Uploaded checkpoint-30000
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1793 -3
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:f74097a448e18def20ac64c3351b03daf39ecbb7772ee021bb94d78882a822e8
|
| 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:bac2bf27f76ba10b8bb7d0aea6171dd39cb800cb897cebf7d6a638cfd591dce7
|
| 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:bc4e81d8710419a2e714a6936530b076f37f0580ef4ada57c8cd6f905915e300
|
| 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 sha256:8b36093e06845c6146f3175c64f0e8bdb441d4f7fc67a6962ed0b80b6725daf1
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 1.262895941734314,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-27500",
|
| 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,
|
|
@@ -19697,6 +19697,1796 @@
|
|
| 19697 |
"eval_samples_per_second": 15.103,
|
| 19698 |
"eval_steps_per_second": 15.103,
|
| 19699 |
"step": 27500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19700 |
}
|
| 19701 |
],
|
| 19702 |
"logging_steps": 10,
|
|
@@ -19704,7 +21494,7 @@
|
|
| 19704 |
"num_input_tokens_seen": 0,
|
| 19705 |
"num_train_epochs": 1,
|
| 19706 |
"save_steps": 2500,
|
| 19707 |
-
"total_flos": 4.
|
| 19708 |
"train_batch_size": 1,
|
| 19709 |
"trial_name": null,
|
| 19710 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 1.262895941734314,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-27500",
|
| 4 |
+
"epoch": 0.75,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 30000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 19697 |
"eval_samples_per_second": 15.103,
|
| 19698 |
"eval_steps_per_second": 15.103,
|
| 19699 |
"step": 27500
|
| 19700 |
+
},
|
| 19701 |
+
{
|
| 19702 |
+
"epoch": 0.69,
|
| 19703 |
+
"grad_norm": 5.283616542816162,
|
| 19704 |
+
"learning_rate": 1.6881355932203391e-06,
|
| 19705 |
+
"loss": 1.3504,
|
| 19706 |
+
"step": 27510
|
| 19707 |
+
},
|
| 19708 |
+
{
|
| 19709 |
+
"epoch": 0.69,
|
| 19710 |
+
"grad_norm": 3.6427159309387207,
|
| 19711 |
+
"learning_rate": 1.6813559322033901e-06,
|
| 19712 |
+
"loss": 1.3028,
|
| 19713 |
+
"step": 27520
|
| 19714 |
+
},
|
| 19715 |
+
{
|
| 19716 |
+
"epoch": 0.69,
|
| 19717 |
+
"grad_norm": 5.701483249664307,
|
| 19718 |
+
"learning_rate": 1.6745762711864409e-06,
|
| 19719 |
+
"loss": 1.4436,
|
| 19720 |
+
"step": 27530
|
| 19721 |
+
},
|
| 19722 |
+
{
|
| 19723 |
+
"epoch": 0.69,
|
| 19724 |
+
"grad_norm": 9.295339584350586,
|
| 19725 |
+
"learning_rate": 1.6677966101694916e-06,
|
| 19726 |
+
"loss": 1.4106,
|
| 19727 |
+
"step": 27540
|
| 19728 |
+
},
|
| 19729 |
+
{
|
| 19730 |
+
"epoch": 0.69,
|
| 19731 |
+
"grad_norm": 3.1163859367370605,
|
| 19732 |
+
"learning_rate": 1.6610169491525424e-06,
|
| 19733 |
+
"loss": 1.2698,
|
| 19734 |
+
"step": 27550
|
| 19735 |
+
},
|
| 19736 |
+
{
|
| 19737 |
+
"epoch": 0.69,
|
| 19738 |
+
"grad_norm": 6.673079967498779,
|
| 19739 |
+
"learning_rate": 1.6542372881355934e-06,
|
| 19740 |
+
"loss": 1.2202,
|
| 19741 |
+
"step": 27560
|
| 19742 |
+
},
|
| 19743 |
+
{
|
| 19744 |
+
"epoch": 0.69,
|
| 19745 |
+
"grad_norm": 4.866211414337158,
|
| 19746 |
+
"learning_rate": 1.6474576271186444e-06,
|
| 19747 |
+
"loss": 1.2538,
|
| 19748 |
+
"step": 27570
|
| 19749 |
+
},
|
| 19750 |
+
{
|
| 19751 |
+
"epoch": 0.69,
|
| 19752 |
+
"grad_norm": 3.319688081741333,
|
| 19753 |
+
"learning_rate": 1.640677966101695e-06,
|
| 19754 |
+
"loss": 1.4981,
|
| 19755 |
+
"step": 27580
|
| 19756 |
+
},
|
| 19757 |
+
{
|
| 19758 |
+
"epoch": 0.69,
|
| 19759 |
+
"grad_norm": 5.170316219329834,
|
| 19760 |
+
"learning_rate": 1.6338983050847459e-06,
|
| 19761 |
+
"loss": 1.3682,
|
| 19762 |
+
"step": 27590
|
| 19763 |
+
},
|
| 19764 |
+
{
|
| 19765 |
+
"epoch": 0.69,
|
| 19766 |
+
"grad_norm": 9.018121719360352,
|
| 19767 |
+
"learning_rate": 1.6271186440677967e-06,
|
| 19768 |
+
"loss": 1.3149,
|
| 19769 |
+
"step": 27600
|
| 19770 |
+
},
|
| 19771 |
+
{
|
| 19772 |
+
"epoch": 0.69,
|
| 19773 |
+
"grad_norm": 6.896349906921387,
|
| 19774 |
+
"learning_rate": 1.6203389830508476e-06,
|
| 19775 |
+
"loss": 1.317,
|
| 19776 |
+
"step": 27610
|
| 19777 |
+
},
|
| 19778 |
+
{
|
| 19779 |
+
"epoch": 0.69,
|
| 19780 |
+
"grad_norm": 7.63593053817749,
|
| 19781 |
+
"learning_rate": 1.6135593220338986e-06,
|
| 19782 |
+
"loss": 1.3238,
|
| 19783 |
+
"step": 27620
|
| 19784 |
+
},
|
| 19785 |
+
{
|
| 19786 |
+
"epoch": 0.69,
|
| 19787 |
+
"grad_norm": 8.823158264160156,
|
| 19788 |
+
"learning_rate": 1.6067796610169492e-06,
|
| 19789 |
+
"loss": 1.4479,
|
| 19790 |
+
"step": 27630
|
| 19791 |
+
},
|
| 19792 |
+
{
|
| 19793 |
+
"epoch": 0.69,
|
| 19794 |
+
"grad_norm": 8.540557861328125,
|
| 19795 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 19796 |
+
"loss": 1.3663,
|
| 19797 |
+
"step": 27640
|
| 19798 |
+
},
|
| 19799 |
+
{
|
| 19800 |
+
"epoch": 0.69,
|
| 19801 |
+
"grad_norm": 8.03847885131836,
|
| 19802 |
+
"learning_rate": 1.593220338983051e-06,
|
| 19803 |
+
"loss": 1.2248,
|
| 19804 |
+
"step": 27650
|
| 19805 |
+
},
|
| 19806 |
+
{
|
| 19807 |
+
"epoch": 0.69,
|
| 19808 |
+
"grad_norm": 5.210377216339111,
|
| 19809 |
+
"learning_rate": 1.5864406779661019e-06,
|
| 19810 |
+
"loss": 1.3373,
|
| 19811 |
+
"step": 27660
|
| 19812 |
+
},
|
| 19813 |
+
{
|
| 19814 |
+
"epoch": 0.69,
|
| 19815 |
+
"grad_norm": 6.994758129119873,
|
| 19816 |
+
"learning_rate": 1.5796610169491526e-06,
|
| 19817 |
+
"loss": 1.3166,
|
| 19818 |
+
"step": 27670
|
| 19819 |
+
},
|
| 19820 |
+
{
|
| 19821 |
+
"epoch": 0.69,
|
| 19822 |
+
"grad_norm": 7.343669414520264,
|
| 19823 |
+
"learning_rate": 1.5728813559322034e-06,
|
| 19824 |
+
"loss": 1.3406,
|
| 19825 |
+
"step": 27680
|
| 19826 |
+
},
|
| 19827 |
+
{
|
| 19828 |
+
"epoch": 0.69,
|
| 19829 |
+
"grad_norm": 12.689948081970215,
|
| 19830 |
+
"learning_rate": 1.5661016949152544e-06,
|
| 19831 |
+
"loss": 1.2073,
|
| 19832 |
+
"step": 27690
|
| 19833 |
+
},
|
| 19834 |
+
{
|
| 19835 |
+
"epoch": 0.69,
|
| 19836 |
+
"grad_norm": 7.018815040588379,
|
| 19837 |
+
"learning_rate": 1.5593220338983054e-06,
|
| 19838 |
+
"loss": 1.2738,
|
| 19839 |
+
"step": 27700
|
| 19840 |
+
},
|
| 19841 |
+
{
|
| 19842 |
+
"epoch": 0.69,
|
| 19843 |
+
"grad_norm": 5.334643840789795,
|
| 19844 |
+
"learning_rate": 1.552542372881356e-06,
|
| 19845 |
+
"loss": 1.2211,
|
| 19846 |
+
"step": 27710
|
| 19847 |
+
},
|
| 19848 |
+
{
|
| 19849 |
+
"epoch": 0.69,
|
| 19850 |
+
"grad_norm": 1.1950844526290894,
|
| 19851 |
+
"learning_rate": 1.545762711864407e-06,
|
| 19852 |
+
"loss": 1.268,
|
| 19853 |
+
"step": 27720
|
| 19854 |
+
},
|
| 19855 |
+
{
|
| 19856 |
+
"epoch": 0.69,
|
| 19857 |
+
"grad_norm": 11.679058074951172,
|
| 19858 |
+
"learning_rate": 1.5389830508474577e-06,
|
| 19859 |
+
"loss": 1.1873,
|
| 19860 |
+
"step": 27730
|
| 19861 |
+
},
|
| 19862 |
+
{
|
| 19863 |
+
"epoch": 0.69,
|
| 19864 |
+
"grad_norm": 7.686078071594238,
|
| 19865 |
+
"learning_rate": 1.5322033898305086e-06,
|
| 19866 |
+
"loss": 1.277,
|
| 19867 |
+
"step": 27740
|
| 19868 |
+
},
|
| 19869 |
+
{
|
| 19870 |
+
"epoch": 0.69,
|
| 19871 |
+
"grad_norm": 2.579845905303955,
|
| 19872 |
+
"learning_rate": 1.5254237288135596e-06,
|
| 19873 |
+
"loss": 1.3752,
|
| 19874 |
+
"step": 27750
|
| 19875 |
+
},
|
| 19876 |
+
{
|
| 19877 |
+
"epoch": 0.69,
|
| 19878 |
+
"grad_norm": 11.227088928222656,
|
| 19879 |
+
"learning_rate": 1.5186440677966102e-06,
|
| 19880 |
+
"loss": 1.2593,
|
| 19881 |
+
"step": 27760
|
| 19882 |
+
},
|
| 19883 |
+
{
|
| 19884 |
+
"epoch": 0.69,
|
| 19885 |
+
"grad_norm": 6.903045654296875,
|
| 19886 |
+
"learning_rate": 1.5118644067796611e-06,
|
| 19887 |
+
"loss": 1.2463,
|
| 19888 |
+
"step": 27770
|
| 19889 |
+
},
|
| 19890 |
+
{
|
| 19891 |
+
"epoch": 0.69,
|
| 19892 |
+
"grad_norm": 12.534771919250488,
|
| 19893 |
+
"learning_rate": 1.505084745762712e-06,
|
| 19894 |
+
"loss": 1.2949,
|
| 19895 |
+
"step": 27780
|
| 19896 |
+
},
|
| 19897 |
+
{
|
| 19898 |
+
"epoch": 0.69,
|
| 19899 |
+
"grad_norm": 4.901615619659424,
|
| 19900 |
+
"learning_rate": 1.4983050847457629e-06,
|
| 19901 |
+
"loss": 1.2975,
|
| 19902 |
+
"step": 27790
|
| 19903 |
+
},
|
| 19904 |
+
{
|
| 19905 |
+
"epoch": 0.69,
|
| 19906 |
+
"grad_norm": 9.669910430908203,
|
| 19907 |
+
"learning_rate": 1.4915254237288139e-06,
|
| 19908 |
+
"loss": 1.3476,
|
| 19909 |
+
"step": 27800
|
| 19910 |
+
},
|
| 19911 |
+
{
|
| 19912 |
+
"epoch": 0.7,
|
| 19913 |
+
"grad_norm": 6.982202529907227,
|
| 19914 |
+
"learning_rate": 1.4847457627118644e-06,
|
| 19915 |
+
"loss": 1.3839,
|
| 19916 |
+
"step": 27810
|
| 19917 |
+
},
|
| 19918 |
+
{
|
| 19919 |
+
"epoch": 0.7,
|
| 19920 |
+
"grad_norm": 9.172724723815918,
|
| 19921 |
+
"learning_rate": 1.4779661016949154e-06,
|
| 19922 |
+
"loss": 1.3519,
|
| 19923 |
+
"step": 27820
|
| 19924 |
+
},
|
| 19925 |
+
{
|
| 19926 |
+
"epoch": 0.7,
|
| 19927 |
+
"grad_norm": 8.561583518981934,
|
| 19928 |
+
"learning_rate": 1.4711864406779664e-06,
|
| 19929 |
+
"loss": 1.3402,
|
| 19930 |
+
"step": 27830
|
| 19931 |
+
},
|
| 19932 |
+
{
|
| 19933 |
+
"epoch": 0.7,
|
| 19934 |
+
"grad_norm": 11.40105152130127,
|
| 19935 |
+
"learning_rate": 1.464406779661017e-06,
|
| 19936 |
+
"loss": 1.2805,
|
| 19937 |
+
"step": 27840
|
| 19938 |
+
},
|
| 19939 |
+
{
|
| 19940 |
+
"epoch": 0.7,
|
| 19941 |
+
"grad_norm": 11.502050399780273,
|
| 19942 |
+
"learning_rate": 1.457627118644068e-06,
|
| 19943 |
+
"loss": 1.4351,
|
| 19944 |
+
"step": 27850
|
| 19945 |
+
},
|
| 19946 |
+
{
|
| 19947 |
+
"epoch": 0.7,
|
| 19948 |
+
"grad_norm": 3.5157549381256104,
|
| 19949 |
+
"learning_rate": 1.4508474576271187e-06,
|
| 19950 |
+
"loss": 1.2948,
|
| 19951 |
+
"step": 27860
|
| 19952 |
+
},
|
| 19953 |
+
{
|
| 19954 |
+
"epoch": 0.7,
|
| 19955 |
+
"grad_norm": 8.663883209228516,
|
| 19956 |
+
"learning_rate": 1.4440677966101696e-06,
|
| 19957 |
+
"loss": 1.3285,
|
| 19958 |
+
"step": 27870
|
| 19959 |
+
},
|
| 19960 |
+
{
|
| 19961 |
+
"epoch": 0.7,
|
| 19962 |
+
"grad_norm": 4.502950191497803,
|
| 19963 |
+
"learning_rate": 1.4372881355932206e-06,
|
| 19964 |
+
"loss": 1.4507,
|
| 19965 |
+
"step": 27880
|
| 19966 |
+
},
|
| 19967 |
+
{
|
| 19968 |
+
"epoch": 0.7,
|
| 19969 |
+
"grad_norm": 2.098318338394165,
|
| 19970 |
+
"learning_rate": 1.4305084745762712e-06,
|
| 19971 |
+
"loss": 1.1753,
|
| 19972 |
+
"step": 27890
|
| 19973 |
+
},
|
| 19974 |
+
{
|
| 19975 |
+
"epoch": 0.7,
|
| 19976 |
+
"grad_norm": 2.721998453140259,
|
| 19977 |
+
"learning_rate": 1.4237288135593222e-06,
|
| 19978 |
+
"loss": 1.3162,
|
| 19979 |
+
"step": 27900
|
| 19980 |
+
},
|
| 19981 |
+
{
|
| 19982 |
+
"epoch": 0.7,
|
| 19983 |
+
"grad_norm": 6.501703262329102,
|
| 19984 |
+
"learning_rate": 1.416949152542373e-06,
|
| 19985 |
+
"loss": 1.2442,
|
| 19986 |
+
"step": 27910
|
| 19987 |
+
},
|
| 19988 |
+
{
|
| 19989 |
+
"epoch": 0.7,
|
| 19990 |
+
"grad_norm": 4.597460746765137,
|
| 19991 |
+
"learning_rate": 1.410169491525424e-06,
|
| 19992 |
+
"loss": 1.2659,
|
| 19993 |
+
"step": 27920
|
| 19994 |
+
},
|
| 19995 |
+
{
|
| 19996 |
+
"epoch": 0.7,
|
| 19997 |
+
"grad_norm": 1.399740219116211,
|
| 19998 |
+
"learning_rate": 1.4033898305084749e-06,
|
| 19999 |
+
"loss": 1.2234,
|
| 20000 |
+
"step": 27930
|
| 20001 |
+
},
|
| 20002 |
+
{
|
| 20003 |
+
"epoch": 0.7,
|
| 20004 |
+
"grad_norm": 1.2826783657073975,
|
| 20005 |
+
"learning_rate": 1.3966101694915254e-06,
|
| 20006 |
+
"loss": 1.3775,
|
| 20007 |
+
"step": 27940
|
| 20008 |
+
},
|
| 20009 |
+
{
|
| 20010 |
+
"epoch": 0.7,
|
| 20011 |
+
"grad_norm": 9.264411926269531,
|
| 20012 |
+
"learning_rate": 1.3898305084745764e-06,
|
| 20013 |
+
"loss": 1.299,
|
| 20014 |
+
"step": 27950
|
| 20015 |
+
},
|
| 20016 |
+
{
|
| 20017 |
+
"epoch": 0.7,
|
| 20018 |
+
"grad_norm": 11.41451644897461,
|
| 20019 |
+
"learning_rate": 1.3830508474576274e-06,
|
| 20020 |
+
"loss": 1.2851,
|
| 20021 |
+
"step": 27960
|
| 20022 |
+
},
|
| 20023 |
+
{
|
| 20024 |
+
"epoch": 0.7,
|
| 20025 |
+
"grad_norm": 4.351644515991211,
|
| 20026 |
+
"learning_rate": 1.376271186440678e-06,
|
| 20027 |
+
"loss": 1.1979,
|
| 20028 |
+
"step": 27970
|
| 20029 |
+
},
|
| 20030 |
+
{
|
| 20031 |
+
"epoch": 0.7,
|
| 20032 |
+
"grad_norm": 9.026026725769043,
|
| 20033 |
+
"learning_rate": 1.369491525423729e-06,
|
| 20034 |
+
"loss": 1.3119,
|
| 20035 |
+
"step": 27980
|
| 20036 |
+
},
|
| 20037 |
+
{
|
| 20038 |
+
"epoch": 0.7,
|
| 20039 |
+
"grad_norm": 6.000504970550537,
|
| 20040 |
+
"learning_rate": 1.3627118644067797e-06,
|
| 20041 |
+
"loss": 1.2543,
|
| 20042 |
+
"step": 27990
|
| 20043 |
+
},
|
| 20044 |
+
{
|
| 20045 |
+
"epoch": 0.7,
|
| 20046 |
+
"grad_norm": 22.265581130981445,
|
| 20047 |
+
"learning_rate": 1.3559322033898307e-06,
|
| 20048 |
+
"loss": 1.34,
|
| 20049 |
+
"step": 28000
|
| 20050 |
+
},
|
| 20051 |
+
{
|
| 20052 |
+
"epoch": 0.7,
|
| 20053 |
+
"eval_loss": 1.3058114051818848,
|
| 20054 |
+
"eval_runtime": 66.2214,
|
| 20055 |
+
"eval_samples_per_second": 15.101,
|
| 20056 |
+
"eval_steps_per_second": 15.101,
|
| 20057 |
+
"step": 28000
|
| 20058 |
+
},
|
| 20059 |
+
{
|
| 20060 |
+
"epoch": 0.7,
|
| 20061 |
+
"grad_norm": 4.122670650482178,
|
| 20062 |
+
"learning_rate": 1.3491525423728816e-06,
|
| 20063 |
+
"loss": 1.2058,
|
| 20064 |
+
"step": 28010
|
| 20065 |
+
},
|
| 20066 |
+
{
|
| 20067 |
+
"epoch": 0.7,
|
| 20068 |
+
"grad_norm": 3.766960859298706,
|
| 20069 |
+
"learning_rate": 1.3423728813559322e-06,
|
| 20070 |
+
"loss": 1.3474,
|
| 20071 |
+
"step": 28020
|
| 20072 |
+
},
|
| 20073 |
+
{
|
| 20074 |
+
"epoch": 0.7,
|
| 20075 |
+
"grad_norm": 2.3636386394500732,
|
| 20076 |
+
"learning_rate": 1.3355932203389832e-06,
|
| 20077 |
+
"loss": 1.1637,
|
| 20078 |
+
"step": 28030
|
| 20079 |
+
},
|
| 20080 |
+
{
|
| 20081 |
+
"epoch": 0.7,
|
| 20082 |
+
"grad_norm": 9.471797943115234,
|
| 20083 |
+
"learning_rate": 1.328813559322034e-06,
|
| 20084 |
+
"loss": 1.3953,
|
| 20085 |
+
"step": 28040
|
| 20086 |
+
},
|
| 20087 |
+
{
|
| 20088 |
+
"epoch": 0.7,
|
| 20089 |
+
"grad_norm": 8.232218742370605,
|
| 20090 |
+
"learning_rate": 1.322033898305085e-06,
|
| 20091 |
+
"loss": 1.4269,
|
| 20092 |
+
"step": 28050
|
| 20093 |
+
},
|
| 20094 |
+
{
|
| 20095 |
+
"epoch": 0.7,
|
| 20096 |
+
"grad_norm": 1.1584389209747314,
|
| 20097 |
+
"learning_rate": 1.3152542372881359e-06,
|
| 20098 |
+
"loss": 1.3231,
|
| 20099 |
+
"step": 28060
|
| 20100 |
+
},
|
| 20101 |
+
{
|
| 20102 |
+
"epoch": 0.7,
|
| 20103 |
+
"grad_norm": 4.911566257476807,
|
| 20104 |
+
"learning_rate": 1.3084745762711864e-06,
|
| 20105 |
+
"loss": 1.241,
|
| 20106 |
+
"step": 28070
|
| 20107 |
+
},
|
| 20108 |
+
{
|
| 20109 |
+
"epoch": 0.7,
|
| 20110 |
+
"grad_norm": 18.741342544555664,
|
| 20111 |
+
"learning_rate": 1.3016949152542374e-06,
|
| 20112 |
+
"loss": 1.3838,
|
| 20113 |
+
"step": 28080
|
| 20114 |
+
},
|
| 20115 |
+
{
|
| 20116 |
+
"epoch": 0.7,
|
| 20117 |
+
"grad_norm": 8.622126579284668,
|
| 20118 |
+
"learning_rate": 1.2949152542372884e-06,
|
| 20119 |
+
"loss": 1.2792,
|
| 20120 |
+
"step": 28090
|
| 20121 |
+
},
|
| 20122 |
+
{
|
| 20123 |
+
"epoch": 0.7,
|
| 20124 |
+
"grad_norm": 4.402095794677734,
|
| 20125 |
+
"learning_rate": 1.288135593220339e-06,
|
| 20126 |
+
"loss": 1.2308,
|
| 20127 |
+
"step": 28100
|
| 20128 |
+
},
|
| 20129 |
+
{
|
| 20130 |
+
"epoch": 0.7,
|
| 20131 |
+
"grad_norm": 5.238518714904785,
|
| 20132 |
+
"learning_rate": 1.28135593220339e-06,
|
| 20133 |
+
"loss": 1.3537,
|
| 20134 |
+
"step": 28110
|
| 20135 |
+
},
|
| 20136 |
+
{
|
| 20137 |
+
"epoch": 0.7,
|
| 20138 |
+
"grad_norm": 5.757815837860107,
|
| 20139 |
+
"learning_rate": 1.2745762711864407e-06,
|
| 20140 |
+
"loss": 1.3987,
|
| 20141 |
+
"step": 28120
|
| 20142 |
+
},
|
| 20143 |
+
{
|
| 20144 |
+
"epoch": 0.7,
|
| 20145 |
+
"grad_norm": 9.999316215515137,
|
| 20146 |
+
"learning_rate": 1.2677966101694917e-06,
|
| 20147 |
+
"loss": 1.1397,
|
| 20148 |
+
"step": 28130
|
| 20149 |
+
},
|
| 20150 |
+
{
|
| 20151 |
+
"epoch": 0.7,
|
| 20152 |
+
"grad_norm": 6.276950359344482,
|
| 20153 |
+
"learning_rate": 1.2610169491525426e-06,
|
| 20154 |
+
"loss": 1.5024,
|
| 20155 |
+
"step": 28140
|
| 20156 |
+
},
|
| 20157 |
+
{
|
| 20158 |
+
"epoch": 0.7,
|
| 20159 |
+
"grad_norm": 5.9979634284973145,
|
| 20160 |
+
"learning_rate": 1.2542372881355932e-06,
|
| 20161 |
+
"loss": 1.4296,
|
| 20162 |
+
"step": 28150
|
| 20163 |
+
},
|
| 20164 |
+
{
|
| 20165 |
+
"epoch": 0.7,
|
| 20166 |
+
"grad_norm": 7.124503135681152,
|
| 20167 |
+
"learning_rate": 1.2474576271186442e-06,
|
| 20168 |
+
"loss": 1.279,
|
| 20169 |
+
"step": 28160
|
| 20170 |
+
},
|
| 20171 |
+
{
|
| 20172 |
+
"epoch": 0.7,
|
| 20173 |
+
"grad_norm": 10.153301239013672,
|
| 20174 |
+
"learning_rate": 1.240677966101695e-06,
|
| 20175 |
+
"loss": 1.3013,
|
| 20176 |
+
"step": 28170
|
| 20177 |
+
},
|
| 20178 |
+
{
|
| 20179 |
+
"epoch": 0.7,
|
| 20180 |
+
"grad_norm": 6.391571521759033,
|
| 20181 |
+
"learning_rate": 1.233898305084746e-06,
|
| 20182 |
+
"loss": 1.2782,
|
| 20183 |
+
"step": 28180
|
| 20184 |
+
},
|
| 20185 |
+
{
|
| 20186 |
+
"epoch": 0.7,
|
| 20187 |
+
"grad_norm": 10.13975715637207,
|
| 20188 |
+
"learning_rate": 1.2271186440677967e-06,
|
| 20189 |
+
"loss": 1.4136,
|
| 20190 |
+
"step": 28190
|
| 20191 |
+
},
|
| 20192 |
+
{
|
| 20193 |
+
"epoch": 0.7,
|
| 20194 |
+
"grad_norm": 5.052265167236328,
|
| 20195 |
+
"learning_rate": 1.2203389830508477e-06,
|
| 20196 |
+
"loss": 1.3246,
|
| 20197 |
+
"step": 28200
|
| 20198 |
+
},
|
| 20199 |
+
{
|
| 20200 |
+
"epoch": 0.71,
|
| 20201 |
+
"grad_norm": 6.8834638595581055,
|
| 20202 |
+
"learning_rate": 1.2135593220338984e-06,
|
| 20203 |
+
"loss": 1.3436,
|
| 20204 |
+
"step": 28210
|
| 20205 |
+
},
|
| 20206 |
+
{
|
| 20207 |
+
"epoch": 0.71,
|
| 20208 |
+
"grad_norm": 10.564448356628418,
|
| 20209 |
+
"learning_rate": 1.2067796610169492e-06,
|
| 20210 |
+
"loss": 1.1619,
|
| 20211 |
+
"step": 28220
|
| 20212 |
+
},
|
| 20213 |
+
{
|
| 20214 |
+
"epoch": 0.71,
|
| 20215 |
+
"grad_norm": 3.8192801475524902,
|
| 20216 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 20217 |
+
"loss": 1.2328,
|
| 20218 |
+
"step": 28230
|
| 20219 |
+
},
|
| 20220 |
+
{
|
| 20221 |
+
"epoch": 0.71,
|
| 20222 |
+
"grad_norm": 5.635645866394043,
|
| 20223 |
+
"learning_rate": 1.193220338983051e-06,
|
| 20224 |
+
"loss": 1.3258,
|
| 20225 |
+
"step": 28240
|
| 20226 |
+
},
|
| 20227 |
+
{
|
| 20228 |
+
"epoch": 0.71,
|
| 20229 |
+
"grad_norm": 6.429792881011963,
|
| 20230 |
+
"learning_rate": 1.186440677966102e-06,
|
| 20231 |
+
"loss": 1.3368,
|
| 20232 |
+
"step": 28250
|
| 20233 |
+
},
|
| 20234 |
+
{
|
| 20235 |
+
"epoch": 0.71,
|
| 20236 |
+
"grad_norm": 11.116402626037598,
|
| 20237 |
+
"learning_rate": 1.1796610169491527e-06,
|
| 20238 |
+
"loss": 1.2527,
|
| 20239 |
+
"step": 28260
|
| 20240 |
+
},
|
| 20241 |
+
{
|
| 20242 |
+
"epoch": 0.71,
|
| 20243 |
+
"grad_norm": 2.7472074031829834,
|
| 20244 |
+
"learning_rate": 1.1728813559322034e-06,
|
| 20245 |
+
"loss": 1.4633,
|
| 20246 |
+
"step": 28270
|
| 20247 |
+
},
|
| 20248 |
+
{
|
| 20249 |
+
"epoch": 0.71,
|
| 20250 |
+
"grad_norm": 4.118687629699707,
|
| 20251 |
+
"learning_rate": 1.1661016949152542e-06,
|
| 20252 |
+
"loss": 1.4852,
|
| 20253 |
+
"step": 28280
|
| 20254 |
+
},
|
| 20255 |
+
{
|
| 20256 |
+
"epoch": 0.71,
|
| 20257 |
+
"grad_norm": 14.502837181091309,
|
| 20258 |
+
"learning_rate": 1.1593220338983052e-06,
|
| 20259 |
+
"loss": 1.1964,
|
| 20260 |
+
"step": 28290
|
| 20261 |
+
},
|
| 20262 |
+
{
|
| 20263 |
+
"epoch": 0.71,
|
| 20264 |
+
"grad_norm": 3.8546032905578613,
|
| 20265 |
+
"learning_rate": 1.152542372881356e-06,
|
| 20266 |
+
"loss": 1.1832,
|
| 20267 |
+
"step": 28300
|
| 20268 |
+
},
|
| 20269 |
+
{
|
| 20270 |
+
"epoch": 0.71,
|
| 20271 |
+
"grad_norm": 2.377305507659912,
|
| 20272 |
+
"learning_rate": 1.145762711864407e-06,
|
| 20273 |
+
"loss": 1.3164,
|
| 20274 |
+
"step": 28310
|
| 20275 |
+
},
|
| 20276 |
+
{
|
| 20277 |
+
"epoch": 0.71,
|
| 20278 |
+
"grad_norm": 2.172879219055176,
|
| 20279 |
+
"learning_rate": 1.1389830508474577e-06,
|
| 20280 |
+
"loss": 1.3163,
|
| 20281 |
+
"step": 28320
|
| 20282 |
+
},
|
| 20283 |
+
{
|
| 20284 |
+
"epoch": 0.71,
|
| 20285 |
+
"grad_norm": 4.290719032287598,
|
| 20286 |
+
"learning_rate": 1.1322033898305087e-06,
|
| 20287 |
+
"loss": 1.2505,
|
| 20288 |
+
"step": 28330
|
| 20289 |
+
},
|
| 20290 |
+
{
|
| 20291 |
+
"epoch": 0.71,
|
| 20292 |
+
"grad_norm": 15.109819412231445,
|
| 20293 |
+
"learning_rate": 1.1254237288135594e-06,
|
| 20294 |
+
"loss": 1.2868,
|
| 20295 |
+
"step": 28340
|
| 20296 |
+
},
|
| 20297 |
+
{
|
| 20298 |
+
"epoch": 0.71,
|
| 20299 |
+
"grad_norm": 11.07304859161377,
|
| 20300 |
+
"learning_rate": 1.1186440677966102e-06,
|
| 20301 |
+
"loss": 1.2118,
|
| 20302 |
+
"step": 28350
|
| 20303 |
+
},
|
| 20304 |
+
{
|
| 20305 |
+
"epoch": 0.71,
|
| 20306 |
+
"grad_norm": 14.64116096496582,
|
| 20307 |
+
"learning_rate": 1.1118644067796612e-06,
|
| 20308 |
+
"loss": 1.2904,
|
| 20309 |
+
"step": 28360
|
| 20310 |
+
},
|
| 20311 |
+
{
|
| 20312 |
+
"epoch": 0.71,
|
| 20313 |
+
"grad_norm": 12.49071216583252,
|
| 20314 |
+
"learning_rate": 1.105084745762712e-06,
|
| 20315 |
+
"loss": 1.286,
|
| 20316 |
+
"step": 28370
|
| 20317 |
+
},
|
| 20318 |
+
{
|
| 20319 |
+
"epoch": 0.71,
|
| 20320 |
+
"grad_norm": 3.750757932662964,
|
| 20321 |
+
"learning_rate": 1.098305084745763e-06,
|
| 20322 |
+
"loss": 1.2723,
|
| 20323 |
+
"step": 28380
|
| 20324 |
+
},
|
| 20325 |
+
{
|
| 20326 |
+
"epoch": 0.71,
|
| 20327 |
+
"grad_norm": 11.984790802001953,
|
| 20328 |
+
"learning_rate": 1.0915254237288137e-06,
|
| 20329 |
+
"loss": 1.1732,
|
| 20330 |
+
"step": 28390
|
| 20331 |
+
},
|
| 20332 |
+
{
|
| 20333 |
+
"epoch": 0.71,
|
| 20334 |
+
"grad_norm": 13.987404823303223,
|
| 20335 |
+
"learning_rate": 1.0847457627118644e-06,
|
| 20336 |
+
"loss": 1.2168,
|
| 20337 |
+
"step": 28400
|
| 20338 |
+
},
|
| 20339 |
+
{
|
| 20340 |
+
"epoch": 0.71,
|
| 20341 |
+
"grad_norm": 10.837672233581543,
|
| 20342 |
+
"learning_rate": 1.0779661016949152e-06,
|
| 20343 |
+
"loss": 1.4036,
|
| 20344 |
+
"step": 28410
|
| 20345 |
+
},
|
| 20346 |
+
{
|
| 20347 |
+
"epoch": 0.71,
|
| 20348 |
+
"grad_norm": 12.291699409484863,
|
| 20349 |
+
"learning_rate": 1.0711864406779662e-06,
|
| 20350 |
+
"loss": 1.1172,
|
| 20351 |
+
"step": 28420
|
| 20352 |
+
},
|
| 20353 |
+
{
|
| 20354 |
+
"epoch": 0.71,
|
| 20355 |
+
"grad_norm": 2.2296297550201416,
|
| 20356 |
+
"learning_rate": 1.064406779661017e-06,
|
| 20357 |
+
"loss": 1.3241,
|
| 20358 |
+
"step": 28430
|
| 20359 |
+
},
|
| 20360 |
+
{
|
| 20361 |
+
"epoch": 0.71,
|
| 20362 |
+
"grad_norm": 6.434119701385498,
|
| 20363 |
+
"learning_rate": 1.057627118644068e-06,
|
| 20364 |
+
"loss": 1.3096,
|
| 20365 |
+
"step": 28440
|
| 20366 |
+
},
|
| 20367 |
+
{
|
| 20368 |
+
"epoch": 0.71,
|
| 20369 |
+
"grad_norm": 3.2857985496520996,
|
| 20370 |
+
"learning_rate": 1.0508474576271187e-06,
|
| 20371 |
+
"loss": 1.3744,
|
| 20372 |
+
"step": 28450
|
| 20373 |
+
},
|
| 20374 |
+
{
|
| 20375 |
+
"epoch": 0.71,
|
| 20376 |
+
"grad_norm": 6.822338104248047,
|
| 20377 |
+
"learning_rate": 1.0440677966101697e-06,
|
| 20378 |
+
"loss": 1.3724,
|
| 20379 |
+
"step": 28460
|
| 20380 |
+
},
|
| 20381 |
+
{
|
| 20382 |
+
"epoch": 0.71,
|
| 20383 |
+
"grad_norm": 6.319363117218018,
|
| 20384 |
+
"learning_rate": 1.0372881355932204e-06,
|
| 20385 |
+
"loss": 1.2181,
|
| 20386 |
+
"step": 28470
|
| 20387 |
+
},
|
| 20388 |
+
{
|
| 20389 |
+
"epoch": 0.71,
|
| 20390 |
+
"grad_norm": 3.777921676635742,
|
| 20391 |
+
"learning_rate": 1.0305084745762712e-06,
|
| 20392 |
+
"loss": 1.3495,
|
| 20393 |
+
"step": 28480
|
| 20394 |
+
},
|
| 20395 |
+
{
|
| 20396 |
+
"epoch": 0.71,
|
| 20397 |
+
"grad_norm": 6.540971755981445,
|
| 20398 |
+
"learning_rate": 1.0237288135593222e-06,
|
| 20399 |
+
"loss": 1.4589,
|
| 20400 |
+
"step": 28490
|
| 20401 |
+
},
|
| 20402 |
+
{
|
| 20403 |
+
"epoch": 0.71,
|
| 20404 |
+
"grad_norm": 7.9154744148254395,
|
| 20405 |
+
"learning_rate": 1.016949152542373e-06,
|
| 20406 |
+
"loss": 1.3015,
|
| 20407 |
+
"step": 28500
|
| 20408 |
+
},
|
| 20409 |
+
{
|
| 20410 |
+
"epoch": 0.71,
|
| 20411 |
+
"eval_loss": 1.3162422180175781,
|
| 20412 |
+
"eval_runtime": 66.2512,
|
| 20413 |
+
"eval_samples_per_second": 15.094,
|
| 20414 |
+
"eval_steps_per_second": 15.094,
|
| 20415 |
+
"step": 28500
|
| 20416 |
+
},
|
| 20417 |
+
{
|
| 20418 |
+
"epoch": 0.71,
|
| 20419 |
+
"grad_norm": 15.148056983947754,
|
| 20420 |
+
"learning_rate": 1.010169491525424e-06,
|
| 20421 |
+
"loss": 1.2363,
|
| 20422 |
+
"step": 28510
|
| 20423 |
+
},
|
| 20424 |
+
{
|
| 20425 |
+
"epoch": 0.71,
|
| 20426 |
+
"grad_norm": 10.743448257446289,
|
| 20427 |
+
"learning_rate": 1.0033898305084747e-06,
|
| 20428 |
+
"loss": 1.1965,
|
| 20429 |
+
"step": 28520
|
| 20430 |
+
},
|
| 20431 |
+
{
|
| 20432 |
+
"epoch": 0.71,
|
| 20433 |
+
"grad_norm": 12.191396713256836,
|
| 20434 |
+
"learning_rate": 9.966101694915254e-07,
|
| 20435 |
+
"loss": 1.4387,
|
| 20436 |
+
"step": 28530
|
| 20437 |
+
},
|
| 20438 |
+
{
|
| 20439 |
+
"epoch": 0.71,
|
| 20440 |
+
"grad_norm": 4.78171968460083,
|
| 20441 |
+
"learning_rate": 9.898305084745762e-07,
|
| 20442 |
+
"loss": 1.4977,
|
| 20443 |
+
"step": 28540
|
| 20444 |
+
},
|
| 20445 |
+
{
|
| 20446 |
+
"epoch": 0.71,
|
| 20447 |
+
"grad_norm": 5.24019718170166,
|
| 20448 |
+
"learning_rate": 9.830508474576272e-07,
|
| 20449 |
+
"loss": 1.5177,
|
| 20450 |
+
"step": 28550
|
| 20451 |
+
},
|
| 20452 |
+
{
|
| 20453 |
+
"epoch": 0.71,
|
| 20454 |
+
"grad_norm": 12.000500679016113,
|
| 20455 |
+
"learning_rate": 9.762711864406782e-07,
|
| 20456 |
+
"loss": 1.1637,
|
| 20457 |
+
"step": 28560
|
| 20458 |
+
},
|
| 20459 |
+
{
|
| 20460 |
+
"epoch": 0.71,
|
| 20461 |
+
"grad_norm": 10.612434387207031,
|
| 20462 |
+
"learning_rate": 9.69491525423729e-07,
|
| 20463 |
+
"loss": 1.334,
|
| 20464 |
+
"step": 28570
|
| 20465 |
+
},
|
| 20466 |
+
{
|
| 20467 |
+
"epoch": 0.71,
|
| 20468 |
+
"grad_norm": 6.343203544616699,
|
| 20469 |
+
"learning_rate": 9.627118644067797e-07,
|
| 20470 |
+
"loss": 1.3152,
|
| 20471 |
+
"step": 28580
|
| 20472 |
+
},
|
| 20473 |
+
{
|
| 20474 |
+
"epoch": 0.71,
|
| 20475 |
+
"grad_norm": 5.986273288726807,
|
| 20476 |
+
"learning_rate": 9.559322033898307e-07,
|
| 20477 |
+
"loss": 1.3724,
|
| 20478 |
+
"step": 28590
|
| 20479 |
+
},
|
| 20480 |
+
{
|
| 20481 |
+
"epoch": 0.71,
|
| 20482 |
+
"grad_norm": 12.447896003723145,
|
| 20483 |
+
"learning_rate": 9.491525423728814e-07,
|
| 20484 |
+
"loss": 1.2404,
|
| 20485 |
+
"step": 28600
|
| 20486 |
+
},
|
| 20487 |
+
{
|
| 20488 |
+
"epoch": 0.72,
|
| 20489 |
+
"grad_norm": 18.211698532104492,
|
| 20490 |
+
"learning_rate": 9.423728813559323e-07,
|
| 20491 |
+
"loss": 1.3091,
|
| 20492 |
+
"step": 28610
|
| 20493 |
+
},
|
| 20494 |
+
{
|
| 20495 |
+
"epoch": 0.72,
|
| 20496 |
+
"grad_norm": 5.227106094360352,
|
| 20497 |
+
"learning_rate": 9.355932203389831e-07,
|
| 20498 |
+
"loss": 1.1275,
|
| 20499 |
+
"step": 28620
|
| 20500 |
+
},
|
| 20501 |
+
{
|
| 20502 |
+
"epoch": 0.72,
|
| 20503 |
+
"grad_norm": 4.874502182006836,
|
| 20504 |
+
"learning_rate": 9.28813559322034e-07,
|
| 20505 |
+
"loss": 1.2959,
|
| 20506 |
+
"step": 28630
|
| 20507 |
+
},
|
| 20508 |
+
{
|
| 20509 |
+
"epoch": 0.72,
|
| 20510 |
+
"grad_norm": 7.699239253997803,
|
| 20511 |
+
"learning_rate": 9.220338983050848e-07,
|
| 20512 |
+
"loss": 1.3687,
|
| 20513 |
+
"step": 28640
|
| 20514 |
+
},
|
| 20515 |
+
{
|
| 20516 |
+
"epoch": 0.72,
|
| 20517 |
+
"grad_norm": 7.712405681610107,
|
| 20518 |
+
"learning_rate": 9.152542372881357e-07,
|
| 20519 |
+
"loss": 1.3746,
|
| 20520 |
+
"step": 28650
|
| 20521 |
+
},
|
| 20522 |
+
{
|
| 20523 |
+
"epoch": 0.72,
|
| 20524 |
+
"grad_norm": 8.854902267456055,
|
| 20525 |
+
"learning_rate": 9.084745762711864e-07,
|
| 20526 |
+
"loss": 1.3399,
|
| 20527 |
+
"step": 28660
|
| 20528 |
+
},
|
| 20529 |
+
{
|
| 20530 |
+
"epoch": 0.72,
|
| 20531 |
+
"grad_norm": 5.26395320892334,
|
| 20532 |
+
"learning_rate": 9.016949152542373e-07,
|
| 20533 |
+
"loss": 1.2296,
|
| 20534 |
+
"step": 28670
|
| 20535 |
+
},
|
| 20536 |
+
{
|
| 20537 |
+
"epoch": 0.72,
|
| 20538 |
+
"grad_norm": 5.966766357421875,
|
| 20539 |
+
"learning_rate": 8.949152542372883e-07,
|
| 20540 |
+
"loss": 1.2806,
|
| 20541 |
+
"step": 28680
|
| 20542 |
+
},
|
| 20543 |
+
{
|
| 20544 |
+
"epoch": 0.72,
|
| 20545 |
+
"grad_norm": 5.988807201385498,
|
| 20546 |
+
"learning_rate": 8.881355932203391e-07,
|
| 20547 |
+
"loss": 1.4652,
|
| 20548 |
+
"step": 28690
|
| 20549 |
+
},
|
| 20550 |
+
{
|
| 20551 |
+
"epoch": 0.72,
|
| 20552 |
+
"grad_norm": 1.523200273513794,
|
| 20553 |
+
"learning_rate": 8.813559322033899e-07,
|
| 20554 |
+
"loss": 1.4848,
|
| 20555 |
+
"step": 28700
|
| 20556 |
+
},
|
| 20557 |
+
{
|
| 20558 |
+
"epoch": 0.72,
|
| 20559 |
+
"grad_norm": 3.4972944259643555,
|
| 20560 |
+
"learning_rate": 8.745762711864407e-07,
|
| 20561 |
+
"loss": 1.296,
|
| 20562 |
+
"step": 28710
|
| 20563 |
+
},
|
| 20564 |
+
{
|
| 20565 |
+
"epoch": 0.72,
|
| 20566 |
+
"grad_norm": 4.062095642089844,
|
| 20567 |
+
"learning_rate": 8.677966101694917e-07,
|
| 20568 |
+
"loss": 1.3716,
|
| 20569 |
+
"step": 28720
|
| 20570 |
+
},
|
| 20571 |
+
{
|
| 20572 |
+
"epoch": 0.72,
|
| 20573 |
+
"grad_norm": 1.065338134765625,
|
| 20574 |
+
"learning_rate": 8.610169491525424e-07,
|
| 20575 |
+
"loss": 1.2375,
|
| 20576 |
+
"step": 28730
|
| 20577 |
+
},
|
| 20578 |
+
{
|
| 20579 |
+
"epoch": 0.72,
|
| 20580 |
+
"grad_norm": 6.1202006340026855,
|
| 20581 |
+
"learning_rate": 8.542372881355933e-07,
|
| 20582 |
+
"loss": 1.3413,
|
| 20583 |
+
"step": 28740
|
| 20584 |
+
},
|
| 20585 |
+
{
|
| 20586 |
+
"epoch": 0.72,
|
| 20587 |
+
"grad_norm": 17.116790771484375,
|
| 20588 |
+
"learning_rate": 8.474576271186441e-07,
|
| 20589 |
+
"loss": 1.2014,
|
| 20590 |
+
"step": 28750
|
| 20591 |
+
},
|
| 20592 |
+
{
|
| 20593 |
+
"epoch": 0.72,
|
| 20594 |
+
"grad_norm": 6.310455799102783,
|
| 20595 |
+
"learning_rate": 8.406779661016951e-07,
|
| 20596 |
+
"loss": 1.2509,
|
| 20597 |
+
"step": 28760
|
| 20598 |
+
},
|
| 20599 |
+
{
|
| 20600 |
+
"epoch": 0.72,
|
| 20601 |
+
"grad_norm": 19.508827209472656,
|
| 20602 |
+
"learning_rate": 8.338983050847458e-07,
|
| 20603 |
+
"loss": 1.4035,
|
| 20604 |
+
"step": 28770
|
| 20605 |
+
},
|
| 20606 |
+
{
|
| 20607 |
+
"epoch": 0.72,
|
| 20608 |
+
"grad_norm": 10.587761878967285,
|
| 20609 |
+
"learning_rate": 8.271186440677967e-07,
|
| 20610 |
+
"loss": 1.3374,
|
| 20611 |
+
"step": 28780
|
| 20612 |
+
},
|
| 20613 |
+
{
|
| 20614 |
+
"epoch": 0.72,
|
| 20615 |
+
"grad_norm": 15.715071678161621,
|
| 20616 |
+
"learning_rate": 8.203389830508475e-07,
|
| 20617 |
+
"loss": 1.3229,
|
| 20618 |
+
"step": 28790
|
| 20619 |
+
},
|
| 20620 |
+
{
|
| 20621 |
+
"epoch": 0.72,
|
| 20622 |
+
"grad_norm": 3.0983753204345703,
|
| 20623 |
+
"learning_rate": 8.135593220338983e-07,
|
| 20624 |
+
"loss": 1.4297,
|
| 20625 |
+
"step": 28800
|
| 20626 |
+
},
|
| 20627 |
+
{
|
| 20628 |
+
"epoch": 0.72,
|
| 20629 |
+
"grad_norm": 4.541349411010742,
|
| 20630 |
+
"learning_rate": 8.067796610169493e-07,
|
| 20631 |
+
"loss": 1.3327,
|
| 20632 |
+
"step": 28810
|
| 20633 |
+
},
|
| 20634 |
+
{
|
| 20635 |
+
"epoch": 0.72,
|
| 20636 |
+
"grad_norm": 4.102433204650879,
|
| 20637 |
+
"learning_rate": 8.000000000000001e-07,
|
| 20638 |
+
"loss": 1.3362,
|
| 20639 |
+
"step": 28820
|
| 20640 |
+
},
|
| 20641 |
+
{
|
| 20642 |
+
"epoch": 0.72,
|
| 20643 |
+
"grad_norm": 6.362680435180664,
|
| 20644 |
+
"learning_rate": 7.932203389830509e-07,
|
| 20645 |
+
"loss": 1.2275,
|
| 20646 |
+
"step": 28830
|
| 20647 |
+
},
|
| 20648 |
+
{
|
| 20649 |
+
"epoch": 0.72,
|
| 20650 |
+
"grad_norm": 9.411408424377441,
|
| 20651 |
+
"learning_rate": 7.864406779661017e-07,
|
| 20652 |
+
"loss": 1.4023,
|
| 20653 |
+
"step": 28840
|
| 20654 |
+
},
|
| 20655 |
+
{
|
| 20656 |
+
"epoch": 0.72,
|
| 20657 |
+
"grad_norm": 6.113147258758545,
|
| 20658 |
+
"learning_rate": 7.796610169491527e-07,
|
| 20659 |
+
"loss": 1.2319,
|
| 20660 |
+
"step": 28850
|
| 20661 |
+
},
|
| 20662 |
+
{
|
| 20663 |
+
"epoch": 0.72,
|
| 20664 |
+
"grad_norm": 6.580341815948486,
|
| 20665 |
+
"learning_rate": 7.728813559322034e-07,
|
| 20666 |
+
"loss": 1.3559,
|
| 20667 |
+
"step": 28860
|
| 20668 |
+
},
|
| 20669 |
+
{
|
| 20670 |
+
"epoch": 0.72,
|
| 20671 |
+
"grad_norm": 10.778473854064941,
|
| 20672 |
+
"learning_rate": 7.661016949152543e-07,
|
| 20673 |
+
"loss": 1.2247,
|
| 20674 |
+
"step": 28870
|
| 20675 |
+
},
|
| 20676 |
+
{
|
| 20677 |
+
"epoch": 0.72,
|
| 20678 |
+
"grad_norm": 4.692105770111084,
|
| 20679 |
+
"learning_rate": 7.593220338983051e-07,
|
| 20680 |
+
"loss": 1.2254,
|
| 20681 |
+
"step": 28880
|
| 20682 |
+
},
|
| 20683 |
+
{
|
| 20684 |
+
"epoch": 0.72,
|
| 20685 |
+
"grad_norm": 8.076458930969238,
|
| 20686 |
+
"learning_rate": 7.52542372881356e-07,
|
| 20687 |
+
"loss": 1.3264,
|
| 20688 |
+
"step": 28890
|
| 20689 |
+
},
|
| 20690 |
+
{
|
| 20691 |
+
"epoch": 0.72,
|
| 20692 |
+
"grad_norm": 11.618599891662598,
|
| 20693 |
+
"learning_rate": 7.457627118644069e-07,
|
| 20694 |
+
"loss": 1.174,
|
| 20695 |
+
"step": 28900
|
| 20696 |
+
},
|
| 20697 |
+
{
|
| 20698 |
+
"epoch": 0.72,
|
| 20699 |
+
"grad_norm": 10.995185852050781,
|
| 20700 |
+
"learning_rate": 7.389830508474577e-07,
|
| 20701 |
+
"loss": 1.21,
|
| 20702 |
+
"step": 28910
|
| 20703 |
+
},
|
| 20704 |
+
{
|
| 20705 |
+
"epoch": 0.72,
|
| 20706 |
+
"grad_norm": 12.011213302612305,
|
| 20707 |
+
"learning_rate": 7.322033898305085e-07,
|
| 20708 |
+
"loss": 1.3272,
|
| 20709 |
+
"step": 28920
|
| 20710 |
+
},
|
| 20711 |
+
{
|
| 20712 |
+
"epoch": 0.72,
|
| 20713 |
+
"grad_norm": 4.367415428161621,
|
| 20714 |
+
"learning_rate": 7.254237288135593e-07,
|
| 20715 |
+
"loss": 1.3997,
|
| 20716 |
+
"step": 28930
|
| 20717 |
+
},
|
| 20718 |
+
{
|
| 20719 |
+
"epoch": 0.72,
|
| 20720 |
+
"grad_norm": 11.401775360107422,
|
| 20721 |
+
"learning_rate": 7.186440677966103e-07,
|
| 20722 |
+
"loss": 1.3507,
|
| 20723 |
+
"step": 28940
|
| 20724 |
+
},
|
| 20725 |
+
{
|
| 20726 |
+
"epoch": 0.72,
|
| 20727 |
+
"grad_norm": 8.856404304504395,
|
| 20728 |
+
"learning_rate": 7.118644067796611e-07,
|
| 20729 |
+
"loss": 1.3274,
|
| 20730 |
+
"step": 28950
|
| 20731 |
+
},
|
| 20732 |
+
{
|
| 20733 |
+
"epoch": 0.72,
|
| 20734 |
+
"grad_norm": 5.873079776763916,
|
| 20735 |
+
"learning_rate": 7.05084745762712e-07,
|
| 20736 |
+
"loss": 1.2987,
|
| 20737 |
+
"step": 28960
|
| 20738 |
+
},
|
| 20739 |
+
{
|
| 20740 |
+
"epoch": 0.72,
|
| 20741 |
+
"grad_norm": 7.142796993255615,
|
| 20742 |
+
"learning_rate": 6.983050847457627e-07,
|
| 20743 |
+
"loss": 1.2049,
|
| 20744 |
+
"step": 28970
|
| 20745 |
+
},
|
| 20746 |
+
{
|
| 20747 |
+
"epoch": 0.72,
|
| 20748 |
+
"grad_norm": 5.9704179763793945,
|
| 20749 |
+
"learning_rate": 6.915254237288137e-07,
|
| 20750 |
+
"loss": 1.3481,
|
| 20751 |
+
"step": 28980
|
| 20752 |
+
},
|
| 20753 |
+
{
|
| 20754 |
+
"epoch": 0.72,
|
| 20755 |
+
"grad_norm": 10.48417854309082,
|
| 20756 |
+
"learning_rate": 6.847457627118645e-07,
|
| 20757 |
+
"loss": 1.3498,
|
| 20758 |
+
"step": 28990
|
| 20759 |
+
},
|
| 20760 |
+
{
|
| 20761 |
+
"epoch": 0.72,
|
| 20762 |
+
"grad_norm": 12.131574630737305,
|
| 20763 |
+
"learning_rate": 6.779661016949153e-07,
|
| 20764 |
+
"loss": 1.3529,
|
| 20765 |
+
"step": 29000
|
| 20766 |
+
},
|
| 20767 |
+
{
|
| 20768 |
+
"epoch": 0.72,
|
| 20769 |
+
"eval_loss": 1.3368879556655884,
|
| 20770 |
+
"eval_runtime": 66.2441,
|
| 20771 |
+
"eval_samples_per_second": 15.096,
|
| 20772 |
+
"eval_steps_per_second": 15.096,
|
| 20773 |
+
"step": 29000
|
| 20774 |
+
},
|
| 20775 |
+
{
|
| 20776 |
+
"epoch": 0.73,
|
| 20777 |
+
"grad_norm": 7.354034423828125,
|
| 20778 |
+
"learning_rate": 6.711864406779661e-07,
|
| 20779 |
+
"loss": 1.2698,
|
| 20780 |
+
"step": 29010
|
| 20781 |
+
},
|
| 20782 |
+
{
|
| 20783 |
+
"epoch": 0.73,
|
| 20784 |
+
"grad_norm": 3.9224977493286133,
|
| 20785 |
+
"learning_rate": 6.64406779661017e-07,
|
| 20786 |
+
"loss": 1.3377,
|
| 20787 |
+
"step": 29020
|
| 20788 |
+
},
|
| 20789 |
+
{
|
| 20790 |
+
"epoch": 0.73,
|
| 20791 |
+
"grad_norm": 8.08880615234375,
|
| 20792 |
+
"learning_rate": 6.576271186440679e-07,
|
| 20793 |
+
"loss": 1.4448,
|
| 20794 |
+
"step": 29030
|
| 20795 |
+
},
|
| 20796 |
+
{
|
| 20797 |
+
"epoch": 0.73,
|
| 20798 |
+
"grad_norm": 3.625216245651245,
|
| 20799 |
+
"learning_rate": 6.508474576271187e-07,
|
| 20800 |
+
"loss": 1.3561,
|
| 20801 |
+
"step": 29040
|
| 20802 |
+
},
|
| 20803 |
+
{
|
| 20804 |
+
"epoch": 0.73,
|
| 20805 |
+
"grad_norm": 13.905680656433105,
|
| 20806 |
+
"learning_rate": 6.440677966101695e-07,
|
| 20807 |
+
"loss": 1.2956,
|
| 20808 |
+
"step": 29050
|
| 20809 |
+
},
|
| 20810 |
+
{
|
| 20811 |
+
"epoch": 0.73,
|
| 20812 |
+
"grad_norm": 3.773040294647217,
|
| 20813 |
+
"learning_rate": 6.372881355932203e-07,
|
| 20814 |
+
"loss": 1.2934,
|
| 20815 |
+
"step": 29060
|
| 20816 |
+
},
|
| 20817 |
+
{
|
| 20818 |
+
"epoch": 0.73,
|
| 20819 |
+
"grad_norm": 5.38674259185791,
|
| 20820 |
+
"learning_rate": 6.305084745762713e-07,
|
| 20821 |
+
"loss": 1.3364,
|
| 20822 |
+
"step": 29070
|
| 20823 |
+
},
|
| 20824 |
+
{
|
| 20825 |
+
"epoch": 0.73,
|
| 20826 |
+
"grad_norm": 4.701632022857666,
|
| 20827 |
+
"learning_rate": 6.237288135593221e-07,
|
| 20828 |
+
"loss": 1.1754,
|
| 20829 |
+
"step": 29080
|
| 20830 |
+
},
|
| 20831 |
+
{
|
| 20832 |
+
"epoch": 0.73,
|
| 20833 |
+
"grad_norm": 8.243915557861328,
|
| 20834 |
+
"learning_rate": 6.16949152542373e-07,
|
| 20835 |
+
"loss": 1.4113,
|
| 20836 |
+
"step": 29090
|
| 20837 |
+
},
|
| 20838 |
+
{
|
| 20839 |
+
"epoch": 0.73,
|
| 20840 |
+
"grad_norm": 4.2352213859558105,
|
| 20841 |
+
"learning_rate": 6.101694915254238e-07,
|
| 20842 |
+
"loss": 1.4673,
|
| 20843 |
+
"step": 29100
|
| 20844 |
+
},
|
| 20845 |
+
{
|
| 20846 |
+
"epoch": 0.73,
|
| 20847 |
+
"grad_norm": 19.758262634277344,
|
| 20848 |
+
"learning_rate": 6.033898305084746e-07,
|
| 20849 |
+
"loss": 1.0756,
|
| 20850 |
+
"step": 29110
|
| 20851 |
+
},
|
| 20852 |
+
{
|
| 20853 |
+
"epoch": 0.73,
|
| 20854 |
+
"grad_norm": 9.024798393249512,
|
| 20855 |
+
"learning_rate": 5.966101694915255e-07,
|
| 20856 |
+
"loss": 1.1835,
|
| 20857 |
+
"step": 29120
|
| 20858 |
+
},
|
| 20859 |
+
{
|
| 20860 |
+
"epoch": 0.73,
|
| 20861 |
+
"grad_norm": 2.7573533058166504,
|
| 20862 |
+
"learning_rate": 5.898305084745763e-07,
|
| 20863 |
+
"loss": 1.3342,
|
| 20864 |
+
"step": 29130
|
| 20865 |
+
},
|
| 20866 |
+
{
|
| 20867 |
+
"epoch": 0.73,
|
| 20868 |
+
"grad_norm": 2.6472461223602295,
|
| 20869 |
+
"learning_rate": 5.830508474576271e-07,
|
| 20870 |
+
"loss": 1.478,
|
| 20871 |
+
"step": 29140
|
| 20872 |
+
},
|
| 20873 |
+
{
|
| 20874 |
+
"epoch": 0.73,
|
| 20875 |
+
"grad_norm": 8.194090843200684,
|
| 20876 |
+
"learning_rate": 5.76271186440678e-07,
|
| 20877 |
+
"loss": 1.2781,
|
| 20878 |
+
"step": 29150
|
| 20879 |
+
},
|
| 20880 |
+
{
|
| 20881 |
+
"epoch": 0.73,
|
| 20882 |
+
"grad_norm": 7.336125373840332,
|
| 20883 |
+
"learning_rate": 5.694915254237288e-07,
|
| 20884 |
+
"loss": 1.3871,
|
| 20885 |
+
"step": 29160
|
| 20886 |
+
},
|
| 20887 |
+
{
|
| 20888 |
+
"epoch": 0.73,
|
| 20889 |
+
"grad_norm": 9.800869941711426,
|
| 20890 |
+
"learning_rate": 5.627118644067797e-07,
|
| 20891 |
+
"loss": 1.237,
|
| 20892 |
+
"step": 29170
|
| 20893 |
+
},
|
| 20894 |
+
{
|
| 20895 |
+
"epoch": 0.73,
|
| 20896 |
+
"grad_norm": 3.7578229904174805,
|
| 20897 |
+
"learning_rate": 5.559322033898306e-07,
|
| 20898 |
+
"loss": 1.3408,
|
| 20899 |
+
"step": 29180
|
| 20900 |
+
},
|
| 20901 |
+
{
|
| 20902 |
+
"epoch": 0.73,
|
| 20903 |
+
"grad_norm": 11.04594612121582,
|
| 20904 |
+
"learning_rate": 5.491525423728815e-07,
|
| 20905 |
+
"loss": 1.4338,
|
| 20906 |
+
"step": 29190
|
| 20907 |
+
},
|
| 20908 |
+
{
|
| 20909 |
+
"epoch": 0.73,
|
| 20910 |
+
"grad_norm": 9.137775421142578,
|
| 20911 |
+
"learning_rate": 5.423728813559322e-07,
|
| 20912 |
+
"loss": 1.2292,
|
| 20913 |
+
"step": 29200
|
| 20914 |
+
},
|
| 20915 |
+
{
|
| 20916 |
+
"epoch": 0.73,
|
| 20917 |
+
"grad_norm": 6.266888618469238,
|
| 20918 |
+
"learning_rate": 5.355932203389831e-07,
|
| 20919 |
+
"loss": 1.3781,
|
| 20920 |
+
"step": 29210
|
| 20921 |
+
},
|
| 20922 |
+
{
|
| 20923 |
+
"epoch": 0.73,
|
| 20924 |
+
"grad_norm": 10.63198184967041,
|
| 20925 |
+
"learning_rate": 5.28813559322034e-07,
|
| 20926 |
+
"loss": 1.2394,
|
| 20927 |
+
"step": 29220
|
| 20928 |
+
},
|
| 20929 |
+
{
|
| 20930 |
+
"epoch": 0.73,
|
| 20931 |
+
"grad_norm": 16.352039337158203,
|
| 20932 |
+
"learning_rate": 5.220338983050848e-07,
|
| 20933 |
+
"loss": 1.3784,
|
| 20934 |
+
"step": 29230
|
| 20935 |
+
},
|
| 20936 |
+
{
|
| 20937 |
+
"epoch": 0.73,
|
| 20938 |
+
"grad_norm": 9.548080444335938,
|
| 20939 |
+
"learning_rate": 5.152542372881356e-07,
|
| 20940 |
+
"loss": 1.3201,
|
| 20941 |
+
"step": 29240
|
| 20942 |
+
},
|
| 20943 |
+
{
|
| 20944 |
+
"epoch": 0.73,
|
| 20945 |
+
"grad_norm": 2.56595516204834,
|
| 20946 |
+
"learning_rate": 5.084745762711865e-07,
|
| 20947 |
+
"loss": 1.2335,
|
| 20948 |
+
"step": 29250
|
| 20949 |
+
},
|
| 20950 |
+
{
|
| 20951 |
+
"epoch": 0.73,
|
| 20952 |
+
"grad_norm": 3.092132568359375,
|
| 20953 |
+
"learning_rate": 5.016949152542373e-07,
|
| 20954 |
+
"loss": 1.2999,
|
| 20955 |
+
"step": 29260
|
| 20956 |
+
},
|
| 20957 |
+
{
|
| 20958 |
+
"epoch": 0.73,
|
| 20959 |
+
"grad_norm": 12.341950416564941,
|
| 20960 |
+
"learning_rate": 4.949152542372881e-07,
|
| 20961 |
+
"loss": 1.2719,
|
| 20962 |
+
"step": 29270
|
| 20963 |
+
},
|
| 20964 |
+
{
|
| 20965 |
+
"epoch": 0.73,
|
| 20966 |
+
"grad_norm": 16.05048179626465,
|
| 20967 |
+
"learning_rate": 4.881355932203391e-07,
|
| 20968 |
+
"loss": 1.1151,
|
| 20969 |
+
"step": 29280
|
| 20970 |
+
},
|
| 20971 |
+
{
|
| 20972 |
+
"epoch": 0.73,
|
| 20973 |
+
"grad_norm": 8.888206481933594,
|
| 20974 |
+
"learning_rate": 4.813559322033898e-07,
|
| 20975 |
+
"loss": 1.0549,
|
| 20976 |
+
"step": 29290
|
| 20977 |
+
},
|
| 20978 |
+
{
|
| 20979 |
+
"epoch": 0.73,
|
| 20980 |
+
"grad_norm": 16.083812713623047,
|
| 20981 |
+
"learning_rate": 4.745762711864407e-07,
|
| 20982 |
+
"loss": 1.2443,
|
| 20983 |
+
"step": 29300
|
| 20984 |
+
},
|
| 20985 |
+
{
|
| 20986 |
+
"epoch": 0.73,
|
| 20987 |
+
"grad_norm": 5.3232741355896,
|
| 20988 |
+
"learning_rate": 4.6779661016949154e-07,
|
| 20989 |
+
"loss": 1.2953,
|
| 20990 |
+
"step": 29310
|
| 20991 |
+
},
|
| 20992 |
+
{
|
| 20993 |
+
"epoch": 0.73,
|
| 20994 |
+
"grad_norm": 15.959102630615234,
|
| 20995 |
+
"learning_rate": 4.610169491525424e-07,
|
| 20996 |
+
"loss": 1.282,
|
| 20997 |
+
"step": 29320
|
| 20998 |
+
},
|
| 20999 |
+
{
|
| 21000 |
+
"epoch": 0.73,
|
| 21001 |
+
"grad_norm": 8.539166450500488,
|
| 21002 |
+
"learning_rate": 4.542372881355932e-07,
|
| 21003 |
+
"loss": 1.2538,
|
| 21004 |
+
"step": 29330
|
| 21005 |
+
},
|
| 21006 |
+
{
|
| 21007 |
+
"epoch": 0.73,
|
| 21008 |
+
"grad_norm": 4.907639980316162,
|
| 21009 |
+
"learning_rate": 4.4745762711864415e-07,
|
| 21010 |
+
"loss": 1.3666,
|
| 21011 |
+
"step": 29340
|
| 21012 |
+
},
|
| 21013 |
+
{
|
| 21014 |
+
"epoch": 0.73,
|
| 21015 |
+
"grad_norm": 2.455517530441284,
|
| 21016 |
+
"learning_rate": 4.4067796610169497e-07,
|
| 21017 |
+
"loss": 1.2269,
|
| 21018 |
+
"step": 29350
|
| 21019 |
+
},
|
| 21020 |
+
{
|
| 21021 |
+
"epoch": 0.73,
|
| 21022 |
+
"grad_norm": 9.032910346984863,
|
| 21023 |
+
"learning_rate": 4.3389830508474584e-07,
|
| 21024 |
+
"loss": 1.2813,
|
| 21025 |
+
"step": 29360
|
| 21026 |
+
},
|
| 21027 |
+
{
|
| 21028 |
+
"epoch": 0.73,
|
| 21029 |
+
"grad_norm": 2.8847286701202393,
|
| 21030 |
+
"learning_rate": 4.2711864406779666e-07,
|
| 21031 |
+
"loss": 1.3336,
|
| 21032 |
+
"step": 29370
|
| 21033 |
+
},
|
| 21034 |
+
{
|
| 21035 |
+
"epoch": 0.73,
|
| 21036 |
+
"grad_norm": 11.284420013427734,
|
| 21037 |
+
"learning_rate": 4.2033898305084753e-07,
|
| 21038 |
+
"loss": 1.286,
|
| 21039 |
+
"step": 29380
|
| 21040 |
+
},
|
| 21041 |
+
{
|
| 21042 |
+
"epoch": 0.73,
|
| 21043 |
+
"grad_norm": 11.412368774414062,
|
| 21044 |
+
"learning_rate": 4.1355932203389835e-07,
|
| 21045 |
+
"loss": 1.3317,
|
| 21046 |
+
"step": 29390
|
| 21047 |
+
},
|
| 21048 |
+
{
|
| 21049 |
+
"epoch": 0.73,
|
| 21050 |
+
"grad_norm": 4.6027302742004395,
|
| 21051 |
+
"learning_rate": 4.0677966101694916e-07,
|
| 21052 |
+
"loss": 1.3263,
|
| 21053 |
+
"step": 29400
|
| 21054 |
+
},
|
| 21055 |
+
{
|
| 21056 |
+
"epoch": 0.74,
|
| 21057 |
+
"grad_norm": 2.39072322845459,
|
| 21058 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 21059 |
+
"loss": 1.3174,
|
| 21060 |
+
"step": 29410
|
| 21061 |
+
},
|
| 21062 |
+
{
|
| 21063 |
+
"epoch": 0.74,
|
| 21064 |
+
"grad_norm": 2.7101926803588867,
|
| 21065 |
+
"learning_rate": 3.9322033898305085e-07,
|
| 21066 |
+
"loss": 1.2673,
|
| 21067 |
+
"step": 29420
|
| 21068 |
+
},
|
| 21069 |
+
{
|
| 21070 |
+
"epoch": 0.74,
|
| 21071 |
+
"grad_norm": 10.152178764343262,
|
| 21072 |
+
"learning_rate": 3.864406779661017e-07,
|
| 21073 |
+
"loss": 1.3241,
|
| 21074 |
+
"step": 29430
|
| 21075 |
+
},
|
| 21076 |
+
{
|
| 21077 |
+
"epoch": 0.74,
|
| 21078 |
+
"grad_norm": 4.905571937561035,
|
| 21079 |
+
"learning_rate": 3.7966101694915254e-07,
|
| 21080 |
+
"loss": 1.2261,
|
| 21081 |
+
"step": 29440
|
| 21082 |
+
},
|
| 21083 |
+
{
|
| 21084 |
+
"epoch": 0.74,
|
| 21085 |
+
"grad_norm": 10.35471248626709,
|
| 21086 |
+
"learning_rate": 3.7288135593220347e-07,
|
| 21087 |
+
"loss": 1.2083,
|
| 21088 |
+
"step": 29450
|
| 21089 |
+
},
|
| 21090 |
+
{
|
| 21091 |
+
"epoch": 0.74,
|
| 21092 |
+
"grad_norm": 11.7713623046875,
|
| 21093 |
+
"learning_rate": 3.6610169491525423e-07,
|
| 21094 |
+
"loss": 1.2647,
|
| 21095 |
+
"step": 29460
|
| 21096 |
+
},
|
| 21097 |
+
{
|
| 21098 |
+
"epoch": 0.74,
|
| 21099 |
+
"grad_norm": 3.181910276412964,
|
| 21100 |
+
"learning_rate": 3.5932203389830516e-07,
|
| 21101 |
+
"loss": 1.4678,
|
| 21102 |
+
"step": 29470
|
| 21103 |
+
},
|
| 21104 |
+
{
|
| 21105 |
+
"epoch": 0.74,
|
| 21106 |
+
"grad_norm": 6.756598472595215,
|
| 21107 |
+
"learning_rate": 3.52542372881356e-07,
|
| 21108 |
+
"loss": 1.273,
|
| 21109 |
+
"step": 29480
|
| 21110 |
+
},
|
| 21111 |
+
{
|
| 21112 |
+
"epoch": 0.74,
|
| 21113 |
+
"grad_norm": 12.660133361816406,
|
| 21114 |
+
"learning_rate": 3.4576271186440684e-07,
|
| 21115 |
+
"loss": 1.3497,
|
| 21116 |
+
"step": 29490
|
| 21117 |
+
},
|
| 21118 |
+
{
|
| 21119 |
+
"epoch": 0.74,
|
| 21120 |
+
"grad_norm": 11.418837547302246,
|
| 21121 |
+
"learning_rate": 3.3898305084745766e-07,
|
| 21122 |
+
"loss": 1.3025,
|
| 21123 |
+
"step": 29500
|
| 21124 |
+
},
|
| 21125 |
+
{
|
| 21126 |
+
"epoch": 0.74,
|
| 21127 |
+
"eval_loss": 1.2983591556549072,
|
| 21128 |
+
"eval_runtime": 66.2327,
|
| 21129 |
+
"eval_samples_per_second": 15.098,
|
| 21130 |
+
"eval_steps_per_second": 15.098,
|
| 21131 |
+
"step": 29500
|
| 21132 |
+
},
|
| 21133 |
+
{
|
| 21134 |
+
"epoch": 0.74,
|
| 21135 |
+
"grad_norm": 5.526192665100098,
|
| 21136 |
+
"learning_rate": 3.322033898305085e-07,
|
| 21137 |
+
"loss": 1.1364,
|
| 21138 |
+
"step": 29510
|
| 21139 |
+
},
|
| 21140 |
+
{
|
| 21141 |
+
"epoch": 0.74,
|
| 21142 |
+
"grad_norm": 4.229588031768799,
|
| 21143 |
+
"learning_rate": 3.2542372881355935e-07,
|
| 21144 |
+
"loss": 1.3351,
|
| 21145 |
+
"step": 29520
|
| 21146 |
+
},
|
| 21147 |
+
{
|
| 21148 |
+
"epoch": 0.74,
|
| 21149 |
+
"grad_norm": 3.8703114986419678,
|
| 21150 |
+
"learning_rate": 3.1864406779661017e-07,
|
| 21151 |
+
"loss": 1.547,
|
| 21152 |
+
"step": 29530
|
| 21153 |
+
},
|
| 21154 |
+
{
|
| 21155 |
+
"epoch": 0.74,
|
| 21156 |
+
"grad_norm": 4.0574445724487305,
|
| 21157 |
+
"learning_rate": 3.1186440677966104e-07,
|
| 21158 |
+
"loss": 1.5269,
|
| 21159 |
+
"step": 29540
|
| 21160 |
+
},
|
| 21161 |
+
{
|
| 21162 |
+
"epoch": 0.74,
|
| 21163 |
+
"grad_norm": 2.609175443649292,
|
| 21164 |
+
"learning_rate": 3.050847457627119e-07,
|
| 21165 |
+
"loss": 1.3574,
|
| 21166 |
+
"step": 29550
|
| 21167 |
+
},
|
| 21168 |
+
{
|
| 21169 |
+
"epoch": 0.74,
|
| 21170 |
+
"grad_norm": 5.445046901702881,
|
| 21171 |
+
"learning_rate": 2.9830508474576273e-07,
|
| 21172 |
+
"loss": 1.3081,
|
| 21173 |
+
"step": 29560
|
| 21174 |
+
},
|
| 21175 |
+
{
|
| 21176 |
+
"epoch": 0.74,
|
| 21177 |
+
"grad_norm": 8.432710647583008,
|
| 21178 |
+
"learning_rate": 2.9152542372881355e-07,
|
| 21179 |
+
"loss": 1.2343,
|
| 21180 |
+
"step": 29570
|
| 21181 |
+
},
|
| 21182 |
+
{
|
| 21183 |
+
"epoch": 0.74,
|
| 21184 |
+
"grad_norm": 10.050646781921387,
|
| 21185 |
+
"learning_rate": 2.847457627118644e-07,
|
| 21186 |
+
"loss": 1.2665,
|
| 21187 |
+
"step": 29580
|
| 21188 |
+
},
|
| 21189 |
+
{
|
| 21190 |
+
"epoch": 0.74,
|
| 21191 |
+
"grad_norm": 6.895621299743652,
|
| 21192 |
+
"learning_rate": 2.779661016949153e-07,
|
| 21193 |
+
"loss": 1.1077,
|
| 21194 |
+
"step": 29590
|
| 21195 |
+
},
|
| 21196 |
+
{
|
| 21197 |
+
"epoch": 0.74,
|
| 21198 |
+
"grad_norm": 6.045702934265137,
|
| 21199 |
+
"learning_rate": 2.711864406779661e-07,
|
| 21200 |
+
"loss": 1.0306,
|
| 21201 |
+
"step": 29600
|
| 21202 |
+
},
|
| 21203 |
+
{
|
| 21204 |
+
"epoch": 0.74,
|
| 21205 |
+
"grad_norm": 2.18096661567688,
|
| 21206 |
+
"learning_rate": 2.64406779661017e-07,
|
| 21207 |
+
"loss": 1.167,
|
| 21208 |
+
"step": 29610
|
| 21209 |
+
},
|
| 21210 |
+
{
|
| 21211 |
+
"epoch": 0.74,
|
| 21212 |
+
"grad_norm": 8.721917152404785,
|
| 21213 |
+
"learning_rate": 2.576271186440678e-07,
|
| 21214 |
+
"loss": 1.2476,
|
| 21215 |
+
"step": 29620
|
| 21216 |
+
},
|
| 21217 |
+
{
|
| 21218 |
+
"epoch": 0.74,
|
| 21219 |
+
"grad_norm": 5.546863079071045,
|
| 21220 |
+
"learning_rate": 2.5084745762711867e-07,
|
| 21221 |
+
"loss": 1.3273,
|
| 21222 |
+
"step": 29630
|
| 21223 |
+
},
|
| 21224 |
+
{
|
| 21225 |
+
"epoch": 0.74,
|
| 21226 |
+
"grad_norm": 7.264005661010742,
|
| 21227 |
+
"learning_rate": 2.4406779661016954e-07,
|
| 21228 |
+
"loss": 1.4333,
|
| 21229 |
+
"step": 29640
|
| 21230 |
+
},
|
| 21231 |
+
{
|
| 21232 |
+
"epoch": 0.74,
|
| 21233 |
+
"grad_norm": 4.217822551727295,
|
| 21234 |
+
"learning_rate": 2.3728813559322036e-07,
|
| 21235 |
+
"loss": 1.5513,
|
| 21236 |
+
"step": 29650
|
| 21237 |
+
},
|
| 21238 |
+
{
|
| 21239 |
+
"epoch": 0.74,
|
| 21240 |
+
"grad_norm": 21.425151824951172,
|
| 21241 |
+
"learning_rate": 2.305084745762712e-07,
|
| 21242 |
+
"loss": 1.15,
|
| 21243 |
+
"step": 29660
|
| 21244 |
+
},
|
| 21245 |
+
{
|
| 21246 |
+
"epoch": 0.74,
|
| 21247 |
+
"grad_norm": 17.36193084716797,
|
| 21248 |
+
"learning_rate": 2.2372881355932207e-07,
|
| 21249 |
+
"loss": 1.4419,
|
| 21250 |
+
"step": 29670
|
| 21251 |
+
},
|
| 21252 |
+
{
|
| 21253 |
+
"epoch": 0.74,
|
| 21254 |
+
"grad_norm": 9.74802017211914,
|
| 21255 |
+
"learning_rate": 2.1694915254237292e-07,
|
| 21256 |
+
"loss": 1.2543,
|
| 21257 |
+
"step": 29680
|
| 21258 |
+
},
|
| 21259 |
+
{
|
| 21260 |
+
"epoch": 0.74,
|
| 21261 |
+
"grad_norm": 12.033125877380371,
|
| 21262 |
+
"learning_rate": 2.1016949152542376e-07,
|
| 21263 |
+
"loss": 1.378,
|
| 21264 |
+
"step": 29690
|
| 21265 |
+
},
|
| 21266 |
+
{
|
| 21267 |
+
"epoch": 0.74,
|
| 21268 |
+
"grad_norm": 14.735278129577637,
|
| 21269 |
+
"learning_rate": 2.0338983050847458e-07,
|
| 21270 |
+
"loss": 1.2712,
|
| 21271 |
+
"step": 29700
|
| 21272 |
+
},
|
| 21273 |
+
{
|
| 21274 |
+
"epoch": 0.74,
|
| 21275 |
+
"grad_norm": 11.541913986206055,
|
| 21276 |
+
"learning_rate": 1.9661016949152543e-07,
|
| 21277 |
+
"loss": 1.0699,
|
| 21278 |
+
"step": 29710
|
| 21279 |
+
},
|
| 21280 |
+
{
|
| 21281 |
+
"epoch": 0.74,
|
| 21282 |
+
"grad_norm": 3.0062718391418457,
|
| 21283 |
+
"learning_rate": 1.8983050847457627e-07,
|
| 21284 |
+
"loss": 1.3865,
|
| 21285 |
+
"step": 29720
|
| 21286 |
+
},
|
| 21287 |
+
{
|
| 21288 |
+
"epoch": 0.74,
|
| 21289 |
+
"grad_norm": 7.821793079376221,
|
| 21290 |
+
"learning_rate": 1.8305084745762712e-07,
|
| 21291 |
+
"loss": 1.1679,
|
| 21292 |
+
"step": 29730
|
| 21293 |
+
},
|
| 21294 |
+
{
|
| 21295 |
+
"epoch": 0.74,
|
| 21296 |
+
"grad_norm": 12.209012031555176,
|
| 21297 |
+
"learning_rate": 1.76271186440678e-07,
|
| 21298 |
+
"loss": 1.1567,
|
| 21299 |
+
"step": 29740
|
| 21300 |
+
},
|
| 21301 |
+
{
|
| 21302 |
+
"epoch": 0.74,
|
| 21303 |
+
"grad_norm": 3.285269021987915,
|
| 21304 |
+
"learning_rate": 1.6949152542372883e-07,
|
| 21305 |
+
"loss": 1.3667,
|
| 21306 |
+
"step": 29750
|
| 21307 |
+
},
|
| 21308 |
+
{
|
| 21309 |
+
"epoch": 0.74,
|
| 21310 |
+
"grad_norm": 5.812708377838135,
|
| 21311 |
+
"learning_rate": 1.6271186440677968e-07,
|
| 21312 |
+
"loss": 1.2521,
|
| 21313 |
+
"step": 29760
|
| 21314 |
+
},
|
| 21315 |
+
{
|
| 21316 |
+
"epoch": 0.74,
|
| 21317 |
+
"grad_norm": 3.4106059074401855,
|
| 21318 |
+
"learning_rate": 1.5593220338983052e-07,
|
| 21319 |
+
"loss": 1.3847,
|
| 21320 |
+
"step": 29770
|
| 21321 |
+
},
|
| 21322 |
+
{
|
| 21323 |
+
"epoch": 0.74,
|
| 21324 |
+
"grad_norm": 9.550897598266602,
|
| 21325 |
+
"learning_rate": 1.4915254237288137e-07,
|
| 21326 |
+
"loss": 1.2853,
|
| 21327 |
+
"step": 29780
|
| 21328 |
+
},
|
| 21329 |
+
{
|
| 21330 |
+
"epoch": 0.74,
|
| 21331 |
+
"grad_norm": 13.644095420837402,
|
| 21332 |
+
"learning_rate": 1.423728813559322e-07,
|
| 21333 |
+
"loss": 1.3305,
|
| 21334 |
+
"step": 29790
|
| 21335 |
+
},
|
| 21336 |
+
{
|
| 21337 |
+
"epoch": 0.74,
|
| 21338 |
+
"grad_norm": 6.250187397003174,
|
| 21339 |
+
"learning_rate": 1.3559322033898305e-07,
|
| 21340 |
+
"loss": 1.2896,
|
| 21341 |
+
"step": 29800
|
| 21342 |
+
},
|
| 21343 |
+
{
|
| 21344 |
+
"epoch": 0.75,
|
| 21345 |
+
"grad_norm": 11.07441234588623,
|
| 21346 |
+
"learning_rate": 1.288135593220339e-07,
|
| 21347 |
+
"loss": 1.2263,
|
| 21348 |
+
"step": 29810
|
| 21349 |
+
},
|
| 21350 |
+
{
|
| 21351 |
+
"epoch": 0.75,
|
| 21352 |
+
"grad_norm": 13.597912788391113,
|
| 21353 |
+
"learning_rate": 1.2203389830508477e-07,
|
| 21354 |
+
"loss": 1.4195,
|
| 21355 |
+
"step": 29820
|
| 21356 |
+
},
|
| 21357 |
+
{
|
| 21358 |
+
"epoch": 0.75,
|
| 21359 |
+
"grad_norm": 4.825064659118652,
|
| 21360 |
+
"learning_rate": 1.152542372881356e-07,
|
| 21361 |
+
"loss": 1.3549,
|
| 21362 |
+
"step": 29830
|
| 21363 |
+
},
|
| 21364 |
+
{
|
| 21365 |
+
"epoch": 0.75,
|
| 21366 |
+
"grad_norm": 3.6931655406951904,
|
| 21367 |
+
"learning_rate": 1.0847457627118646e-07,
|
| 21368 |
+
"loss": 1.3675,
|
| 21369 |
+
"step": 29840
|
| 21370 |
+
},
|
| 21371 |
+
{
|
| 21372 |
+
"epoch": 0.75,
|
| 21373 |
+
"grad_norm": 5.718669414520264,
|
| 21374 |
+
"learning_rate": 1.0169491525423729e-07,
|
| 21375 |
+
"loss": 1.4715,
|
| 21376 |
+
"step": 29850
|
| 21377 |
+
},
|
| 21378 |
+
{
|
| 21379 |
+
"epoch": 0.75,
|
| 21380 |
+
"grad_norm": 6.913152694702148,
|
| 21381 |
+
"learning_rate": 9.491525423728814e-08,
|
| 21382 |
+
"loss": 1.4853,
|
| 21383 |
+
"step": 29860
|
| 21384 |
+
},
|
| 21385 |
+
{
|
| 21386 |
+
"epoch": 0.75,
|
| 21387 |
+
"grad_norm": 5.6454644203186035,
|
| 21388 |
+
"learning_rate": 8.8135593220339e-08,
|
| 21389 |
+
"loss": 1.3638,
|
| 21390 |
+
"step": 29870
|
| 21391 |
+
},
|
| 21392 |
+
{
|
| 21393 |
+
"epoch": 0.75,
|
| 21394 |
+
"grad_norm": 7.006107330322266,
|
| 21395 |
+
"learning_rate": 8.135593220338984e-08,
|
| 21396 |
+
"loss": 1.387,
|
| 21397 |
+
"step": 29880
|
| 21398 |
+
},
|
| 21399 |
+
{
|
| 21400 |
+
"epoch": 0.75,
|
| 21401 |
+
"grad_norm": 7.425577163696289,
|
| 21402 |
+
"learning_rate": 7.457627118644068e-08,
|
| 21403 |
+
"loss": 1.314,
|
| 21404 |
+
"step": 29890
|
| 21405 |
+
},
|
| 21406 |
+
{
|
| 21407 |
+
"epoch": 0.75,
|
| 21408 |
+
"grad_norm": 12.527627944946289,
|
| 21409 |
+
"learning_rate": 6.779661016949153e-08,
|
| 21410 |
+
"loss": 1.3822,
|
| 21411 |
+
"step": 29900
|
| 21412 |
+
},
|
| 21413 |
+
{
|
| 21414 |
+
"epoch": 0.75,
|
| 21415 |
+
"grad_norm": 7.276275634765625,
|
| 21416 |
+
"learning_rate": 6.101694915254239e-08,
|
| 21417 |
+
"loss": 1.1887,
|
| 21418 |
+
"step": 29910
|
| 21419 |
+
},
|
| 21420 |
+
{
|
| 21421 |
+
"epoch": 0.75,
|
| 21422 |
+
"grad_norm": 1.8828225135803223,
|
| 21423 |
+
"learning_rate": 5.423728813559323e-08,
|
| 21424 |
+
"loss": 1.3262,
|
| 21425 |
+
"step": 29920
|
| 21426 |
+
},
|
| 21427 |
+
{
|
| 21428 |
+
"epoch": 0.75,
|
| 21429 |
+
"grad_norm": 3.2337329387664795,
|
| 21430 |
+
"learning_rate": 4.745762711864407e-08,
|
| 21431 |
+
"loss": 1.3487,
|
| 21432 |
+
"step": 29930
|
| 21433 |
+
},
|
| 21434 |
+
{
|
| 21435 |
+
"epoch": 0.75,
|
| 21436 |
+
"grad_norm": 6.47509765625,
|
| 21437 |
+
"learning_rate": 4.067796610169492e-08,
|
| 21438 |
+
"loss": 1.3835,
|
| 21439 |
+
"step": 29940
|
| 21440 |
+
},
|
| 21441 |
+
{
|
| 21442 |
+
"epoch": 0.75,
|
| 21443 |
+
"grad_norm": 8.018141746520996,
|
| 21444 |
+
"learning_rate": 3.3898305084745764e-08,
|
| 21445 |
+
"loss": 1.1818,
|
| 21446 |
+
"step": 29950
|
| 21447 |
+
},
|
| 21448 |
+
{
|
| 21449 |
+
"epoch": 0.75,
|
| 21450 |
+
"grad_norm": 6.676591396331787,
|
| 21451 |
+
"learning_rate": 2.7118644067796615e-08,
|
| 21452 |
+
"loss": 1.2067,
|
| 21453 |
+
"step": 29960
|
| 21454 |
+
},
|
| 21455 |
+
{
|
| 21456 |
+
"epoch": 0.75,
|
| 21457 |
+
"grad_norm": 3.872598648071289,
|
| 21458 |
+
"learning_rate": 2.033898305084746e-08,
|
| 21459 |
+
"loss": 1.2989,
|
| 21460 |
+
"step": 29970
|
| 21461 |
+
},
|
| 21462 |
+
{
|
| 21463 |
+
"epoch": 0.75,
|
| 21464 |
+
"grad_norm": 2.7663283348083496,
|
| 21465 |
+
"learning_rate": 1.3559322033898307e-08,
|
| 21466 |
+
"loss": 1.321,
|
| 21467 |
+
"step": 29980
|
| 21468 |
+
},
|
| 21469 |
+
{
|
| 21470 |
+
"epoch": 0.75,
|
| 21471 |
+
"grad_norm": 6.447133541107178,
|
| 21472 |
+
"learning_rate": 6.779661016949154e-09,
|
| 21473 |
+
"loss": 1.255,
|
| 21474 |
+
"step": 29990
|
| 21475 |
+
},
|
| 21476 |
+
{
|
| 21477 |
+
"epoch": 0.75,
|
| 21478 |
+
"grad_norm": 5.224142074584961,
|
| 21479 |
+
"learning_rate": 0.0,
|
| 21480 |
+
"loss": 1.2453,
|
| 21481 |
+
"step": 30000
|
| 21482 |
+
},
|
| 21483 |
+
{
|
| 21484 |
+
"epoch": 0.75,
|
| 21485 |
+
"eval_loss": 1.307248592376709,
|
| 21486 |
+
"eval_runtime": 66.2202,
|
| 21487 |
+
"eval_samples_per_second": 15.101,
|
| 21488 |
+
"eval_steps_per_second": 15.101,
|
| 21489 |
+
"step": 30000
|
| 21490 |
}
|
| 21491 |
],
|
| 21492 |
"logging_steps": 10,
|
|
|
|
| 21494 |
"num_input_tokens_seen": 0,
|
| 21495 |
"num_train_epochs": 1,
|
| 21496 |
"save_steps": 2500,
|
| 21497 |
+
"total_flos": 4.8306377981952e+17,
|
| 21498 |
"train_batch_size": 1,
|
| 21499 |
"trial_name": null,
|
| 21500 |
"trial_params": null
|