Uploaded checkpoint-17500
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 sha256:162241111a35b4dd2bd8251eb44c4f9ed485c39f432082deec2b913318be26b3
|
| 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 sha256:2640cb6f400e7439fff7ee437394f2676dd7a329f43b4ef033bc2e958e48c385
|
| 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:b6ca59958578f29e636921c5d01edf609d279634685f5c1700ffd019a9a229f9
|
| 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:eb29930cc2c8e69e7c76b92867840499fb9c566b9d6b348753e567d4e680bb99
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 1.3866758346557617,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-12500",
|
| 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,
|
|
@@ -10747,6 +10747,1796 @@
|
|
| 10747 |
"eval_samples_per_second": 15.121,
|
| 10748 |
"eval_steps_per_second": 15.121,
|
| 10749 |
"step": 15000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10750 |
}
|
| 10751 |
],
|
| 10752 |
"logging_steps": 10,
|
|
@@ -10754,7 +12544,7 @@
|
|
| 10754 |
"num_input_tokens_seen": 0,
|
| 10755 |
"num_train_epochs": 1,
|
| 10756 |
"save_steps": 2500,
|
| 10757 |
-
"total_flos": 2.
|
| 10758 |
"train_batch_size": 1,
|
| 10759 |
"trial_name": null,
|
| 10760 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 1.3866758346557617,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-12500",
|
| 4 |
+
"epoch": 0.4375,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 17500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 10747 |
"eval_samples_per_second": 15.121,
|
| 10748 |
"eval_steps_per_second": 15.121,
|
| 10749 |
"step": 15000
|
| 10750 |
+
},
|
| 10751 |
+
{
|
| 10752 |
+
"epoch": 0.38,
|
| 10753 |
+
"grad_norm": 3.327223777770996,
|
| 10754 |
+
"learning_rate": 1.016271186440678e-05,
|
| 10755 |
+
"loss": 1.4933,
|
| 10756 |
+
"step": 15010
|
| 10757 |
+
},
|
| 10758 |
+
{
|
| 10759 |
+
"epoch": 0.38,
|
| 10760 |
+
"grad_norm": 5.010586738586426,
|
| 10761 |
+
"learning_rate": 1.015593220338983e-05,
|
| 10762 |
+
"loss": 1.2914,
|
| 10763 |
+
"step": 15020
|
| 10764 |
+
},
|
| 10765 |
+
{
|
| 10766 |
+
"epoch": 0.38,
|
| 10767 |
+
"grad_norm": 4.52363395690918,
|
| 10768 |
+
"learning_rate": 1.0149152542372882e-05,
|
| 10769 |
+
"loss": 1.3007,
|
| 10770 |
+
"step": 15030
|
| 10771 |
+
},
|
| 10772 |
+
{
|
| 10773 |
+
"epoch": 0.38,
|
| 10774 |
+
"grad_norm": 4.301648139953613,
|
| 10775 |
+
"learning_rate": 1.0142372881355933e-05,
|
| 10776 |
+
"loss": 1.3652,
|
| 10777 |
+
"step": 15040
|
| 10778 |
+
},
|
| 10779 |
+
{
|
| 10780 |
+
"epoch": 0.38,
|
| 10781 |
+
"grad_norm": 8.124075889587402,
|
| 10782 |
+
"learning_rate": 1.0135593220338985e-05,
|
| 10783 |
+
"loss": 1.3957,
|
| 10784 |
+
"step": 15050
|
| 10785 |
+
},
|
| 10786 |
+
{
|
| 10787 |
+
"epoch": 0.38,
|
| 10788 |
+
"grad_norm": 4.588447093963623,
|
| 10789 |
+
"learning_rate": 1.0128813559322034e-05,
|
| 10790 |
+
"loss": 1.1597,
|
| 10791 |
+
"step": 15060
|
| 10792 |
+
},
|
| 10793 |
+
{
|
| 10794 |
+
"epoch": 0.38,
|
| 10795 |
+
"grad_norm": 3.6226577758789062,
|
| 10796 |
+
"learning_rate": 1.0122033898305086e-05,
|
| 10797 |
+
"loss": 1.5604,
|
| 10798 |
+
"step": 15070
|
| 10799 |
+
},
|
| 10800 |
+
{
|
| 10801 |
+
"epoch": 0.38,
|
| 10802 |
+
"grad_norm": 7.039438724517822,
|
| 10803 |
+
"learning_rate": 1.0115254237288137e-05,
|
| 10804 |
+
"loss": 1.2913,
|
| 10805 |
+
"step": 15080
|
| 10806 |
+
},
|
| 10807 |
+
{
|
| 10808 |
+
"epoch": 0.38,
|
| 10809 |
+
"grad_norm": 1.9853217601776123,
|
| 10810 |
+
"learning_rate": 1.0108474576271189e-05,
|
| 10811 |
+
"loss": 1.2087,
|
| 10812 |
+
"step": 15090
|
| 10813 |
+
},
|
| 10814 |
+
{
|
| 10815 |
+
"epoch": 0.38,
|
| 10816 |
+
"grad_norm": 6.037110805511475,
|
| 10817 |
+
"learning_rate": 1.0101694915254238e-05,
|
| 10818 |
+
"loss": 1.3358,
|
| 10819 |
+
"step": 15100
|
| 10820 |
+
},
|
| 10821 |
+
{
|
| 10822 |
+
"epoch": 0.38,
|
| 10823 |
+
"grad_norm": 1.2350672483444214,
|
| 10824 |
+
"learning_rate": 1.009491525423729e-05,
|
| 10825 |
+
"loss": 1.4253,
|
| 10826 |
+
"step": 15110
|
| 10827 |
+
},
|
| 10828 |
+
{
|
| 10829 |
+
"epoch": 0.38,
|
| 10830 |
+
"grad_norm": 4.601665019989014,
|
| 10831 |
+
"learning_rate": 1.0088135593220341e-05,
|
| 10832 |
+
"loss": 1.3471,
|
| 10833 |
+
"step": 15120
|
| 10834 |
+
},
|
| 10835 |
+
{
|
| 10836 |
+
"epoch": 0.38,
|
| 10837 |
+
"grad_norm": 7.822099208831787,
|
| 10838 |
+
"learning_rate": 1.008135593220339e-05,
|
| 10839 |
+
"loss": 1.4255,
|
| 10840 |
+
"step": 15130
|
| 10841 |
+
},
|
| 10842 |
+
{
|
| 10843 |
+
"epoch": 0.38,
|
| 10844 |
+
"grad_norm": 2.1710550785064697,
|
| 10845 |
+
"learning_rate": 1.0074576271186442e-05,
|
| 10846 |
+
"loss": 1.4419,
|
| 10847 |
+
"step": 15140
|
| 10848 |
+
},
|
| 10849 |
+
{
|
| 10850 |
+
"epoch": 0.38,
|
| 10851 |
+
"grad_norm": 6.896235942840576,
|
| 10852 |
+
"learning_rate": 1.0067796610169492e-05,
|
| 10853 |
+
"loss": 1.3166,
|
| 10854 |
+
"step": 15150
|
| 10855 |
+
},
|
| 10856 |
+
{
|
| 10857 |
+
"epoch": 0.38,
|
| 10858 |
+
"grad_norm": 2.7912566661834717,
|
| 10859 |
+
"learning_rate": 1.0061016949152542e-05,
|
| 10860 |
+
"loss": 1.3577,
|
| 10861 |
+
"step": 15160
|
| 10862 |
+
},
|
| 10863 |
+
{
|
| 10864 |
+
"epoch": 0.38,
|
| 10865 |
+
"grad_norm": 2.8911972045898438,
|
| 10866 |
+
"learning_rate": 1.0054237288135593e-05,
|
| 10867 |
+
"loss": 1.3415,
|
| 10868 |
+
"step": 15170
|
| 10869 |
+
},
|
| 10870 |
+
{
|
| 10871 |
+
"epoch": 0.38,
|
| 10872 |
+
"grad_norm": 1.8997740745544434,
|
| 10873 |
+
"learning_rate": 1.0047457627118644e-05,
|
| 10874 |
+
"loss": 1.3933,
|
| 10875 |
+
"step": 15180
|
| 10876 |
+
},
|
| 10877 |
+
{
|
| 10878 |
+
"epoch": 0.38,
|
| 10879 |
+
"grad_norm": 5.039108753204346,
|
| 10880 |
+
"learning_rate": 1.0040677966101696e-05,
|
| 10881 |
+
"loss": 1.3395,
|
| 10882 |
+
"step": 15190
|
| 10883 |
+
},
|
| 10884 |
+
{
|
| 10885 |
+
"epoch": 0.38,
|
| 10886 |
+
"grad_norm": 5.744147300720215,
|
| 10887 |
+
"learning_rate": 1.0033898305084746e-05,
|
| 10888 |
+
"loss": 1.4215,
|
| 10889 |
+
"step": 15200
|
| 10890 |
+
},
|
| 10891 |
+
{
|
| 10892 |
+
"epoch": 0.38,
|
| 10893 |
+
"grad_norm": 18.610450744628906,
|
| 10894 |
+
"learning_rate": 1.0027118644067797e-05,
|
| 10895 |
+
"loss": 1.4639,
|
| 10896 |
+
"step": 15210
|
| 10897 |
+
},
|
| 10898 |
+
{
|
| 10899 |
+
"epoch": 0.38,
|
| 10900 |
+
"grad_norm": 4.890042304992676,
|
| 10901 |
+
"learning_rate": 1.0020338983050848e-05,
|
| 10902 |
+
"loss": 1.34,
|
| 10903 |
+
"step": 15220
|
| 10904 |
+
},
|
| 10905 |
+
{
|
| 10906 |
+
"epoch": 0.38,
|
| 10907 |
+
"grad_norm": 2.5182347297668457,
|
| 10908 |
+
"learning_rate": 1.00135593220339e-05,
|
| 10909 |
+
"loss": 1.5222,
|
| 10910 |
+
"step": 15230
|
| 10911 |
+
},
|
| 10912 |
+
{
|
| 10913 |
+
"epoch": 0.38,
|
| 10914 |
+
"grad_norm": 2.647080183029175,
|
| 10915 |
+
"learning_rate": 1.000677966101695e-05,
|
| 10916 |
+
"loss": 1.3716,
|
| 10917 |
+
"step": 15240
|
| 10918 |
+
},
|
| 10919 |
+
{
|
| 10920 |
+
"epoch": 0.38,
|
| 10921 |
+
"grad_norm": 7.10032320022583,
|
| 10922 |
+
"learning_rate": 1e-05,
|
| 10923 |
+
"loss": 1.4433,
|
| 10924 |
+
"step": 15250
|
| 10925 |
+
},
|
| 10926 |
+
{
|
| 10927 |
+
"epoch": 0.38,
|
| 10928 |
+
"grad_norm": 10.124835968017578,
|
| 10929 |
+
"learning_rate": 9.993220338983052e-06,
|
| 10930 |
+
"loss": 1.402,
|
| 10931 |
+
"step": 15260
|
| 10932 |
+
},
|
| 10933 |
+
{
|
| 10934 |
+
"epoch": 0.38,
|
| 10935 |
+
"grad_norm": 13.186834335327148,
|
| 10936 |
+
"learning_rate": 9.986440677966102e-06,
|
| 10937 |
+
"loss": 1.4211,
|
| 10938 |
+
"step": 15270
|
| 10939 |
+
},
|
| 10940 |
+
{
|
| 10941 |
+
"epoch": 0.38,
|
| 10942 |
+
"grad_norm": 3.1523687839508057,
|
| 10943 |
+
"learning_rate": 9.979661016949153e-06,
|
| 10944 |
+
"loss": 1.3585,
|
| 10945 |
+
"step": 15280
|
| 10946 |
+
},
|
| 10947 |
+
{
|
| 10948 |
+
"epoch": 0.38,
|
| 10949 |
+
"grad_norm": 5.048830032348633,
|
| 10950 |
+
"learning_rate": 9.972881355932205e-06,
|
| 10951 |
+
"loss": 1.4079,
|
| 10952 |
+
"step": 15290
|
| 10953 |
+
},
|
| 10954 |
+
{
|
| 10955 |
+
"epoch": 0.38,
|
| 10956 |
+
"grad_norm": 9.165743827819824,
|
| 10957 |
+
"learning_rate": 9.966101694915256e-06,
|
| 10958 |
+
"loss": 1.4847,
|
| 10959 |
+
"step": 15300
|
| 10960 |
+
},
|
| 10961 |
+
{
|
| 10962 |
+
"epoch": 0.38,
|
| 10963 |
+
"grad_norm": 10.89134693145752,
|
| 10964 |
+
"learning_rate": 9.959322033898306e-06,
|
| 10965 |
+
"loss": 1.4922,
|
| 10966 |
+
"step": 15310
|
| 10967 |
+
},
|
| 10968 |
+
{
|
| 10969 |
+
"epoch": 0.38,
|
| 10970 |
+
"grad_norm": 2.4734840393066406,
|
| 10971 |
+
"learning_rate": 9.952542372881356e-06,
|
| 10972 |
+
"loss": 1.496,
|
| 10973 |
+
"step": 15320
|
| 10974 |
+
},
|
| 10975 |
+
{
|
| 10976 |
+
"epoch": 0.38,
|
| 10977 |
+
"grad_norm": 2.6290500164031982,
|
| 10978 |
+
"learning_rate": 9.945762711864407e-06,
|
| 10979 |
+
"loss": 1.4658,
|
| 10980 |
+
"step": 15330
|
| 10981 |
+
},
|
| 10982 |
+
{
|
| 10983 |
+
"epoch": 0.38,
|
| 10984 |
+
"grad_norm": 2.3270649909973145,
|
| 10985 |
+
"learning_rate": 9.938983050847458e-06,
|
| 10986 |
+
"loss": 1.3203,
|
| 10987 |
+
"step": 15340
|
| 10988 |
+
},
|
| 10989 |
+
{
|
| 10990 |
+
"epoch": 0.38,
|
| 10991 |
+
"grad_norm": 3.256084442138672,
|
| 10992 |
+
"learning_rate": 9.93220338983051e-06,
|
| 10993 |
+
"loss": 1.3317,
|
| 10994 |
+
"step": 15350
|
| 10995 |
+
},
|
| 10996 |
+
{
|
| 10997 |
+
"epoch": 0.38,
|
| 10998 |
+
"grad_norm": 7.33984899520874,
|
| 10999 |
+
"learning_rate": 9.92542372881356e-06,
|
| 11000 |
+
"loss": 1.2692,
|
| 11001 |
+
"step": 15360
|
| 11002 |
+
},
|
| 11003 |
+
{
|
| 11004 |
+
"epoch": 0.38,
|
| 11005 |
+
"grad_norm": 5.8543877601623535,
|
| 11006 |
+
"learning_rate": 9.918644067796611e-06,
|
| 11007 |
+
"loss": 1.4074,
|
| 11008 |
+
"step": 15370
|
| 11009 |
+
},
|
| 11010 |
+
{
|
| 11011 |
+
"epoch": 0.38,
|
| 11012 |
+
"grad_norm": 3.914440870285034,
|
| 11013 |
+
"learning_rate": 9.911864406779662e-06,
|
| 11014 |
+
"loss": 1.4998,
|
| 11015 |
+
"step": 15380
|
| 11016 |
+
},
|
| 11017 |
+
{
|
| 11018 |
+
"epoch": 0.38,
|
| 11019 |
+
"grad_norm": 4.911335468292236,
|
| 11020 |
+
"learning_rate": 9.905084745762714e-06,
|
| 11021 |
+
"loss": 1.2676,
|
| 11022 |
+
"step": 15390
|
| 11023 |
+
},
|
| 11024 |
+
{
|
| 11025 |
+
"epoch": 0.39,
|
| 11026 |
+
"grad_norm": 2.9057512283325195,
|
| 11027 |
+
"learning_rate": 9.898305084745763e-06,
|
| 11028 |
+
"loss": 1.4411,
|
| 11029 |
+
"step": 15400
|
| 11030 |
+
},
|
| 11031 |
+
{
|
| 11032 |
+
"epoch": 0.39,
|
| 11033 |
+
"grad_norm": 6.507194519042969,
|
| 11034 |
+
"learning_rate": 9.891525423728813e-06,
|
| 11035 |
+
"loss": 1.3075,
|
| 11036 |
+
"step": 15410
|
| 11037 |
+
},
|
| 11038 |
+
{
|
| 11039 |
+
"epoch": 0.39,
|
| 11040 |
+
"grad_norm": 2.6351258754730225,
|
| 11041 |
+
"learning_rate": 9.884745762711864e-06,
|
| 11042 |
+
"loss": 1.368,
|
| 11043 |
+
"step": 15420
|
| 11044 |
+
},
|
| 11045 |
+
{
|
| 11046 |
+
"epoch": 0.39,
|
| 11047 |
+
"grad_norm": 10.542738914489746,
|
| 11048 |
+
"learning_rate": 9.877966101694916e-06,
|
| 11049 |
+
"loss": 1.3388,
|
| 11050 |
+
"step": 15430
|
| 11051 |
+
},
|
| 11052 |
+
{
|
| 11053 |
+
"epoch": 0.39,
|
| 11054 |
+
"grad_norm": 5.773244857788086,
|
| 11055 |
+
"learning_rate": 9.871186440677967e-06,
|
| 11056 |
+
"loss": 1.3478,
|
| 11057 |
+
"step": 15440
|
| 11058 |
+
},
|
| 11059 |
+
{
|
| 11060 |
+
"epoch": 0.39,
|
| 11061 |
+
"grad_norm": 3.678858757019043,
|
| 11062 |
+
"learning_rate": 9.864406779661017e-06,
|
| 11063 |
+
"loss": 1.3866,
|
| 11064 |
+
"step": 15450
|
| 11065 |
+
},
|
| 11066 |
+
{
|
| 11067 |
+
"epoch": 0.39,
|
| 11068 |
+
"grad_norm": 6.312896251678467,
|
| 11069 |
+
"learning_rate": 9.857627118644068e-06,
|
| 11070 |
+
"loss": 1.418,
|
| 11071 |
+
"step": 15460
|
| 11072 |
+
},
|
| 11073 |
+
{
|
| 11074 |
+
"epoch": 0.39,
|
| 11075 |
+
"grad_norm": 6.026124954223633,
|
| 11076 |
+
"learning_rate": 9.85084745762712e-06,
|
| 11077 |
+
"loss": 1.584,
|
| 11078 |
+
"step": 15470
|
| 11079 |
+
},
|
| 11080 |
+
{
|
| 11081 |
+
"epoch": 0.39,
|
| 11082 |
+
"grad_norm": 3.8132076263427734,
|
| 11083 |
+
"learning_rate": 9.844067796610171e-06,
|
| 11084 |
+
"loss": 1.4433,
|
| 11085 |
+
"step": 15480
|
| 11086 |
+
},
|
| 11087 |
+
{
|
| 11088 |
+
"epoch": 0.39,
|
| 11089 |
+
"grad_norm": 3.9010612964630127,
|
| 11090 |
+
"learning_rate": 9.837288135593221e-06,
|
| 11091 |
+
"loss": 1.3962,
|
| 11092 |
+
"step": 15490
|
| 11093 |
+
},
|
| 11094 |
+
{
|
| 11095 |
+
"epoch": 0.39,
|
| 11096 |
+
"grad_norm": 2.6610724925994873,
|
| 11097 |
+
"learning_rate": 9.830508474576272e-06,
|
| 11098 |
+
"loss": 1.3506,
|
| 11099 |
+
"step": 15500
|
| 11100 |
+
},
|
| 11101 |
+
{
|
| 11102 |
+
"epoch": 0.39,
|
| 11103 |
+
"eval_loss": 1.4206969738006592,
|
| 11104 |
+
"eval_runtime": 66.1225,
|
| 11105 |
+
"eval_samples_per_second": 15.123,
|
| 11106 |
+
"eval_steps_per_second": 15.123,
|
| 11107 |
+
"step": 15500
|
| 11108 |
+
},
|
| 11109 |
+
{
|
| 11110 |
+
"epoch": 0.39,
|
| 11111 |
+
"grad_norm": 10.500436782836914,
|
| 11112 |
+
"learning_rate": 9.823728813559322e-06,
|
| 11113 |
+
"loss": 1.4373,
|
| 11114 |
+
"step": 15510
|
| 11115 |
+
},
|
| 11116 |
+
{
|
| 11117 |
+
"epoch": 0.39,
|
| 11118 |
+
"grad_norm": 4.49819278717041,
|
| 11119 |
+
"learning_rate": 9.816949152542373e-06,
|
| 11120 |
+
"loss": 1.3353,
|
| 11121 |
+
"step": 15520
|
| 11122 |
+
},
|
| 11123 |
+
{
|
| 11124 |
+
"epoch": 0.39,
|
| 11125 |
+
"grad_norm": 4.517664909362793,
|
| 11126 |
+
"learning_rate": 9.810169491525425e-06,
|
| 11127 |
+
"loss": 1.4062,
|
| 11128 |
+
"step": 15530
|
| 11129 |
+
},
|
| 11130 |
+
{
|
| 11131 |
+
"epoch": 0.39,
|
| 11132 |
+
"grad_norm": 11.253808975219727,
|
| 11133 |
+
"learning_rate": 9.803389830508474e-06,
|
| 11134 |
+
"loss": 1.4507,
|
| 11135 |
+
"step": 15540
|
| 11136 |
+
},
|
| 11137 |
+
{
|
| 11138 |
+
"epoch": 0.39,
|
| 11139 |
+
"grad_norm": 2.5719611644744873,
|
| 11140 |
+
"learning_rate": 9.796610169491526e-06,
|
| 11141 |
+
"loss": 1.2784,
|
| 11142 |
+
"step": 15550
|
| 11143 |
+
},
|
| 11144 |
+
{
|
| 11145 |
+
"epoch": 0.39,
|
| 11146 |
+
"grad_norm": 2.9650001525878906,
|
| 11147 |
+
"learning_rate": 9.789830508474577e-06,
|
| 11148 |
+
"loss": 1.3021,
|
| 11149 |
+
"step": 15560
|
| 11150 |
+
},
|
| 11151 |
+
{
|
| 11152 |
+
"epoch": 0.39,
|
| 11153 |
+
"grad_norm": 4.1970319747924805,
|
| 11154 |
+
"learning_rate": 9.783050847457629e-06,
|
| 11155 |
+
"loss": 1.4252,
|
| 11156 |
+
"step": 15570
|
| 11157 |
+
},
|
| 11158 |
+
{
|
| 11159 |
+
"epoch": 0.39,
|
| 11160 |
+
"grad_norm": 1.926850438117981,
|
| 11161 |
+
"learning_rate": 9.776271186440678e-06,
|
| 11162 |
+
"loss": 1.2455,
|
| 11163 |
+
"step": 15580
|
| 11164 |
+
},
|
| 11165 |
+
{
|
| 11166 |
+
"epoch": 0.39,
|
| 11167 |
+
"grad_norm": 8.753373146057129,
|
| 11168 |
+
"learning_rate": 9.76949152542373e-06,
|
| 11169 |
+
"loss": 1.1841,
|
| 11170 |
+
"step": 15590
|
| 11171 |
+
},
|
| 11172 |
+
{
|
| 11173 |
+
"epoch": 0.39,
|
| 11174 |
+
"grad_norm": 11.411351203918457,
|
| 11175 |
+
"learning_rate": 9.762711864406781e-06,
|
| 11176 |
+
"loss": 1.391,
|
| 11177 |
+
"step": 15600
|
| 11178 |
+
},
|
| 11179 |
+
{
|
| 11180 |
+
"epoch": 0.39,
|
| 11181 |
+
"grad_norm": 9.503700256347656,
|
| 11182 |
+
"learning_rate": 9.755932203389833e-06,
|
| 11183 |
+
"loss": 1.4071,
|
| 11184 |
+
"step": 15610
|
| 11185 |
+
},
|
| 11186 |
+
{
|
| 11187 |
+
"epoch": 0.39,
|
| 11188 |
+
"grad_norm": 3.8297741413116455,
|
| 11189 |
+
"learning_rate": 9.749152542372882e-06,
|
| 11190 |
+
"loss": 1.4042,
|
| 11191 |
+
"step": 15620
|
| 11192 |
+
},
|
| 11193 |
+
{
|
| 11194 |
+
"epoch": 0.39,
|
| 11195 |
+
"grad_norm": 7.153141498565674,
|
| 11196 |
+
"learning_rate": 9.742372881355932e-06,
|
| 11197 |
+
"loss": 1.3474,
|
| 11198 |
+
"step": 15630
|
| 11199 |
+
},
|
| 11200 |
+
{
|
| 11201 |
+
"epoch": 0.39,
|
| 11202 |
+
"grad_norm": 2.806121826171875,
|
| 11203 |
+
"learning_rate": 9.735593220338983e-06,
|
| 11204 |
+
"loss": 1.4126,
|
| 11205 |
+
"step": 15640
|
| 11206 |
+
},
|
| 11207 |
+
{
|
| 11208 |
+
"epoch": 0.39,
|
| 11209 |
+
"grad_norm": 7.874058246612549,
|
| 11210 |
+
"learning_rate": 9.728813559322035e-06,
|
| 11211 |
+
"loss": 1.5157,
|
| 11212 |
+
"step": 15650
|
| 11213 |
+
},
|
| 11214 |
+
{
|
| 11215 |
+
"epoch": 0.39,
|
| 11216 |
+
"grad_norm": 3.6346802711486816,
|
| 11217 |
+
"learning_rate": 9.722033898305086e-06,
|
| 11218 |
+
"loss": 1.3617,
|
| 11219 |
+
"step": 15660
|
| 11220 |
+
},
|
| 11221 |
+
{
|
| 11222 |
+
"epoch": 0.39,
|
| 11223 |
+
"grad_norm": 19.185550689697266,
|
| 11224 |
+
"learning_rate": 9.715254237288136e-06,
|
| 11225 |
+
"loss": 1.4801,
|
| 11226 |
+
"step": 15670
|
| 11227 |
+
},
|
| 11228 |
+
{
|
| 11229 |
+
"epoch": 0.39,
|
| 11230 |
+
"grad_norm": 15.9682035446167,
|
| 11231 |
+
"learning_rate": 9.708474576271187e-06,
|
| 11232 |
+
"loss": 1.1923,
|
| 11233 |
+
"step": 15680
|
| 11234 |
+
},
|
| 11235 |
+
{
|
| 11236 |
+
"epoch": 0.39,
|
| 11237 |
+
"grad_norm": 2.7673888206481934,
|
| 11238 |
+
"learning_rate": 9.701694915254239e-06,
|
| 11239 |
+
"loss": 1.5568,
|
| 11240 |
+
"step": 15690
|
| 11241 |
+
},
|
| 11242 |
+
{
|
| 11243 |
+
"epoch": 0.39,
|
| 11244 |
+
"grad_norm": 7.633556842803955,
|
| 11245 |
+
"learning_rate": 9.69491525423729e-06,
|
| 11246 |
+
"loss": 1.4391,
|
| 11247 |
+
"step": 15700
|
| 11248 |
+
},
|
| 11249 |
+
{
|
| 11250 |
+
"epoch": 0.39,
|
| 11251 |
+
"grad_norm": 11.961231231689453,
|
| 11252 |
+
"learning_rate": 9.68813559322034e-06,
|
| 11253 |
+
"loss": 1.4212,
|
| 11254 |
+
"step": 15710
|
| 11255 |
+
},
|
| 11256 |
+
{
|
| 11257 |
+
"epoch": 0.39,
|
| 11258 |
+
"grad_norm": 2.426793098449707,
|
| 11259 |
+
"learning_rate": 9.68135593220339e-06,
|
| 11260 |
+
"loss": 1.1991,
|
| 11261 |
+
"step": 15720
|
| 11262 |
+
},
|
| 11263 |
+
{
|
| 11264 |
+
"epoch": 0.39,
|
| 11265 |
+
"grad_norm": 7.8566107749938965,
|
| 11266 |
+
"learning_rate": 9.674576271186441e-06,
|
| 11267 |
+
"loss": 1.2991,
|
| 11268 |
+
"step": 15730
|
| 11269 |
+
},
|
| 11270 |
+
{
|
| 11271 |
+
"epoch": 0.39,
|
| 11272 |
+
"grad_norm": 10.158808708190918,
|
| 11273 |
+
"learning_rate": 9.667796610169492e-06,
|
| 11274 |
+
"loss": 1.4826,
|
| 11275 |
+
"step": 15740
|
| 11276 |
+
},
|
| 11277 |
+
{
|
| 11278 |
+
"epoch": 0.39,
|
| 11279 |
+
"grad_norm": 6.697884559631348,
|
| 11280 |
+
"learning_rate": 9.661016949152544e-06,
|
| 11281 |
+
"loss": 1.4448,
|
| 11282 |
+
"step": 15750
|
| 11283 |
+
},
|
| 11284 |
+
{
|
| 11285 |
+
"epoch": 0.39,
|
| 11286 |
+
"grad_norm": 3.4867446422576904,
|
| 11287 |
+
"learning_rate": 9.654237288135593e-06,
|
| 11288 |
+
"loss": 1.4839,
|
| 11289 |
+
"step": 15760
|
| 11290 |
+
},
|
| 11291 |
+
{
|
| 11292 |
+
"epoch": 0.39,
|
| 11293 |
+
"grad_norm": 1.6158361434936523,
|
| 11294 |
+
"learning_rate": 9.647457627118645e-06,
|
| 11295 |
+
"loss": 1.2674,
|
| 11296 |
+
"step": 15770
|
| 11297 |
+
},
|
| 11298 |
+
{
|
| 11299 |
+
"epoch": 0.39,
|
| 11300 |
+
"grad_norm": 3.913695812225342,
|
| 11301 |
+
"learning_rate": 9.640677966101696e-06,
|
| 11302 |
+
"loss": 1.3351,
|
| 11303 |
+
"step": 15780
|
| 11304 |
+
},
|
| 11305 |
+
{
|
| 11306 |
+
"epoch": 0.39,
|
| 11307 |
+
"grad_norm": 1.9007806777954102,
|
| 11308 |
+
"learning_rate": 9.633898305084746e-06,
|
| 11309 |
+
"loss": 1.3478,
|
| 11310 |
+
"step": 15790
|
| 11311 |
+
},
|
| 11312 |
+
{
|
| 11313 |
+
"epoch": 0.4,
|
| 11314 |
+
"grad_norm": 2.1645798683166504,
|
| 11315 |
+
"learning_rate": 9.627118644067797e-06,
|
| 11316 |
+
"loss": 1.3018,
|
| 11317 |
+
"step": 15800
|
| 11318 |
+
},
|
| 11319 |
+
{
|
| 11320 |
+
"epoch": 0.4,
|
| 11321 |
+
"grad_norm": 3.865454912185669,
|
| 11322 |
+
"learning_rate": 9.620338983050849e-06,
|
| 11323 |
+
"loss": 1.5398,
|
| 11324 |
+
"step": 15810
|
| 11325 |
+
},
|
| 11326 |
+
{
|
| 11327 |
+
"epoch": 0.4,
|
| 11328 |
+
"grad_norm": 4.120809555053711,
|
| 11329 |
+
"learning_rate": 9.6135593220339e-06,
|
| 11330 |
+
"loss": 1.4546,
|
| 11331 |
+
"step": 15820
|
| 11332 |
+
},
|
| 11333 |
+
{
|
| 11334 |
+
"epoch": 0.4,
|
| 11335 |
+
"grad_norm": 5.410675525665283,
|
| 11336 |
+
"learning_rate": 9.60677966101695e-06,
|
| 11337 |
+
"loss": 1.4648,
|
| 11338 |
+
"step": 15830
|
| 11339 |
+
},
|
| 11340 |
+
{
|
| 11341 |
+
"epoch": 0.4,
|
| 11342 |
+
"grad_norm": 8.03393268585205,
|
| 11343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 11344 |
+
"loss": 1.2891,
|
| 11345 |
+
"step": 15840
|
| 11346 |
+
},
|
| 11347 |
+
{
|
| 11348 |
+
"epoch": 0.4,
|
| 11349 |
+
"grad_norm": 7.1069655418396,
|
| 11350 |
+
"learning_rate": 9.593220338983051e-06,
|
| 11351 |
+
"loss": 1.314,
|
| 11352 |
+
"step": 15850
|
| 11353 |
+
},
|
| 11354 |
+
{
|
| 11355 |
+
"epoch": 0.4,
|
| 11356 |
+
"grad_norm": 6.312065601348877,
|
| 11357 |
+
"learning_rate": 9.586440677966102e-06,
|
| 11358 |
+
"loss": 1.2608,
|
| 11359 |
+
"step": 15860
|
| 11360 |
+
},
|
| 11361 |
+
{
|
| 11362 |
+
"epoch": 0.4,
|
| 11363 |
+
"grad_norm": 3.499418258666992,
|
| 11364 |
+
"learning_rate": 9.579661016949154e-06,
|
| 11365 |
+
"loss": 1.3816,
|
| 11366 |
+
"step": 15870
|
| 11367 |
+
},
|
| 11368 |
+
{
|
| 11369 |
+
"epoch": 0.4,
|
| 11370 |
+
"grad_norm": 5.211281776428223,
|
| 11371 |
+
"learning_rate": 9.572881355932203e-06,
|
| 11372 |
+
"loss": 1.1021,
|
| 11373 |
+
"step": 15880
|
| 11374 |
+
},
|
| 11375 |
+
{
|
| 11376 |
+
"epoch": 0.4,
|
| 11377 |
+
"grad_norm": 4.372194290161133,
|
| 11378 |
+
"learning_rate": 9.566101694915255e-06,
|
| 11379 |
+
"loss": 1.4179,
|
| 11380 |
+
"step": 15890
|
| 11381 |
+
},
|
| 11382 |
+
{
|
| 11383 |
+
"epoch": 0.4,
|
| 11384 |
+
"grad_norm": 5.44555139541626,
|
| 11385 |
+
"learning_rate": 9.559322033898306e-06,
|
| 11386 |
+
"loss": 1.4848,
|
| 11387 |
+
"step": 15900
|
| 11388 |
+
},
|
| 11389 |
+
{
|
| 11390 |
+
"epoch": 0.4,
|
| 11391 |
+
"grad_norm": 3.483454942703247,
|
| 11392 |
+
"learning_rate": 9.552542372881358e-06,
|
| 11393 |
+
"loss": 1.2985,
|
| 11394 |
+
"step": 15910
|
| 11395 |
+
},
|
| 11396 |
+
{
|
| 11397 |
+
"epoch": 0.4,
|
| 11398 |
+
"grad_norm": 3.8721625804901123,
|
| 11399 |
+
"learning_rate": 9.545762711864407e-06,
|
| 11400 |
+
"loss": 1.4731,
|
| 11401 |
+
"step": 15920
|
| 11402 |
+
},
|
| 11403 |
+
{
|
| 11404 |
+
"epoch": 0.4,
|
| 11405 |
+
"grad_norm": 4.317465305328369,
|
| 11406 |
+
"learning_rate": 9.538983050847457e-06,
|
| 11407 |
+
"loss": 1.501,
|
| 11408 |
+
"step": 15930
|
| 11409 |
+
},
|
| 11410 |
+
{
|
| 11411 |
+
"epoch": 0.4,
|
| 11412 |
+
"grad_norm": 7.075793743133545,
|
| 11413 |
+
"learning_rate": 9.532203389830508e-06,
|
| 11414 |
+
"loss": 1.5138,
|
| 11415 |
+
"step": 15940
|
| 11416 |
+
},
|
| 11417 |
+
{
|
| 11418 |
+
"epoch": 0.4,
|
| 11419 |
+
"grad_norm": 5.606925964355469,
|
| 11420 |
+
"learning_rate": 9.52542372881356e-06,
|
| 11421 |
+
"loss": 1.3771,
|
| 11422 |
+
"step": 15950
|
| 11423 |
+
},
|
| 11424 |
+
{
|
| 11425 |
+
"epoch": 0.4,
|
| 11426 |
+
"grad_norm": 2.937833547592163,
|
| 11427 |
+
"learning_rate": 9.518644067796611e-06,
|
| 11428 |
+
"loss": 1.4764,
|
| 11429 |
+
"step": 15960
|
| 11430 |
+
},
|
| 11431 |
+
{
|
| 11432 |
+
"epoch": 0.4,
|
| 11433 |
+
"grad_norm": 4.39876651763916,
|
| 11434 |
+
"learning_rate": 9.511864406779661e-06,
|
| 11435 |
+
"loss": 1.4022,
|
| 11436 |
+
"step": 15970
|
| 11437 |
+
},
|
| 11438 |
+
{
|
| 11439 |
+
"epoch": 0.4,
|
| 11440 |
+
"grad_norm": 6.455230712890625,
|
| 11441 |
+
"learning_rate": 9.505084745762712e-06,
|
| 11442 |
+
"loss": 1.3569,
|
| 11443 |
+
"step": 15980
|
| 11444 |
+
},
|
| 11445 |
+
{
|
| 11446 |
+
"epoch": 0.4,
|
| 11447 |
+
"grad_norm": 4.89456033706665,
|
| 11448 |
+
"learning_rate": 9.498305084745764e-06,
|
| 11449 |
+
"loss": 1.3426,
|
| 11450 |
+
"step": 15990
|
| 11451 |
+
},
|
| 11452 |
+
{
|
| 11453 |
+
"epoch": 0.4,
|
| 11454 |
+
"grad_norm": 5.696287631988525,
|
| 11455 |
+
"learning_rate": 9.491525423728815e-06,
|
| 11456 |
+
"loss": 1.5381,
|
| 11457 |
+
"step": 16000
|
| 11458 |
+
},
|
| 11459 |
+
{
|
| 11460 |
+
"epoch": 0.4,
|
| 11461 |
+
"eval_loss": 1.3579893112182617,
|
| 11462 |
+
"eval_runtime": 66.2433,
|
| 11463 |
+
"eval_samples_per_second": 15.096,
|
| 11464 |
+
"eval_steps_per_second": 15.096,
|
| 11465 |
+
"step": 16000
|
| 11466 |
+
},
|
| 11467 |
+
{
|
| 11468 |
+
"epoch": 0.4,
|
| 11469 |
+
"grad_norm": 6.06011962890625,
|
| 11470 |
+
"learning_rate": 9.484745762711865e-06,
|
| 11471 |
+
"loss": 1.4422,
|
| 11472 |
+
"step": 16010
|
| 11473 |
+
},
|
| 11474 |
+
{
|
| 11475 |
+
"epoch": 0.4,
|
| 11476 |
+
"grad_norm": 0.9439699053764343,
|
| 11477 |
+
"learning_rate": 9.477966101694916e-06,
|
| 11478 |
+
"loss": 1.3661,
|
| 11479 |
+
"step": 16020
|
| 11480 |
+
},
|
| 11481 |
+
{
|
| 11482 |
+
"epoch": 0.4,
|
| 11483 |
+
"grad_norm": 2.9158008098602295,
|
| 11484 |
+
"learning_rate": 9.471186440677966e-06,
|
| 11485 |
+
"loss": 1.1349,
|
| 11486 |
+
"step": 16030
|
| 11487 |
+
},
|
| 11488 |
+
{
|
| 11489 |
+
"epoch": 0.4,
|
| 11490 |
+
"grad_norm": 3.3317673206329346,
|
| 11491 |
+
"learning_rate": 9.464406779661017e-06,
|
| 11492 |
+
"loss": 1.4161,
|
| 11493 |
+
"step": 16040
|
| 11494 |
+
},
|
| 11495 |
+
{
|
| 11496 |
+
"epoch": 0.4,
|
| 11497 |
+
"grad_norm": 2.866821050643921,
|
| 11498 |
+
"learning_rate": 9.457627118644069e-06,
|
| 11499 |
+
"loss": 1.2952,
|
| 11500 |
+
"step": 16050
|
| 11501 |
+
},
|
| 11502 |
+
{
|
| 11503 |
+
"epoch": 0.4,
|
| 11504 |
+
"grad_norm": 8.00590991973877,
|
| 11505 |
+
"learning_rate": 9.450847457627119e-06,
|
| 11506 |
+
"loss": 1.3493,
|
| 11507 |
+
"step": 16060
|
| 11508 |
+
},
|
| 11509 |
+
{
|
| 11510 |
+
"epoch": 0.4,
|
| 11511 |
+
"grad_norm": 2.1362667083740234,
|
| 11512 |
+
"learning_rate": 9.44406779661017e-06,
|
| 11513 |
+
"loss": 1.3385,
|
| 11514 |
+
"step": 16070
|
| 11515 |
+
},
|
| 11516 |
+
{
|
| 11517 |
+
"epoch": 0.4,
|
| 11518 |
+
"grad_norm": 4.567724227905273,
|
| 11519 |
+
"learning_rate": 9.437288135593221e-06,
|
| 11520 |
+
"loss": 1.2988,
|
| 11521 |
+
"step": 16080
|
| 11522 |
+
},
|
| 11523 |
+
{
|
| 11524 |
+
"epoch": 0.4,
|
| 11525 |
+
"grad_norm": 3.382605791091919,
|
| 11526 |
+
"learning_rate": 9.430508474576273e-06,
|
| 11527 |
+
"loss": 1.2131,
|
| 11528 |
+
"step": 16090
|
| 11529 |
+
},
|
| 11530 |
+
{
|
| 11531 |
+
"epoch": 0.4,
|
| 11532 |
+
"grad_norm": 13.581380844116211,
|
| 11533 |
+
"learning_rate": 9.423728813559322e-06,
|
| 11534 |
+
"loss": 1.3962,
|
| 11535 |
+
"step": 16100
|
| 11536 |
+
},
|
| 11537 |
+
{
|
| 11538 |
+
"epoch": 0.4,
|
| 11539 |
+
"grad_norm": 6.365314483642578,
|
| 11540 |
+
"learning_rate": 9.416949152542374e-06,
|
| 11541 |
+
"loss": 1.4928,
|
| 11542 |
+
"step": 16110
|
| 11543 |
+
},
|
| 11544 |
+
{
|
| 11545 |
+
"epoch": 0.4,
|
| 11546 |
+
"grad_norm": 5.413196563720703,
|
| 11547 |
+
"learning_rate": 9.410169491525425e-06,
|
| 11548 |
+
"loss": 1.2728,
|
| 11549 |
+
"step": 16120
|
| 11550 |
+
},
|
| 11551 |
+
{
|
| 11552 |
+
"epoch": 0.4,
|
| 11553 |
+
"grad_norm": 3.6787874698638916,
|
| 11554 |
+
"learning_rate": 9.403389830508477e-06,
|
| 11555 |
+
"loss": 1.1831,
|
| 11556 |
+
"step": 16130
|
| 11557 |
+
},
|
| 11558 |
+
{
|
| 11559 |
+
"epoch": 0.4,
|
| 11560 |
+
"grad_norm": 2.8850882053375244,
|
| 11561 |
+
"learning_rate": 9.396610169491526e-06,
|
| 11562 |
+
"loss": 1.3025,
|
| 11563 |
+
"step": 16140
|
| 11564 |
+
},
|
| 11565 |
+
{
|
| 11566 |
+
"epoch": 0.4,
|
| 11567 |
+
"grad_norm": 3.6025898456573486,
|
| 11568 |
+
"learning_rate": 9.389830508474576e-06,
|
| 11569 |
+
"loss": 1.3553,
|
| 11570 |
+
"step": 16150
|
| 11571 |
+
},
|
| 11572 |
+
{
|
| 11573 |
+
"epoch": 0.4,
|
| 11574 |
+
"grad_norm": 3.00011944770813,
|
| 11575 |
+
"learning_rate": 9.383050847457627e-06,
|
| 11576 |
+
"loss": 1.5127,
|
| 11577 |
+
"step": 16160
|
| 11578 |
+
},
|
| 11579 |
+
{
|
| 11580 |
+
"epoch": 0.4,
|
| 11581 |
+
"grad_norm": 4.33295202255249,
|
| 11582 |
+
"learning_rate": 9.376271186440679e-06,
|
| 11583 |
+
"loss": 1.5001,
|
| 11584 |
+
"step": 16170
|
| 11585 |
+
},
|
| 11586 |
+
{
|
| 11587 |
+
"epoch": 0.4,
|
| 11588 |
+
"grad_norm": 3.6982882022857666,
|
| 11589 |
+
"learning_rate": 9.36949152542373e-06,
|
| 11590 |
+
"loss": 1.4729,
|
| 11591 |
+
"step": 16180
|
| 11592 |
+
},
|
| 11593 |
+
{
|
| 11594 |
+
"epoch": 0.4,
|
| 11595 |
+
"grad_norm": 7.049408435821533,
|
| 11596 |
+
"learning_rate": 9.36271186440678e-06,
|
| 11597 |
+
"loss": 1.276,
|
| 11598 |
+
"step": 16190
|
| 11599 |
+
},
|
| 11600 |
+
{
|
| 11601 |
+
"epoch": 0.41,
|
| 11602 |
+
"grad_norm": 1.6569970846176147,
|
| 11603 |
+
"learning_rate": 9.355932203389831e-06,
|
| 11604 |
+
"loss": 1.3321,
|
| 11605 |
+
"step": 16200
|
| 11606 |
+
},
|
| 11607 |
+
{
|
| 11608 |
+
"epoch": 0.41,
|
| 11609 |
+
"grad_norm": 2.4531731605529785,
|
| 11610 |
+
"learning_rate": 9.349152542372883e-06,
|
| 11611 |
+
"loss": 1.435,
|
| 11612 |
+
"step": 16210
|
| 11613 |
+
},
|
| 11614 |
+
{
|
| 11615 |
+
"epoch": 0.41,
|
| 11616 |
+
"grad_norm": 2.776325225830078,
|
| 11617 |
+
"learning_rate": 9.342372881355934e-06,
|
| 11618 |
+
"loss": 1.2586,
|
| 11619 |
+
"step": 16220
|
| 11620 |
+
},
|
| 11621 |
+
{
|
| 11622 |
+
"epoch": 0.41,
|
| 11623 |
+
"grad_norm": 2.805823564529419,
|
| 11624 |
+
"learning_rate": 9.335593220338984e-06,
|
| 11625 |
+
"loss": 1.3393,
|
| 11626 |
+
"step": 16230
|
| 11627 |
+
},
|
| 11628 |
+
{
|
| 11629 |
+
"epoch": 0.41,
|
| 11630 |
+
"grad_norm": 2.3886725902557373,
|
| 11631 |
+
"learning_rate": 9.328813559322034e-06,
|
| 11632 |
+
"loss": 1.4241,
|
| 11633 |
+
"step": 16240
|
| 11634 |
+
},
|
| 11635 |
+
{
|
| 11636 |
+
"epoch": 0.41,
|
| 11637 |
+
"grad_norm": 10.983514785766602,
|
| 11638 |
+
"learning_rate": 9.322033898305085e-06,
|
| 11639 |
+
"loss": 1.5032,
|
| 11640 |
+
"step": 16250
|
| 11641 |
+
},
|
| 11642 |
+
{
|
| 11643 |
+
"epoch": 0.41,
|
| 11644 |
+
"grad_norm": 2.922609329223633,
|
| 11645 |
+
"learning_rate": 9.315254237288136e-06,
|
| 11646 |
+
"loss": 1.2289,
|
| 11647 |
+
"step": 16260
|
| 11648 |
+
},
|
| 11649 |
+
{
|
| 11650 |
+
"epoch": 0.41,
|
| 11651 |
+
"grad_norm": 3.623088836669922,
|
| 11652 |
+
"learning_rate": 9.308474576271188e-06,
|
| 11653 |
+
"loss": 1.2854,
|
| 11654 |
+
"step": 16270
|
| 11655 |
+
},
|
| 11656 |
+
{
|
| 11657 |
+
"epoch": 0.41,
|
| 11658 |
+
"grad_norm": 3.6580920219421387,
|
| 11659 |
+
"learning_rate": 9.301694915254237e-06,
|
| 11660 |
+
"loss": 1.3371,
|
| 11661 |
+
"step": 16280
|
| 11662 |
+
},
|
| 11663 |
+
{
|
| 11664 |
+
"epoch": 0.41,
|
| 11665 |
+
"grad_norm": 3.010573625564575,
|
| 11666 |
+
"learning_rate": 9.294915254237289e-06,
|
| 11667 |
+
"loss": 1.465,
|
| 11668 |
+
"step": 16290
|
| 11669 |
+
},
|
| 11670 |
+
{
|
| 11671 |
+
"epoch": 0.41,
|
| 11672 |
+
"grad_norm": 3.1502161026000977,
|
| 11673 |
+
"learning_rate": 9.28813559322034e-06,
|
| 11674 |
+
"loss": 1.2245,
|
| 11675 |
+
"step": 16300
|
| 11676 |
+
},
|
| 11677 |
+
{
|
| 11678 |
+
"epoch": 0.41,
|
| 11679 |
+
"grad_norm": 2.7607476711273193,
|
| 11680 |
+
"learning_rate": 9.28135593220339e-06,
|
| 11681 |
+
"loss": 1.4668,
|
| 11682 |
+
"step": 16310
|
| 11683 |
+
},
|
| 11684 |
+
{
|
| 11685 |
+
"epoch": 0.41,
|
| 11686 |
+
"grad_norm": 6.022114276885986,
|
| 11687 |
+
"learning_rate": 9.274576271186441e-06,
|
| 11688 |
+
"loss": 1.1924,
|
| 11689 |
+
"step": 16320
|
| 11690 |
+
},
|
| 11691 |
+
{
|
| 11692 |
+
"epoch": 0.41,
|
| 11693 |
+
"grad_norm": 5.824821949005127,
|
| 11694 |
+
"learning_rate": 9.267796610169493e-06,
|
| 11695 |
+
"loss": 1.3773,
|
| 11696 |
+
"step": 16330
|
| 11697 |
+
},
|
| 11698 |
+
{
|
| 11699 |
+
"epoch": 0.41,
|
| 11700 |
+
"grad_norm": 3.7949423789978027,
|
| 11701 |
+
"learning_rate": 9.261016949152544e-06,
|
| 11702 |
+
"loss": 1.3334,
|
| 11703 |
+
"step": 16340
|
| 11704 |
+
},
|
| 11705 |
+
{
|
| 11706 |
+
"epoch": 0.41,
|
| 11707 |
+
"grad_norm": 1.7965031862258911,
|
| 11708 |
+
"learning_rate": 9.254237288135594e-06,
|
| 11709 |
+
"loss": 1.5022,
|
| 11710 |
+
"step": 16350
|
| 11711 |
+
},
|
| 11712 |
+
{
|
| 11713 |
+
"epoch": 0.41,
|
| 11714 |
+
"grad_norm": 4.245846271514893,
|
| 11715 |
+
"learning_rate": 9.247457627118645e-06,
|
| 11716 |
+
"loss": 1.2312,
|
| 11717 |
+
"step": 16360
|
| 11718 |
+
},
|
| 11719 |
+
{
|
| 11720 |
+
"epoch": 0.41,
|
| 11721 |
+
"grad_norm": 7.283750057220459,
|
| 11722 |
+
"learning_rate": 9.240677966101695e-06,
|
| 11723 |
+
"loss": 1.2448,
|
| 11724 |
+
"step": 16370
|
| 11725 |
+
},
|
| 11726 |
+
{
|
| 11727 |
+
"epoch": 0.41,
|
| 11728 |
+
"grad_norm": 3.51558780670166,
|
| 11729 |
+
"learning_rate": 9.233898305084746e-06,
|
| 11730 |
+
"loss": 1.4823,
|
| 11731 |
+
"step": 16380
|
| 11732 |
+
},
|
| 11733 |
+
{
|
| 11734 |
+
"epoch": 0.41,
|
| 11735 |
+
"grad_norm": 7.842286586761475,
|
| 11736 |
+
"learning_rate": 9.227118644067798e-06,
|
| 11737 |
+
"loss": 1.3931,
|
| 11738 |
+
"step": 16390
|
| 11739 |
+
},
|
| 11740 |
+
{
|
| 11741 |
+
"epoch": 0.41,
|
| 11742 |
+
"grad_norm": 2.6128389835357666,
|
| 11743 |
+
"learning_rate": 9.220338983050847e-06,
|
| 11744 |
+
"loss": 1.4359,
|
| 11745 |
+
"step": 16400
|
| 11746 |
+
},
|
| 11747 |
+
{
|
| 11748 |
+
"epoch": 0.41,
|
| 11749 |
+
"grad_norm": 2.5427112579345703,
|
| 11750 |
+
"learning_rate": 9.213559322033899e-06,
|
| 11751 |
+
"loss": 1.4364,
|
| 11752 |
+
"step": 16410
|
| 11753 |
+
},
|
| 11754 |
+
{
|
| 11755 |
+
"epoch": 0.41,
|
| 11756 |
+
"grad_norm": 3.0635411739349365,
|
| 11757 |
+
"learning_rate": 9.20677966101695e-06,
|
| 11758 |
+
"loss": 1.4804,
|
| 11759 |
+
"step": 16420
|
| 11760 |
+
},
|
| 11761 |
+
{
|
| 11762 |
+
"epoch": 0.41,
|
| 11763 |
+
"grad_norm": 4.00955057144165,
|
| 11764 |
+
"learning_rate": 9.200000000000002e-06,
|
| 11765 |
+
"loss": 1.1819,
|
| 11766 |
+
"step": 16430
|
| 11767 |
+
},
|
| 11768 |
+
{
|
| 11769 |
+
"epoch": 0.41,
|
| 11770 |
+
"grad_norm": 2.0021607875823975,
|
| 11771 |
+
"learning_rate": 9.193220338983051e-06,
|
| 11772 |
+
"loss": 1.312,
|
| 11773 |
+
"step": 16440
|
| 11774 |
+
},
|
| 11775 |
+
{
|
| 11776 |
+
"epoch": 0.41,
|
| 11777 |
+
"grad_norm": 6.373915672302246,
|
| 11778 |
+
"learning_rate": 9.186440677966101e-06,
|
| 11779 |
+
"loss": 1.5638,
|
| 11780 |
+
"step": 16450
|
| 11781 |
+
},
|
| 11782 |
+
{
|
| 11783 |
+
"epoch": 0.41,
|
| 11784 |
+
"grad_norm": 7.486032962799072,
|
| 11785 |
+
"learning_rate": 9.179661016949153e-06,
|
| 11786 |
+
"loss": 1.5156,
|
| 11787 |
+
"step": 16460
|
| 11788 |
+
},
|
| 11789 |
+
{
|
| 11790 |
+
"epoch": 0.41,
|
| 11791 |
+
"grad_norm": 7.894711971282959,
|
| 11792 |
+
"learning_rate": 9.172881355932204e-06,
|
| 11793 |
+
"loss": 1.3042,
|
| 11794 |
+
"step": 16470
|
| 11795 |
+
},
|
| 11796 |
+
{
|
| 11797 |
+
"epoch": 0.41,
|
| 11798 |
+
"grad_norm": 4.769121170043945,
|
| 11799 |
+
"learning_rate": 9.166101694915255e-06,
|
| 11800 |
+
"loss": 1.5031,
|
| 11801 |
+
"step": 16480
|
| 11802 |
+
},
|
| 11803 |
+
{
|
| 11804 |
+
"epoch": 0.41,
|
| 11805 |
+
"grad_norm": 5.979970455169678,
|
| 11806 |
+
"learning_rate": 9.159322033898305e-06,
|
| 11807 |
+
"loss": 1.2542,
|
| 11808 |
+
"step": 16490
|
| 11809 |
+
},
|
| 11810 |
+
{
|
| 11811 |
+
"epoch": 0.41,
|
| 11812 |
+
"grad_norm": 6.4072651863098145,
|
| 11813 |
+
"learning_rate": 9.152542372881356e-06,
|
| 11814 |
+
"loss": 1.319,
|
| 11815 |
+
"step": 16500
|
| 11816 |
+
},
|
| 11817 |
+
{
|
| 11818 |
+
"epoch": 0.41,
|
| 11819 |
+
"eval_loss": 1.3969231843948364,
|
| 11820 |
+
"eval_runtime": 66.1684,
|
| 11821 |
+
"eval_samples_per_second": 15.113,
|
| 11822 |
+
"eval_steps_per_second": 15.113,
|
| 11823 |
+
"step": 16500
|
| 11824 |
+
},
|
| 11825 |
+
{
|
| 11826 |
+
"epoch": 0.41,
|
| 11827 |
+
"grad_norm": 6.439510822296143,
|
| 11828 |
+
"learning_rate": 9.145762711864408e-06,
|
| 11829 |
+
"loss": 1.6106,
|
| 11830 |
+
"step": 16510
|
| 11831 |
+
},
|
| 11832 |
+
{
|
| 11833 |
+
"epoch": 0.41,
|
| 11834 |
+
"grad_norm": 3.3735623359680176,
|
| 11835 |
+
"learning_rate": 9.13898305084746e-06,
|
| 11836 |
+
"loss": 1.3179,
|
| 11837 |
+
"step": 16520
|
| 11838 |
+
},
|
| 11839 |
+
{
|
| 11840 |
+
"epoch": 0.41,
|
| 11841 |
+
"grad_norm": 4.679867267608643,
|
| 11842 |
+
"learning_rate": 9.132203389830509e-06,
|
| 11843 |
+
"loss": 1.465,
|
| 11844 |
+
"step": 16530
|
| 11845 |
+
},
|
| 11846 |
+
{
|
| 11847 |
+
"epoch": 0.41,
|
| 11848 |
+
"grad_norm": 3.878396987915039,
|
| 11849 |
+
"learning_rate": 9.12542372881356e-06,
|
| 11850 |
+
"loss": 1.2999,
|
| 11851 |
+
"step": 16540
|
| 11852 |
+
},
|
| 11853 |
+
{
|
| 11854 |
+
"epoch": 0.41,
|
| 11855 |
+
"grad_norm": 5.397853374481201,
|
| 11856 |
+
"learning_rate": 9.11864406779661e-06,
|
| 11857 |
+
"loss": 1.3575,
|
| 11858 |
+
"step": 16550
|
| 11859 |
+
},
|
| 11860 |
+
{
|
| 11861 |
+
"epoch": 0.41,
|
| 11862 |
+
"grad_norm": 3.2240700721740723,
|
| 11863 |
+
"learning_rate": 9.111864406779661e-06,
|
| 11864 |
+
"loss": 1.5086,
|
| 11865 |
+
"step": 16560
|
| 11866 |
+
},
|
| 11867 |
+
{
|
| 11868 |
+
"epoch": 0.41,
|
| 11869 |
+
"grad_norm": 4.239155292510986,
|
| 11870 |
+
"learning_rate": 9.105084745762713e-06,
|
| 11871 |
+
"loss": 1.3695,
|
| 11872 |
+
"step": 16570
|
| 11873 |
+
},
|
| 11874 |
+
{
|
| 11875 |
+
"epoch": 0.41,
|
| 11876 |
+
"grad_norm": 9.791718482971191,
|
| 11877 |
+
"learning_rate": 9.098305084745763e-06,
|
| 11878 |
+
"loss": 1.4928,
|
| 11879 |
+
"step": 16580
|
| 11880 |
+
},
|
| 11881 |
+
{
|
| 11882 |
+
"epoch": 0.41,
|
| 11883 |
+
"grad_norm": 4.839406967163086,
|
| 11884 |
+
"learning_rate": 9.091525423728814e-06,
|
| 11885 |
+
"loss": 1.4183,
|
| 11886 |
+
"step": 16590
|
| 11887 |
+
},
|
| 11888 |
+
{
|
| 11889 |
+
"epoch": 0.41,
|
| 11890 |
+
"grad_norm": 3.728440046310425,
|
| 11891 |
+
"learning_rate": 9.084745762711865e-06,
|
| 11892 |
+
"loss": 1.2819,
|
| 11893 |
+
"step": 16600
|
| 11894 |
+
},
|
| 11895 |
+
{
|
| 11896 |
+
"epoch": 0.42,
|
| 11897 |
+
"grad_norm": 2.5298233032226562,
|
| 11898 |
+
"learning_rate": 9.077966101694917e-06,
|
| 11899 |
+
"loss": 1.3922,
|
| 11900 |
+
"step": 16610
|
| 11901 |
+
},
|
| 11902 |
+
{
|
| 11903 |
+
"epoch": 0.42,
|
| 11904 |
+
"grad_norm": 4.599859714508057,
|
| 11905 |
+
"learning_rate": 9.071186440677966e-06,
|
| 11906 |
+
"loss": 1.4627,
|
| 11907 |
+
"step": 16620
|
| 11908 |
+
},
|
| 11909 |
+
{
|
| 11910 |
+
"epoch": 0.42,
|
| 11911 |
+
"grad_norm": 2.5592544078826904,
|
| 11912 |
+
"learning_rate": 9.064406779661018e-06,
|
| 11913 |
+
"loss": 1.544,
|
| 11914 |
+
"step": 16630
|
| 11915 |
+
},
|
| 11916 |
+
{
|
| 11917 |
+
"epoch": 0.42,
|
| 11918 |
+
"grad_norm": 8.281981468200684,
|
| 11919 |
+
"learning_rate": 9.05762711864407e-06,
|
| 11920 |
+
"loss": 1.4084,
|
| 11921 |
+
"step": 16640
|
| 11922 |
+
},
|
| 11923 |
+
{
|
| 11924 |
+
"epoch": 0.42,
|
| 11925 |
+
"grad_norm": 5.964704990386963,
|
| 11926 |
+
"learning_rate": 9.05084745762712e-06,
|
| 11927 |
+
"loss": 1.4186,
|
| 11928 |
+
"step": 16650
|
| 11929 |
+
},
|
| 11930 |
+
{
|
| 11931 |
+
"epoch": 0.42,
|
| 11932 |
+
"grad_norm": 4.577818393707275,
|
| 11933 |
+
"learning_rate": 9.04406779661017e-06,
|
| 11934 |
+
"loss": 1.2691,
|
| 11935 |
+
"step": 16660
|
| 11936 |
+
},
|
| 11937 |
+
{
|
| 11938 |
+
"epoch": 0.42,
|
| 11939 |
+
"grad_norm": 3.4889256954193115,
|
| 11940 |
+
"learning_rate": 9.03728813559322e-06,
|
| 11941 |
+
"loss": 1.3494,
|
| 11942 |
+
"step": 16670
|
| 11943 |
+
},
|
| 11944 |
+
{
|
| 11945 |
+
"epoch": 0.42,
|
| 11946 |
+
"grad_norm": 4.458433628082275,
|
| 11947 |
+
"learning_rate": 9.030508474576271e-06,
|
| 11948 |
+
"loss": 1.4639,
|
| 11949 |
+
"step": 16680
|
| 11950 |
+
},
|
| 11951 |
+
{
|
| 11952 |
+
"epoch": 0.42,
|
| 11953 |
+
"grad_norm": 5.813331127166748,
|
| 11954 |
+
"learning_rate": 9.023728813559323e-06,
|
| 11955 |
+
"loss": 1.2206,
|
| 11956 |
+
"step": 16690
|
| 11957 |
+
},
|
| 11958 |
+
{
|
| 11959 |
+
"epoch": 0.42,
|
| 11960 |
+
"grad_norm": 2.583115339279175,
|
| 11961 |
+
"learning_rate": 9.016949152542374e-06,
|
| 11962 |
+
"loss": 1.4299,
|
| 11963 |
+
"step": 16700
|
| 11964 |
+
},
|
| 11965 |
+
{
|
| 11966 |
+
"epoch": 0.42,
|
| 11967 |
+
"grad_norm": 7.3467206954956055,
|
| 11968 |
+
"learning_rate": 9.010169491525424e-06,
|
| 11969 |
+
"loss": 1.3001,
|
| 11970 |
+
"step": 16710
|
| 11971 |
+
},
|
| 11972 |
+
{
|
| 11973 |
+
"epoch": 0.42,
|
| 11974 |
+
"grad_norm": 3.896238088607788,
|
| 11975 |
+
"learning_rate": 9.003389830508475e-06,
|
| 11976 |
+
"loss": 1.3541,
|
| 11977 |
+
"step": 16720
|
| 11978 |
+
},
|
| 11979 |
+
{
|
| 11980 |
+
"epoch": 0.42,
|
| 11981 |
+
"grad_norm": 3.967496156692505,
|
| 11982 |
+
"learning_rate": 8.996610169491527e-06,
|
| 11983 |
+
"loss": 1.3315,
|
| 11984 |
+
"step": 16730
|
| 11985 |
+
},
|
| 11986 |
+
{
|
| 11987 |
+
"epoch": 0.42,
|
| 11988 |
+
"grad_norm": 2.9316656589508057,
|
| 11989 |
+
"learning_rate": 8.989830508474578e-06,
|
| 11990 |
+
"loss": 1.3935,
|
| 11991 |
+
"step": 16740
|
| 11992 |
+
},
|
| 11993 |
+
{
|
| 11994 |
+
"epoch": 0.42,
|
| 11995 |
+
"grad_norm": 1.619555950164795,
|
| 11996 |
+
"learning_rate": 8.983050847457628e-06,
|
| 11997 |
+
"loss": 1.2879,
|
| 11998 |
+
"step": 16750
|
| 11999 |
+
},
|
| 12000 |
+
{
|
| 12001 |
+
"epoch": 0.42,
|
| 12002 |
+
"grad_norm": 3.6497581005096436,
|
| 12003 |
+
"learning_rate": 8.976271186440678e-06,
|
| 12004 |
+
"loss": 1.5195,
|
| 12005 |
+
"step": 16760
|
| 12006 |
+
},
|
| 12007 |
+
{
|
| 12008 |
+
"epoch": 0.42,
|
| 12009 |
+
"grad_norm": 5.33650016784668,
|
| 12010 |
+
"learning_rate": 8.969491525423729e-06,
|
| 12011 |
+
"loss": 1.4568,
|
| 12012 |
+
"step": 16770
|
| 12013 |
+
},
|
| 12014 |
+
{
|
| 12015 |
+
"epoch": 0.42,
|
| 12016 |
+
"grad_norm": 7.370872974395752,
|
| 12017 |
+
"learning_rate": 8.96271186440678e-06,
|
| 12018 |
+
"loss": 1.5122,
|
| 12019 |
+
"step": 16780
|
| 12020 |
+
},
|
| 12021 |
+
{
|
| 12022 |
+
"epoch": 0.42,
|
| 12023 |
+
"grad_norm": 6.295271873474121,
|
| 12024 |
+
"learning_rate": 8.955932203389832e-06,
|
| 12025 |
+
"loss": 1.177,
|
| 12026 |
+
"step": 16790
|
| 12027 |
+
},
|
| 12028 |
+
{
|
| 12029 |
+
"epoch": 0.42,
|
| 12030 |
+
"grad_norm": 7.165903091430664,
|
| 12031 |
+
"learning_rate": 8.949152542372881e-06,
|
| 12032 |
+
"loss": 1.3597,
|
| 12033 |
+
"step": 16800
|
| 12034 |
+
},
|
| 12035 |
+
{
|
| 12036 |
+
"epoch": 0.42,
|
| 12037 |
+
"grad_norm": 4.152050018310547,
|
| 12038 |
+
"learning_rate": 8.942372881355933e-06,
|
| 12039 |
+
"loss": 1.3837,
|
| 12040 |
+
"step": 16810
|
| 12041 |
+
},
|
| 12042 |
+
{
|
| 12043 |
+
"epoch": 0.42,
|
| 12044 |
+
"grad_norm": 2.2240993976593018,
|
| 12045 |
+
"learning_rate": 8.935593220338984e-06,
|
| 12046 |
+
"loss": 1.3974,
|
| 12047 |
+
"step": 16820
|
| 12048 |
+
},
|
| 12049 |
+
{
|
| 12050 |
+
"epoch": 0.42,
|
| 12051 |
+
"grad_norm": 2.854851245880127,
|
| 12052 |
+
"learning_rate": 8.928813559322036e-06,
|
| 12053 |
+
"loss": 1.4478,
|
| 12054 |
+
"step": 16830
|
| 12055 |
+
},
|
| 12056 |
+
{
|
| 12057 |
+
"epoch": 0.42,
|
| 12058 |
+
"grad_norm": 4.584380149841309,
|
| 12059 |
+
"learning_rate": 8.922033898305085e-06,
|
| 12060 |
+
"loss": 1.4808,
|
| 12061 |
+
"step": 16840
|
| 12062 |
+
},
|
| 12063 |
+
{
|
| 12064 |
+
"epoch": 0.42,
|
| 12065 |
+
"grad_norm": 5.496336460113525,
|
| 12066 |
+
"learning_rate": 8.915254237288137e-06,
|
| 12067 |
+
"loss": 1.4587,
|
| 12068 |
+
"step": 16850
|
| 12069 |
+
},
|
| 12070 |
+
{
|
| 12071 |
+
"epoch": 0.42,
|
| 12072 |
+
"grad_norm": 9.56225872039795,
|
| 12073 |
+
"learning_rate": 8.908474576271188e-06,
|
| 12074 |
+
"loss": 1.2726,
|
| 12075 |
+
"step": 16860
|
| 12076 |
+
},
|
| 12077 |
+
{
|
| 12078 |
+
"epoch": 0.42,
|
| 12079 |
+
"grad_norm": 3.6302268505096436,
|
| 12080 |
+
"learning_rate": 8.901694915254238e-06,
|
| 12081 |
+
"loss": 1.4753,
|
| 12082 |
+
"step": 16870
|
| 12083 |
+
},
|
| 12084 |
+
{
|
| 12085 |
+
"epoch": 0.42,
|
| 12086 |
+
"grad_norm": 9.815473556518555,
|
| 12087 |
+
"learning_rate": 8.89491525423729e-06,
|
| 12088 |
+
"loss": 1.3163,
|
| 12089 |
+
"step": 16880
|
| 12090 |
+
},
|
| 12091 |
+
{
|
| 12092 |
+
"epoch": 0.42,
|
| 12093 |
+
"grad_norm": 6.138949871063232,
|
| 12094 |
+
"learning_rate": 8.888135593220339e-06,
|
| 12095 |
+
"loss": 1.3637,
|
| 12096 |
+
"step": 16890
|
| 12097 |
+
},
|
| 12098 |
+
{
|
| 12099 |
+
"epoch": 0.42,
|
| 12100 |
+
"grad_norm": 5.0360212326049805,
|
| 12101 |
+
"learning_rate": 8.88135593220339e-06,
|
| 12102 |
+
"loss": 1.5961,
|
| 12103 |
+
"step": 16900
|
| 12104 |
+
},
|
| 12105 |
+
{
|
| 12106 |
+
"epoch": 0.42,
|
| 12107 |
+
"grad_norm": 11.983102798461914,
|
| 12108 |
+
"learning_rate": 8.874576271186442e-06,
|
| 12109 |
+
"loss": 1.3359,
|
| 12110 |
+
"step": 16910
|
| 12111 |
+
},
|
| 12112 |
+
{
|
| 12113 |
+
"epoch": 0.42,
|
| 12114 |
+
"grad_norm": 3.8704030513763428,
|
| 12115 |
+
"learning_rate": 8.867796610169492e-06,
|
| 12116 |
+
"loss": 1.2624,
|
| 12117 |
+
"step": 16920
|
| 12118 |
+
},
|
| 12119 |
+
{
|
| 12120 |
+
"epoch": 0.42,
|
| 12121 |
+
"grad_norm": 8.61357593536377,
|
| 12122 |
+
"learning_rate": 8.861016949152543e-06,
|
| 12123 |
+
"loss": 1.4325,
|
| 12124 |
+
"step": 16930
|
| 12125 |
+
},
|
| 12126 |
+
{
|
| 12127 |
+
"epoch": 0.42,
|
| 12128 |
+
"grad_norm": 1.4079731702804565,
|
| 12129 |
+
"learning_rate": 8.854237288135594e-06,
|
| 12130 |
+
"loss": 1.3984,
|
| 12131 |
+
"step": 16940
|
| 12132 |
+
},
|
| 12133 |
+
{
|
| 12134 |
+
"epoch": 0.42,
|
| 12135 |
+
"grad_norm": 4.425031661987305,
|
| 12136 |
+
"learning_rate": 8.847457627118646e-06,
|
| 12137 |
+
"loss": 1.3202,
|
| 12138 |
+
"step": 16950
|
| 12139 |
+
},
|
| 12140 |
+
{
|
| 12141 |
+
"epoch": 0.42,
|
| 12142 |
+
"grad_norm": 7.565074920654297,
|
| 12143 |
+
"learning_rate": 8.840677966101695e-06,
|
| 12144 |
+
"loss": 1.4561,
|
| 12145 |
+
"step": 16960
|
| 12146 |
+
},
|
| 12147 |
+
{
|
| 12148 |
+
"epoch": 0.42,
|
| 12149 |
+
"grad_norm": 4.513836860656738,
|
| 12150 |
+
"learning_rate": 8.833898305084747e-06,
|
| 12151 |
+
"loss": 1.52,
|
| 12152 |
+
"step": 16970
|
| 12153 |
+
},
|
| 12154 |
+
{
|
| 12155 |
+
"epoch": 0.42,
|
| 12156 |
+
"grad_norm": 7.830351829528809,
|
| 12157 |
+
"learning_rate": 8.827118644067797e-06,
|
| 12158 |
+
"loss": 1.4321,
|
| 12159 |
+
"step": 16980
|
| 12160 |
+
},
|
| 12161 |
+
{
|
| 12162 |
+
"epoch": 0.42,
|
| 12163 |
+
"grad_norm": 4.269097328186035,
|
| 12164 |
+
"learning_rate": 8.820338983050848e-06,
|
| 12165 |
+
"loss": 1.4247,
|
| 12166 |
+
"step": 16990
|
| 12167 |
+
},
|
| 12168 |
+
{
|
| 12169 |
+
"epoch": 0.42,
|
| 12170 |
+
"grad_norm": 3.2125535011291504,
|
| 12171 |
+
"learning_rate": 8.8135593220339e-06,
|
| 12172 |
+
"loss": 1.1334,
|
| 12173 |
+
"step": 17000
|
| 12174 |
+
},
|
| 12175 |
+
{
|
| 12176 |
+
"epoch": 0.42,
|
| 12177 |
+
"eval_loss": 1.3314954042434692,
|
| 12178 |
+
"eval_runtime": 66.1036,
|
| 12179 |
+
"eval_samples_per_second": 15.128,
|
| 12180 |
+
"eval_steps_per_second": 15.128,
|
| 12181 |
+
"step": 17000
|
| 12182 |
+
},
|
| 12183 |
+
{
|
| 12184 |
+
"epoch": 0.43,
|
| 12185 |
+
"grad_norm": 3.5944836139678955,
|
| 12186 |
+
"learning_rate": 8.806779661016949e-06,
|
| 12187 |
+
"loss": 1.2363,
|
| 12188 |
+
"step": 17010
|
| 12189 |
+
},
|
| 12190 |
+
{
|
| 12191 |
+
"epoch": 0.43,
|
| 12192 |
+
"grad_norm": 3.680687189102173,
|
| 12193 |
+
"learning_rate": 8.8e-06,
|
| 12194 |
+
"loss": 1.2968,
|
| 12195 |
+
"step": 17020
|
| 12196 |
+
},
|
| 12197 |
+
{
|
| 12198 |
+
"epoch": 0.43,
|
| 12199 |
+
"grad_norm": 9.756001472473145,
|
| 12200 |
+
"learning_rate": 8.793220338983052e-06,
|
| 12201 |
+
"loss": 1.378,
|
| 12202 |
+
"step": 17030
|
| 12203 |
+
},
|
| 12204 |
+
{
|
| 12205 |
+
"epoch": 0.43,
|
| 12206 |
+
"grad_norm": 3.8217949867248535,
|
| 12207 |
+
"learning_rate": 8.786440677966103e-06,
|
| 12208 |
+
"loss": 1.1933,
|
| 12209 |
+
"step": 17040
|
| 12210 |
+
},
|
| 12211 |
+
{
|
| 12212 |
+
"epoch": 0.43,
|
| 12213 |
+
"grad_norm": 4.635326862335205,
|
| 12214 |
+
"learning_rate": 8.779661016949153e-06,
|
| 12215 |
+
"loss": 1.4635,
|
| 12216 |
+
"step": 17050
|
| 12217 |
+
},
|
| 12218 |
+
{
|
| 12219 |
+
"epoch": 0.43,
|
| 12220 |
+
"grad_norm": 12.349688529968262,
|
| 12221 |
+
"learning_rate": 8.772881355932204e-06,
|
| 12222 |
+
"loss": 1.1233,
|
| 12223 |
+
"step": 17060
|
| 12224 |
+
},
|
| 12225 |
+
{
|
| 12226 |
+
"epoch": 0.43,
|
| 12227 |
+
"grad_norm": 2.0350751876831055,
|
| 12228 |
+
"learning_rate": 8.766101694915254e-06,
|
| 12229 |
+
"loss": 1.5517,
|
| 12230 |
+
"step": 17070
|
| 12231 |
+
},
|
| 12232 |
+
{
|
| 12233 |
+
"epoch": 0.43,
|
| 12234 |
+
"grad_norm": 2.533076524734497,
|
| 12235 |
+
"learning_rate": 8.759322033898305e-06,
|
| 12236 |
+
"loss": 1.2468,
|
| 12237 |
+
"step": 17080
|
| 12238 |
+
},
|
| 12239 |
+
{
|
| 12240 |
+
"epoch": 0.43,
|
| 12241 |
+
"grad_norm": 5.4851813316345215,
|
| 12242 |
+
"learning_rate": 8.752542372881357e-06,
|
| 12243 |
+
"loss": 1.3057,
|
| 12244 |
+
"step": 17090
|
| 12245 |
+
},
|
| 12246 |
+
{
|
| 12247 |
+
"epoch": 0.43,
|
| 12248 |
+
"grad_norm": 5.9718499183654785,
|
| 12249 |
+
"learning_rate": 8.745762711864407e-06,
|
| 12250 |
+
"loss": 1.3685,
|
| 12251 |
+
"step": 17100
|
| 12252 |
+
},
|
| 12253 |
+
{
|
| 12254 |
+
"epoch": 0.43,
|
| 12255 |
+
"grad_norm": 6.759610176086426,
|
| 12256 |
+
"learning_rate": 8.738983050847458e-06,
|
| 12257 |
+
"loss": 1.2429,
|
| 12258 |
+
"step": 17110
|
| 12259 |
+
},
|
| 12260 |
+
{
|
| 12261 |
+
"epoch": 0.43,
|
| 12262 |
+
"grad_norm": 10.515771865844727,
|
| 12263 |
+
"learning_rate": 8.73220338983051e-06,
|
| 12264 |
+
"loss": 1.1885,
|
| 12265 |
+
"step": 17120
|
| 12266 |
+
},
|
| 12267 |
+
{
|
| 12268 |
+
"epoch": 0.43,
|
| 12269 |
+
"grad_norm": 5.439294815063477,
|
| 12270 |
+
"learning_rate": 8.72542372881356e-06,
|
| 12271 |
+
"loss": 1.4804,
|
| 12272 |
+
"step": 17130
|
| 12273 |
+
},
|
| 12274 |
+
{
|
| 12275 |
+
"epoch": 0.43,
|
| 12276 |
+
"grad_norm": 3.2483952045440674,
|
| 12277 |
+
"learning_rate": 8.71864406779661e-06,
|
| 12278 |
+
"loss": 1.2879,
|
| 12279 |
+
"step": 17140
|
| 12280 |
+
},
|
| 12281 |
+
{
|
| 12282 |
+
"epoch": 0.43,
|
| 12283 |
+
"grad_norm": 12.16537857055664,
|
| 12284 |
+
"learning_rate": 8.711864406779662e-06,
|
| 12285 |
+
"loss": 1.365,
|
| 12286 |
+
"step": 17150
|
| 12287 |
+
},
|
| 12288 |
+
{
|
| 12289 |
+
"epoch": 0.43,
|
| 12290 |
+
"grad_norm": 7.272519588470459,
|
| 12291 |
+
"learning_rate": 8.705084745762713e-06,
|
| 12292 |
+
"loss": 1.2356,
|
| 12293 |
+
"step": 17160
|
| 12294 |
+
},
|
| 12295 |
+
{
|
| 12296 |
+
"epoch": 0.43,
|
| 12297 |
+
"grad_norm": 5.10579252243042,
|
| 12298 |
+
"learning_rate": 8.698305084745765e-06,
|
| 12299 |
+
"loss": 1.2475,
|
| 12300 |
+
"step": 17170
|
| 12301 |
+
},
|
| 12302 |
+
{
|
| 12303 |
+
"epoch": 0.43,
|
| 12304 |
+
"grad_norm": 2.060873031616211,
|
| 12305 |
+
"learning_rate": 8.691525423728814e-06,
|
| 12306 |
+
"loss": 1.2728,
|
| 12307 |
+
"step": 17180
|
| 12308 |
+
},
|
| 12309 |
+
{
|
| 12310 |
+
"epoch": 0.43,
|
| 12311 |
+
"grad_norm": 4.275463581085205,
|
| 12312 |
+
"learning_rate": 8.684745762711864e-06,
|
| 12313 |
+
"loss": 1.3372,
|
| 12314 |
+
"step": 17190
|
| 12315 |
+
},
|
| 12316 |
+
{
|
| 12317 |
+
"epoch": 0.43,
|
| 12318 |
+
"grad_norm": 7.369024753570557,
|
| 12319 |
+
"learning_rate": 8.677966101694915e-06,
|
| 12320 |
+
"loss": 1.3319,
|
| 12321 |
+
"step": 17200
|
| 12322 |
+
},
|
| 12323 |
+
{
|
| 12324 |
+
"epoch": 0.43,
|
| 12325 |
+
"grad_norm": 4.748632431030273,
|
| 12326 |
+
"learning_rate": 8.671186440677967e-06,
|
| 12327 |
+
"loss": 1.3954,
|
| 12328 |
+
"step": 17210
|
| 12329 |
+
},
|
| 12330 |
+
{
|
| 12331 |
+
"epoch": 0.43,
|
| 12332 |
+
"grad_norm": 6.593578815460205,
|
| 12333 |
+
"learning_rate": 8.664406779661018e-06,
|
| 12334 |
+
"loss": 1.218,
|
| 12335 |
+
"step": 17220
|
| 12336 |
+
},
|
| 12337 |
+
{
|
| 12338 |
+
"epoch": 0.43,
|
| 12339 |
+
"grad_norm": 9.88740348815918,
|
| 12340 |
+
"learning_rate": 8.657627118644068e-06,
|
| 12341 |
+
"loss": 1.3813,
|
| 12342 |
+
"step": 17230
|
| 12343 |
+
},
|
| 12344 |
+
{
|
| 12345 |
+
"epoch": 0.43,
|
| 12346 |
+
"grad_norm": 5.244362831115723,
|
| 12347 |
+
"learning_rate": 8.65084745762712e-06,
|
| 12348 |
+
"loss": 1.3307,
|
| 12349 |
+
"step": 17240
|
| 12350 |
+
},
|
| 12351 |
+
{
|
| 12352 |
+
"epoch": 0.43,
|
| 12353 |
+
"grad_norm": 6.388607025146484,
|
| 12354 |
+
"learning_rate": 8.64406779661017e-06,
|
| 12355 |
+
"loss": 1.3975,
|
| 12356 |
+
"step": 17250
|
| 12357 |
+
},
|
| 12358 |
+
{
|
| 12359 |
+
"epoch": 0.43,
|
| 12360 |
+
"grad_norm": 5.940032005310059,
|
| 12361 |
+
"learning_rate": 8.637288135593222e-06,
|
| 12362 |
+
"loss": 1.452,
|
| 12363 |
+
"step": 17260
|
| 12364 |
+
},
|
| 12365 |
+
{
|
| 12366 |
+
"epoch": 0.43,
|
| 12367 |
+
"grad_norm": 1.8261200189590454,
|
| 12368 |
+
"learning_rate": 8.630508474576272e-06,
|
| 12369 |
+
"loss": 1.3114,
|
| 12370 |
+
"step": 17270
|
| 12371 |
+
},
|
| 12372 |
+
{
|
| 12373 |
+
"epoch": 0.43,
|
| 12374 |
+
"grad_norm": 4.878362655639648,
|
| 12375 |
+
"learning_rate": 8.623728813559322e-06,
|
| 12376 |
+
"loss": 1.3152,
|
| 12377 |
+
"step": 17280
|
| 12378 |
+
},
|
| 12379 |
+
{
|
| 12380 |
+
"epoch": 0.43,
|
| 12381 |
+
"grad_norm": 5.558933258056641,
|
| 12382 |
+
"learning_rate": 8.616949152542373e-06,
|
| 12383 |
+
"loss": 1.5567,
|
| 12384 |
+
"step": 17290
|
| 12385 |
+
},
|
| 12386 |
+
{
|
| 12387 |
+
"epoch": 0.43,
|
| 12388 |
+
"grad_norm": 2.346237897872925,
|
| 12389 |
+
"learning_rate": 8.610169491525424e-06,
|
| 12390 |
+
"loss": 1.4462,
|
| 12391 |
+
"step": 17300
|
| 12392 |
+
},
|
| 12393 |
+
{
|
| 12394 |
+
"epoch": 0.43,
|
| 12395 |
+
"grad_norm": 5.712830066680908,
|
| 12396 |
+
"learning_rate": 8.603389830508476e-06,
|
| 12397 |
+
"loss": 1.2658,
|
| 12398 |
+
"step": 17310
|
| 12399 |
+
},
|
| 12400 |
+
{
|
| 12401 |
+
"epoch": 0.43,
|
| 12402 |
+
"grad_norm": 6.277383804321289,
|
| 12403 |
+
"learning_rate": 8.596610169491526e-06,
|
| 12404 |
+
"loss": 1.3562,
|
| 12405 |
+
"step": 17320
|
| 12406 |
+
},
|
| 12407 |
+
{
|
| 12408 |
+
"epoch": 0.43,
|
| 12409 |
+
"grad_norm": 2.8792171478271484,
|
| 12410 |
+
"learning_rate": 8.589830508474577e-06,
|
| 12411 |
+
"loss": 1.2409,
|
| 12412 |
+
"step": 17330
|
| 12413 |
+
},
|
| 12414 |
+
{
|
| 12415 |
+
"epoch": 0.43,
|
| 12416 |
+
"grad_norm": 4.000193119049072,
|
| 12417 |
+
"learning_rate": 8.583050847457628e-06,
|
| 12418 |
+
"loss": 1.3503,
|
| 12419 |
+
"step": 17340
|
| 12420 |
+
},
|
| 12421 |
+
{
|
| 12422 |
+
"epoch": 0.43,
|
| 12423 |
+
"grad_norm": 6.255504131317139,
|
| 12424 |
+
"learning_rate": 8.57627118644068e-06,
|
| 12425 |
+
"loss": 1.3411,
|
| 12426 |
+
"step": 17350
|
| 12427 |
+
},
|
| 12428 |
+
{
|
| 12429 |
+
"epoch": 0.43,
|
| 12430 |
+
"grad_norm": 5.366970539093018,
|
| 12431 |
+
"learning_rate": 8.56949152542373e-06,
|
| 12432 |
+
"loss": 1.2918,
|
| 12433 |
+
"step": 17360
|
| 12434 |
+
},
|
| 12435 |
+
{
|
| 12436 |
+
"epoch": 0.43,
|
| 12437 |
+
"grad_norm": 1.9505482912063599,
|
| 12438 |
+
"learning_rate": 8.56271186440678e-06,
|
| 12439 |
+
"loss": 1.3139,
|
| 12440 |
+
"step": 17370
|
| 12441 |
+
},
|
| 12442 |
+
{
|
| 12443 |
+
"epoch": 0.43,
|
| 12444 |
+
"grad_norm": 8.022522926330566,
|
| 12445 |
+
"learning_rate": 8.55593220338983e-06,
|
| 12446 |
+
"loss": 1.515,
|
| 12447 |
+
"step": 17380
|
| 12448 |
+
},
|
| 12449 |
+
{
|
| 12450 |
+
"epoch": 0.43,
|
| 12451 |
+
"grad_norm": 5.483621120452881,
|
| 12452 |
+
"learning_rate": 8.549152542372882e-06,
|
| 12453 |
+
"loss": 1.4464,
|
| 12454 |
+
"step": 17390
|
| 12455 |
+
},
|
| 12456 |
+
{
|
| 12457 |
+
"epoch": 0.43,
|
| 12458 |
+
"grad_norm": 5.349776268005371,
|
| 12459 |
+
"learning_rate": 8.542372881355933e-06,
|
| 12460 |
+
"loss": 1.5864,
|
| 12461 |
+
"step": 17400
|
| 12462 |
+
},
|
| 12463 |
+
{
|
| 12464 |
+
"epoch": 0.44,
|
| 12465 |
+
"grad_norm": 3.4136993885040283,
|
| 12466 |
+
"learning_rate": 8.535593220338983e-06,
|
| 12467 |
+
"loss": 1.2964,
|
| 12468 |
+
"step": 17410
|
| 12469 |
+
},
|
| 12470 |
+
{
|
| 12471 |
+
"epoch": 0.44,
|
| 12472 |
+
"grad_norm": 4.270326614379883,
|
| 12473 |
+
"learning_rate": 8.528813559322034e-06,
|
| 12474 |
+
"loss": 1.3736,
|
| 12475 |
+
"step": 17420
|
| 12476 |
+
},
|
| 12477 |
+
{
|
| 12478 |
+
"epoch": 0.44,
|
| 12479 |
+
"grad_norm": 2.886993169784546,
|
| 12480 |
+
"learning_rate": 8.522033898305086e-06,
|
| 12481 |
+
"loss": 1.3513,
|
| 12482 |
+
"step": 17430
|
| 12483 |
+
},
|
| 12484 |
+
{
|
| 12485 |
+
"epoch": 0.44,
|
| 12486 |
+
"grad_norm": 4.42819881439209,
|
| 12487 |
+
"learning_rate": 8.515254237288136e-06,
|
| 12488 |
+
"loss": 1.4226,
|
| 12489 |
+
"step": 17440
|
| 12490 |
+
},
|
| 12491 |
+
{
|
| 12492 |
+
"epoch": 0.44,
|
| 12493 |
+
"grad_norm": 5.844710350036621,
|
| 12494 |
+
"learning_rate": 8.508474576271187e-06,
|
| 12495 |
+
"loss": 1.3135,
|
| 12496 |
+
"step": 17450
|
| 12497 |
+
},
|
| 12498 |
+
{
|
| 12499 |
+
"epoch": 0.44,
|
| 12500 |
+
"grad_norm": 2.4220869541168213,
|
| 12501 |
+
"learning_rate": 8.501694915254238e-06,
|
| 12502 |
+
"loss": 1.165,
|
| 12503 |
+
"step": 17460
|
| 12504 |
+
},
|
| 12505 |
+
{
|
| 12506 |
+
"epoch": 0.44,
|
| 12507 |
+
"grad_norm": 4.976444244384766,
|
| 12508 |
+
"learning_rate": 8.49491525423729e-06,
|
| 12509 |
+
"loss": 1.1964,
|
| 12510 |
+
"step": 17470
|
| 12511 |
+
},
|
| 12512 |
+
{
|
| 12513 |
+
"epoch": 0.44,
|
| 12514 |
+
"grad_norm": 12.55490779876709,
|
| 12515 |
+
"learning_rate": 8.48813559322034e-06,
|
| 12516 |
+
"loss": 1.5891,
|
| 12517 |
+
"step": 17480
|
| 12518 |
+
},
|
| 12519 |
+
{
|
| 12520 |
+
"epoch": 0.44,
|
| 12521 |
+
"grad_norm": 4.102578163146973,
|
| 12522 |
+
"learning_rate": 8.481355932203391e-06,
|
| 12523 |
+
"loss": 1.445,
|
| 12524 |
+
"step": 17490
|
| 12525 |
+
},
|
| 12526 |
+
{
|
| 12527 |
+
"epoch": 0.44,
|
| 12528 |
+
"grad_norm": 2.6767678260803223,
|
| 12529 |
+
"learning_rate": 8.47457627118644e-06,
|
| 12530 |
+
"loss": 1.536,
|
| 12531 |
+
"step": 17500
|
| 12532 |
+
},
|
| 12533 |
+
{
|
| 12534 |
+
"epoch": 0.44,
|
| 12535 |
+
"eval_loss": 1.388473391532898,
|
| 12536 |
+
"eval_runtime": 66.17,
|
| 12537 |
+
"eval_samples_per_second": 15.113,
|
| 12538 |
+
"eval_steps_per_second": 15.113,
|
| 12539 |
+
"step": 17500
|
| 12540 |
}
|
| 12541 |
],
|
| 12542 |
"logging_steps": 10,
|
|
|
|
| 12544 |
"num_input_tokens_seen": 0,
|
| 12545 |
"num_train_epochs": 1,
|
| 12546 |
"save_steps": 2500,
|
| 12547 |
+
"total_flos": 2.8178720489472e+17,
|
| 12548 |
"train_batch_size": 1,
|
| 12549 |
"trial_name": null,
|
| 12550 |
"trial_params": null
|