+
"epoch": 0.56,
|
| 16038 |
+
"grad_norm": 2.515448570251465,
|
| 16039 |
+
"learning_rate": 5.152542372881356e-06,
|
| 16040 |
+
"loss": 1.2665,
|
| 16041 |
+
"step": 22400
|
| 16042 |
+
},
|
| 16043 |
+
{
|
| 16044 |
+
"epoch": 0.56,
|
| 16045 |
+
"grad_norm": 9.47354507446289,
|
| 16046 |
+
"learning_rate": 5.145762711864407e-06,
|
| 16047 |
+
"loss": 1.3086,
|
| 16048 |
+
"step": 22410
|
| 16049 |
+
},
|
| 16050 |
+
{
|
| 16051 |
+
"epoch": 0.56,
|
| 16052 |
+
"grad_norm": 6.550108432769775,
|
| 16053 |
+
"learning_rate": 5.138983050847458e-06,
|
| 16054 |
+
"loss": 1.3129,
|
| 16055 |
+
"step": 22420
|
| 16056 |
+
},
|
| 16057 |
+
{
|
| 16058 |
+
"epoch": 0.56,
|
| 16059 |
+
"grad_norm": 3.860886335372925,
|
| 16060 |
+
"learning_rate": 5.132203389830509e-06,
|
| 16061 |
+
"loss": 1.4046,
|
| 16062 |
+
"step": 22430
|
| 16063 |
+
},
|
| 16064 |
+
{
|
| 16065 |
+
"epoch": 0.56,
|
| 16066 |
+
"grad_norm": 8.062068939208984,
|
| 16067 |
+
"learning_rate": 5.12542372881356e-06,
|
| 16068 |
+
"loss": 1.207,
|
| 16069 |
+
"step": 22440
|
| 16070 |
+
},
|
| 16071 |
+
{
|
| 16072 |
+
"epoch": 0.56,
|
| 16073 |
+
"grad_norm": 5.823556423187256,
|
| 16074 |
+
"learning_rate": 5.118644067796611e-06,
|
| 16075 |
+
"loss": 1.2547,
|
| 16076 |
+
"step": 22450
|
| 16077 |
+
},
|
| 16078 |
+
{
|
| 16079 |
+
"epoch": 0.56,
|
| 16080 |
+
"grad_norm": 7.168029308319092,
|
| 16081 |
+
"learning_rate": 5.111864406779662e-06,
|
| 16082 |
+
"loss": 1.2772,
|
| 16083 |
+
"step": 22460
|
| 16084 |
+
},
|
| 16085 |
+
{
|
| 16086 |
+
"epoch": 0.56,
|
| 16087 |
+
"grad_norm": 9.224081039428711,
|
| 16088 |
+
"learning_rate": 5.105084745762711e-06,
|
| 16089 |
+
"loss": 1.3066,
|
| 16090 |
+
"step": 22470
|
| 16091 |
+
},
|
| 16092 |
+
{
|
| 16093 |
+
"epoch": 0.56,
|
| 16094 |
+
"grad_norm": 3.878537178039551,
|
| 16095 |
+
"learning_rate": 5.098305084745763e-06,
|
| 16096 |
+
"loss": 1.2413,
|
| 16097 |
+
"step": 22480
|
| 16098 |
+
},
|
| 16099 |
+
{
|
| 16100 |
+
"epoch": 0.56,
|
| 16101 |
+
"grad_norm": 5.821982383728027,
|
| 16102 |
+
"learning_rate": 5.091525423728813e-06,
|
| 16103 |
+
"loss": 1.3609,
|
| 16104 |
+
"step": 22490
|
| 16105 |
+
},
|
| 16106 |
+
{
|
| 16107 |
+
"epoch": 0.56,
|
| 16108 |
+
"grad_norm": 9.057456970214844,
|
| 16109 |
+
"learning_rate": 5.084745762711865e-06,
|
| 16110 |
+
"loss": 1.2255,
|
| 16111 |
+
"step": 22500
|
| 16112 |
+
},
|
| 16113 |
+
{
|
| 16114 |
+
"epoch": 0.56,
|
| 16115 |
+
"eval_loss": 1.3439137935638428,
|
| 16116 |
+
"eval_runtime": 66.2088,
|
| 16117 |
+
"eval_samples_per_second": 15.104,
|
| 16118 |
+
"eval_steps_per_second": 15.104,
|
| 16119 |
+
"step": 22500
|
| 16120 |
}
|
| 16121 |
],
|
| 16122 |
"logging_steps": 10,
|
|
|
|
| 16124 |
"num_input_tokens_seen": 0,
|
| 16125 |
"num_train_epochs": 1,
|
| 16126 |
"save_steps": 2500,
|
| 16127 |
+
"total_flos": 3.6229783486464e+17,
|
| 16128 |
"train_batch_size": 1,
|
| 16129 |
"trial_name": null,
|
| 16130 |
"trial_params": null
|