Uploaded checkpoint-27500
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:d97fc9daed8fc42b8286be9a22db8d8f0c98b367d6f684f8724075b9c509868c
|
| 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:da5ec1b3aa5d12330bda5fc479f69a184c2145592be7a96f6bd1ace39646aaf8
|
| 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 sha256:42ed1734c5823abfe806343a4de18dcccd1e9ad5af5349e08097c7bde2aa7437
|
| 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:bae572518ab53ddc674f52a5ef01613875bea64a8d9c53d4b7d4a9aedc712f19
|
| 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,
|
|
@@ -17907,6 +17907,1796 @@
|
|
| 17907 |
"eval_samples_per_second": 15.125,
|
| 17908 |
"eval_steps_per_second": 15.125,
|
| 17909 |
"step": 25000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17910 |
}
|
| 17911 |
],
|
| 17912 |
"logging_steps": 10,
|
|
@@ -17914,7 +19704,7 @@
|
|
| 17914 |
"num_input_tokens_seen": 0,
|
| 17915 |
"num_train_epochs": 1,
|
| 17916 |
"save_steps": 2500,
|
| 17917 |
-
"total_flos": 4.
|
| 17918 |
"train_batch_size": 1,
|
| 17919 |
"trial_name": null,
|
| 17920 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.262895941734314,
|
| 3 |
+
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-27500",
|
| 4 |
+
"epoch": 0.6875,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 27500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 17907 |
"eval_samples_per_second": 15.125,
|
| 17908 |
"eval_steps_per_second": 15.125,
|
| 17909 |
"step": 25000
|
| 17910 |
+
},
|
| 17911 |
+
{
|
| 17912 |
+
"epoch": 0.63,
|
| 17913 |
+
"grad_norm": 7.775777339935303,
|
| 17914 |
+
"learning_rate": 3.3830508474576273e-06,
|
| 17915 |
+
"loss": 1.5704,
|
| 17916 |
+
"step": 25010
|
| 17917 |
+
},
|
| 17918 |
+
{
|
| 17919 |
+
"epoch": 0.63,
|
| 17920 |
+
"grad_norm": 6.8910698890686035,
|
| 17921 |
+
"learning_rate": 3.3762711864406783e-06,
|
| 17922 |
+
"loss": 1.2296,
|
| 17923 |
+
"step": 25020
|
| 17924 |
+
},
|
| 17925 |
+
{
|
| 17926 |
+
"epoch": 0.63,
|
| 17927 |
+
"grad_norm": 3.11647367477417,
|
| 17928 |
+
"learning_rate": 3.3694915254237292e-06,
|
| 17929 |
+
"loss": 1.1465,
|
| 17930 |
+
"step": 25030
|
| 17931 |
+
},
|
| 17932 |
+
{
|
| 17933 |
+
"epoch": 0.63,
|
| 17934 |
+
"grad_norm": 4.026238441467285,
|
| 17935 |
+
"learning_rate": 3.3627118644067802e-06,
|
| 17936 |
+
"loss": 1.4266,
|
| 17937 |
+
"step": 25040
|
| 17938 |
+
},
|
| 17939 |
+
{
|
| 17940 |
+
"epoch": 0.63,
|
| 17941 |
+
"grad_norm": 3.3097872734069824,
|
| 17942 |
+
"learning_rate": 3.3559322033898308e-06,
|
| 17943 |
+
"loss": 1.4,
|
| 17944 |
+
"step": 25050
|
| 17945 |
+
},
|
| 17946 |
+
{
|
| 17947 |
+
"epoch": 0.63,
|
| 17948 |
+
"grad_norm": 3.2518560886383057,
|
| 17949 |
+
"learning_rate": 3.3491525423728817e-06,
|
| 17950 |
+
"loss": 1.3535,
|
| 17951 |
+
"step": 25060
|
| 17952 |
+
},
|
| 17953 |
+
{
|
| 17954 |
+
"epoch": 0.63,
|
| 17955 |
+
"grad_norm": 4.915974140167236,
|
| 17956 |
+
"learning_rate": 3.3423728813559327e-06,
|
| 17957 |
+
"loss": 1.3441,
|
| 17958 |
+
"step": 25070
|
| 17959 |
+
},
|
| 17960 |
+
{
|
| 17961 |
+
"epoch": 0.63,
|
| 17962 |
+
"grad_norm": 4.099907398223877,
|
| 17963 |
+
"learning_rate": 3.3355932203389833e-06,
|
| 17964 |
+
"loss": 1.4649,
|
| 17965 |
+
"step": 25080
|
| 17966 |
+
},
|
| 17967 |
+
{
|
| 17968 |
+
"epoch": 0.63,
|
| 17969 |
+
"grad_norm": 7.00651741027832,
|
| 17970 |
+
"learning_rate": 3.3288135593220343e-06,
|
| 17971 |
+
"loss": 1.4445,
|
| 17972 |
+
"step": 25090
|
| 17973 |
+
},
|
| 17974 |
+
{
|
| 17975 |
+
"epoch": 0.63,
|
| 17976 |
+
"grad_norm": 2.7813143730163574,
|
| 17977 |
+
"learning_rate": 3.322033898305085e-06,
|
| 17978 |
+
"loss": 1.2997,
|
| 17979 |
+
"step": 25100
|
| 17980 |
+
},
|
| 17981 |
+
{
|
| 17982 |
+
"epoch": 0.63,
|
| 17983 |
+
"grad_norm": 4.583659648895264,
|
| 17984 |
+
"learning_rate": 3.3152542372881358e-06,
|
| 17985 |
+
"loss": 1.3112,
|
| 17986 |
+
"step": 25110
|
| 17987 |
+
},
|
| 17988 |
+
{
|
| 17989 |
+
"epoch": 0.63,
|
| 17990 |
+
"grad_norm": 9.5046968460083,
|
| 17991 |
+
"learning_rate": 3.3084745762711868e-06,
|
| 17992 |
+
"loss": 1.3488,
|
| 17993 |
+
"step": 25120
|
| 17994 |
+
},
|
| 17995 |
+
{
|
| 17996 |
+
"epoch": 0.63,
|
| 17997 |
+
"grad_norm": 4.829354286193848,
|
| 17998 |
+
"learning_rate": 3.3016949152542377e-06,
|
| 17999 |
+
"loss": 1.2801,
|
| 18000 |
+
"step": 25130
|
| 18001 |
+
},
|
| 18002 |
+
{
|
| 18003 |
+
"epoch": 0.63,
|
| 18004 |
+
"grad_norm": 8.664487838745117,
|
| 18005 |
+
"learning_rate": 3.2949152542372887e-06,
|
| 18006 |
+
"loss": 1.3001,
|
| 18007 |
+
"step": 25140
|
| 18008 |
+
},
|
| 18009 |
+
{
|
| 18010 |
+
"epoch": 0.63,
|
| 18011 |
+
"grad_norm": 1.5184565782546997,
|
| 18012 |
+
"learning_rate": 3.288135593220339e-06,
|
| 18013 |
+
"loss": 1.3178,
|
| 18014 |
+
"step": 25150
|
| 18015 |
+
},
|
| 18016 |
+
{
|
| 18017 |
+
"epoch": 0.63,
|
| 18018 |
+
"grad_norm": 2.527219295501709,
|
| 18019 |
+
"learning_rate": 3.28135593220339e-06,
|
| 18020 |
+
"loss": 1.1742,
|
| 18021 |
+
"step": 25160
|
| 18022 |
+
},
|
| 18023 |
+
{
|
| 18024 |
+
"epoch": 0.63,
|
| 18025 |
+
"grad_norm": 1.9469351768493652,
|
| 18026 |
+
"learning_rate": 3.274576271186441e-06,
|
| 18027 |
+
"loss": 1.3338,
|
| 18028 |
+
"step": 25170
|
| 18029 |
+
},
|
| 18030 |
+
{
|
| 18031 |
+
"epoch": 0.63,
|
| 18032 |
+
"grad_norm": 9.183582305908203,
|
| 18033 |
+
"learning_rate": 3.2677966101694918e-06,
|
| 18034 |
+
"loss": 1.2689,
|
| 18035 |
+
"step": 25180
|
| 18036 |
+
},
|
| 18037 |
+
{
|
| 18038 |
+
"epoch": 0.63,
|
| 18039 |
+
"grad_norm": 8.95080280303955,
|
| 18040 |
+
"learning_rate": 3.2610169491525428e-06,
|
| 18041 |
+
"loss": 1.4399,
|
| 18042 |
+
"step": 25190
|
| 18043 |
+
},
|
| 18044 |
+
{
|
| 18045 |
+
"epoch": 0.63,
|
| 18046 |
+
"grad_norm": 3.5885729789733887,
|
| 18047 |
+
"learning_rate": 3.2542372881355933e-06,
|
| 18048 |
+
"loss": 1.2411,
|
| 18049 |
+
"step": 25200
|
| 18050 |
+
},
|
| 18051 |
+
{
|
| 18052 |
+
"epoch": 0.63,
|
| 18053 |
+
"grad_norm": 6.78897762298584,
|
| 18054 |
+
"learning_rate": 3.2474576271186443e-06,
|
| 18055 |
+
"loss": 1.1516,
|
| 18056 |
+
"step": 25210
|
| 18057 |
+
},
|
| 18058 |
+
{
|
| 18059 |
+
"epoch": 0.63,
|
| 18060 |
+
"grad_norm": 16.358980178833008,
|
| 18061 |
+
"learning_rate": 3.2406779661016953e-06,
|
| 18062 |
+
"loss": 1.3401,
|
| 18063 |
+
"step": 25220
|
| 18064 |
+
},
|
| 18065 |
+
{
|
| 18066 |
+
"epoch": 0.63,
|
| 18067 |
+
"grad_norm": 13.469403266906738,
|
| 18068 |
+
"learning_rate": 3.2338983050847462e-06,
|
| 18069 |
+
"loss": 1.1778,
|
| 18070 |
+
"step": 25230
|
| 18071 |
+
},
|
| 18072 |
+
{
|
| 18073 |
+
"epoch": 0.63,
|
| 18074 |
+
"grad_norm": 9.235873222351074,
|
| 18075 |
+
"learning_rate": 3.2271186440677972e-06,
|
| 18076 |
+
"loss": 1.463,
|
| 18077 |
+
"step": 25240
|
| 18078 |
+
},
|
| 18079 |
+
{
|
| 18080 |
+
"epoch": 0.63,
|
| 18081 |
+
"grad_norm": 5.634754657745361,
|
| 18082 |
+
"learning_rate": 3.2203389830508473e-06,
|
| 18083 |
+
"loss": 1.4592,
|
| 18084 |
+
"step": 25250
|
| 18085 |
+
},
|
| 18086 |
+
{
|
| 18087 |
+
"epoch": 0.63,
|
| 18088 |
+
"grad_norm": 7.916177749633789,
|
| 18089 |
+
"learning_rate": 3.2135593220338983e-06,
|
| 18090 |
+
"loss": 1.2284,
|
| 18091 |
+
"step": 25260
|
| 18092 |
+
},
|
| 18093 |
+
{
|
| 18094 |
+
"epoch": 0.63,
|
| 18095 |
+
"grad_norm": 3.810818672180176,
|
| 18096 |
+
"learning_rate": 3.2067796610169493e-06,
|
| 18097 |
+
"loss": 1.3948,
|
| 18098 |
+
"step": 25270
|
| 18099 |
+
},
|
| 18100 |
+
{
|
| 18101 |
+
"epoch": 0.63,
|
| 18102 |
+
"grad_norm": 3.151276111602783,
|
| 18103 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 18104 |
+
"loss": 1.3252,
|
| 18105 |
+
"step": 25280
|
| 18106 |
+
},
|
| 18107 |
+
{
|
| 18108 |
+
"epoch": 0.63,
|
| 18109 |
+
"grad_norm": 4.654315948486328,
|
| 18110 |
+
"learning_rate": 3.1932203389830513e-06,
|
| 18111 |
+
"loss": 1.3577,
|
| 18112 |
+
"step": 25290
|
| 18113 |
+
},
|
| 18114 |
+
{
|
| 18115 |
+
"epoch": 0.63,
|
| 18116 |
+
"grad_norm": 7.808940410614014,
|
| 18117 |
+
"learning_rate": 3.186440677966102e-06,
|
| 18118 |
+
"loss": 1.419,
|
| 18119 |
+
"step": 25300
|
| 18120 |
+
},
|
| 18121 |
+
{
|
| 18122 |
+
"epoch": 0.63,
|
| 18123 |
+
"grad_norm": 6.512801647186279,
|
| 18124 |
+
"learning_rate": 3.1796610169491528e-06,
|
| 18125 |
+
"loss": 1.2751,
|
| 18126 |
+
"step": 25310
|
| 18127 |
+
},
|
| 18128 |
+
{
|
| 18129 |
+
"epoch": 0.63,
|
| 18130 |
+
"grad_norm": 8.710541725158691,
|
| 18131 |
+
"learning_rate": 3.1728813559322038e-06,
|
| 18132 |
+
"loss": 1.2923,
|
| 18133 |
+
"step": 25320
|
| 18134 |
+
},
|
| 18135 |
+
{
|
| 18136 |
+
"epoch": 0.63,
|
| 18137 |
+
"grad_norm": 3.160752296447754,
|
| 18138 |
+
"learning_rate": 3.1661016949152547e-06,
|
| 18139 |
+
"loss": 1.3357,
|
| 18140 |
+
"step": 25330
|
| 18141 |
+
},
|
| 18142 |
+
{
|
| 18143 |
+
"epoch": 0.63,
|
| 18144 |
+
"grad_norm": 2.4420571327209473,
|
| 18145 |
+
"learning_rate": 3.1593220338983053e-06,
|
| 18146 |
+
"loss": 1.4122,
|
| 18147 |
+
"step": 25340
|
| 18148 |
+
},
|
| 18149 |
+
{
|
| 18150 |
+
"epoch": 0.63,
|
| 18151 |
+
"grad_norm": 12.090134620666504,
|
| 18152 |
+
"learning_rate": 3.1525423728813563e-06,
|
| 18153 |
+
"loss": 1.1703,
|
| 18154 |
+
"step": 25350
|
| 18155 |
+
},
|
| 18156 |
+
{
|
| 18157 |
+
"epoch": 0.63,
|
| 18158 |
+
"grad_norm": 5.109529972076416,
|
| 18159 |
+
"learning_rate": 3.145762711864407e-06,
|
| 18160 |
+
"loss": 1.4391,
|
| 18161 |
+
"step": 25360
|
| 18162 |
+
},
|
| 18163 |
+
{
|
| 18164 |
+
"epoch": 0.63,
|
| 18165 |
+
"grad_norm": 3.0033655166625977,
|
| 18166 |
+
"learning_rate": 3.138983050847458e-06,
|
| 18167 |
+
"loss": 1.2308,
|
| 18168 |
+
"step": 25370
|
| 18169 |
+
},
|
| 18170 |
+
{
|
| 18171 |
+
"epoch": 0.63,
|
| 18172 |
+
"grad_norm": 6.253593921661377,
|
| 18173 |
+
"learning_rate": 3.1322033898305088e-06,
|
| 18174 |
+
"loss": 1.1566,
|
| 18175 |
+
"step": 25380
|
| 18176 |
+
},
|
| 18177 |
+
{
|
| 18178 |
+
"epoch": 0.63,
|
| 18179 |
+
"grad_norm": 5.59571647644043,
|
| 18180 |
+
"learning_rate": 3.1254237288135598e-06,
|
| 18181 |
+
"loss": 1.3234,
|
| 18182 |
+
"step": 25390
|
| 18183 |
+
},
|
| 18184 |
+
{
|
| 18185 |
+
"epoch": 0.64,
|
| 18186 |
+
"grad_norm": 15.45341968536377,
|
| 18187 |
+
"learning_rate": 3.1186440677966107e-06,
|
| 18188 |
+
"loss": 1.3298,
|
| 18189 |
+
"step": 25400
|
| 18190 |
+
},
|
| 18191 |
+
{
|
| 18192 |
+
"epoch": 0.64,
|
| 18193 |
+
"grad_norm": 4.012663841247559,
|
| 18194 |
+
"learning_rate": 3.111864406779661e-06,
|
| 18195 |
+
"loss": 1.3515,
|
| 18196 |
+
"step": 25410
|
| 18197 |
+
},
|
| 18198 |
+
{
|
| 18199 |
+
"epoch": 0.64,
|
| 18200 |
+
"grad_norm": 11.210050582885742,
|
| 18201 |
+
"learning_rate": 3.105084745762712e-06,
|
| 18202 |
+
"loss": 1.5219,
|
| 18203 |
+
"step": 25420
|
| 18204 |
+
},
|
| 18205 |
+
{
|
| 18206 |
+
"epoch": 0.64,
|
| 18207 |
+
"grad_norm": 12.784728050231934,
|
| 18208 |
+
"learning_rate": 3.098305084745763e-06,
|
| 18209 |
+
"loss": 1.1988,
|
| 18210 |
+
"step": 25430
|
| 18211 |
+
},
|
| 18212 |
+
{
|
| 18213 |
+
"epoch": 0.64,
|
| 18214 |
+
"grad_norm": 4.20169734954834,
|
| 18215 |
+
"learning_rate": 3.091525423728814e-06,
|
| 18216 |
+
"loss": 1.421,
|
| 18217 |
+
"step": 25440
|
| 18218 |
+
},
|
| 18219 |
+
{
|
| 18220 |
+
"epoch": 0.64,
|
| 18221 |
+
"grad_norm": 6.456515312194824,
|
| 18222 |
+
"learning_rate": 3.0847457627118648e-06,
|
| 18223 |
+
"loss": 1.1633,
|
| 18224 |
+
"step": 25450
|
| 18225 |
+
},
|
| 18226 |
+
{
|
| 18227 |
+
"epoch": 0.64,
|
| 18228 |
+
"grad_norm": 6.611431121826172,
|
| 18229 |
+
"learning_rate": 3.0779661016949153e-06,
|
| 18230 |
+
"loss": 1.2853,
|
| 18231 |
+
"step": 25460
|
| 18232 |
+
},
|
| 18233 |
+
{
|
| 18234 |
+
"epoch": 0.64,
|
| 18235 |
+
"grad_norm": 4.3997015953063965,
|
| 18236 |
+
"learning_rate": 3.0711864406779663e-06,
|
| 18237 |
+
"loss": 1.2609,
|
| 18238 |
+
"step": 25470
|
| 18239 |
+
},
|
| 18240 |
+
{
|
| 18241 |
+
"epoch": 0.64,
|
| 18242 |
+
"grad_norm": 12.016313552856445,
|
| 18243 |
+
"learning_rate": 3.0644067796610173e-06,
|
| 18244 |
+
"loss": 1.1744,
|
| 18245 |
+
"step": 25480
|
| 18246 |
+
},
|
| 18247 |
+
{
|
| 18248 |
+
"epoch": 0.64,
|
| 18249 |
+
"grad_norm": 5.508640289306641,
|
| 18250 |
+
"learning_rate": 3.0576271186440683e-06,
|
| 18251 |
+
"loss": 1.2545,
|
| 18252 |
+
"step": 25490
|
| 18253 |
+
},
|
| 18254 |
+
{
|
| 18255 |
+
"epoch": 0.64,
|
| 18256 |
+
"grad_norm": 7.7658305168151855,
|
| 18257 |
+
"learning_rate": 3.0508474576271192e-06,
|
| 18258 |
+
"loss": 1.132,
|
| 18259 |
+
"step": 25500
|
| 18260 |
+
},
|
| 18261 |
+
{
|
| 18262 |
+
"epoch": 0.64,
|
| 18263 |
+
"eval_loss": 1.3264555931091309,
|
| 18264 |
+
"eval_runtime": 66.1247,
|
| 18265 |
+
"eval_samples_per_second": 15.123,
|
| 18266 |
+
"eval_steps_per_second": 15.123,
|
| 18267 |
+
"step": 25500
|
| 18268 |
+
},
|
| 18269 |
+
{
|
| 18270 |
+
"epoch": 0.64,
|
| 18271 |
+
"grad_norm": 4.297203063964844,
|
| 18272 |
+
"learning_rate": 3.0440677966101694e-06,
|
| 18273 |
+
"loss": 1.0663,
|
| 18274 |
+
"step": 25510
|
| 18275 |
+
},
|
| 18276 |
+
{
|
| 18277 |
+
"epoch": 0.64,
|
| 18278 |
+
"grad_norm": 12.412454605102539,
|
| 18279 |
+
"learning_rate": 3.0372881355932203e-06,
|
| 18280 |
+
"loss": 1.1782,
|
| 18281 |
+
"step": 25520
|
| 18282 |
+
},
|
| 18283 |
+
{
|
| 18284 |
+
"epoch": 0.64,
|
| 18285 |
+
"grad_norm": 10.640459060668945,
|
| 18286 |
+
"learning_rate": 3.0305084745762713e-06,
|
| 18287 |
+
"loss": 1.4982,
|
| 18288 |
+
"step": 25530
|
| 18289 |
+
},
|
| 18290 |
+
{
|
| 18291 |
+
"epoch": 0.64,
|
| 18292 |
+
"grad_norm": 12.518424987792969,
|
| 18293 |
+
"learning_rate": 3.0237288135593223e-06,
|
| 18294 |
+
"loss": 1.4771,
|
| 18295 |
+
"step": 25540
|
| 18296 |
+
},
|
| 18297 |
+
{
|
| 18298 |
+
"epoch": 0.64,
|
| 18299 |
+
"grad_norm": 7.445399284362793,
|
| 18300 |
+
"learning_rate": 3.0169491525423733e-06,
|
| 18301 |
+
"loss": 1.2959,
|
| 18302 |
+
"step": 25550
|
| 18303 |
+
},
|
| 18304 |
+
{
|
| 18305 |
+
"epoch": 0.64,
|
| 18306 |
+
"grad_norm": 3.960470676422119,
|
| 18307 |
+
"learning_rate": 3.010169491525424e-06,
|
| 18308 |
+
"loss": 1.4464,
|
| 18309 |
+
"step": 25560
|
| 18310 |
+
},
|
| 18311 |
+
{
|
| 18312 |
+
"epoch": 0.64,
|
| 18313 |
+
"grad_norm": 11.259561538696289,
|
| 18314 |
+
"learning_rate": 3.003389830508475e-06,
|
| 18315 |
+
"loss": 1.2178,
|
| 18316 |
+
"step": 25570
|
| 18317 |
+
},
|
| 18318 |
+
{
|
| 18319 |
+
"epoch": 0.64,
|
| 18320 |
+
"grad_norm": 7.605580806732178,
|
| 18321 |
+
"learning_rate": 2.9966101694915258e-06,
|
| 18322 |
+
"loss": 1.3286,
|
| 18323 |
+
"step": 25580
|
| 18324 |
+
},
|
| 18325 |
+
{
|
| 18326 |
+
"epoch": 0.64,
|
| 18327 |
+
"grad_norm": 4.82462215423584,
|
| 18328 |
+
"learning_rate": 2.9898305084745768e-06,
|
| 18329 |
+
"loss": 1.4727,
|
| 18330 |
+
"step": 25590
|
| 18331 |
+
},
|
| 18332 |
+
{
|
| 18333 |
+
"epoch": 0.64,
|
| 18334 |
+
"grad_norm": 4.511000633239746,
|
| 18335 |
+
"learning_rate": 2.9830508474576277e-06,
|
| 18336 |
+
"loss": 1.3572,
|
| 18337 |
+
"step": 25600
|
| 18338 |
+
},
|
| 18339 |
+
{
|
| 18340 |
+
"epoch": 0.64,
|
| 18341 |
+
"grad_norm": 6.090604305267334,
|
| 18342 |
+
"learning_rate": 2.9762711864406783e-06,
|
| 18343 |
+
"loss": 1.4802,
|
| 18344 |
+
"step": 25610
|
| 18345 |
+
},
|
| 18346 |
+
{
|
| 18347 |
+
"epoch": 0.64,
|
| 18348 |
+
"grad_norm": 9.687544822692871,
|
| 18349 |
+
"learning_rate": 2.969491525423729e-06,
|
| 18350 |
+
"loss": 1.3652,
|
| 18351 |
+
"step": 25620
|
| 18352 |
+
},
|
| 18353 |
+
{
|
| 18354 |
+
"epoch": 0.64,
|
| 18355 |
+
"grad_norm": 12.760902404785156,
|
| 18356 |
+
"learning_rate": 2.96271186440678e-06,
|
| 18357 |
+
"loss": 1.3634,
|
| 18358 |
+
"step": 25630
|
| 18359 |
+
},
|
| 18360 |
+
{
|
| 18361 |
+
"epoch": 0.64,
|
| 18362 |
+
"grad_norm": 6.688235759735107,
|
| 18363 |
+
"learning_rate": 2.955932203389831e-06,
|
| 18364 |
+
"loss": 1.209,
|
| 18365 |
+
"step": 25640
|
| 18366 |
+
},
|
| 18367 |
+
{
|
| 18368 |
+
"epoch": 0.64,
|
| 18369 |
+
"grad_norm": 5.238013744354248,
|
| 18370 |
+
"learning_rate": 2.9491525423728818e-06,
|
| 18371 |
+
"loss": 1.5462,
|
| 18372 |
+
"step": 25650
|
| 18373 |
+
},
|
| 18374 |
+
{
|
| 18375 |
+
"epoch": 0.64,
|
| 18376 |
+
"grad_norm": 18.444650650024414,
|
| 18377 |
+
"learning_rate": 2.9423728813559327e-06,
|
| 18378 |
+
"loss": 1.2963,
|
| 18379 |
+
"step": 25660
|
| 18380 |
+
},
|
| 18381 |
+
{
|
| 18382 |
+
"epoch": 0.64,
|
| 18383 |
+
"grad_norm": 5.625192642211914,
|
| 18384 |
+
"learning_rate": 2.935593220338983e-06,
|
| 18385 |
+
"loss": 1.3514,
|
| 18386 |
+
"step": 25670
|
| 18387 |
+
},
|
| 18388 |
+
{
|
| 18389 |
+
"epoch": 0.64,
|
| 18390 |
+
"grad_norm": 5.223139762878418,
|
| 18391 |
+
"learning_rate": 2.928813559322034e-06,
|
| 18392 |
+
"loss": 1.2339,
|
| 18393 |
+
"step": 25680
|
| 18394 |
+
},
|
| 18395 |
+
{
|
| 18396 |
+
"epoch": 0.64,
|
| 18397 |
+
"grad_norm": 4.5006422996521,
|
| 18398 |
+
"learning_rate": 2.922033898305085e-06,
|
| 18399 |
+
"loss": 1.4424,
|
| 18400 |
+
"step": 25690
|
| 18401 |
+
},
|
| 18402 |
+
{
|
| 18403 |
+
"epoch": 0.64,
|
| 18404 |
+
"grad_norm": 9.127293586730957,
|
| 18405 |
+
"learning_rate": 2.915254237288136e-06,
|
| 18406 |
+
"loss": 1.2334,
|
| 18407 |
+
"step": 25700
|
| 18408 |
+
},
|
| 18409 |
+
{
|
| 18410 |
+
"epoch": 0.64,
|
| 18411 |
+
"grad_norm": 5.78049373626709,
|
| 18412 |
+
"learning_rate": 2.9084745762711868e-06,
|
| 18413 |
+
"loss": 1.301,
|
| 18414 |
+
"step": 25710
|
| 18415 |
+
},
|
| 18416 |
+
{
|
| 18417 |
+
"epoch": 0.64,
|
| 18418 |
+
"grad_norm": 5.395204544067383,
|
| 18419 |
+
"learning_rate": 2.9016949152542373e-06,
|
| 18420 |
+
"loss": 1.1889,
|
| 18421 |
+
"step": 25720
|
| 18422 |
+
},
|
| 18423 |
+
{
|
| 18424 |
+
"epoch": 0.64,
|
| 18425 |
+
"grad_norm": 2.3502519130706787,
|
| 18426 |
+
"learning_rate": 2.8949152542372883e-06,
|
| 18427 |
+
"loss": 1.41,
|
| 18428 |
+
"step": 25730
|
| 18429 |
+
},
|
| 18430 |
+
{
|
| 18431 |
+
"epoch": 0.64,
|
| 18432 |
+
"grad_norm": 3.3766727447509766,
|
| 18433 |
+
"learning_rate": 2.8881355932203393e-06,
|
| 18434 |
+
"loss": 1.2768,
|
| 18435 |
+
"step": 25740
|
| 18436 |
+
},
|
| 18437 |
+
{
|
| 18438 |
+
"epoch": 0.64,
|
| 18439 |
+
"grad_norm": 6.30269718170166,
|
| 18440 |
+
"learning_rate": 2.8813559322033903e-06,
|
| 18441 |
+
"loss": 1.3432,
|
| 18442 |
+
"step": 25750
|
| 18443 |
+
},
|
| 18444 |
+
{
|
| 18445 |
+
"epoch": 0.64,
|
| 18446 |
+
"grad_norm": 7.727334022521973,
|
| 18447 |
+
"learning_rate": 2.8745762711864412e-06,
|
| 18448 |
+
"loss": 1.3465,
|
| 18449 |
+
"step": 25760
|
| 18450 |
+
},
|
| 18451 |
+
{
|
| 18452 |
+
"epoch": 0.64,
|
| 18453 |
+
"grad_norm": 8.983928680419922,
|
| 18454 |
+
"learning_rate": 2.8677966101694914e-06,
|
| 18455 |
+
"loss": 1.1202,
|
| 18456 |
+
"step": 25770
|
| 18457 |
+
},
|
| 18458 |
+
{
|
| 18459 |
+
"epoch": 0.64,
|
| 18460 |
+
"grad_norm": 4.910791397094727,
|
| 18461 |
+
"learning_rate": 2.8610169491525424e-06,
|
| 18462 |
+
"loss": 1.2038,
|
| 18463 |
+
"step": 25780
|
| 18464 |
+
},
|
| 18465 |
+
{
|
| 18466 |
+
"epoch": 0.64,
|
| 18467 |
+
"grad_norm": 12.079176902770996,
|
| 18468 |
+
"learning_rate": 2.8542372881355933e-06,
|
| 18469 |
+
"loss": 1.1599,
|
| 18470 |
+
"step": 25790
|
| 18471 |
+
},
|
| 18472 |
+
{
|
| 18473 |
+
"epoch": 0.65,
|
| 18474 |
+
"grad_norm": 3.6942214965820312,
|
| 18475 |
+
"learning_rate": 2.8474576271186443e-06,
|
| 18476 |
+
"loss": 1.2242,
|
| 18477 |
+
"step": 25800
|
| 18478 |
+
},
|
| 18479 |
+
{
|
| 18480 |
+
"epoch": 0.65,
|
| 18481 |
+
"grad_norm": 2.4002459049224854,
|
| 18482 |
+
"learning_rate": 2.8406779661016953e-06,
|
| 18483 |
+
"loss": 1.2288,
|
| 18484 |
+
"step": 25810
|
| 18485 |
+
},
|
| 18486 |
+
{
|
| 18487 |
+
"epoch": 0.65,
|
| 18488 |
+
"grad_norm": 4.8168840408325195,
|
| 18489 |
+
"learning_rate": 2.833898305084746e-06,
|
| 18490 |
+
"loss": 1.4489,
|
| 18491 |
+
"step": 25820
|
| 18492 |
+
},
|
| 18493 |
+
{
|
| 18494 |
+
"epoch": 0.65,
|
| 18495 |
+
"grad_norm": 10.90778636932373,
|
| 18496 |
+
"learning_rate": 2.827118644067797e-06,
|
| 18497 |
+
"loss": 1.4108,
|
| 18498 |
+
"step": 25830
|
| 18499 |
+
},
|
| 18500 |
+
{
|
| 18501 |
+
"epoch": 0.65,
|
| 18502 |
+
"grad_norm": 9.852614402770996,
|
| 18503 |
+
"learning_rate": 2.820338983050848e-06,
|
| 18504 |
+
"loss": 1.3685,
|
| 18505 |
+
"step": 25840
|
| 18506 |
+
},
|
| 18507 |
+
{
|
| 18508 |
+
"epoch": 0.65,
|
| 18509 |
+
"grad_norm": 8.391303062438965,
|
| 18510 |
+
"learning_rate": 2.8135593220338988e-06,
|
| 18511 |
+
"loss": 1.2921,
|
| 18512 |
+
"step": 25850
|
| 18513 |
+
},
|
| 18514 |
+
{
|
| 18515 |
+
"epoch": 0.65,
|
| 18516 |
+
"grad_norm": 5.348249912261963,
|
| 18517 |
+
"learning_rate": 2.8067796610169497e-06,
|
| 18518 |
+
"loss": 1.3463,
|
| 18519 |
+
"step": 25860
|
| 18520 |
+
},
|
| 18521 |
+
{
|
| 18522 |
+
"epoch": 0.65,
|
| 18523 |
+
"grad_norm": 3.3569209575653076,
|
| 18524 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 18525 |
+
"loss": 1.2877,
|
| 18526 |
+
"step": 25870
|
| 18527 |
+
},
|
| 18528 |
+
{
|
| 18529 |
+
"epoch": 0.65,
|
| 18530 |
+
"grad_norm": 9.246468544006348,
|
| 18531 |
+
"learning_rate": 2.793220338983051e-06,
|
| 18532 |
+
"loss": 1.3478,
|
| 18533 |
+
"step": 25880
|
| 18534 |
+
},
|
| 18535 |
+
{
|
| 18536 |
+
"epoch": 0.65,
|
| 18537 |
+
"grad_norm": 7.3121771812438965,
|
| 18538 |
+
"learning_rate": 2.786440677966102e-06,
|
| 18539 |
+
"loss": 1.3116,
|
| 18540 |
+
"step": 25890
|
| 18541 |
+
},
|
| 18542 |
+
{
|
| 18543 |
+
"epoch": 0.65,
|
| 18544 |
+
"grad_norm": 7.324717044830322,
|
| 18545 |
+
"learning_rate": 2.779661016949153e-06,
|
| 18546 |
+
"loss": 1.4986,
|
| 18547 |
+
"step": 25900
|
| 18548 |
+
},
|
| 18549 |
+
{
|
| 18550 |
+
"epoch": 0.65,
|
| 18551 |
+
"grad_norm": 6.7373738288879395,
|
| 18552 |
+
"learning_rate": 2.7728813559322038e-06,
|
| 18553 |
+
"loss": 1.5417,
|
| 18554 |
+
"step": 25910
|
| 18555 |
+
},
|
| 18556 |
+
{
|
| 18557 |
+
"epoch": 0.65,
|
| 18558 |
+
"grad_norm": 10.05711555480957,
|
| 18559 |
+
"learning_rate": 2.7661016949152548e-06,
|
| 18560 |
+
"loss": 1.3439,
|
| 18561 |
+
"step": 25920
|
| 18562 |
+
},
|
| 18563 |
+
{
|
| 18564 |
+
"epoch": 0.65,
|
| 18565 |
+
"grad_norm": 3.843493700027466,
|
| 18566 |
+
"learning_rate": 2.7593220338983053e-06,
|
| 18567 |
+
"loss": 1.1626,
|
| 18568 |
+
"step": 25930
|
| 18569 |
+
},
|
| 18570 |
+
{
|
| 18571 |
+
"epoch": 0.65,
|
| 18572 |
+
"grad_norm": 9.513094902038574,
|
| 18573 |
+
"learning_rate": 2.752542372881356e-06,
|
| 18574 |
+
"loss": 1.3412,
|
| 18575 |
+
"step": 25940
|
| 18576 |
+
},
|
| 18577 |
+
{
|
| 18578 |
+
"epoch": 0.65,
|
| 18579 |
+
"grad_norm": 8.064391136169434,
|
| 18580 |
+
"learning_rate": 2.745762711864407e-06,
|
| 18581 |
+
"loss": 1.3483,
|
| 18582 |
+
"step": 25950
|
| 18583 |
+
},
|
| 18584 |
+
{
|
| 18585 |
+
"epoch": 0.65,
|
| 18586 |
+
"grad_norm": 10.491897583007812,
|
| 18587 |
+
"learning_rate": 2.738983050847458e-06,
|
| 18588 |
+
"loss": 1.3214,
|
| 18589 |
+
"step": 25960
|
| 18590 |
+
},
|
| 18591 |
+
{
|
| 18592 |
+
"epoch": 0.65,
|
| 18593 |
+
"grad_norm": 3.9402971267700195,
|
| 18594 |
+
"learning_rate": 2.732203389830509e-06,
|
| 18595 |
+
"loss": 1.3803,
|
| 18596 |
+
"step": 25970
|
| 18597 |
+
},
|
| 18598 |
+
{
|
| 18599 |
+
"epoch": 0.65,
|
| 18600 |
+
"grad_norm": 9.979386329650879,
|
| 18601 |
+
"learning_rate": 2.7254237288135593e-06,
|
| 18602 |
+
"loss": 1.1408,
|
| 18603 |
+
"step": 25980
|
| 18604 |
+
},
|
| 18605 |
+
{
|
| 18606 |
+
"epoch": 0.65,
|
| 18607 |
+
"grad_norm": 7.461246490478516,
|
| 18608 |
+
"learning_rate": 2.7186440677966103e-06,
|
| 18609 |
+
"loss": 1.3098,
|
| 18610 |
+
"step": 25990
|
| 18611 |
+
},
|
| 18612 |
+
{
|
| 18613 |
+
"epoch": 0.65,
|
| 18614 |
+
"grad_norm": 4.412223815917969,
|
| 18615 |
+
"learning_rate": 2.7118644067796613e-06,
|
| 18616 |
+
"loss": 1.457,
|
| 18617 |
+
"step": 26000
|
| 18618 |
+
},
|
| 18619 |
+
{
|
| 18620 |
+
"epoch": 0.65,
|
| 18621 |
+
"eval_loss": 1.3418877124786377,
|
| 18622 |
+
"eval_runtime": 66.1458,
|
| 18623 |
+
"eval_samples_per_second": 15.118,
|
| 18624 |
+
"eval_steps_per_second": 15.118,
|
| 18625 |
+
"step": 26000
|
| 18626 |
+
},
|
| 18627 |
+
{
|
| 18628 |
+
"epoch": 0.65,
|
| 18629 |
+
"grad_norm": 4.992754936218262,
|
| 18630 |
+
"learning_rate": 2.7050847457627123e-06,
|
| 18631 |
+
"loss": 1.1937,
|
| 18632 |
+
"step": 26010
|
| 18633 |
+
},
|
| 18634 |
+
{
|
| 18635 |
+
"epoch": 0.65,
|
| 18636 |
+
"grad_norm": 5.497439861297607,
|
| 18637 |
+
"learning_rate": 2.6983050847457633e-06,
|
| 18638 |
+
"loss": 1.3435,
|
| 18639 |
+
"step": 26020
|
| 18640 |
+
},
|
| 18641 |
+
{
|
| 18642 |
+
"epoch": 0.65,
|
| 18643 |
+
"grad_norm": 4.283881664276123,
|
| 18644 |
+
"learning_rate": 2.6915254237288134e-06,
|
| 18645 |
+
"loss": 1.3997,
|
| 18646 |
+
"step": 26030
|
| 18647 |
+
},
|
| 18648 |
+
{
|
| 18649 |
+
"epoch": 0.65,
|
| 18650 |
+
"grad_norm": 6.579436779022217,
|
| 18651 |
+
"learning_rate": 2.6847457627118644e-06,
|
| 18652 |
+
"loss": 1.4047,
|
| 18653 |
+
"step": 26040
|
| 18654 |
+
},
|
| 18655 |
+
{
|
| 18656 |
+
"epoch": 0.65,
|
| 18657 |
+
"grad_norm": 4.805014133453369,
|
| 18658 |
+
"learning_rate": 2.6779661016949153e-06,
|
| 18659 |
+
"loss": 1.4021,
|
| 18660 |
+
"step": 26050
|
| 18661 |
+
},
|
| 18662 |
+
{
|
| 18663 |
+
"epoch": 0.65,
|
| 18664 |
+
"grad_norm": 4.568453788757324,
|
| 18665 |
+
"learning_rate": 2.6711864406779663e-06,
|
| 18666 |
+
"loss": 1.4761,
|
| 18667 |
+
"step": 26060
|
| 18668 |
+
},
|
| 18669 |
+
{
|
| 18670 |
+
"epoch": 0.65,
|
| 18671 |
+
"grad_norm": 5.559253692626953,
|
| 18672 |
+
"learning_rate": 2.6644067796610173e-06,
|
| 18673 |
+
"loss": 1.2958,
|
| 18674 |
+
"step": 26070
|
| 18675 |
+
},
|
| 18676 |
+
{
|
| 18677 |
+
"epoch": 0.65,
|
| 18678 |
+
"grad_norm": 14.279343605041504,
|
| 18679 |
+
"learning_rate": 2.657627118644068e-06,
|
| 18680 |
+
"loss": 1.194,
|
| 18681 |
+
"step": 26080
|
| 18682 |
+
},
|
| 18683 |
+
{
|
| 18684 |
+
"epoch": 0.65,
|
| 18685 |
+
"grad_norm": 15.45864486694336,
|
| 18686 |
+
"learning_rate": 2.650847457627119e-06,
|
| 18687 |
+
"loss": 1.2134,
|
| 18688 |
+
"step": 26090
|
| 18689 |
+
},
|
| 18690 |
+
{
|
| 18691 |
+
"epoch": 0.65,
|
| 18692 |
+
"grad_norm": 17.310068130493164,
|
| 18693 |
+
"learning_rate": 2.64406779661017e-06,
|
| 18694 |
+
"loss": 1.2946,
|
| 18695 |
+
"step": 26100
|
| 18696 |
+
},
|
| 18697 |
+
{
|
| 18698 |
+
"epoch": 0.65,
|
| 18699 |
+
"grad_norm": 4.31659460067749,
|
| 18700 |
+
"learning_rate": 2.6372881355932208e-06,
|
| 18701 |
+
"loss": 1.3783,
|
| 18702 |
+
"step": 26110
|
| 18703 |
+
},
|
| 18704 |
+
{
|
| 18705 |
+
"epoch": 0.65,
|
| 18706 |
+
"grad_norm": 14.684484481811523,
|
| 18707 |
+
"learning_rate": 2.6305084745762718e-06,
|
| 18708 |
+
"loss": 1.3414,
|
| 18709 |
+
"step": 26120
|
| 18710 |
+
},
|
| 18711 |
+
{
|
| 18712 |
+
"epoch": 0.65,
|
| 18713 |
+
"grad_norm": 5.934599876403809,
|
| 18714 |
+
"learning_rate": 2.6237288135593223e-06,
|
| 18715 |
+
"loss": 1.3266,
|
| 18716 |
+
"step": 26130
|
| 18717 |
+
},
|
| 18718 |
+
{
|
| 18719 |
+
"epoch": 0.65,
|
| 18720 |
+
"grad_norm": 7.535737037658691,
|
| 18721 |
+
"learning_rate": 2.616949152542373e-06,
|
| 18722 |
+
"loss": 1.2579,
|
| 18723 |
+
"step": 26140
|
| 18724 |
+
},
|
| 18725 |
+
{
|
| 18726 |
+
"epoch": 0.65,
|
| 18727 |
+
"grad_norm": 5.8215556144714355,
|
| 18728 |
+
"learning_rate": 2.610169491525424e-06,
|
| 18729 |
+
"loss": 1.304,
|
| 18730 |
+
"step": 26150
|
| 18731 |
+
},
|
| 18732 |
+
{
|
| 18733 |
+
"epoch": 0.65,
|
| 18734 |
+
"grad_norm": 5.646538734436035,
|
| 18735 |
+
"learning_rate": 2.603389830508475e-06,
|
| 18736 |
+
"loss": 1.2792,
|
| 18737 |
+
"step": 26160
|
| 18738 |
+
},
|
| 18739 |
+
{
|
| 18740 |
+
"epoch": 0.65,
|
| 18741 |
+
"grad_norm": 3.834282398223877,
|
| 18742 |
+
"learning_rate": 2.596610169491526e-06,
|
| 18743 |
+
"loss": 1.4717,
|
| 18744 |
+
"step": 26170
|
| 18745 |
+
},
|
| 18746 |
+
{
|
| 18747 |
+
"epoch": 0.65,
|
| 18748 |
+
"grad_norm": 3.313835620880127,
|
| 18749 |
+
"learning_rate": 2.5898305084745768e-06,
|
| 18750 |
+
"loss": 1.3312,
|
| 18751 |
+
"step": 26180
|
| 18752 |
+
},
|
| 18753 |
+
{
|
| 18754 |
+
"epoch": 0.65,
|
| 18755 |
+
"grad_norm": 20.453479766845703,
|
| 18756 |
+
"learning_rate": 2.5830508474576273e-06,
|
| 18757 |
+
"loss": 1.3784,
|
| 18758 |
+
"step": 26190
|
| 18759 |
+
},
|
| 18760 |
+
{
|
| 18761 |
+
"epoch": 0.66,
|
| 18762 |
+
"grad_norm": 1.0731091499328613,
|
| 18763 |
+
"learning_rate": 2.576271186440678e-06,
|
| 18764 |
+
"loss": 1.4162,
|
| 18765 |
+
"step": 26200
|
| 18766 |
+
},
|
| 18767 |
+
{
|
| 18768 |
+
"epoch": 0.66,
|
| 18769 |
+
"grad_norm": 7.5899577140808105,
|
| 18770 |
+
"learning_rate": 2.569491525423729e-06,
|
| 18771 |
+
"loss": 1.2536,
|
| 18772 |
+
"step": 26210
|
| 18773 |
+
},
|
| 18774 |
+
{
|
| 18775 |
+
"epoch": 0.66,
|
| 18776 |
+
"grad_norm": 4.527303218841553,
|
| 18777 |
+
"learning_rate": 2.56271186440678e-06,
|
| 18778 |
+
"loss": 1.5048,
|
| 18779 |
+
"step": 26220
|
| 18780 |
+
},
|
| 18781 |
+
{
|
| 18782 |
+
"epoch": 0.66,
|
| 18783 |
+
"grad_norm": 3.702897787094116,
|
| 18784 |
+
"learning_rate": 2.555932203389831e-06,
|
| 18785 |
+
"loss": 1.1666,
|
| 18786 |
+
"step": 26230
|
| 18787 |
+
},
|
| 18788 |
+
{
|
| 18789 |
+
"epoch": 0.66,
|
| 18790 |
+
"grad_norm": 3.448979139328003,
|
| 18791 |
+
"learning_rate": 2.5491525423728814e-06,
|
| 18792 |
+
"loss": 1.2812,
|
| 18793 |
+
"step": 26240
|
| 18794 |
+
},
|
| 18795 |
+
{
|
| 18796 |
+
"epoch": 0.66,
|
| 18797 |
+
"grad_norm": 15.188081741333008,
|
| 18798 |
+
"learning_rate": 2.5423728813559323e-06,
|
| 18799 |
+
"loss": 1.3521,
|
| 18800 |
+
"step": 26250
|
| 18801 |
+
},
|
| 18802 |
+
{
|
| 18803 |
+
"epoch": 0.66,
|
| 18804 |
+
"grad_norm": 8.675294876098633,
|
| 18805 |
+
"learning_rate": 2.5355932203389833e-06,
|
| 18806 |
+
"loss": 1.4845,
|
| 18807 |
+
"step": 26260
|
| 18808 |
+
},
|
| 18809 |
+
{
|
| 18810 |
+
"epoch": 0.66,
|
| 18811 |
+
"grad_norm": 5.673226356506348,
|
| 18812 |
+
"learning_rate": 2.5288135593220343e-06,
|
| 18813 |
+
"loss": 1.2264,
|
| 18814 |
+
"step": 26270
|
| 18815 |
+
},
|
| 18816 |
+
{
|
| 18817 |
+
"epoch": 0.66,
|
| 18818 |
+
"grad_norm": 6.015854358673096,
|
| 18819 |
+
"learning_rate": 2.5220338983050853e-06,
|
| 18820 |
+
"loss": 1.3036,
|
| 18821 |
+
"step": 26280
|
| 18822 |
+
},
|
| 18823 |
+
{
|
| 18824 |
+
"epoch": 0.66,
|
| 18825 |
+
"grad_norm": 12.492931365966797,
|
| 18826 |
+
"learning_rate": 2.5152542372881354e-06,
|
| 18827 |
+
"loss": 1.4434,
|
| 18828 |
+
"step": 26290
|
| 18829 |
+
},
|
| 18830 |
+
{
|
| 18831 |
+
"epoch": 0.66,
|
| 18832 |
+
"grad_norm": 5.928922176361084,
|
| 18833 |
+
"learning_rate": 2.5084745762711864e-06,
|
| 18834 |
+
"loss": 1.3118,
|
| 18835 |
+
"step": 26300
|
| 18836 |
+
},
|
| 18837 |
+
{
|
| 18838 |
+
"epoch": 0.66,
|
| 18839 |
+
"grad_norm": 5.230995178222656,
|
| 18840 |
+
"learning_rate": 2.5016949152542374e-06,
|
| 18841 |
+
"loss": 1.3524,
|
| 18842 |
+
"step": 26310
|
| 18843 |
+
},
|
| 18844 |
+
{
|
| 18845 |
+
"epoch": 0.66,
|
| 18846 |
+
"grad_norm": 6.7296671867370605,
|
| 18847 |
+
"learning_rate": 2.4949152542372883e-06,
|
| 18848 |
+
"loss": 1.3651,
|
| 18849 |
+
"step": 26320
|
| 18850 |
+
},
|
| 18851 |
+
{
|
| 18852 |
+
"epoch": 0.66,
|
| 18853 |
+
"grad_norm": 2.378596544265747,
|
| 18854 |
+
"learning_rate": 2.488135593220339e-06,
|
| 18855 |
+
"loss": 1.4236,
|
| 18856 |
+
"step": 26330
|
| 18857 |
+
},
|
| 18858 |
+
{
|
| 18859 |
+
"epoch": 0.66,
|
| 18860 |
+
"grad_norm": 5.858238697052002,
|
| 18861 |
+
"learning_rate": 2.48135593220339e-06,
|
| 18862 |
+
"loss": 1.1553,
|
| 18863 |
+
"step": 26340
|
| 18864 |
+
},
|
| 18865 |
+
{
|
| 18866 |
+
"epoch": 0.66,
|
| 18867 |
+
"grad_norm": 3.7590560913085938,
|
| 18868 |
+
"learning_rate": 2.474576271186441e-06,
|
| 18869 |
+
"loss": 1.4302,
|
| 18870 |
+
"step": 26350
|
| 18871 |
+
},
|
| 18872 |
+
{
|
| 18873 |
+
"epoch": 0.66,
|
| 18874 |
+
"grad_norm": 10.852778434753418,
|
| 18875 |
+
"learning_rate": 2.467796610169492e-06,
|
| 18876 |
+
"loss": 1.1943,
|
| 18877 |
+
"step": 26360
|
| 18878 |
+
},
|
| 18879 |
+
{
|
| 18880 |
+
"epoch": 0.66,
|
| 18881 |
+
"grad_norm": 11.347558975219727,
|
| 18882 |
+
"learning_rate": 2.461016949152543e-06,
|
| 18883 |
+
"loss": 1.2559,
|
| 18884 |
+
"step": 26370
|
| 18885 |
+
},
|
| 18886 |
+
{
|
| 18887 |
+
"epoch": 0.66,
|
| 18888 |
+
"grad_norm": 6.081493854522705,
|
| 18889 |
+
"learning_rate": 2.4542372881355933e-06,
|
| 18890 |
+
"loss": 1.3532,
|
| 18891 |
+
"step": 26380
|
| 18892 |
+
},
|
| 18893 |
+
{
|
| 18894 |
+
"epoch": 0.66,
|
| 18895 |
+
"grad_norm": 8.263628959655762,
|
| 18896 |
+
"learning_rate": 2.4474576271186443e-06,
|
| 18897 |
+
"loss": 1.2041,
|
| 18898 |
+
"step": 26390
|
| 18899 |
+
},
|
| 18900 |
+
{
|
| 18901 |
+
"epoch": 0.66,
|
| 18902 |
+
"grad_norm": 7.144092082977295,
|
| 18903 |
+
"learning_rate": 2.4406779661016953e-06,
|
| 18904 |
+
"loss": 1.2193,
|
| 18905 |
+
"step": 26400
|
| 18906 |
+
},
|
| 18907 |
+
{
|
| 18908 |
+
"epoch": 0.66,
|
| 18909 |
+
"grad_norm": 6.578696250915527,
|
| 18910 |
+
"learning_rate": 2.433898305084746e-06,
|
| 18911 |
+
"loss": 1.3139,
|
| 18912 |
+
"step": 26410
|
| 18913 |
+
},
|
| 18914 |
+
{
|
| 18915 |
+
"epoch": 0.66,
|
| 18916 |
+
"grad_norm": 3.019033193588257,
|
| 18917 |
+
"learning_rate": 2.427118644067797e-06,
|
| 18918 |
+
"loss": 1.4642,
|
| 18919 |
+
"step": 26420
|
| 18920 |
+
},
|
| 18921 |
+
{
|
| 18922 |
+
"epoch": 0.66,
|
| 18923 |
+
"grad_norm": 5.443304538726807,
|
| 18924 |
+
"learning_rate": 2.4203389830508474e-06,
|
| 18925 |
+
"loss": 1.3738,
|
| 18926 |
+
"step": 26430
|
| 18927 |
+
},
|
| 18928 |
+
{
|
| 18929 |
+
"epoch": 0.66,
|
| 18930 |
+
"grad_norm": 10.051207542419434,
|
| 18931 |
+
"learning_rate": 2.4135593220338984e-06,
|
| 18932 |
+
"loss": 1.4396,
|
| 18933 |
+
"step": 26440
|
| 18934 |
+
},
|
| 18935 |
+
{
|
| 18936 |
+
"epoch": 0.66,
|
| 18937 |
+
"grad_norm": 10.604415893554688,
|
| 18938 |
+
"learning_rate": 2.4067796610169493e-06,
|
| 18939 |
+
"loss": 1.2971,
|
| 18940 |
+
"step": 26450
|
| 18941 |
+
},
|
| 18942 |
+
{
|
| 18943 |
+
"epoch": 0.66,
|
| 18944 |
+
"grad_norm": 9.754619598388672,
|
| 18945 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 18946 |
+
"loss": 1.3421,
|
| 18947 |
+
"step": 26460
|
| 18948 |
+
},
|
| 18949 |
+
{
|
| 18950 |
+
"epoch": 0.66,
|
| 18951 |
+
"grad_norm": 6.159849166870117,
|
| 18952 |
+
"learning_rate": 2.393220338983051e-06,
|
| 18953 |
+
"loss": 1.3461,
|
| 18954 |
+
"step": 26470
|
| 18955 |
+
},
|
| 18956 |
+
{
|
| 18957 |
+
"epoch": 0.66,
|
| 18958 |
+
"grad_norm": 1.3210265636444092,
|
| 18959 |
+
"learning_rate": 2.386440677966102e-06,
|
| 18960 |
+
"loss": 1.2615,
|
| 18961 |
+
"step": 26480
|
| 18962 |
+
},
|
| 18963 |
+
{
|
| 18964 |
+
"epoch": 0.66,
|
| 18965 |
+
"grad_norm": 5.67736291885376,
|
| 18966 |
+
"learning_rate": 2.379661016949153e-06,
|
| 18967 |
+
"loss": 1.3296,
|
| 18968 |
+
"step": 26490
|
| 18969 |
+
},
|
| 18970 |
+
{
|
| 18971 |
+
"epoch": 0.66,
|
| 18972 |
+
"grad_norm": 4.055379867553711,
|
| 18973 |
+
"learning_rate": 2.372881355932204e-06,
|
| 18974 |
+
"loss": 1.4705,
|
| 18975 |
+
"step": 26500
|
| 18976 |
+
},
|
| 18977 |
+
{
|
| 18978 |
+
"epoch": 0.66,
|
| 18979 |
+
"eval_loss": 1.2714667320251465,
|
| 18980 |
+
"eval_runtime": 66.1328,
|
| 18981 |
+
"eval_samples_per_second": 15.121,
|
| 18982 |
+
"eval_steps_per_second": 15.121,
|
| 18983 |
+
"step": 26500
|
| 18984 |
+
},
|
| 18985 |
+
{
|
| 18986 |
+
"epoch": 0.66,
|
| 18987 |
+
"grad_norm": 7.19535493850708,
|
| 18988 |
+
"learning_rate": 2.3661016949152544e-06,
|
| 18989 |
+
"loss": 1.2956,
|
| 18990 |
+
"step": 26510
|
| 18991 |
+
},
|
| 18992 |
+
{
|
| 18993 |
+
"epoch": 0.66,
|
| 18994 |
+
"grad_norm": 7.952252388000488,
|
| 18995 |
+
"learning_rate": 2.3593220338983053e-06,
|
| 18996 |
+
"loss": 1.2952,
|
| 18997 |
+
"step": 26520
|
| 18998 |
+
},
|
| 18999 |
+
{
|
| 19000 |
+
"epoch": 0.66,
|
| 19001 |
+
"grad_norm": 1.3878239393234253,
|
| 19002 |
+
"learning_rate": 2.3525423728813563e-06,
|
| 19003 |
+
"loss": 1.3957,
|
| 19004 |
+
"step": 26530
|
| 19005 |
+
},
|
| 19006 |
+
{
|
| 19007 |
+
"epoch": 0.66,
|
| 19008 |
+
"grad_norm": 9.453435897827148,
|
| 19009 |
+
"learning_rate": 2.345762711864407e-06,
|
| 19010 |
+
"loss": 1.3314,
|
| 19011 |
+
"step": 26540
|
| 19012 |
+
},
|
| 19013 |
+
{
|
| 19014 |
+
"epoch": 0.66,
|
| 19015 |
+
"grad_norm": 2.3859434127807617,
|
| 19016 |
+
"learning_rate": 2.338983050847458e-06,
|
| 19017 |
+
"loss": 1.3959,
|
| 19018 |
+
"step": 26550
|
| 19019 |
+
},
|
| 19020 |
+
{
|
| 19021 |
+
"epoch": 0.66,
|
| 19022 |
+
"grad_norm": 4.052861213684082,
|
| 19023 |
+
"learning_rate": 2.3322033898305084e-06,
|
| 19024 |
+
"loss": 1.3097,
|
| 19025 |
+
"step": 26560
|
| 19026 |
+
},
|
| 19027 |
+
{
|
| 19028 |
+
"epoch": 0.66,
|
| 19029 |
+
"grad_norm": 14.907791137695312,
|
| 19030 |
+
"learning_rate": 2.3254237288135594e-06,
|
| 19031 |
+
"loss": 1.3706,
|
| 19032 |
+
"step": 26570
|
| 19033 |
+
},
|
| 19034 |
+
{
|
| 19035 |
+
"epoch": 0.66,
|
| 19036 |
+
"grad_norm": 7.020768642425537,
|
| 19037 |
+
"learning_rate": 2.3186440677966103e-06,
|
| 19038 |
+
"loss": 1.3536,
|
| 19039 |
+
"step": 26580
|
| 19040 |
+
},
|
| 19041 |
+
{
|
| 19042 |
+
"epoch": 0.66,
|
| 19043 |
+
"grad_norm": 15.225879669189453,
|
| 19044 |
+
"learning_rate": 2.3118644067796613e-06,
|
| 19045 |
+
"loss": 1.354,
|
| 19046 |
+
"step": 26590
|
| 19047 |
+
},
|
| 19048 |
+
{
|
| 19049 |
+
"epoch": 0.67,
|
| 19050 |
+
"grad_norm": 6.249582767486572,
|
| 19051 |
+
"learning_rate": 2.305084745762712e-06,
|
| 19052 |
+
"loss": 1.2721,
|
| 19053 |
+
"step": 26600
|
| 19054 |
+
},
|
| 19055 |
+
{
|
| 19056 |
+
"epoch": 0.67,
|
| 19057 |
+
"grad_norm": 4.756400108337402,
|
| 19058 |
+
"learning_rate": 2.298305084745763e-06,
|
| 19059 |
+
"loss": 1.3044,
|
| 19060 |
+
"step": 26610
|
| 19061 |
+
},
|
| 19062 |
+
{
|
| 19063 |
+
"epoch": 0.67,
|
| 19064 |
+
"grad_norm": 9.848575592041016,
|
| 19065 |
+
"learning_rate": 2.291525423728814e-06,
|
| 19066 |
+
"loss": 1.2471,
|
| 19067 |
+
"step": 26620
|
| 19068 |
+
},
|
| 19069 |
+
{
|
| 19070 |
+
"epoch": 0.67,
|
| 19071 |
+
"grad_norm": 9.997851371765137,
|
| 19072 |
+
"learning_rate": 2.284745762711865e-06,
|
| 19073 |
+
"loss": 1.1673,
|
| 19074 |
+
"step": 26630
|
| 19075 |
+
},
|
| 19076 |
+
{
|
| 19077 |
+
"epoch": 0.67,
|
| 19078 |
+
"grad_norm": 11.57490062713623,
|
| 19079 |
+
"learning_rate": 2.2779661016949154e-06,
|
| 19080 |
+
"loss": 1.3241,
|
| 19081 |
+
"step": 26640
|
| 19082 |
+
},
|
| 19083 |
+
{
|
| 19084 |
+
"epoch": 0.67,
|
| 19085 |
+
"grad_norm": 8.282011032104492,
|
| 19086 |
+
"learning_rate": 2.2711864406779663e-06,
|
| 19087 |
+
"loss": 1.2821,
|
| 19088 |
+
"step": 26650
|
| 19089 |
+
},
|
| 19090 |
+
{
|
| 19091 |
+
"epoch": 0.67,
|
| 19092 |
+
"grad_norm": 6.8680644035339355,
|
| 19093 |
+
"learning_rate": 2.2644067796610173e-06,
|
| 19094 |
+
"loss": 1.2355,
|
| 19095 |
+
"step": 26660
|
| 19096 |
+
},
|
| 19097 |
+
{
|
| 19098 |
+
"epoch": 0.67,
|
| 19099 |
+
"grad_norm": 15.327707290649414,
|
| 19100 |
+
"learning_rate": 2.257627118644068e-06,
|
| 19101 |
+
"loss": 1.2088,
|
| 19102 |
+
"step": 26670
|
| 19103 |
+
},
|
| 19104 |
+
{
|
| 19105 |
+
"epoch": 0.67,
|
| 19106 |
+
"grad_norm": 6.978293418884277,
|
| 19107 |
+
"learning_rate": 2.250847457627119e-06,
|
| 19108 |
+
"loss": 1.3372,
|
| 19109 |
+
"step": 26680
|
| 19110 |
+
},
|
| 19111 |
+
{
|
| 19112 |
+
"epoch": 0.67,
|
| 19113 |
+
"grad_norm": 10.799981117248535,
|
| 19114 |
+
"learning_rate": 2.2440677966101694e-06,
|
| 19115 |
+
"loss": 1.4839,
|
| 19116 |
+
"step": 26690
|
| 19117 |
+
},
|
| 19118 |
+
{
|
| 19119 |
+
"epoch": 0.67,
|
| 19120 |
+
"grad_norm": 4.861376762390137,
|
| 19121 |
+
"learning_rate": 2.2372881355932204e-06,
|
| 19122 |
+
"loss": 1.3361,
|
| 19123 |
+
"step": 26700
|
| 19124 |
+
},
|
| 19125 |
+
{
|
| 19126 |
+
"epoch": 0.67,
|
| 19127 |
+
"grad_norm": 5.07356595993042,
|
| 19128 |
+
"learning_rate": 2.2305084745762714e-06,
|
| 19129 |
+
"loss": 1.1581,
|
| 19130 |
+
"step": 26710
|
| 19131 |
+
},
|
| 19132 |
+
{
|
| 19133 |
+
"epoch": 0.67,
|
| 19134 |
+
"grad_norm": 16.509151458740234,
|
| 19135 |
+
"learning_rate": 2.2237288135593223e-06,
|
| 19136 |
+
"loss": 1.3014,
|
| 19137 |
+
"step": 26720
|
| 19138 |
+
},
|
| 19139 |
+
{
|
| 19140 |
+
"epoch": 0.67,
|
| 19141 |
+
"grad_norm": 8.232512474060059,
|
| 19142 |
+
"learning_rate": 2.216949152542373e-06,
|
| 19143 |
+
"loss": 1.4513,
|
| 19144 |
+
"step": 26730
|
| 19145 |
+
},
|
| 19146 |
+
{
|
| 19147 |
+
"epoch": 0.67,
|
| 19148 |
+
"grad_norm": 4.673596382141113,
|
| 19149 |
+
"learning_rate": 2.210169491525424e-06,
|
| 19150 |
+
"loss": 1.3162,
|
| 19151 |
+
"step": 26740
|
| 19152 |
+
},
|
| 19153 |
+
{
|
| 19154 |
+
"epoch": 0.67,
|
| 19155 |
+
"grad_norm": 3.565079927444458,
|
| 19156 |
+
"learning_rate": 2.203389830508475e-06,
|
| 19157 |
+
"loss": 1.3423,
|
| 19158 |
+
"step": 26750
|
| 19159 |
+
},
|
| 19160 |
+
{
|
| 19161 |
+
"epoch": 0.67,
|
| 19162 |
+
"grad_norm": 8.415205001831055,
|
| 19163 |
+
"learning_rate": 2.196610169491526e-06,
|
| 19164 |
+
"loss": 1.3341,
|
| 19165 |
+
"step": 26760
|
| 19166 |
+
},
|
| 19167 |
+
{
|
| 19168 |
+
"epoch": 0.67,
|
| 19169 |
+
"grad_norm": 13.892396926879883,
|
| 19170 |
+
"learning_rate": 2.1898305084745764e-06,
|
| 19171 |
+
"loss": 1.3268,
|
| 19172 |
+
"step": 26770
|
| 19173 |
+
},
|
| 19174 |
+
{
|
| 19175 |
+
"epoch": 0.67,
|
| 19176 |
+
"grad_norm": 12.167963981628418,
|
| 19177 |
+
"learning_rate": 2.1830508474576273e-06,
|
| 19178 |
+
"loss": 1.3763,
|
| 19179 |
+
"step": 26780
|
| 19180 |
+
},
|
| 19181 |
+
{
|
| 19182 |
+
"epoch": 0.67,
|
| 19183 |
+
"grad_norm": 9.574051856994629,
|
| 19184 |
+
"learning_rate": 2.1762711864406783e-06,
|
| 19185 |
+
"loss": 1.5376,
|
| 19186 |
+
"step": 26790
|
| 19187 |
+
},
|
| 19188 |
+
{
|
| 19189 |
+
"epoch": 0.67,
|
| 19190 |
+
"grad_norm": 12.77506160736084,
|
| 19191 |
+
"learning_rate": 2.169491525423729e-06,
|
| 19192 |
+
"loss": 1.2457,
|
| 19193 |
+
"step": 26800
|
| 19194 |
+
},
|
| 19195 |
+
{
|
| 19196 |
+
"epoch": 0.67,
|
| 19197 |
+
"grad_norm": 7.312320232391357,
|
| 19198 |
+
"learning_rate": 2.16271186440678e-06,
|
| 19199 |
+
"loss": 1.3077,
|
| 19200 |
+
"step": 26810
|
| 19201 |
+
},
|
| 19202 |
+
{
|
| 19203 |
+
"epoch": 0.67,
|
| 19204 |
+
"grad_norm": 10.53618335723877,
|
| 19205 |
+
"learning_rate": 2.1559322033898304e-06,
|
| 19206 |
+
"loss": 1.3706,
|
| 19207 |
+
"step": 26820
|
| 19208 |
+
},
|
| 19209 |
+
{
|
| 19210 |
+
"epoch": 0.67,
|
| 19211 |
+
"grad_norm": 4.399540424346924,
|
| 19212 |
+
"learning_rate": 2.1491525423728814e-06,
|
| 19213 |
+
"loss": 1.3237,
|
| 19214 |
+
"step": 26830
|
| 19215 |
+
},
|
| 19216 |
+
{
|
| 19217 |
+
"epoch": 0.67,
|
| 19218 |
+
"grad_norm": 8.757078170776367,
|
| 19219 |
+
"learning_rate": 2.1423728813559324e-06,
|
| 19220 |
+
"loss": 1.2385,
|
| 19221 |
+
"step": 26840
|
| 19222 |
+
},
|
| 19223 |
+
{
|
| 19224 |
+
"epoch": 0.67,
|
| 19225 |
+
"grad_norm": 3.866237163543701,
|
| 19226 |
+
"learning_rate": 2.1355932203389833e-06,
|
| 19227 |
+
"loss": 1.4038,
|
| 19228 |
+
"step": 26850
|
| 19229 |
+
},
|
| 19230 |
+
{
|
| 19231 |
+
"epoch": 0.67,
|
| 19232 |
+
"grad_norm": 4.188089370727539,
|
| 19233 |
+
"learning_rate": 2.128813559322034e-06,
|
| 19234 |
+
"loss": 1.5465,
|
| 19235 |
+
"step": 26860
|
| 19236 |
+
},
|
| 19237 |
+
{
|
| 19238 |
+
"epoch": 0.67,
|
| 19239 |
+
"grad_norm": 10.617339134216309,
|
| 19240 |
+
"learning_rate": 2.122033898305085e-06,
|
| 19241 |
+
"loss": 1.1873,
|
| 19242 |
+
"step": 26870
|
| 19243 |
+
},
|
| 19244 |
+
{
|
| 19245 |
+
"epoch": 0.67,
|
| 19246 |
+
"grad_norm": 3.4754912853240967,
|
| 19247 |
+
"learning_rate": 2.115254237288136e-06,
|
| 19248 |
+
"loss": 1.2539,
|
| 19249 |
+
"step": 26880
|
| 19250 |
+
},
|
| 19251 |
+
{
|
| 19252 |
+
"epoch": 0.67,
|
| 19253 |
+
"grad_norm": 6.057491302490234,
|
| 19254 |
+
"learning_rate": 2.108474576271187e-06,
|
| 19255 |
+
"loss": 1.2034,
|
| 19256 |
+
"step": 26890
|
| 19257 |
+
},
|
| 19258 |
+
{
|
| 19259 |
+
"epoch": 0.67,
|
| 19260 |
+
"grad_norm": 9.13399887084961,
|
| 19261 |
+
"learning_rate": 2.1016949152542374e-06,
|
| 19262 |
+
"loss": 1.4333,
|
| 19263 |
+
"step": 26900
|
| 19264 |
+
},
|
| 19265 |
+
{
|
| 19266 |
+
"epoch": 0.67,
|
| 19267 |
+
"grad_norm": 14.797123908996582,
|
| 19268 |
+
"learning_rate": 2.0949152542372883e-06,
|
| 19269 |
+
"loss": 1.1961,
|
| 19270 |
+
"step": 26910
|
| 19271 |
+
},
|
| 19272 |
+
{
|
| 19273 |
+
"epoch": 0.67,
|
| 19274 |
+
"grad_norm": 6.3039398193359375,
|
| 19275 |
+
"learning_rate": 2.0881355932203393e-06,
|
| 19276 |
+
"loss": 1.0776,
|
| 19277 |
+
"step": 26920
|
| 19278 |
+
},
|
| 19279 |
+
{
|
| 19280 |
+
"epoch": 0.67,
|
| 19281 |
+
"grad_norm": 5.700479507446289,
|
| 19282 |
+
"learning_rate": 2.08135593220339e-06,
|
| 19283 |
+
"loss": 1.5808,
|
| 19284 |
+
"step": 26930
|
| 19285 |
+
},
|
| 19286 |
+
{
|
| 19287 |
+
"epoch": 0.67,
|
| 19288 |
+
"grad_norm": 17.28398895263672,
|
| 19289 |
+
"learning_rate": 2.074576271186441e-06,
|
| 19290 |
+
"loss": 1.4131,
|
| 19291 |
+
"step": 26940
|
| 19292 |
+
},
|
| 19293 |
+
{
|
| 19294 |
+
"epoch": 0.67,
|
| 19295 |
+
"grad_norm": 5.036632061004639,
|
| 19296 |
+
"learning_rate": 2.0677966101694914e-06,
|
| 19297 |
+
"loss": 1.1761,
|
| 19298 |
+
"step": 26950
|
| 19299 |
+
},
|
| 19300 |
+
{
|
| 19301 |
+
"epoch": 0.67,
|
| 19302 |
+
"grad_norm": 4.72507381439209,
|
| 19303 |
+
"learning_rate": 2.0610169491525424e-06,
|
| 19304 |
+
"loss": 1.3633,
|
| 19305 |
+
"step": 26960
|
| 19306 |
+
},
|
| 19307 |
+
{
|
| 19308 |
+
"epoch": 0.67,
|
| 19309 |
+
"grad_norm": 18.702543258666992,
|
| 19310 |
+
"learning_rate": 2.0542372881355934e-06,
|
| 19311 |
+
"loss": 1.3922,
|
| 19312 |
+
"step": 26970
|
| 19313 |
+
},
|
| 19314 |
+
{
|
| 19315 |
+
"epoch": 0.67,
|
| 19316 |
+
"grad_norm": 11.218758583068848,
|
| 19317 |
+
"learning_rate": 2.0474576271186443e-06,
|
| 19318 |
+
"loss": 1.3427,
|
| 19319 |
+
"step": 26980
|
| 19320 |
+
},
|
| 19321 |
+
{
|
| 19322 |
+
"epoch": 0.67,
|
| 19323 |
+
"grad_norm": 5.716217517852783,
|
| 19324 |
+
"learning_rate": 2.0406779661016953e-06,
|
| 19325 |
+
"loss": 1.2603,
|
| 19326 |
+
"step": 26990
|
| 19327 |
+
},
|
| 19328 |
+
{
|
| 19329 |
+
"epoch": 0.68,
|
| 19330 |
+
"grad_norm": 8.085409164428711,
|
| 19331 |
+
"learning_rate": 2.033898305084746e-06,
|
| 19332 |
+
"loss": 1.4003,
|
| 19333 |
+
"step": 27000
|
| 19334 |
+
},
|
| 19335 |
+
{
|
| 19336 |
+
"epoch": 0.68,
|
| 19337 |
+
"eval_loss": 1.3196759223937988,
|
| 19338 |
+
"eval_runtime": 66.1008,
|
| 19339 |
+
"eval_samples_per_second": 15.128,
|
| 19340 |
+
"eval_steps_per_second": 15.128,
|
| 19341 |
+
"step": 27000
|
| 19342 |
+
},
|
| 19343 |
+
{
|
| 19344 |
+
"epoch": 0.68,
|
| 19345 |
+
"grad_norm": 2.983642578125,
|
| 19346 |
+
"learning_rate": 2.027118644067797e-06,
|
| 19347 |
+
"loss": 1.2987,
|
| 19348 |
+
"step": 27010
|
| 19349 |
+
},
|
| 19350 |
+
{
|
| 19351 |
+
"epoch": 0.68,
|
| 19352 |
+
"grad_norm": 3.0962276458740234,
|
| 19353 |
+
"learning_rate": 2.020338983050848e-06,
|
| 19354 |
+
"loss": 1.3189,
|
| 19355 |
+
"step": 27020
|
| 19356 |
+
},
|
| 19357 |
+
{
|
| 19358 |
+
"epoch": 0.68,
|
| 19359 |
+
"grad_norm": 4.857142925262451,
|
| 19360 |
+
"learning_rate": 2.0135593220338984e-06,
|
| 19361 |
+
"loss": 1.3651,
|
| 19362 |
+
"step": 27030
|
| 19363 |
+
},
|
| 19364 |
+
{
|
| 19365 |
+
"epoch": 0.68,
|
| 19366 |
+
"grad_norm": 4.3937482833862305,
|
| 19367 |
+
"learning_rate": 2.0067796610169494e-06,
|
| 19368 |
+
"loss": 1.4407,
|
| 19369 |
+
"step": 27040
|
| 19370 |
+
},
|
| 19371 |
+
{
|
| 19372 |
+
"epoch": 0.68,
|
| 19373 |
+
"grad_norm": 4.091692924499512,
|
| 19374 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19375 |
+
"loss": 1.2065,
|
| 19376 |
+
"step": 27050
|
| 19377 |
+
},
|
| 19378 |
+
{
|
| 19379 |
+
"epoch": 0.68,
|
| 19380 |
+
"grad_norm": 7.725543975830078,
|
| 19381 |
+
"learning_rate": 1.993220338983051e-06,
|
| 19382 |
+
"loss": 1.1836,
|
| 19383 |
+
"step": 27060
|
| 19384 |
+
},
|
| 19385 |
+
{
|
| 19386 |
+
"epoch": 0.68,
|
| 19387 |
+
"grad_norm": 5.124769687652588,
|
| 19388 |
+
"learning_rate": 1.986440677966102e-06,
|
| 19389 |
+
"loss": 1.3262,
|
| 19390 |
+
"step": 27070
|
| 19391 |
+
},
|
| 19392 |
+
{
|
| 19393 |
+
"epoch": 0.68,
|
| 19394 |
+
"grad_norm": 5.097384452819824,
|
| 19395 |
+
"learning_rate": 1.9796610169491524e-06,
|
| 19396 |
+
"loss": 1.3057,
|
| 19397 |
+
"step": 27080
|
| 19398 |
+
},
|
| 19399 |
+
{
|
| 19400 |
+
"epoch": 0.68,
|
| 19401 |
+
"grad_norm": 3.207469940185547,
|
| 19402 |
+
"learning_rate": 1.9728813559322034e-06,
|
| 19403 |
+
"loss": 1.241,
|
| 19404 |
+
"step": 27090
|
| 19405 |
+
},
|
| 19406 |
+
{
|
| 19407 |
+
"epoch": 0.68,
|
| 19408 |
+
"grad_norm": 8.80706787109375,
|
| 19409 |
+
"learning_rate": 1.9661016949152544e-06,
|
| 19410 |
+
"loss": 1.4066,
|
| 19411 |
+
"step": 27100
|
| 19412 |
+
},
|
| 19413 |
+
{
|
| 19414 |
+
"epoch": 0.68,
|
| 19415 |
+
"grad_norm": 2.1298069953918457,
|
| 19416 |
+
"learning_rate": 1.9593220338983053e-06,
|
| 19417 |
+
"loss": 1.3394,
|
| 19418 |
+
"step": 27110
|
| 19419 |
+
},
|
| 19420 |
+
{
|
| 19421 |
+
"epoch": 0.68,
|
| 19422 |
+
"grad_norm": 9.768424034118652,
|
| 19423 |
+
"learning_rate": 1.9525423728813563e-06,
|
| 19424 |
+
"loss": 1.4714,
|
| 19425 |
+
"step": 27120
|
| 19426 |
+
},
|
| 19427 |
+
{
|
| 19428 |
+
"epoch": 0.68,
|
| 19429 |
+
"grad_norm": 16.757766723632812,
|
| 19430 |
+
"learning_rate": 1.945762711864407e-06,
|
| 19431 |
+
"loss": 1.4322,
|
| 19432 |
+
"step": 27130
|
| 19433 |
+
},
|
| 19434 |
+
{
|
| 19435 |
+
"epoch": 0.68,
|
| 19436 |
+
"grad_norm": 4.268533229827881,
|
| 19437 |
+
"learning_rate": 1.938983050847458e-06,
|
| 19438 |
+
"loss": 1.3988,
|
| 19439 |
+
"step": 27140
|
| 19440 |
+
},
|
| 19441 |
+
{
|
| 19442 |
+
"epoch": 0.68,
|
| 19443 |
+
"grad_norm": 7.853205680847168,
|
| 19444 |
+
"learning_rate": 1.932203389830509e-06,
|
| 19445 |
+
"loss": 1.2837,
|
| 19446 |
+
"step": 27150
|
| 19447 |
+
},
|
| 19448 |
+
{
|
| 19449 |
+
"epoch": 0.68,
|
| 19450 |
+
"grad_norm": 3.1779301166534424,
|
| 19451 |
+
"learning_rate": 1.9254237288135594e-06,
|
| 19452 |
+
"loss": 1.3075,
|
| 19453 |
+
"step": 27160
|
| 19454 |
+
},
|
| 19455 |
+
{
|
| 19456 |
+
"epoch": 0.68,
|
| 19457 |
+
"grad_norm": 7.716470718383789,
|
| 19458 |
+
"learning_rate": 1.9186440677966104e-06,
|
| 19459 |
+
"loss": 1.2207,
|
| 19460 |
+
"step": 27170
|
| 19461 |
+
},
|
| 19462 |
+
{
|
| 19463 |
+
"epoch": 0.68,
|
| 19464 |
+
"grad_norm": 4.0963029861450195,
|
| 19465 |
+
"learning_rate": 1.9118644067796613e-06,
|
| 19466 |
+
"loss": 1.3241,
|
| 19467 |
+
"step": 27180
|
| 19468 |
+
},
|
| 19469 |
+
{
|
| 19470 |
+
"epoch": 0.68,
|
| 19471 |
+
"grad_norm": 10.319551467895508,
|
| 19472 |
+
"learning_rate": 1.9050847457627119e-06,
|
| 19473 |
+
"loss": 1.4133,
|
| 19474 |
+
"step": 27190
|
| 19475 |
+
},
|
| 19476 |
+
{
|
| 19477 |
+
"epoch": 0.68,
|
| 19478 |
+
"grad_norm": 4.663591384887695,
|
| 19479 |
+
"learning_rate": 1.8983050847457629e-06,
|
| 19480 |
+
"loss": 1.2505,
|
| 19481 |
+
"step": 27200
|
| 19482 |
+
},
|
| 19483 |
+
{
|
| 19484 |
+
"epoch": 0.68,
|
| 19485 |
+
"grad_norm": 8.912657737731934,
|
| 19486 |
+
"learning_rate": 1.8915254237288136e-06,
|
| 19487 |
+
"loss": 1.2057,
|
| 19488 |
+
"step": 27210
|
| 19489 |
+
},
|
| 19490 |
+
{
|
| 19491 |
+
"epoch": 0.68,
|
| 19492 |
+
"grad_norm": 11.967384338378906,
|
| 19493 |
+
"learning_rate": 1.8847457627118646e-06,
|
| 19494 |
+
"loss": 1.4043,
|
| 19495 |
+
"step": 27220
|
| 19496 |
+
},
|
| 19497 |
+
{
|
| 19498 |
+
"epoch": 0.68,
|
| 19499 |
+
"grad_norm": 19.304821014404297,
|
| 19500 |
+
"learning_rate": 1.8779661016949156e-06,
|
| 19501 |
+
"loss": 1.3083,
|
| 19502 |
+
"step": 27230
|
| 19503 |
+
},
|
| 19504 |
+
{
|
| 19505 |
+
"epoch": 0.68,
|
| 19506 |
+
"grad_norm": 4.1671600341796875,
|
| 19507 |
+
"learning_rate": 1.8711864406779661e-06,
|
| 19508 |
+
"loss": 1.415,
|
| 19509 |
+
"step": 27240
|
| 19510 |
+
},
|
| 19511 |
+
{
|
| 19512 |
+
"epoch": 0.68,
|
| 19513 |
+
"grad_norm": 3.430443525314331,
|
| 19514 |
+
"learning_rate": 1.8644067796610171e-06,
|
| 19515 |
+
"loss": 1.3154,
|
| 19516 |
+
"step": 27250
|
| 19517 |
+
},
|
| 19518 |
+
{
|
| 19519 |
+
"epoch": 0.68,
|
| 19520 |
+
"grad_norm": 8.876851081848145,
|
| 19521 |
+
"learning_rate": 1.857627118644068e-06,
|
| 19522 |
+
"loss": 1.2701,
|
| 19523 |
+
"step": 27260
|
| 19524 |
+
},
|
| 19525 |
+
{
|
| 19526 |
+
"epoch": 0.68,
|
| 19527 |
+
"grad_norm": 12.077353477478027,
|
| 19528 |
+
"learning_rate": 1.8508474576271189e-06,
|
| 19529 |
+
"loss": 1.3744,
|
| 19530 |
+
"step": 27270
|
| 19531 |
+
},
|
| 19532 |
+
{
|
| 19533 |
+
"epoch": 0.68,
|
| 19534 |
+
"grad_norm": 1.86147940158844,
|
| 19535 |
+
"learning_rate": 1.8440677966101696e-06,
|
| 19536 |
+
"loss": 1.4091,
|
| 19537 |
+
"step": 27280
|
| 19538 |
+
},
|
| 19539 |
+
{
|
| 19540 |
+
"epoch": 0.68,
|
| 19541 |
+
"grad_norm": 6.873291969299316,
|
| 19542 |
+
"learning_rate": 1.8372881355932204e-06,
|
| 19543 |
+
"loss": 1.3488,
|
| 19544 |
+
"step": 27290
|
| 19545 |
+
},
|
| 19546 |
+
{
|
| 19547 |
+
"epoch": 0.68,
|
| 19548 |
+
"grad_norm": 5.088196754455566,
|
| 19549 |
+
"learning_rate": 1.8305084745762714e-06,
|
| 19550 |
+
"loss": 1.437,
|
| 19551 |
+
"step": 27300
|
| 19552 |
+
},
|
| 19553 |
+
{
|
| 19554 |
+
"epoch": 0.68,
|
| 19555 |
+
"grad_norm": 3.1845881938934326,
|
| 19556 |
+
"learning_rate": 1.8237288135593223e-06,
|
| 19557 |
+
"loss": 1.3961,
|
| 19558 |
+
"step": 27310
|
| 19559 |
+
},
|
| 19560 |
+
{
|
| 19561 |
+
"epoch": 0.68,
|
| 19562 |
+
"grad_norm": 8.228582382202148,
|
| 19563 |
+
"learning_rate": 1.816949152542373e-06,
|
| 19564 |
+
"loss": 1.1022,
|
| 19565 |
+
"step": 27320
|
| 19566 |
+
},
|
| 19567 |
+
{
|
| 19568 |
+
"epoch": 0.68,
|
| 19569 |
+
"grad_norm": 15.932997703552246,
|
| 19570 |
+
"learning_rate": 1.8101694915254239e-06,
|
| 19571 |
+
"loss": 1.3529,
|
| 19572 |
+
"step": 27330
|
| 19573 |
+
},
|
| 19574 |
+
{
|
| 19575 |
+
"epoch": 0.68,
|
| 19576 |
+
"grad_norm": 1.3720171451568604,
|
| 19577 |
+
"learning_rate": 1.8033898305084746e-06,
|
| 19578 |
+
"loss": 1.2631,
|
| 19579 |
+
"step": 27340
|
| 19580 |
+
},
|
| 19581 |
+
{
|
| 19582 |
+
"epoch": 0.68,
|
| 19583 |
+
"grad_norm": 3.995771646499634,
|
| 19584 |
+
"learning_rate": 1.7966101694915256e-06,
|
| 19585 |
+
"loss": 1.2678,
|
| 19586 |
+
"step": 27350
|
| 19587 |
+
},
|
| 19588 |
+
{
|
| 19589 |
+
"epoch": 0.68,
|
| 19590 |
+
"grad_norm": 5.7671895027160645,
|
| 19591 |
+
"learning_rate": 1.7898305084745766e-06,
|
| 19592 |
+
"loss": 1.329,
|
| 19593 |
+
"step": 27360
|
| 19594 |
+
},
|
| 19595 |
+
{
|
| 19596 |
+
"epoch": 0.68,
|
| 19597 |
+
"grad_norm": 3.57239031791687,
|
| 19598 |
+
"learning_rate": 1.7830508474576271e-06,
|
| 19599 |
+
"loss": 1.2557,
|
| 19600 |
+
"step": 27370
|
| 19601 |
+
},
|
| 19602 |
+
{
|
| 19603 |
+
"epoch": 0.68,
|
| 19604 |
+
"grad_norm": 8.127071380615234,
|
| 19605 |
+
"learning_rate": 1.7762711864406781e-06,
|
| 19606 |
+
"loss": 1.3917,
|
| 19607 |
+
"step": 27380
|
| 19608 |
+
},
|
| 19609 |
+
{
|
| 19610 |
+
"epoch": 0.68,
|
| 19611 |
+
"grad_norm": 11.008809089660645,
|
| 19612 |
+
"learning_rate": 1.769491525423729e-06,
|
| 19613 |
+
"loss": 1.4125,
|
| 19614 |
+
"step": 27390
|
| 19615 |
+
},
|
| 19616 |
+
{
|
| 19617 |
+
"epoch": 0.69,
|
| 19618 |
+
"grad_norm": 3.305449962615967,
|
| 19619 |
+
"learning_rate": 1.7627118644067799e-06,
|
| 19620 |
+
"loss": 1.4367,
|
| 19621 |
+
"step": 27400
|
| 19622 |
+
},
|
| 19623 |
+
{
|
| 19624 |
+
"epoch": 0.69,
|
| 19625 |
+
"grad_norm": 4.471165180206299,
|
| 19626 |
+
"learning_rate": 1.7559322033898306e-06,
|
| 19627 |
+
"loss": 1.1336,
|
| 19628 |
+
"step": 27410
|
| 19629 |
+
},
|
| 19630 |
+
{
|
| 19631 |
+
"epoch": 0.69,
|
| 19632 |
+
"grad_norm": 4.836517810821533,
|
| 19633 |
+
"learning_rate": 1.7491525423728814e-06,
|
| 19634 |
+
"loss": 1.37,
|
| 19635 |
+
"step": 27420
|
| 19636 |
+
},
|
| 19637 |
+
{
|
| 19638 |
+
"epoch": 0.69,
|
| 19639 |
+
"grad_norm": 6.102312088012695,
|
| 19640 |
+
"learning_rate": 1.7423728813559324e-06,
|
| 19641 |
+
"loss": 1.3035,
|
| 19642 |
+
"step": 27430
|
| 19643 |
+
},
|
| 19644 |
+
{
|
| 19645 |
+
"epoch": 0.69,
|
| 19646 |
+
"grad_norm": 4.232480049133301,
|
| 19647 |
+
"learning_rate": 1.7355932203389834e-06,
|
| 19648 |
+
"loss": 1.2197,
|
| 19649 |
+
"step": 27440
|
| 19650 |
+
},
|
| 19651 |
+
{
|
| 19652 |
+
"epoch": 0.69,
|
| 19653 |
+
"grad_norm": 9.582504272460938,
|
| 19654 |
+
"learning_rate": 1.728813559322034e-06,
|
| 19655 |
+
"loss": 1.2699,
|
| 19656 |
+
"step": 27450
|
| 19657 |
+
},
|
| 19658 |
+
{
|
| 19659 |
+
"epoch": 0.69,
|
| 19660 |
+
"grad_norm": 9.52001667022705,
|
| 19661 |
+
"learning_rate": 1.7220338983050849e-06,
|
| 19662 |
+
"loss": 1.356,
|
| 19663 |
+
"step": 27460
|
| 19664 |
+
},
|
| 19665 |
+
{
|
| 19666 |
+
"epoch": 0.69,
|
| 19667 |
+
"grad_norm": 13.017762184143066,
|
| 19668 |
+
"learning_rate": 1.7152542372881356e-06,
|
| 19669 |
+
"loss": 1.2246,
|
| 19670 |
+
"step": 27470
|
| 19671 |
+
},
|
| 19672 |
+
{
|
| 19673 |
+
"epoch": 0.69,
|
| 19674 |
+
"grad_norm": 7.059675693511963,
|
| 19675 |
+
"learning_rate": 1.7084745762711866e-06,
|
| 19676 |
+
"loss": 1.3978,
|
| 19677 |
+
"step": 27480
|
| 19678 |
+
},
|
| 19679 |
+
{
|
| 19680 |
+
"epoch": 0.69,
|
| 19681 |
+
"grad_norm": 4.05330753326416,
|
| 19682 |
+
"learning_rate": 1.7016949152542376e-06,
|
| 19683 |
+
"loss": 1.3678,
|
| 19684 |
+
"step": 27490
|
| 19685 |
+
},
|
| 19686 |
+
{
|
| 19687 |
+
"epoch": 0.69,
|
| 19688 |
+
"grad_norm": 8.122435569763184,
|
| 19689 |
+
"learning_rate": 1.6949152542372882e-06,
|
| 19690 |
+
"loss": 1.3062,
|
| 19691 |
+
"step": 27500
|
| 19692 |
+
},
|
| 19693 |
+
{
|
| 19694 |
+
"epoch": 0.69,
|
| 19695 |
+
"eval_loss": 1.262895941734314,
|
| 19696 |
+
"eval_runtime": 66.2138,
|
| 19697 |
+
"eval_samples_per_second": 15.103,
|
| 19698 |
+
"eval_steps_per_second": 15.103,
|
| 19699 |
+
"step": 27500
|
| 19700 |
}
|
| 19701 |
],
|
| 19702 |
"logging_steps": 10,
|
|
|
|
| 19704 |
"num_input_tokens_seen": 0,
|
| 19705 |
"num_train_epochs": 1,
|
| 19706 |
"save_steps": 2500,
|
| 19707 |
+
"total_flos": 4.4280846483456e+17,
|
| 19708 |
"train_batch_size": 1,
|
| 19709 |
"trial_name": null,
|
| 19710 |
"trial_params": null
|