Uploaded checkpoint-25000
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 sha256:58ea76fcf16a2912a570cf295dd1757cd9562cb7f7f8e74d37938855d31dc866
|
| 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 sha256:231c6b24970ef58291d1980aeb742ace763101289d628ec3f4ac808335924d18
|
| 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:c2bd031f30ceb89483d2d8b5eb187850133dcc5a689162e8975b2cc0e61b4001
|
| 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:e3bdbaa37c77733a3ea9eb90a36bc290f4f5b9f56abe23cc6586cbaa459f92c6
|
| 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,
|
|
@@ -16117,6 +16117,1796 @@
|
|
| 16117 |
"eval_samples_per_second": 15.104,
|
| 16118 |
"eval_steps_per_second": 15.104,
|
| 16119 |
"step": 22500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16120 |
}
|
| 16121 |
],
|
| 16122 |
"logging_steps": 10,
|
|
@@ -16124,7 +17914,7 @@
|
|
| 16124 |
"num_input_tokens_seen": 0,
|
| 16125 |
"num_train_epochs": 1,
|
| 16126 |
"save_steps": 2500,
|
| 16127 |
-
"total_flos":
|
| 16128 |
"train_batch_size": 1,
|
| 16129 |
"trial_name": null,
|
| 16130 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.3409814834594727,
|
| 3 |
+
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-25000",
|
| 4 |
+
"epoch": 0.625,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 25000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 16117 |
"eval_samples_per_second": 15.104,
|
| 16118 |
"eval_steps_per_second": 15.104,
|
| 16119 |
"step": 22500
|
| 16120 |
+
},
|
| 16121 |
+
{
|
| 16122 |
+
"epoch": 0.56,
|
| 16123 |
+
"grad_norm": 3.232191562652588,
|
| 16124 |
+
"learning_rate": 5.077966101694915e-06,
|
| 16125 |
+
"loss": 1.3644,
|
| 16126 |
+
"step": 22510
|
| 16127 |
+
},
|
| 16128 |
+
{
|
| 16129 |
+
"epoch": 0.56,
|
| 16130 |
+
"grad_norm": 5.3384528160095215,
|
| 16131 |
+
"learning_rate": 5.071186440677967e-06,
|
| 16132 |
+
"loss": 1.3093,
|
| 16133 |
+
"step": 22520
|
| 16134 |
+
},
|
| 16135 |
+
{
|
| 16136 |
+
"epoch": 0.56,
|
| 16137 |
+
"grad_norm": 6.911621570587158,
|
| 16138 |
+
"learning_rate": 5.064406779661017e-06,
|
| 16139 |
+
"loss": 1.2571,
|
| 16140 |
+
"step": 22530
|
| 16141 |
+
},
|
| 16142 |
+
{
|
| 16143 |
+
"epoch": 0.56,
|
| 16144 |
+
"grad_norm": 6.84785270690918,
|
| 16145 |
+
"learning_rate": 5.057627118644069e-06,
|
| 16146 |
+
"loss": 1.324,
|
| 16147 |
+
"step": 22540
|
| 16148 |
+
},
|
| 16149 |
+
{
|
| 16150 |
+
"epoch": 0.56,
|
| 16151 |
+
"grad_norm": 4.634193420410156,
|
| 16152 |
+
"learning_rate": 5.050847457627119e-06,
|
| 16153 |
+
"loss": 1.4672,
|
| 16154 |
+
"step": 22550
|
| 16155 |
+
},
|
| 16156 |
+
{
|
| 16157 |
+
"epoch": 0.56,
|
| 16158 |
+
"grad_norm": 3.4189460277557373,
|
| 16159 |
+
"learning_rate": 5.0440677966101705e-06,
|
| 16160 |
+
"loss": 1.325,
|
| 16161 |
+
"step": 22560
|
| 16162 |
+
},
|
| 16163 |
+
{
|
| 16164 |
+
"epoch": 0.56,
|
| 16165 |
+
"grad_norm": 7.2140793800354,
|
| 16166 |
+
"learning_rate": 5.037288135593221e-06,
|
| 16167 |
+
"loss": 1.3923,
|
| 16168 |
+
"step": 22570
|
| 16169 |
+
},
|
| 16170 |
+
{
|
| 16171 |
+
"epoch": 0.56,
|
| 16172 |
+
"grad_norm": 3.916853189468384,
|
| 16173 |
+
"learning_rate": 5.030508474576271e-06,
|
| 16174 |
+
"loss": 1.3431,
|
| 16175 |
+
"step": 22580
|
| 16176 |
+
},
|
| 16177 |
+
{
|
| 16178 |
+
"epoch": 0.56,
|
| 16179 |
+
"grad_norm": 6.434349536895752,
|
| 16180 |
+
"learning_rate": 5.023728813559322e-06,
|
| 16181 |
+
"loss": 1.4615,
|
| 16182 |
+
"step": 22590
|
| 16183 |
+
},
|
| 16184 |
+
{
|
| 16185 |
+
"epoch": 0.56,
|
| 16186 |
+
"grad_norm": 6.851098537445068,
|
| 16187 |
+
"learning_rate": 5.016949152542373e-06,
|
| 16188 |
+
"loss": 1.3013,
|
| 16189 |
+
"step": 22600
|
| 16190 |
+
},
|
| 16191 |
+
{
|
| 16192 |
+
"epoch": 0.57,
|
| 16193 |
+
"grad_norm": 7.541562080383301,
|
| 16194 |
+
"learning_rate": 5.010169491525424e-06,
|
| 16195 |
+
"loss": 1.3696,
|
| 16196 |
+
"step": 22610
|
| 16197 |
+
},
|
| 16198 |
+
{
|
| 16199 |
+
"epoch": 0.57,
|
| 16200 |
+
"grad_norm": 4.100308895111084,
|
| 16201 |
+
"learning_rate": 5.003389830508475e-06,
|
| 16202 |
+
"loss": 1.3802,
|
| 16203 |
+
"step": 22620
|
| 16204 |
+
},
|
| 16205 |
+
{
|
| 16206 |
+
"epoch": 0.57,
|
| 16207 |
+
"grad_norm": 12.694953918457031,
|
| 16208 |
+
"learning_rate": 4.996610169491526e-06,
|
| 16209 |
+
"loss": 1.2328,
|
| 16210 |
+
"step": 22630
|
| 16211 |
+
},
|
| 16212 |
+
{
|
| 16213 |
+
"epoch": 0.57,
|
| 16214 |
+
"grad_norm": 12.468879699707031,
|
| 16215 |
+
"learning_rate": 4.989830508474577e-06,
|
| 16216 |
+
"loss": 1.3995,
|
| 16217 |
+
"step": 22640
|
| 16218 |
+
},
|
| 16219 |
+
{
|
| 16220 |
+
"epoch": 0.57,
|
| 16221 |
+
"grad_norm": 3.142810821533203,
|
| 16222 |
+
"learning_rate": 4.983050847457628e-06,
|
| 16223 |
+
"loss": 1.2403,
|
| 16224 |
+
"step": 22650
|
| 16225 |
+
},
|
| 16226 |
+
{
|
| 16227 |
+
"epoch": 0.57,
|
| 16228 |
+
"grad_norm": 7.71440315246582,
|
| 16229 |
+
"learning_rate": 4.976271186440678e-06,
|
| 16230 |
+
"loss": 1.2902,
|
| 16231 |
+
"step": 22660
|
| 16232 |
+
},
|
| 16233 |
+
{
|
| 16234 |
+
"epoch": 0.57,
|
| 16235 |
+
"grad_norm": 4.5355634689331055,
|
| 16236 |
+
"learning_rate": 4.969491525423729e-06,
|
| 16237 |
+
"loss": 1.3243,
|
| 16238 |
+
"step": 22670
|
| 16239 |
+
},
|
| 16240 |
+
{
|
| 16241 |
+
"epoch": 0.57,
|
| 16242 |
+
"grad_norm": 1.5578103065490723,
|
| 16243 |
+
"learning_rate": 4.96271186440678e-06,
|
| 16244 |
+
"loss": 1.2796,
|
| 16245 |
+
"step": 22680
|
| 16246 |
+
},
|
| 16247 |
+
{
|
| 16248 |
+
"epoch": 0.57,
|
| 16249 |
+
"grad_norm": 4.997007846832275,
|
| 16250 |
+
"learning_rate": 4.955932203389831e-06,
|
| 16251 |
+
"loss": 1.1288,
|
| 16252 |
+
"step": 22690
|
| 16253 |
+
},
|
| 16254 |
+
{
|
| 16255 |
+
"epoch": 0.57,
|
| 16256 |
+
"grad_norm": 4.240728855133057,
|
| 16257 |
+
"learning_rate": 4.949152542372882e-06,
|
| 16258 |
+
"loss": 1.408,
|
| 16259 |
+
"step": 22700
|
| 16260 |
+
},
|
| 16261 |
+
{
|
| 16262 |
+
"epoch": 0.57,
|
| 16263 |
+
"grad_norm": 7.517406940460205,
|
| 16264 |
+
"learning_rate": 4.942372881355932e-06,
|
| 16265 |
+
"loss": 1.3879,
|
| 16266 |
+
"step": 22710
|
| 16267 |
+
},
|
| 16268 |
+
{
|
| 16269 |
+
"epoch": 0.57,
|
| 16270 |
+
"grad_norm": 1.1581978797912598,
|
| 16271 |
+
"learning_rate": 4.935593220338984e-06,
|
| 16272 |
+
"loss": 1.3547,
|
| 16273 |
+
"step": 22720
|
| 16274 |
+
},
|
| 16275 |
+
{
|
| 16276 |
+
"epoch": 0.57,
|
| 16277 |
+
"grad_norm": 5.073269367218018,
|
| 16278 |
+
"learning_rate": 4.928813559322034e-06,
|
| 16279 |
+
"loss": 1.3664,
|
| 16280 |
+
"step": 22730
|
| 16281 |
+
},
|
| 16282 |
+
{
|
| 16283 |
+
"epoch": 0.57,
|
| 16284 |
+
"grad_norm": 3.7317769527435303,
|
| 16285 |
+
"learning_rate": 4.922033898305086e-06,
|
| 16286 |
+
"loss": 1.3942,
|
| 16287 |
+
"step": 22740
|
| 16288 |
+
},
|
| 16289 |
+
{
|
| 16290 |
+
"epoch": 0.57,
|
| 16291 |
+
"grad_norm": 4.843672275543213,
|
| 16292 |
+
"learning_rate": 4.915254237288136e-06,
|
| 16293 |
+
"loss": 1.3723,
|
| 16294 |
+
"step": 22750
|
| 16295 |
+
},
|
| 16296 |
+
{
|
| 16297 |
+
"epoch": 0.57,
|
| 16298 |
+
"grad_norm": 7.751224040985107,
|
| 16299 |
+
"learning_rate": 4.908474576271187e-06,
|
| 16300 |
+
"loss": 1.2057,
|
| 16301 |
+
"step": 22760
|
| 16302 |
+
},
|
| 16303 |
+
{
|
| 16304 |
+
"epoch": 0.57,
|
| 16305 |
+
"grad_norm": 2.4519495964050293,
|
| 16306 |
+
"learning_rate": 4.901694915254237e-06,
|
| 16307 |
+
"loss": 1.3836,
|
| 16308 |
+
"step": 22770
|
| 16309 |
+
},
|
| 16310 |
+
{
|
| 16311 |
+
"epoch": 0.57,
|
| 16312 |
+
"grad_norm": 8.7233304977417,
|
| 16313 |
+
"learning_rate": 4.894915254237289e-06,
|
| 16314 |
+
"loss": 1.3842,
|
| 16315 |
+
"step": 22780
|
| 16316 |
+
},
|
| 16317 |
+
{
|
| 16318 |
+
"epoch": 0.57,
|
| 16319 |
+
"grad_norm": 2.717367172241211,
|
| 16320 |
+
"learning_rate": 4.888135593220339e-06,
|
| 16321 |
+
"loss": 1.4519,
|
| 16322 |
+
"step": 22790
|
| 16323 |
+
},
|
| 16324 |
+
{
|
| 16325 |
+
"epoch": 0.57,
|
| 16326 |
+
"grad_norm": 4.797736167907715,
|
| 16327 |
+
"learning_rate": 4.881355932203391e-06,
|
| 16328 |
+
"loss": 1.1785,
|
| 16329 |
+
"step": 22800
|
| 16330 |
+
},
|
| 16331 |
+
{
|
| 16332 |
+
"epoch": 0.57,
|
| 16333 |
+
"grad_norm": 11.28987979888916,
|
| 16334 |
+
"learning_rate": 4.874576271186441e-06,
|
| 16335 |
+
"loss": 1.3058,
|
| 16336 |
+
"step": 22810
|
| 16337 |
+
},
|
| 16338 |
+
{
|
| 16339 |
+
"epoch": 0.57,
|
| 16340 |
+
"grad_norm": 5.097863674163818,
|
| 16341 |
+
"learning_rate": 4.867796610169492e-06,
|
| 16342 |
+
"loss": 1.232,
|
| 16343 |
+
"step": 22820
|
| 16344 |
+
},
|
| 16345 |
+
{
|
| 16346 |
+
"epoch": 0.57,
|
| 16347 |
+
"grad_norm": 5.479716777801514,
|
| 16348 |
+
"learning_rate": 4.861016949152543e-06,
|
| 16349 |
+
"loss": 1.4677,
|
| 16350 |
+
"step": 22830
|
| 16351 |
+
},
|
| 16352 |
+
{
|
| 16353 |
+
"epoch": 0.57,
|
| 16354 |
+
"grad_norm": 11.921456336975098,
|
| 16355 |
+
"learning_rate": 4.854237288135594e-06,
|
| 16356 |
+
"loss": 1.5442,
|
| 16357 |
+
"step": 22840
|
| 16358 |
+
},
|
| 16359 |
+
{
|
| 16360 |
+
"epoch": 0.57,
|
| 16361 |
+
"grad_norm": 9.934676170349121,
|
| 16362 |
+
"learning_rate": 4.847457627118645e-06,
|
| 16363 |
+
"loss": 1.4147,
|
| 16364 |
+
"step": 22850
|
| 16365 |
+
},
|
| 16366 |
+
{
|
| 16367 |
+
"epoch": 0.57,
|
| 16368 |
+
"grad_norm": 3.3140487670898438,
|
| 16369 |
+
"learning_rate": 4.840677966101695e-06,
|
| 16370 |
+
"loss": 1.2146,
|
| 16371 |
+
"step": 22860
|
| 16372 |
+
},
|
| 16373 |
+
{
|
| 16374 |
+
"epoch": 0.57,
|
| 16375 |
+
"grad_norm": 2.1303317546844482,
|
| 16376 |
+
"learning_rate": 4.833898305084746e-06,
|
| 16377 |
+
"loss": 1.3021,
|
| 16378 |
+
"step": 22870
|
| 16379 |
+
},
|
| 16380 |
+
{
|
| 16381 |
+
"epoch": 0.57,
|
| 16382 |
+
"grad_norm": 3.943474769592285,
|
| 16383 |
+
"learning_rate": 4.827118644067797e-06,
|
| 16384 |
+
"loss": 1.2028,
|
| 16385 |
+
"step": 22880
|
| 16386 |
+
},
|
| 16387 |
+
{
|
| 16388 |
+
"epoch": 0.57,
|
| 16389 |
+
"grad_norm": 5.444009780883789,
|
| 16390 |
+
"learning_rate": 4.820338983050848e-06,
|
| 16391 |
+
"loss": 1.3873,
|
| 16392 |
+
"step": 22890
|
| 16393 |
+
},
|
| 16394 |
+
{
|
| 16395 |
+
"epoch": 0.57,
|
| 16396 |
+
"grad_norm": 5.230558395385742,
|
| 16397 |
+
"learning_rate": 4.813559322033899e-06,
|
| 16398 |
+
"loss": 1.2907,
|
| 16399 |
+
"step": 22900
|
| 16400 |
+
},
|
| 16401 |
+
{
|
| 16402 |
+
"epoch": 0.57,
|
| 16403 |
+
"grad_norm": 2.652158498764038,
|
| 16404 |
+
"learning_rate": 4.80677966101695e-06,
|
| 16405 |
+
"loss": 1.27,
|
| 16406 |
+
"step": 22910
|
| 16407 |
+
},
|
| 16408 |
+
{
|
| 16409 |
+
"epoch": 0.57,
|
| 16410 |
+
"grad_norm": 5.423113822937012,
|
| 16411 |
+
"learning_rate": 4.800000000000001e-06,
|
| 16412 |
+
"loss": 1.3488,
|
| 16413 |
+
"step": 22920
|
| 16414 |
+
},
|
| 16415 |
+
{
|
| 16416 |
+
"epoch": 0.57,
|
| 16417 |
+
"grad_norm": 6.731577396392822,
|
| 16418 |
+
"learning_rate": 4.793220338983051e-06,
|
| 16419 |
+
"loss": 1.3411,
|
| 16420 |
+
"step": 22930
|
| 16421 |
+
},
|
| 16422 |
+
{
|
| 16423 |
+
"epoch": 0.57,
|
| 16424 |
+
"grad_norm": 6.142993927001953,
|
| 16425 |
+
"learning_rate": 4.786440677966102e-06,
|
| 16426 |
+
"loss": 1.4032,
|
| 16427 |
+
"step": 22940
|
| 16428 |
+
},
|
| 16429 |
+
{
|
| 16430 |
+
"epoch": 0.57,
|
| 16431 |
+
"grad_norm": 7.845600605010986,
|
| 16432 |
+
"learning_rate": 4.779661016949153e-06,
|
| 16433 |
+
"loss": 1.3567,
|
| 16434 |
+
"step": 22950
|
| 16435 |
+
},
|
| 16436 |
+
{
|
| 16437 |
+
"epoch": 0.57,
|
| 16438 |
+
"grad_norm": 3.123938798904419,
|
| 16439 |
+
"learning_rate": 4.772881355932204e-06,
|
| 16440 |
+
"loss": 1.2106,
|
| 16441 |
+
"step": 22960
|
| 16442 |
+
},
|
| 16443 |
+
{
|
| 16444 |
+
"epoch": 0.57,
|
| 16445 |
+
"grad_norm": 4.684544086456299,
|
| 16446 |
+
"learning_rate": 4.766101694915254e-06,
|
| 16447 |
+
"loss": 1.3407,
|
| 16448 |
+
"step": 22970
|
| 16449 |
+
},
|
| 16450 |
+
{
|
| 16451 |
+
"epoch": 0.57,
|
| 16452 |
+
"grad_norm": 11.232462882995605,
|
| 16453 |
+
"learning_rate": 4.759322033898306e-06,
|
| 16454 |
+
"loss": 1.136,
|
| 16455 |
+
"step": 22980
|
| 16456 |
+
},
|
| 16457 |
+
{
|
| 16458 |
+
"epoch": 0.57,
|
| 16459 |
+
"grad_norm": 3.3728113174438477,
|
| 16460 |
+
"learning_rate": 4.752542372881356e-06,
|
| 16461 |
+
"loss": 1.199,
|
| 16462 |
+
"step": 22990
|
| 16463 |
+
},
|
| 16464 |
+
{
|
| 16465 |
+
"epoch": 0.57,
|
| 16466 |
+
"grad_norm": 3.5851376056671143,
|
| 16467 |
+
"learning_rate": 4.745762711864408e-06,
|
| 16468 |
+
"loss": 1.463,
|
| 16469 |
+
"step": 23000
|
| 16470 |
+
},
|
| 16471 |
+
{
|
| 16472 |
+
"epoch": 0.57,
|
| 16473 |
+
"eval_loss": 1.3161565065383911,
|
| 16474 |
+
"eval_runtime": 66.103,
|
| 16475 |
+
"eval_samples_per_second": 15.128,
|
| 16476 |
+
"eval_steps_per_second": 15.128,
|
| 16477 |
+
"step": 23000
|
| 16478 |
+
},
|
| 16479 |
+
{
|
| 16480 |
+
"epoch": 0.58,
|
| 16481 |
+
"grad_norm": 5.720829010009766,
|
| 16482 |
+
"learning_rate": 4.738983050847458e-06,
|
| 16483 |
+
"loss": 1.4393,
|
| 16484 |
+
"step": 23010
|
| 16485 |
+
},
|
| 16486 |
+
{
|
| 16487 |
+
"epoch": 0.58,
|
| 16488 |
+
"grad_norm": 11.663517951965332,
|
| 16489 |
+
"learning_rate": 4.732203389830509e-06,
|
| 16490 |
+
"loss": 1.3158,
|
| 16491 |
+
"step": 23020
|
| 16492 |
+
},
|
| 16493 |
+
{
|
| 16494 |
+
"epoch": 0.58,
|
| 16495 |
+
"grad_norm": 10.158949851989746,
|
| 16496 |
+
"learning_rate": 4.725423728813559e-06,
|
| 16497 |
+
"loss": 1.4086,
|
| 16498 |
+
"step": 23030
|
| 16499 |
+
},
|
| 16500 |
+
{
|
| 16501 |
+
"epoch": 0.58,
|
| 16502 |
+
"grad_norm": 6.925542831420898,
|
| 16503 |
+
"learning_rate": 4.718644067796611e-06,
|
| 16504 |
+
"loss": 1.3655,
|
| 16505 |
+
"step": 23040
|
| 16506 |
+
},
|
| 16507 |
+
{
|
| 16508 |
+
"epoch": 0.58,
|
| 16509 |
+
"grad_norm": 1.4193698167800903,
|
| 16510 |
+
"learning_rate": 4.711864406779661e-06,
|
| 16511 |
+
"loss": 1.2348,
|
| 16512 |
+
"step": 23050
|
| 16513 |
+
},
|
| 16514 |
+
{
|
| 16515 |
+
"epoch": 0.58,
|
| 16516 |
+
"grad_norm": 6.884500980377197,
|
| 16517 |
+
"learning_rate": 4.705084745762713e-06,
|
| 16518 |
+
"loss": 1.4053,
|
| 16519 |
+
"step": 23060
|
| 16520 |
+
},
|
| 16521 |
+
{
|
| 16522 |
+
"epoch": 0.58,
|
| 16523 |
+
"grad_norm": 4.412232875823975,
|
| 16524 |
+
"learning_rate": 4.698305084745763e-06,
|
| 16525 |
+
"loss": 1.2476,
|
| 16526 |
+
"step": 23070
|
| 16527 |
+
},
|
| 16528 |
+
{
|
| 16529 |
+
"epoch": 0.58,
|
| 16530 |
+
"grad_norm": 2.403428077697754,
|
| 16531 |
+
"learning_rate": 4.691525423728814e-06,
|
| 16532 |
+
"loss": 1.4168,
|
| 16533 |
+
"step": 23080
|
| 16534 |
+
},
|
| 16535 |
+
{
|
| 16536 |
+
"epoch": 0.58,
|
| 16537 |
+
"grad_norm": 7.022388458251953,
|
| 16538 |
+
"learning_rate": 4.684745762711865e-06,
|
| 16539 |
+
"loss": 1.359,
|
| 16540 |
+
"step": 23090
|
| 16541 |
+
},
|
| 16542 |
+
{
|
| 16543 |
+
"epoch": 0.58,
|
| 16544 |
+
"grad_norm": 5.728121280670166,
|
| 16545 |
+
"learning_rate": 4.677966101694916e-06,
|
| 16546 |
+
"loss": 1.3916,
|
| 16547 |
+
"step": 23100
|
| 16548 |
+
},
|
| 16549 |
+
{
|
| 16550 |
+
"epoch": 0.58,
|
| 16551 |
+
"grad_norm": 14.995932579040527,
|
| 16552 |
+
"learning_rate": 4.671186440677967e-06,
|
| 16553 |
+
"loss": 1.4541,
|
| 16554 |
+
"step": 23110
|
| 16555 |
+
},
|
| 16556 |
+
{
|
| 16557 |
+
"epoch": 0.58,
|
| 16558 |
+
"grad_norm": 12.448729515075684,
|
| 16559 |
+
"learning_rate": 4.664406779661017e-06,
|
| 16560 |
+
"loss": 1.2989,
|
| 16561 |
+
"step": 23120
|
| 16562 |
+
},
|
| 16563 |
+
{
|
| 16564 |
+
"epoch": 0.58,
|
| 16565 |
+
"grad_norm": 2.5807507038116455,
|
| 16566 |
+
"learning_rate": 4.657627118644068e-06,
|
| 16567 |
+
"loss": 1.4881,
|
| 16568 |
+
"step": 23130
|
| 16569 |
+
},
|
| 16570 |
+
{
|
| 16571 |
+
"epoch": 0.58,
|
| 16572 |
+
"grad_norm": 4.670041561126709,
|
| 16573 |
+
"learning_rate": 4.650847457627119e-06,
|
| 16574 |
+
"loss": 1.4595,
|
| 16575 |
+
"step": 23140
|
| 16576 |
+
},
|
| 16577 |
+
{
|
| 16578 |
+
"epoch": 0.58,
|
| 16579 |
+
"grad_norm": 3.8450145721435547,
|
| 16580 |
+
"learning_rate": 4.64406779661017e-06,
|
| 16581 |
+
"loss": 1.245,
|
| 16582 |
+
"step": 23150
|
| 16583 |
+
},
|
| 16584 |
+
{
|
| 16585 |
+
"epoch": 0.58,
|
| 16586 |
+
"grad_norm": 15.546969413757324,
|
| 16587 |
+
"learning_rate": 4.637288135593221e-06,
|
| 16588 |
+
"loss": 1.2199,
|
| 16589 |
+
"step": 23160
|
| 16590 |
+
},
|
| 16591 |
+
{
|
| 16592 |
+
"epoch": 0.58,
|
| 16593 |
+
"grad_norm": 8.563859939575195,
|
| 16594 |
+
"learning_rate": 4.630508474576272e-06,
|
| 16595 |
+
"loss": 1.3752,
|
| 16596 |
+
"step": 23170
|
| 16597 |
+
},
|
| 16598 |
+
{
|
| 16599 |
+
"epoch": 0.58,
|
| 16600 |
+
"grad_norm": 8.742653846740723,
|
| 16601 |
+
"learning_rate": 4.623728813559323e-06,
|
| 16602 |
+
"loss": 1.2802,
|
| 16603 |
+
"step": 23180
|
| 16604 |
+
},
|
| 16605 |
+
{
|
| 16606 |
+
"epoch": 0.58,
|
| 16607 |
+
"grad_norm": 6.253279685974121,
|
| 16608 |
+
"learning_rate": 4.616949152542373e-06,
|
| 16609 |
+
"loss": 1.4559,
|
| 16610 |
+
"step": 23190
|
| 16611 |
+
},
|
| 16612 |
+
{
|
| 16613 |
+
"epoch": 0.58,
|
| 16614 |
+
"grad_norm": 5.3615827560424805,
|
| 16615 |
+
"learning_rate": 4.610169491525424e-06,
|
| 16616 |
+
"loss": 1.1649,
|
| 16617 |
+
"step": 23200
|
| 16618 |
+
},
|
| 16619 |
+
{
|
| 16620 |
+
"epoch": 0.58,
|
| 16621 |
+
"grad_norm": 9.165109634399414,
|
| 16622 |
+
"learning_rate": 4.603389830508475e-06,
|
| 16623 |
+
"loss": 1.4435,
|
| 16624 |
+
"step": 23210
|
| 16625 |
+
},
|
| 16626 |
+
{
|
| 16627 |
+
"epoch": 0.58,
|
| 16628 |
+
"grad_norm": 6.625391483306885,
|
| 16629 |
+
"learning_rate": 4.596610169491526e-06,
|
| 16630 |
+
"loss": 1.2911,
|
| 16631 |
+
"step": 23220
|
| 16632 |
+
},
|
| 16633 |
+
{
|
| 16634 |
+
"epoch": 0.58,
|
| 16635 |
+
"grad_norm": 3.898466110229492,
|
| 16636 |
+
"learning_rate": 4.589830508474576e-06,
|
| 16637 |
+
"loss": 1.2653,
|
| 16638 |
+
"step": 23230
|
| 16639 |
+
},
|
| 16640 |
+
{
|
| 16641 |
+
"epoch": 0.58,
|
| 16642 |
+
"grad_norm": 1.5483791828155518,
|
| 16643 |
+
"learning_rate": 4.583050847457628e-06,
|
| 16644 |
+
"loss": 1.1665,
|
| 16645 |
+
"step": 23240
|
| 16646 |
+
},
|
| 16647 |
+
{
|
| 16648 |
+
"epoch": 0.58,
|
| 16649 |
+
"grad_norm": 4.248947620391846,
|
| 16650 |
+
"learning_rate": 4.576271186440678e-06,
|
| 16651 |
+
"loss": 1.2275,
|
| 16652 |
+
"step": 23250
|
| 16653 |
+
},
|
| 16654 |
+
{
|
| 16655 |
+
"epoch": 0.58,
|
| 16656 |
+
"grad_norm": 3.522451639175415,
|
| 16657 |
+
"learning_rate": 4.56949152542373e-06,
|
| 16658 |
+
"loss": 1.3427,
|
| 16659 |
+
"step": 23260
|
| 16660 |
+
},
|
| 16661 |
+
{
|
| 16662 |
+
"epoch": 0.58,
|
| 16663 |
+
"grad_norm": 5.79518461227417,
|
| 16664 |
+
"learning_rate": 4.56271186440678e-06,
|
| 16665 |
+
"loss": 1.2545,
|
| 16666 |
+
"step": 23270
|
| 16667 |
+
},
|
| 16668 |
+
{
|
| 16669 |
+
"epoch": 0.58,
|
| 16670 |
+
"grad_norm": 7.211407661437988,
|
| 16671 |
+
"learning_rate": 4.555932203389831e-06,
|
| 16672 |
+
"loss": 1.3256,
|
| 16673 |
+
"step": 23280
|
| 16674 |
+
},
|
| 16675 |
+
{
|
| 16676 |
+
"epoch": 0.58,
|
| 16677 |
+
"grad_norm": 2.218186855316162,
|
| 16678 |
+
"learning_rate": 4.549152542372881e-06,
|
| 16679 |
+
"loss": 1.327,
|
| 16680 |
+
"step": 23290
|
| 16681 |
+
},
|
| 16682 |
+
{
|
| 16683 |
+
"epoch": 0.58,
|
| 16684 |
+
"grad_norm": 3.0725247859954834,
|
| 16685 |
+
"learning_rate": 4.542372881355933e-06,
|
| 16686 |
+
"loss": 1.1152,
|
| 16687 |
+
"step": 23300
|
| 16688 |
+
},
|
| 16689 |
+
{
|
| 16690 |
+
"epoch": 0.58,
|
| 16691 |
+
"grad_norm": 10.184927940368652,
|
| 16692 |
+
"learning_rate": 4.535593220338983e-06,
|
| 16693 |
+
"loss": 1.3153,
|
| 16694 |
+
"step": 23310
|
| 16695 |
+
},
|
| 16696 |
+
{
|
| 16697 |
+
"epoch": 0.58,
|
| 16698 |
+
"grad_norm": 6.575405597686768,
|
| 16699 |
+
"learning_rate": 4.528813559322035e-06,
|
| 16700 |
+
"loss": 1.2879,
|
| 16701 |
+
"step": 23320
|
| 16702 |
+
},
|
| 16703 |
+
{
|
| 16704 |
+
"epoch": 0.58,
|
| 16705 |
+
"grad_norm": 5.2145094871521,
|
| 16706 |
+
"learning_rate": 4.522033898305085e-06,
|
| 16707 |
+
"loss": 1.2836,
|
| 16708 |
+
"step": 23330
|
| 16709 |
+
},
|
| 16710 |
+
{
|
| 16711 |
+
"epoch": 0.58,
|
| 16712 |
+
"grad_norm": 16.586687088012695,
|
| 16713 |
+
"learning_rate": 4.515254237288136e-06,
|
| 16714 |
+
"loss": 1.245,
|
| 16715 |
+
"step": 23340
|
| 16716 |
+
},
|
| 16717 |
+
{
|
| 16718 |
+
"epoch": 0.58,
|
| 16719 |
+
"grad_norm": 5.264496803283691,
|
| 16720 |
+
"learning_rate": 4.508474576271187e-06,
|
| 16721 |
+
"loss": 1.1731,
|
| 16722 |
+
"step": 23350
|
| 16723 |
+
},
|
| 16724 |
+
{
|
| 16725 |
+
"epoch": 0.58,
|
| 16726 |
+
"grad_norm": 5.267670631408691,
|
| 16727 |
+
"learning_rate": 4.501694915254238e-06,
|
| 16728 |
+
"loss": 1.3662,
|
| 16729 |
+
"step": 23360
|
| 16730 |
+
},
|
| 16731 |
+
{
|
| 16732 |
+
"epoch": 0.58,
|
| 16733 |
+
"grad_norm": 7.335075855255127,
|
| 16734 |
+
"learning_rate": 4.494915254237289e-06,
|
| 16735 |
+
"loss": 1.3261,
|
| 16736 |
+
"step": 23370
|
| 16737 |
+
},
|
| 16738 |
+
{
|
| 16739 |
+
"epoch": 0.58,
|
| 16740 |
+
"grad_norm": 7.068376064300537,
|
| 16741 |
+
"learning_rate": 4.488135593220339e-06,
|
| 16742 |
+
"loss": 1.2805,
|
| 16743 |
+
"step": 23380
|
| 16744 |
+
},
|
| 16745 |
+
{
|
| 16746 |
+
"epoch": 0.58,
|
| 16747 |
+
"grad_norm": 15.14303207397461,
|
| 16748 |
+
"learning_rate": 4.48135593220339e-06,
|
| 16749 |
+
"loss": 1.3605,
|
| 16750 |
+
"step": 23390
|
| 16751 |
+
},
|
| 16752 |
+
{
|
| 16753 |
+
"epoch": 0.58,
|
| 16754 |
+
"grad_norm": 12.552229881286621,
|
| 16755 |
+
"learning_rate": 4.474576271186441e-06,
|
| 16756 |
+
"loss": 1.4804,
|
| 16757 |
+
"step": 23400
|
| 16758 |
+
},
|
| 16759 |
+
{
|
| 16760 |
+
"epoch": 0.59,
|
| 16761 |
+
"grad_norm": 6.760104656219482,
|
| 16762 |
+
"learning_rate": 4.467796610169492e-06,
|
| 16763 |
+
"loss": 1.2604,
|
| 16764 |
+
"step": 23410
|
| 16765 |
+
},
|
| 16766 |
+
{
|
| 16767 |
+
"epoch": 0.59,
|
| 16768 |
+
"grad_norm": 6.444414138793945,
|
| 16769 |
+
"learning_rate": 4.461016949152543e-06,
|
| 16770 |
+
"loss": 1.3677,
|
| 16771 |
+
"step": 23420
|
| 16772 |
+
},
|
| 16773 |
+
{
|
| 16774 |
+
"epoch": 0.59,
|
| 16775 |
+
"grad_norm": 2.223396062850952,
|
| 16776 |
+
"learning_rate": 4.454237288135594e-06,
|
| 16777 |
+
"loss": 1.4585,
|
| 16778 |
+
"step": 23430
|
| 16779 |
+
},
|
| 16780 |
+
{
|
| 16781 |
+
"epoch": 0.59,
|
| 16782 |
+
"grad_norm": 3.0469980239868164,
|
| 16783 |
+
"learning_rate": 4.447457627118645e-06,
|
| 16784 |
+
"loss": 1.2125,
|
| 16785 |
+
"step": 23440
|
| 16786 |
+
},
|
| 16787 |
+
{
|
| 16788 |
+
"epoch": 0.59,
|
| 16789 |
+
"grad_norm": 9.140281677246094,
|
| 16790 |
+
"learning_rate": 4.440677966101695e-06,
|
| 16791 |
+
"loss": 1.2847,
|
| 16792 |
+
"step": 23450
|
| 16793 |
+
},
|
| 16794 |
+
{
|
| 16795 |
+
"epoch": 0.59,
|
| 16796 |
+
"grad_norm": 10.596829414367676,
|
| 16797 |
+
"learning_rate": 4.433898305084746e-06,
|
| 16798 |
+
"loss": 1.1919,
|
| 16799 |
+
"step": 23460
|
| 16800 |
+
},
|
| 16801 |
+
{
|
| 16802 |
+
"epoch": 0.59,
|
| 16803 |
+
"grad_norm": 6.769688129425049,
|
| 16804 |
+
"learning_rate": 4.427118644067797e-06,
|
| 16805 |
+
"loss": 1.3911,
|
| 16806 |
+
"step": 23470
|
| 16807 |
+
},
|
| 16808 |
+
{
|
| 16809 |
+
"epoch": 0.59,
|
| 16810 |
+
"grad_norm": 8.3526029586792,
|
| 16811 |
+
"learning_rate": 4.420338983050848e-06,
|
| 16812 |
+
"loss": 1.3243,
|
| 16813 |
+
"step": 23480
|
| 16814 |
+
},
|
| 16815 |
+
{
|
| 16816 |
+
"epoch": 0.59,
|
| 16817 |
+
"grad_norm": 1.516774296760559,
|
| 16818 |
+
"learning_rate": 4.413559322033898e-06,
|
| 16819 |
+
"loss": 1.4328,
|
| 16820 |
+
"step": 23490
|
| 16821 |
+
},
|
| 16822 |
+
{
|
| 16823 |
+
"epoch": 0.59,
|
| 16824 |
+
"grad_norm": 5.758790493011475,
|
| 16825 |
+
"learning_rate": 4.40677966101695e-06,
|
| 16826 |
+
"loss": 1.1158,
|
| 16827 |
+
"step": 23500
|
| 16828 |
+
},
|
| 16829 |
+
{
|
| 16830 |
+
"epoch": 0.59,
|
| 16831 |
+
"eval_loss": 1.3372619152069092,
|
| 16832 |
+
"eval_runtime": 66.1468,
|
| 16833 |
+
"eval_samples_per_second": 15.118,
|
| 16834 |
+
"eval_steps_per_second": 15.118,
|
| 16835 |
+
"step": 23500
|
| 16836 |
+
},
|
| 16837 |
+
{
|
| 16838 |
+
"epoch": 0.59,
|
| 16839 |
+
"grad_norm": 10.471700668334961,
|
| 16840 |
+
"learning_rate": 4.4e-06,
|
| 16841 |
+
"loss": 1.1029,
|
| 16842 |
+
"step": 23510
|
| 16843 |
+
},
|
| 16844 |
+
{
|
| 16845 |
+
"epoch": 0.59,
|
| 16846 |
+
"grad_norm": 6.78934383392334,
|
| 16847 |
+
"learning_rate": 4.393220338983052e-06,
|
| 16848 |
+
"loss": 1.2838,
|
| 16849 |
+
"step": 23520
|
| 16850 |
+
},
|
| 16851 |
+
{
|
| 16852 |
+
"epoch": 0.59,
|
| 16853 |
+
"grad_norm": 4.890566825866699,
|
| 16854 |
+
"learning_rate": 4.386440677966102e-06,
|
| 16855 |
+
"loss": 1.2861,
|
| 16856 |
+
"step": 23530
|
| 16857 |
+
},
|
| 16858 |
+
{
|
| 16859 |
+
"epoch": 0.59,
|
| 16860 |
+
"grad_norm": 9.901065826416016,
|
| 16861 |
+
"learning_rate": 4.379661016949153e-06,
|
| 16862 |
+
"loss": 1.3043,
|
| 16863 |
+
"step": 23540
|
| 16864 |
+
},
|
| 16865 |
+
{
|
| 16866 |
+
"epoch": 0.59,
|
| 16867 |
+
"grad_norm": 3.332019805908203,
|
| 16868 |
+
"learning_rate": 4.372881355932203e-06,
|
| 16869 |
+
"loss": 1.5098,
|
| 16870 |
+
"step": 23550
|
| 16871 |
+
},
|
| 16872 |
+
{
|
| 16873 |
+
"epoch": 0.59,
|
| 16874 |
+
"grad_norm": 9.1102876663208,
|
| 16875 |
+
"learning_rate": 4.366101694915255e-06,
|
| 16876 |
+
"loss": 1.2222,
|
| 16877 |
+
"step": 23560
|
| 16878 |
+
},
|
| 16879 |
+
{
|
| 16880 |
+
"epoch": 0.59,
|
| 16881 |
+
"grad_norm": 2.800964832305908,
|
| 16882 |
+
"learning_rate": 4.359322033898305e-06,
|
| 16883 |
+
"loss": 1.1916,
|
| 16884 |
+
"step": 23570
|
| 16885 |
+
},
|
| 16886 |
+
{
|
| 16887 |
+
"epoch": 0.59,
|
| 16888 |
+
"grad_norm": 4.45274019241333,
|
| 16889 |
+
"learning_rate": 4.352542372881357e-06,
|
| 16890 |
+
"loss": 1.496,
|
| 16891 |
+
"step": 23580
|
| 16892 |
+
},
|
| 16893 |
+
{
|
| 16894 |
+
"epoch": 0.59,
|
| 16895 |
+
"grad_norm": 7.7979350090026855,
|
| 16896 |
+
"learning_rate": 4.345762711864407e-06,
|
| 16897 |
+
"loss": 1.3483,
|
| 16898 |
+
"step": 23590
|
| 16899 |
+
},
|
| 16900 |
+
{
|
| 16901 |
+
"epoch": 0.59,
|
| 16902 |
+
"grad_norm": 5.517279148101807,
|
| 16903 |
+
"learning_rate": 4.338983050847458e-06,
|
| 16904 |
+
"loss": 1.1602,
|
| 16905 |
+
"step": 23600
|
| 16906 |
+
},
|
| 16907 |
+
{
|
| 16908 |
+
"epoch": 0.59,
|
| 16909 |
+
"grad_norm": 8.224603652954102,
|
| 16910 |
+
"learning_rate": 4.332203389830509e-06,
|
| 16911 |
+
"loss": 1.2216,
|
| 16912 |
+
"step": 23610
|
| 16913 |
+
},
|
| 16914 |
+
{
|
| 16915 |
+
"epoch": 0.59,
|
| 16916 |
+
"grad_norm": 3.9079153537750244,
|
| 16917 |
+
"learning_rate": 4.32542372881356e-06,
|
| 16918 |
+
"loss": 1.4094,
|
| 16919 |
+
"step": 23620
|
| 16920 |
+
},
|
| 16921 |
+
{
|
| 16922 |
+
"epoch": 0.59,
|
| 16923 |
+
"grad_norm": 7.209962844848633,
|
| 16924 |
+
"learning_rate": 4.318644067796611e-06,
|
| 16925 |
+
"loss": 1.4033,
|
| 16926 |
+
"step": 23630
|
| 16927 |
+
},
|
| 16928 |
+
{
|
| 16929 |
+
"epoch": 0.59,
|
| 16930 |
+
"grad_norm": 6.915498733520508,
|
| 16931 |
+
"learning_rate": 4.311864406779661e-06,
|
| 16932 |
+
"loss": 1.3988,
|
| 16933 |
+
"step": 23640
|
| 16934 |
+
},
|
| 16935 |
+
{
|
| 16936 |
+
"epoch": 0.59,
|
| 16937 |
+
"grad_norm": 6.8702778816223145,
|
| 16938 |
+
"learning_rate": 4.305084745762712e-06,
|
| 16939 |
+
"loss": 1.3308,
|
| 16940 |
+
"step": 23650
|
| 16941 |
+
},
|
| 16942 |
+
{
|
| 16943 |
+
"epoch": 0.59,
|
| 16944 |
+
"grad_norm": 6.673946380615234,
|
| 16945 |
+
"learning_rate": 4.298305084745763e-06,
|
| 16946 |
+
"loss": 1.2756,
|
| 16947 |
+
"step": 23660
|
| 16948 |
+
},
|
| 16949 |
+
{
|
| 16950 |
+
"epoch": 0.59,
|
| 16951 |
+
"grad_norm": 2.729367971420288,
|
| 16952 |
+
"learning_rate": 4.291525423728814e-06,
|
| 16953 |
+
"loss": 1.2702,
|
| 16954 |
+
"step": 23670
|
| 16955 |
+
},
|
| 16956 |
+
{
|
| 16957 |
+
"epoch": 0.59,
|
| 16958 |
+
"grad_norm": 4.333055019378662,
|
| 16959 |
+
"learning_rate": 4.284745762711865e-06,
|
| 16960 |
+
"loss": 1.3365,
|
| 16961 |
+
"step": 23680
|
| 16962 |
+
},
|
| 16963 |
+
{
|
| 16964 |
+
"epoch": 0.59,
|
| 16965 |
+
"grad_norm": 8.36184024810791,
|
| 16966 |
+
"learning_rate": 4.277966101694915e-06,
|
| 16967 |
+
"loss": 1.2783,
|
| 16968 |
+
"step": 23690
|
| 16969 |
+
},
|
| 16970 |
+
{
|
| 16971 |
+
"epoch": 0.59,
|
| 16972 |
+
"grad_norm": 4.62699031829834,
|
| 16973 |
+
"learning_rate": 4.271186440677967e-06,
|
| 16974 |
+
"loss": 1.4385,
|
| 16975 |
+
"step": 23700
|
| 16976 |
+
},
|
| 16977 |
+
{
|
| 16978 |
+
"epoch": 0.59,
|
| 16979 |
+
"grad_norm": 3.193026304244995,
|
| 16980 |
+
"learning_rate": 4.264406779661017e-06,
|
| 16981 |
+
"loss": 1.4843,
|
| 16982 |
+
"step": 23710
|
| 16983 |
+
},
|
| 16984 |
+
{
|
| 16985 |
+
"epoch": 0.59,
|
| 16986 |
+
"grad_norm": 8.289533615112305,
|
| 16987 |
+
"learning_rate": 4.257627118644068e-06,
|
| 16988 |
+
"loss": 1.5263,
|
| 16989 |
+
"step": 23720
|
| 16990 |
+
},
|
| 16991 |
+
{
|
| 16992 |
+
"epoch": 0.59,
|
| 16993 |
+
"grad_norm": 3.887775182723999,
|
| 16994 |
+
"learning_rate": 4.250847457627119e-06,
|
| 16995 |
+
"loss": 1.228,
|
| 16996 |
+
"step": 23730
|
| 16997 |
+
},
|
| 16998 |
+
{
|
| 16999 |
+
"epoch": 0.59,
|
| 17000 |
+
"grad_norm": 10.728804588317871,
|
| 17001 |
+
"learning_rate": 4.24406779661017e-06,
|
| 17002 |
+
"loss": 1.4097,
|
| 17003 |
+
"step": 23740
|
| 17004 |
+
},
|
| 17005 |
+
{
|
| 17006 |
+
"epoch": 0.59,
|
| 17007 |
+
"grad_norm": 5.405580997467041,
|
| 17008 |
+
"learning_rate": 4.23728813559322e-06,
|
| 17009 |
+
"loss": 1.227,
|
| 17010 |
+
"step": 23750
|
| 17011 |
+
},
|
| 17012 |
+
{
|
| 17013 |
+
"epoch": 0.59,
|
| 17014 |
+
"grad_norm": 2.104985237121582,
|
| 17015 |
+
"learning_rate": 4.230508474576272e-06,
|
| 17016 |
+
"loss": 1.3258,
|
| 17017 |
+
"step": 23760
|
| 17018 |
+
},
|
| 17019 |
+
{
|
| 17020 |
+
"epoch": 0.59,
|
| 17021 |
+
"grad_norm": 11.678805351257324,
|
| 17022 |
+
"learning_rate": 4.223728813559322e-06,
|
| 17023 |
+
"loss": 1.1797,
|
| 17024 |
+
"step": 23770
|
| 17025 |
+
},
|
| 17026 |
+
{
|
| 17027 |
+
"epoch": 0.59,
|
| 17028 |
+
"grad_norm": 12.024051666259766,
|
| 17029 |
+
"learning_rate": 4.216949152542374e-06,
|
| 17030 |
+
"loss": 1.3278,
|
| 17031 |
+
"step": 23780
|
| 17032 |
+
},
|
| 17033 |
+
{
|
| 17034 |
+
"epoch": 0.59,
|
| 17035 |
+
"grad_norm": 12.879485130310059,
|
| 17036 |
+
"learning_rate": 4.210169491525424e-06,
|
| 17037 |
+
"loss": 1.2552,
|
| 17038 |
+
"step": 23790
|
| 17039 |
+
},
|
| 17040 |
+
{
|
| 17041 |
+
"epoch": 0.59,
|
| 17042 |
+
"grad_norm": 6.001992702484131,
|
| 17043 |
+
"learning_rate": 4.203389830508475e-06,
|
| 17044 |
+
"loss": 1.4639,
|
| 17045 |
+
"step": 23800
|
| 17046 |
+
},
|
| 17047 |
+
{
|
| 17048 |
+
"epoch": 0.6,
|
| 17049 |
+
"grad_norm": 7.713657855987549,
|
| 17050 |
+
"learning_rate": 4.196610169491525e-06,
|
| 17051 |
+
"loss": 1.3542,
|
| 17052 |
+
"step": 23810
|
| 17053 |
+
},
|
| 17054 |
+
{
|
| 17055 |
+
"epoch": 0.6,
|
| 17056 |
+
"grad_norm": 25.137435913085938,
|
| 17057 |
+
"learning_rate": 4.189830508474577e-06,
|
| 17058 |
+
"loss": 1.2694,
|
| 17059 |
+
"step": 23820
|
| 17060 |
+
},
|
| 17061 |
+
{
|
| 17062 |
+
"epoch": 0.6,
|
| 17063 |
+
"grad_norm": 13.080780029296875,
|
| 17064 |
+
"learning_rate": 4.183050847457627e-06,
|
| 17065 |
+
"loss": 1.5512,
|
| 17066 |
+
"step": 23830
|
| 17067 |
+
},
|
| 17068 |
+
{
|
| 17069 |
+
"epoch": 0.6,
|
| 17070 |
+
"grad_norm": 3.648967981338501,
|
| 17071 |
+
"learning_rate": 4.176271186440679e-06,
|
| 17072 |
+
"loss": 1.4919,
|
| 17073 |
+
"step": 23840
|
| 17074 |
+
},
|
| 17075 |
+
{
|
| 17076 |
+
"epoch": 0.6,
|
| 17077 |
+
"grad_norm": 2.8366498947143555,
|
| 17078 |
+
"learning_rate": 4.169491525423729e-06,
|
| 17079 |
+
"loss": 1.3528,
|
| 17080 |
+
"step": 23850
|
| 17081 |
+
},
|
| 17082 |
+
{
|
| 17083 |
+
"epoch": 0.6,
|
| 17084 |
+
"grad_norm": 2.3198916912078857,
|
| 17085 |
+
"learning_rate": 4.16271186440678e-06,
|
| 17086 |
+
"loss": 1.4445,
|
| 17087 |
+
"step": 23860
|
| 17088 |
+
},
|
| 17089 |
+
{
|
| 17090 |
+
"epoch": 0.6,
|
| 17091 |
+
"grad_norm": 9.170830726623535,
|
| 17092 |
+
"learning_rate": 4.155932203389831e-06,
|
| 17093 |
+
"loss": 1.1801,
|
| 17094 |
+
"step": 23870
|
| 17095 |
+
},
|
| 17096 |
+
{
|
| 17097 |
+
"epoch": 0.6,
|
| 17098 |
+
"grad_norm": 7.985089302062988,
|
| 17099 |
+
"learning_rate": 4.149152542372882e-06,
|
| 17100 |
+
"loss": 1.3195,
|
| 17101 |
+
"step": 23880
|
| 17102 |
+
},
|
| 17103 |
+
{
|
| 17104 |
+
"epoch": 0.6,
|
| 17105 |
+
"grad_norm": 10.688753128051758,
|
| 17106 |
+
"learning_rate": 4.142372881355933e-06,
|
| 17107 |
+
"loss": 1.4527,
|
| 17108 |
+
"step": 23890
|
| 17109 |
+
},
|
| 17110 |
+
{
|
| 17111 |
+
"epoch": 0.6,
|
| 17112 |
+
"grad_norm": 12.181285858154297,
|
| 17113 |
+
"learning_rate": 4.135593220338983e-06,
|
| 17114 |
+
"loss": 1.3875,
|
| 17115 |
+
"step": 23900
|
| 17116 |
+
},
|
| 17117 |
+
{
|
| 17118 |
+
"epoch": 0.6,
|
| 17119 |
+
"grad_norm": 10.353550910949707,
|
| 17120 |
+
"learning_rate": 4.128813559322034e-06,
|
| 17121 |
+
"loss": 1.3937,
|
| 17122 |
+
"step": 23910
|
| 17123 |
+
},
|
| 17124 |
+
{
|
| 17125 |
+
"epoch": 0.6,
|
| 17126 |
+
"grad_norm": 3.3962326049804688,
|
| 17127 |
+
"learning_rate": 4.122033898305085e-06,
|
| 17128 |
+
"loss": 1.2212,
|
| 17129 |
+
"step": 23920
|
| 17130 |
+
},
|
| 17131 |
+
{
|
| 17132 |
+
"epoch": 0.6,
|
| 17133 |
+
"grad_norm": 9.191743850708008,
|
| 17134 |
+
"learning_rate": 4.115254237288136e-06,
|
| 17135 |
+
"loss": 1.3884,
|
| 17136 |
+
"step": 23930
|
| 17137 |
+
},
|
| 17138 |
+
{
|
| 17139 |
+
"epoch": 0.6,
|
| 17140 |
+
"grad_norm": 8.74504566192627,
|
| 17141 |
+
"learning_rate": 4.108474576271187e-06,
|
| 17142 |
+
"loss": 1.369,
|
| 17143 |
+
"step": 23940
|
| 17144 |
+
},
|
| 17145 |
+
{
|
| 17146 |
+
"epoch": 0.6,
|
| 17147 |
+
"grad_norm": 15.484914779663086,
|
| 17148 |
+
"learning_rate": 4.101694915254237e-06,
|
| 17149 |
+
"loss": 1.3607,
|
| 17150 |
+
"step": 23950
|
| 17151 |
+
},
|
| 17152 |
+
{
|
| 17153 |
+
"epoch": 0.6,
|
| 17154 |
+
"grad_norm": 8.069631576538086,
|
| 17155 |
+
"learning_rate": 4.094915254237289e-06,
|
| 17156 |
+
"loss": 1.1674,
|
| 17157 |
+
"step": 23960
|
| 17158 |
+
},
|
| 17159 |
+
{
|
| 17160 |
+
"epoch": 0.6,
|
| 17161 |
+
"grad_norm": 5.688279151916504,
|
| 17162 |
+
"learning_rate": 4.088135593220339e-06,
|
| 17163 |
+
"loss": 1.2652,
|
| 17164 |
+
"step": 23970
|
| 17165 |
+
},
|
| 17166 |
+
{
|
| 17167 |
+
"epoch": 0.6,
|
| 17168 |
+
"grad_norm": 2.326960325241089,
|
| 17169 |
+
"learning_rate": 4.081355932203391e-06,
|
| 17170 |
+
"loss": 1.2149,
|
| 17171 |
+
"step": 23980
|
| 17172 |
+
},
|
| 17173 |
+
{
|
| 17174 |
+
"epoch": 0.6,
|
| 17175 |
+
"grad_norm": 9.749725341796875,
|
| 17176 |
+
"learning_rate": 4.074576271186441e-06,
|
| 17177 |
+
"loss": 1.2378,
|
| 17178 |
+
"step": 23990
|
| 17179 |
+
},
|
| 17180 |
+
{
|
| 17181 |
+
"epoch": 0.6,
|
| 17182 |
+
"grad_norm": 5.552289962768555,
|
| 17183 |
+
"learning_rate": 4.067796610169492e-06,
|
| 17184 |
+
"loss": 1.2699,
|
| 17185 |
+
"step": 24000
|
| 17186 |
+
},
|
| 17187 |
+
{
|
| 17188 |
+
"epoch": 0.6,
|
| 17189 |
+
"eval_loss": 1.3411681652069092,
|
| 17190 |
+
"eval_runtime": 66.1396,
|
| 17191 |
+
"eval_samples_per_second": 15.12,
|
| 17192 |
+
"eval_steps_per_second": 15.12,
|
| 17193 |
+
"step": 24000
|
| 17194 |
+
},
|
| 17195 |
+
{
|
| 17196 |
+
"epoch": 0.6,
|
| 17197 |
+
"grad_norm": 8.768118858337402,
|
| 17198 |
+
"learning_rate": 4.061016949152542e-06,
|
| 17199 |
+
"loss": 1.1869,
|
| 17200 |
+
"step": 24010
|
| 17201 |
+
},
|
| 17202 |
+
{
|
| 17203 |
+
"epoch": 0.6,
|
| 17204 |
+
"grad_norm": 1.0193852186203003,
|
| 17205 |
+
"learning_rate": 4.054237288135594e-06,
|
| 17206 |
+
"loss": 1.1477,
|
| 17207 |
+
"step": 24020
|
| 17208 |
+
},
|
| 17209 |
+
{
|
| 17210 |
+
"epoch": 0.6,
|
| 17211 |
+
"grad_norm": 11.04339599609375,
|
| 17212 |
+
"learning_rate": 4.047457627118644e-06,
|
| 17213 |
+
"loss": 1.2492,
|
| 17214 |
+
"step": 24030
|
| 17215 |
+
},
|
| 17216 |
+
{
|
| 17217 |
+
"epoch": 0.6,
|
| 17218 |
+
"grad_norm": 2.5347607135772705,
|
| 17219 |
+
"learning_rate": 4.040677966101696e-06,
|
| 17220 |
+
"loss": 1.2424,
|
| 17221 |
+
"step": 24040
|
| 17222 |
+
},
|
| 17223 |
+
{
|
| 17224 |
+
"epoch": 0.6,
|
| 17225 |
+
"grad_norm": 5.121871471405029,
|
| 17226 |
+
"learning_rate": 4.033898305084746e-06,
|
| 17227 |
+
"loss": 1.1773,
|
| 17228 |
+
"step": 24050
|
| 17229 |
+
},
|
| 17230 |
+
{
|
| 17231 |
+
"epoch": 0.6,
|
| 17232 |
+
"grad_norm": 8.53433609008789,
|
| 17233 |
+
"learning_rate": 4.027118644067797e-06,
|
| 17234 |
+
"loss": 1.4407,
|
| 17235 |
+
"step": 24060
|
| 17236 |
+
},
|
| 17237 |
+
{
|
| 17238 |
+
"epoch": 0.6,
|
| 17239 |
+
"grad_norm": 11.311376571655273,
|
| 17240 |
+
"learning_rate": 4.020338983050847e-06,
|
| 17241 |
+
"loss": 1.4095,
|
| 17242 |
+
"step": 24070
|
| 17243 |
+
},
|
| 17244 |
+
{
|
| 17245 |
+
"epoch": 0.6,
|
| 17246 |
+
"grad_norm": 2.8956375122070312,
|
| 17247 |
+
"learning_rate": 4.013559322033899e-06,
|
| 17248 |
+
"loss": 1.3076,
|
| 17249 |
+
"step": 24080
|
| 17250 |
+
},
|
| 17251 |
+
{
|
| 17252 |
+
"epoch": 0.6,
|
| 17253 |
+
"grad_norm": 3.6406021118164062,
|
| 17254 |
+
"learning_rate": 4.006779661016949e-06,
|
| 17255 |
+
"loss": 1.1801,
|
| 17256 |
+
"step": 24090
|
| 17257 |
+
},
|
| 17258 |
+
{
|
| 17259 |
+
"epoch": 0.6,
|
| 17260 |
+
"grad_norm": 4.67333459854126,
|
| 17261 |
+
"learning_rate": 4.000000000000001e-06,
|
| 17262 |
+
"loss": 1.3627,
|
| 17263 |
+
"step": 24100
|
| 17264 |
+
},
|
| 17265 |
+
{
|
| 17266 |
+
"epoch": 0.6,
|
| 17267 |
+
"grad_norm": 4.243159294128418,
|
| 17268 |
+
"learning_rate": 3.993220338983051e-06,
|
| 17269 |
+
"loss": 1.2065,
|
| 17270 |
+
"step": 24110
|
| 17271 |
+
},
|
| 17272 |
+
{
|
| 17273 |
+
"epoch": 0.6,
|
| 17274 |
+
"grad_norm": 4.570652484893799,
|
| 17275 |
+
"learning_rate": 3.986440677966102e-06,
|
| 17276 |
+
"loss": 1.4576,
|
| 17277 |
+
"step": 24120
|
| 17278 |
+
},
|
| 17279 |
+
{
|
| 17280 |
+
"epoch": 0.6,
|
| 17281 |
+
"grad_norm": 10.30574893951416,
|
| 17282 |
+
"learning_rate": 3.979661016949153e-06,
|
| 17283 |
+
"loss": 1.3798,
|
| 17284 |
+
"step": 24130
|
| 17285 |
+
},
|
| 17286 |
+
{
|
| 17287 |
+
"epoch": 0.6,
|
| 17288 |
+
"grad_norm": 1.7883845567703247,
|
| 17289 |
+
"learning_rate": 3.972881355932204e-06,
|
| 17290 |
+
"loss": 1.1474,
|
| 17291 |
+
"step": 24140
|
| 17292 |
+
},
|
| 17293 |
+
{
|
| 17294 |
+
"epoch": 0.6,
|
| 17295 |
+
"grad_norm": 2.429614305496216,
|
| 17296 |
+
"learning_rate": 3.966101694915255e-06,
|
| 17297 |
+
"loss": 1.3992,
|
| 17298 |
+
"step": 24150
|
| 17299 |
+
},
|
| 17300 |
+
{
|
| 17301 |
+
"epoch": 0.6,
|
| 17302 |
+
"grad_norm": 5.791226863861084,
|
| 17303 |
+
"learning_rate": 3.959322033898305e-06,
|
| 17304 |
+
"loss": 1.4164,
|
| 17305 |
+
"step": 24160
|
| 17306 |
+
},
|
| 17307 |
+
{
|
| 17308 |
+
"epoch": 0.6,
|
| 17309 |
+
"grad_norm": 6.212001800537109,
|
| 17310 |
+
"learning_rate": 3.952542372881356e-06,
|
| 17311 |
+
"loss": 1.4647,
|
| 17312 |
+
"step": 24170
|
| 17313 |
+
},
|
| 17314 |
+
{
|
| 17315 |
+
"epoch": 0.6,
|
| 17316 |
+
"grad_norm": 4.9569292068481445,
|
| 17317 |
+
"learning_rate": 3.945762711864407e-06,
|
| 17318 |
+
"loss": 1.2948,
|
| 17319 |
+
"step": 24180
|
| 17320 |
+
},
|
| 17321 |
+
{
|
| 17322 |
+
"epoch": 0.6,
|
| 17323 |
+
"grad_norm": 2.2119970321655273,
|
| 17324 |
+
"learning_rate": 3.938983050847458e-06,
|
| 17325 |
+
"loss": 1.3955,
|
| 17326 |
+
"step": 24190
|
| 17327 |
+
},
|
| 17328 |
+
{
|
| 17329 |
+
"epoch": 0.6,
|
| 17330 |
+
"grad_norm": 10.280770301818848,
|
| 17331 |
+
"learning_rate": 3.932203389830509e-06,
|
| 17332 |
+
"loss": 1.4461,
|
| 17333 |
+
"step": 24200
|
| 17334 |
+
},
|
| 17335 |
+
{
|
| 17336 |
+
"epoch": 0.61,
|
| 17337 |
+
"grad_norm": 3.701272487640381,
|
| 17338 |
+
"learning_rate": 3.925423728813559e-06,
|
| 17339 |
+
"loss": 1.348,
|
| 17340 |
+
"step": 24210
|
| 17341 |
+
},
|
| 17342 |
+
{
|
| 17343 |
+
"epoch": 0.61,
|
| 17344 |
+
"grad_norm": 8.827926635742188,
|
| 17345 |
+
"learning_rate": 3.918644067796611e-06,
|
| 17346 |
+
"loss": 1.3398,
|
| 17347 |
+
"step": 24220
|
| 17348 |
+
},
|
| 17349 |
+
{
|
| 17350 |
+
"epoch": 0.61,
|
| 17351 |
+
"grad_norm": 6.997286319732666,
|
| 17352 |
+
"learning_rate": 3.911864406779661e-06,
|
| 17353 |
+
"loss": 1.4724,
|
| 17354 |
+
"step": 24230
|
| 17355 |
+
},
|
| 17356 |
+
{
|
| 17357 |
+
"epoch": 0.61,
|
| 17358 |
+
"grad_norm": 5.5268449783325195,
|
| 17359 |
+
"learning_rate": 3.905084745762713e-06,
|
| 17360 |
+
"loss": 1.388,
|
| 17361 |
+
"step": 24240
|
| 17362 |
+
},
|
| 17363 |
+
{
|
| 17364 |
+
"epoch": 0.61,
|
| 17365 |
+
"grad_norm": 8.842992782592773,
|
| 17366 |
+
"learning_rate": 3.898305084745763e-06,
|
| 17367 |
+
"loss": 1.2382,
|
| 17368 |
+
"step": 24250
|
| 17369 |
+
},
|
| 17370 |
+
{
|
| 17371 |
+
"epoch": 0.61,
|
| 17372 |
+
"grad_norm": 11.24975872039795,
|
| 17373 |
+
"learning_rate": 3.891525423728814e-06,
|
| 17374 |
+
"loss": 1.2194,
|
| 17375 |
+
"step": 24260
|
| 17376 |
+
},
|
| 17377 |
+
{
|
| 17378 |
+
"epoch": 0.61,
|
| 17379 |
+
"grad_norm": 2.875722646713257,
|
| 17380 |
+
"learning_rate": 3.884745762711864e-06,
|
| 17381 |
+
"loss": 1.3792,
|
| 17382 |
+
"step": 24270
|
| 17383 |
+
},
|
| 17384 |
+
{
|
| 17385 |
+
"epoch": 0.61,
|
| 17386 |
+
"grad_norm": 8.459474563598633,
|
| 17387 |
+
"learning_rate": 3.877966101694916e-06,
|
| 17388 |
+
"loss": 1.3021,
|
| 17389 |
+
"step": 24280
|
| 17390 |
+
},
|
| 17391 |
+
{
|
| 17392 |
+
"epoch": 0.61,
|
| 17393 |
+
"grad_norm": 3.315873861312866,
|
| 17394 |
+
"learning_rate": 3.871186440677966e-06,
|
| 17395 |
+
"loss": 1.2976,
|
| 17396 |
+
"step": 24290
|
| 17397 |
+
},
|
| 17398 |
+
{
|
| 17399 |
+
"epoch": 0.61,
|
| 17400 |
+
"grad_norm": 6.280729293823242,
|
| 17401 |
+
"learning_rate": 3.864406779661018e-06,
|
| 17402 |
+
"loss": 1.3294,
|
| 17403 |
+
"step": 24300
|
| 17404 |
+
},
|
| 17405 |
+
{
|
| 17406 |
+
"epoch": 0.61,
|
| 17407 |
+
"grad_norm": 6.004711627960205,
|
| 17408 |
+
"learning_rate": 3.857627118644068e-06,
|
| 17409 |
+
"loss": 1.3178,
|
| 17410 |
+
"step": 24310
|
| 17411 |
+
},
|
| 17412 |
+
{
|
| 17413 |
+
"epoch": 0.61,
|
| 17414 |
+
"grad_norm": 8.207845687866211,
|
| 17415 |
+
"learning_rate": 3.850847457627119e-06,
|
| 17416 |
+
"loss": 1.424,
|
| 17417 |
+
"step": 24320
|
| 17418 |
+
},
|
| 17419 |
+
{
|
| 17420 |
+
"epoch": 0.61,
|
| 17421 |
+
"grad_norm": 8.01065444946289,
|
| 17422 |
+
"learning_rate": 3.844067796610169e-06,
|
| 17423 |
+
"loss": 1.2842,
|
| 17424 |
+
"step": 24330
|
| 17425 |
+
},
|
| 17426 |
+
{
|
| 17427 |
+
"epoch": 0.61,
|
| 17428 |
+
"grad_norm": 9.126721382141113,
|
| 17429 |
+
"learning_rate": 3.837288135593221e-06,
|
| 17430 |
+
"loss": 1.368,
|
| 17431 |
+
"step": 24340
|
| 17432 |
+
},
|
| 17433 |
+
{
|
| 17434 |
+
"epoch": 0.61,
|
| 17435 |
+
"grad_norm": 11.590188026428223,
|
| 17436 |
+
"learning_rate": 3.830508474576271e-06,
|
| 17437 |
+
"loss": 1.375,
|
| 17438 |
+
"step": 24350
|
| 17439 |
+
},
|
| 17440 |
+
{
|
| 17441 |
+
"epoch": 0.61,
|
| 17442 |
+
"grad_norm": 7.325139045715332,
|
| 17443 |
+
"learning_rate": 3.823728813559323e-06,
|
| 17444 |
+
"loss": 1.2593,
|
| 17445 |
+
"step": 24360
|
| 17446 |
+
},
|
| 17447 |
+
{
|
| 17448 |
+
"epoch": 0.61,
|
| 17449 |
+
"grad_norm": 6.3924760818481445,
|
| 17450 |
+
"learning_rate": 3.816949152542373e-06,
|
| 17451 |
+
"loss": 1.3238,
|
| 17452 |
+
"step": 24370
|
| 17453 |
+
},
|
| 17454 |
+
{
|
| 17455 |
+
"epoch": 0.61,
|
| 17456 |
+
"grad_norm": 5.093543529510498,
|
| 17457 |
+
"learning_rate": 3.8101694915254238e-06,
|
| 17458 |
+
"loss": 1.3398,
|
| 17459 |
+
"step": 24380
|
| 17460 |
+
},
|
| 17461 |
+
{
|
| 17462 |
+
"epoch": 0.61,
|
| 17463 |
+
"grad_norm": 4.488302707672119,
|
| 17464 |
+
"learning_rate": 3.8033898305084748e-06,
|
| 17465 |
+
"loss": 1.2532,
|
| 17466 |
+
"step": 24390
|
| 17467 |
+
},
|
| 17468 |
+
{
|
| 17469 |
+
"epoch": 0.61,
|
| 17470 |
+
"grad_norm": 3.5369062423706055,
|
| 17471 |
+
"learning_rate": 3.7966101694915257e-06,
|
| 17472 |
+
"loss": 1.1723,
|
| 17473 |
+
"step": 24400
|
| 17474 |
+
},
|
| 17475 |
+
{
|
| 17476 |
+
"epoch": 0.61,
|
| 17477 |
+
"grad_norm": 3.2012510299682617,
|
| 17478 |
+
"learning_rate": 3.7898305084745767e-06,
|
| 17479 |
+
"loss": 1.3348,
|
| 17480 |
+
"step": 24410
|
| 17481 |
+
},
|
| 17482 |
+
{
|
| 17483 |
+
"epoch": 0.61,
|
| 17484 |
+
"grad_norm": 1.5665017366409302,
|
| 17485 |
+
"learning_rate": 3.7830508474576273e-06,
|
| 17486 |
+
"loss": 1.4159,
|
| 17487 |
+
"step": 24420
|
| 17488 |
+
},
|
| 17489 |
+
{
|
| 17490 |
+
"epoch": 0.61,
|
| 17491 |
+
"grad_norm": 12.912787437438965,
|
| 17492 |
+
"learning_rate": 3.7762711864406782e-06,
|
| 17493 |
+
"loss": 1.2406,
|
| 17494 |
+
"step": 24430
|
| 17495 |
+
},
|
| 17496 |
+
{
|
| 17497 |
+
"epoch": 0.61,
|
| 17498 |
+
"grad_norm": 6.572142124176025,
|
| 17499 |
+
"learning_rate": 3.7694915254237292e-06,
|
| 17500 |
+
"loss": 1.1544,
|
| 17501 |
+
"step": 24440
|
| 17502 |
+
},
|
| 17503 |
+
{
|
| 17504 |
+
"epoch": 0.61,
|
| 17505 |
+
"grad_norm": 4.999161720275879,
|
| 17506 |
+
"learning_rate": 3.76271186440678e-06,
|
| 17507 |
+
"loss": 1.1744,
|
| 17508 |
+
"step": 24450
|
| 17509 |
+
},
|
| 17510 |
+
{
|
| 17511 |
+
"epoch": 0.61,
|
| 17512 |
+
"grad_norm": 3.322866439819336,
|
| 17513 |
+
"learning_rate": 3.755932203389831e-06,
|
| 17514 |
+
"loss": 1.3713,
|
| 17515 |
+
"step": 24460
|
| 17516 |
+
},
|
| 17517 |
+
{
|
| 17518 |
+
"epoch": 0.61,
|
| 17519 |
+
"grad_norm": 5.197652816772461,
|
| 17520 |
+
"learning_rate": 3.7491525423728813e-06,
|
| 17521 |
+
"loss": 1.1858,
|
| 17522 |
+
"step": 24470
|
| 17523 |
+
},
|
| 17524 |
+
{
|
| 17525 |
+
"epoch": 0.61,
|
| 17526 |
+
"grad_norm": 6.7361369132995605,
|
| 17527 |
+
"learning_rate": 3.7423728813559323e-06,
|
| 17528 |
+
"loss": 1.2778,
|
| 17529 |
+
"step": 24480
|
| 17530 |
+
},
|
| 17531 |
+
{
|
| 17532 |
+
"epoch": 0.61,
|
| 17533 |
+
"grad_norm": 1.1276848316192627,
|
| 17534 |
+
"learning_rate": 3.7355932203389833e-06,
|
| 17535 |
+
"loss": 1.3971,
|
| 17536 |
+
"step": 24490
|
| 17537 |
+
},
|
| 17538 |
+
{
|
| 17539 |
+
"epoch": 0.61,
|
| 17540 |
+
"grad_norm": 4.62593412399292,
|
| 17541 |
+
"learning_rate": 3.7288135593220342e-06,
|
| 17542 |
+
"loss": 1.479,
|
| 17543 |
+
"step": 24500
|
| 17544 |
+
},
|
| 17545 |
+
{
|
| 17546 |
+
"epoch": 0.61,
|
| 17547 |
+
"eval_loss": 1.3212531805038452,
|
| 17548 |
+
"eval_runtime": 66.1318,
|
| 17549 |
+
"eval_samples_per_second": 15.121,
|
| 17550 |
+
"eval_steps_per_second": 15.121,
|
| 17551 |
+
"step": 24500
|
| 17552 |
+
},
|
| 17553 |
+
{
|
| 17554 |
+
"epoch": 0.61,
|
| 17555 |
+
"grad_norm": 7.978855133056641,
|
| 17556 |
+
"learning_rate": 3.7220338983050852e-06,
|
| 17557 |
+
"loss": 1.1919,
|
| 17558 |
+
"step": 24510
|
| 17559 |
+
},
|
| 17560 |
+
{
|
| 17561 |
+
"epoch": 0.61,
|
| 17562 |
+
"grad_norm": 6.347212314605713,
|
| 17563 |
+
"learning_rate": 3.715254237288136e-06,
|
| 17564 |
+
"loss": 1.2493,
|
| 17565 |
+
"step": 24520
|
| 17566 |
+
},
|
| 17567 |
+
{
|
| 17568 |
+
"epoch": 0.61,
|
| 17569 |
+
"grad_norm": 6.2206573486328125,
|
| 17570 |
+
"learning_rate": 3.7084745762711867e-06,
|
| 17571 |
+
"loss": 1.2258,
|
| 17572 |
+
"step": 24530
|
| 17573 |
+
},
|
| 17574 |
+
{
|
| 17575 |
+
"epoch": 0.61,
|
| 17576 |
+
"grad_norm": 2.548797607421875,
|
| 17577 |
+
"learning_rate": 3.7016949152542377e-06,
|
| 17578 |
+
"loss": 1.262,
|
| 17579 |
+
"step": 24540
|
| 17580 |
+
},
|
| 17581 |
+
{
|
| 17582 |
+
"epoch": 0.61,
|
| 17583 |
+
"grad_norm": 9.5992431640625,
|
| 17584 |
+
"learning_rate": 3.6949152542372883e-06,
|
| 17585 |
+
"loss": 1.4298,
|
| 17586 |
+
"step": 24550
|
| 17587 |
+
},
|
| 17588 |
+
{
|
| 17589 |
+
"epoch": 0.61,
|
| 17590 |
+
"grad_norm": 2.236175298690796,
|
| 17591 |
+
"learning_rate": 3.6881355932203393e-06,
|
| 17592 |
+
"loss": 1.3466,
|
| 17593 |
+
"step": 24560
|
| 17594 |
+
},
|
| 17595 |
+
{
|
| 17596 |
+
"epoch": 0.61,
|
| 17597 |
+
"grad_norm": 7.134004592895508,
|
| 17598 |
+
"learning_rate": 3.6813559322033902e-06,
|
| 17599 |
+
"loss": 1.2642,
|
| 17600 |
+
"step": 24570
|
| 17601 |
+
},
|
| 17602 |
+
{
|
| 17603 |
+
"epoch": 0.61,
|
| 17604 |
+
"grad_norm": 12.453125,
|
| 17605 |
+
"learning_rate": 3.6745762711864408e-06,
|
| 17606 |
+
"loss": 1.369,
|
| 17607 |
+
"step": 24580
|
| 17608 |
+
},
|
| 17609 |
+
{
|
| 17610 |
+
"epoch": 0.61,
|
| 17611 |
+
"grad_norm": 16.243106842041016,
|
| 17612 |
+
"learning_rate": 3.6677966101694918e-06,
|
| 17613 |
+
"loss": 1.3269,
|
| 17614 |
+
"step": 24590
|
| 17615 |
+
},
|
| 17616 |
+
{
|
| 17617 |
+
"epoch": 0.61,
|
| 17618 |
+
"grad_norm": 11.406882286071777,
|
| 17619 |
+
"learning_rate": 3.6610169491525427e-06,
|
| 17620 |
+
"loss": 1.4339,
|
| 17621 |
+
"step": 24600
|
| 17622 |
+
},
|
| 17623 |
+
{
|
| 17624 |
+
"epoch": 0.62,
|
| 17625 |
+
"grad_norm": 6.334946632385254,
|
| 17626 |
+
"learning_rate": 3.6542372881355937e-06,
|
| 17627 |
+
"loss": 1.2445,
|
| 17628 |
+
"step": 24610
|
| 17629 |
+
},
|
| 17630 |
+
{
|
| 17631 |
+
"epoch": 0.62,
|
| 17632 |
+
"grad_norm": 9.451517105102539,
|
| 17633 |
+
"learning_rate": 3.6474576271186447e-06,
|
| 17634 |
+
"loss": 1.1902,
|
| 17635 |
+
"step": 24620
|
| 17636 |
+
},
|
| 17637 |
+
{
|
| 17638 |
+
"epoch": 0.62,
|
| 17639 |
+
"grad_norm": 1.9071747064590454,
|
| 17640 |
+
"learning_rate": 3.640677966101695e-06,
|
| 17641 |
+
"loss": 1.3273,
|
| 17642 |
+
"step": 24630
|
| 17643 |
+
},
|
| 17644 |
+
{
|
| 17645 |
+
"epoch": 0.62,
|
| 17646 |
+
"grad_norm": 4.054659843444824,
|
| 17647 |
+
"learning_rate": 3.633898305084746e-06,
|
| 17648 |
+
"loss": 1.2808,
|
| 17649 |
+
"step": 24640
|
| 17650 |
+
},
|
| 17651 |
+
{
|
| 17652 |
+
"epoch": 0.62,
|
| 17653 |
+
"grad_norm": 6.314877986907959,
|
| 17654 |
+
"learning_rate": 3.6271186440677968e-06,
|
| 17655 |
+
"loss": 1.4351,
|
| 17656 |
+
"step": 24650
|
| 17657 |
+
},
|
| 17658 |
+
{
|
| 17659 |
+
"epoch": 0.62,
|
| 17660 |
+
"grad_norm": 6.971933841705322,
|
| 17661 |
+
"learning_rate": 3.6203389830508478e-06,
|
| 17662 |
+
"loss": 1.2973,
|
| 17663 |
+
"step": 24660
|
| 17664 |
+
},
|
| 17665 |
+
{
|
| 17666 |
+
"epoch": 0.62,
|
| 17667 |
+
"grad_norm": 2.7064402103424072,
|
| 17668 |
+
"learning_rate": 3.6135593220338987e-06,
|
| 17669 |
+
"loss": 1.4335,
|
| 17670 |
+
"step": 24670
|
| 17671 |
+
},
|
| 17672 |
+
{
|
| 17673 |
+
"epoch": 0.62,
|
| 17674 |
+
"grad_norm": 5.698015213012695,
|
| 17675 |
+
"learning_rate": 3.6067796610169493e-06,
|
| 17676 |
+
"loss": 1.3042,
|
| 17677 |
+
"step": 24680
|
| 17678 |
+
},
|
| 17679 |
+
{
|
| 17680 |
+
"epoch": 0.62,
|
| 17681 |
+
"grad_norm": 5.647088050842285,
|
| 17682 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 17683 |
+
"loss": 1.3832,
|
| 17684 |
+
"step": 24690
|
| 17685 |
+
},
|
| 17686 |
+
{
|
| 17687 |
+
"epoch": 0.62,
|
| 17688 |
+
"grad_norm": 10.945414543151855,
|
| 17689 |
+
"learning_rate": 3.5932203389830512e-06,
|
| 17690 |
+
"loss": 1.2363,
|
| 17691 |
+
"step": 24700
|
| 17692 |
+
},
|
| 17693 |
+
{
|
| 17694 |
+
"epoch": 0.62,
|
| 17695 |
+
"grad_norm": 7.516660213470459,
|
| 17696 |
+
"learning_rate": 3.5864406779661022e-06,
|
| 17697 |
+
"loss": 1.2615,
|
| 17698 |
+
"step": 24710
|
| 17699 |
+
},
|
| 17700 |
+
{
|
| 17701 |
+
"epoch": 0.62,
|
| 17702 |
+
"grad_norm": 3.6117703914642334,
|
| 17703 |
+
"learning_rate": 3.579661016949153e-06,
|
| 17704 |
+
"loss": 1.3533,
|
| 17705 |
+
"step": 24720
|
| 17706 |
+
},
|
| 17707 |
+
{
|
| 17708 |
+
"epoch": 0.62,
|
| 17709 |
+
"grad_norm": 10.817008972167969,
|
| 17710 |
+
"learning_rate": 3.5728813559322033e-06,
|
| 17711 |
+
"loss": 1.3425,
|
| 17712 |
+
"step": 24730
|
| 17713 |
+
},
|
| 17714 |
+
{
|
| 17715 |
+
"epoch": 0.62,
|
| 17716 |
+
"grad_norm": 7.7072858810424805,
|
| 17717 |
+
"learning_rate": 3.5661016949152543e-06,
|
| 17718 |
+
"loss": 1.2762,
|
| 17719 |
+
"step": 24740
|
| 17720 |
+
},
|
| 17721 |
+
{
|
| 17722 |
+
"epoch": 0.62,
|
| 17723 |
+
"grad_norm": 5.9887495040893555,
|
| 17724 |
+
"learning_rate": 3.5593220338983053e-06,
|
| 17725 |
+
"loss": 1.3516,
|
| 17726 |
+
"step": 24750
|
| 17727 |
+
},
|
| 17728 |
+
{
|
| 17729 |
+
"epoch": 0.62,
|
| 17730 |
+
"grad_norm": 3.4481067657470703,
|
| 17731 |
+
"learning_rate": 3.5525423728813563e-06,
|
| 17732 |
+
"loss": 1.0325,
|
| 17733 |
+
"step": 24760
|
| 17734 |
+
},
|
| 17735 |
+
{
|
| 17736 |
+
"epoch": 0.62,
|
| 17737 |
+
"grad_norm": 5.2485551834106445,
|
| 17738 |
+
"learning_rate": 3.5457627118644072e-06,
|
| 17739 |
+
"loss": 1.4088,
|
| 17740 |
+
"step": 24770
|
| 17741 |
+
},
|
| 17742 |
+
{
|
| 17743 |
+
"epoch": 0.62,
|
| 17744 |
+
"grad_norm": 6.970777988433838,
|
| 17745 |
+
"learning_rate": 3.538983050847458e-06,
|
| 17746 |
+
"loss": 1.4962,
|
| 17747 |
+
"step": 24780
|
| 17748 |
+
},
|
| 17749 |
+
{
|
| 17750 |
+
"epoch": 0.62,
|
| 17751 |
+
"grad_norm": 6.037806034088135,
|
| 17752 |
+
"learning_rate": 3.5322033898305088e-06,
|
| 17753 |
+
"loss": 1.3806,
|
| 17754 |
+
"step": 24790
|
| 17755 |
+
},
|
| 17756 |
+
{
|
| 17757 |
+
"epoch": 0.62,
|
| 17758 |
+
"grad_norm": 4.5726518630981445,
|
| 17759 |
+
"learning_rate": 3.5254237288135597e-06,
|
| 17760 |
+
"loss": 1.3086,
|
| 17761 |
+
"step": 24800
|
| 17762 |
+
},
|
| 17763 |
+
{
|
| 17764 |
+
"epoch": 0.62,
|
| 17765 |
+
"grad_norm": 4.184850692749023,
|
| 17766 |
+
"learning_rate": 3.5186440677966103e-06,
|
| 17767 |
+
"loss": 1.4004,
|
| 17768 |
+
"step": 24810
|
| 17769 |
+
},
|
| 17770 |
+
{
|
| 17771 |
+
"epoch": 0.62,
|
| 17772 |
+
"grad_norm": 9.181009292602539,
|
| 17773 |
+
"learning_rate": 3.5118644067796613e-06,
|
| 17774 |
+
"loss": 1.4348,
|
| 17775 |
+
"step": 24820
|
| 17776 |
+
},
|
| 17777 |
+
{
|
| 17778 |
+
"epoch": 0.62,
|
| 17779 |
+
"grad_norm": 5.124319076538086,
|
| 17780 |
+
"learning_rate": 3.5050847457627122e-06,
|
| 17781 |
+
"loss": 1.2272,
|
| 17782 |
+
"step": 24830
|
| 17783 |
+
},
|
| 17784 |
+
{
|
| 17785 |
+
"epoch": 0.62,
|
| 17786 |
+
"grad_norm": 7.30942440032959,
|
| 17787 |
+
"learning_rate": 3.498305084745763e-06,
|
| 17788 |
+
"loss": 1.3819,
|
| 17789 |
+
"step": 24840
|
| 17790 |
+
},
|
| 17791 |
+
{
|
| 17792 |
+
"epoch": 0.62,
|
| 17793 |
+
"grad_norm": 7.727287769317627,
|
| 17794 |
+
"learning_rate": 3.4915254237288138e-06,
|
| 17795 |
+
"loss": 1.1956,
|
| 17796 |
+
"step": 24850
|
| 17797 |
+
},
|
| 17798 |
+
{
|
| 17799 |
+
"epoch": 0.62,
|
| 17800 |
+
"grad_norm": 5.2934112548828125,
|
| 17801 |
+
"learning_rate": 3.4847457627118648e-06,
|
| 17802 |
+
"loss": 1.3141,
|
| 17803 |
+
"step": 24860
|
| 17804 |
+
},
|
| 17805 |
+
{
|
| 17806 |
+
"epoch": 0.62,
|
| 17807 |
+
"grad_norm": 7.7370500564575195,
|
| 17808 |
+
"learning_rate": 3.4779661016949157e-06,
|
| 17809 |
+
"loss": 1.2418,
|
| 17810 |
+
"step": 24870
|
| 17811 |
+
},
|
| 17812 |
+
{
|
| 17813 |
+
"epoch": 0.62,
|
| 17814 |
+
"grad_norm": 5.379338264465332,
|
| 17815 |
+
"learning_rate": 3.4711864406779667e-06,
|
| 17816 |
+
"loss": 1.435,
|
| 17817 |
+
"step": 24880
|
| 17818 |
+
},
|
| 17819 |
+
{
|
| 17820 |
+
"epoch": 0.62,
|
| 17821 |
+
"grad_norm": 5.592279434204102,
|
| 17822 |
+
"learning_rate": 3.464406779661017e-06,
|
| 17823 |
+
"loss": 1.324,
|
| 17824 |
+
"step": 24890
|
| 17825 |
+
},
|
| 17826 |
+
{
|
| 17827 |
+
"epoch": 0.62,
|
| 17828 |
+
"grad_norm": 4.178751468658447,
|
| 17829 |
+
"learning_rate": 3.457627118644068e-06,
|
| 17830 |
+
"loss": 1.4095,
|
| 17831 |
+
"step": 24900
|
| 17832 |
+
},
|
| 17833 |
+
{
|
| 17834 |
+
"epoch": 0.62,
|
| 17835 |
+
"grad_norm": 2.7852957248687744,
|
| 17836 |
+
"learning_rate": 3.450847457627119e-06,
|
| 17837 |
+
"loss": 1.1834,
|
| 17838 |
+
"step": 24910
|
| 17839 |
+
},
|
| 17840 |
+
{
|
| 17841 |
+
"epoch": 0.62,
|
| 17842 |
+
"grad_norm": 12.001542091369629,
|
| 17843 |
+
"learning_rate": 3.4440677966101698e-06,
|
| 17844 |
+
"loss": 1.3294,
|
| 17845 |
+
"step": 24920
|
| 17846 |
+
},
|
| 17847 |
+
{
|
| 17848 |
+
"epoch": 0.62,
|
| 17849 |
+
"grad_norm": 4.010140895843506,
|
| 17850 |
+
"learning_rate": 3.4372881355932207e-06,
|
| 17851 |
+
"loss": 1.4615,
|
| 17852 |
+
"step": 24930
|
| 17853 |
+
},
|
| 17854 |
+
{
|
| 17855 |
+
"epoch": 0.62,
|
| 17856 |
+
"grad_norm": 2.7130627632141113,
|
| 17857 |
+
"learning_rate": 3.4305084745762713e-06,
|
| 17858 |
+
"loss": 1.4789,
|
| 17859 |
+
"step": 24940
|
| 17860 |
+
},
|
| 17861 |
+
{
|
| 17862 |
+
"epoch": 0.62,
|
| 17863 |
+
"grad_norm": 5.209987640380859,
|
| 17864 |
+
"learning_rate": 3.4237288135593223e-06,
|
| 17865 |
+
"loss": 1.2389,
|
| 17866 |
+
"step": 24950
|
| 17867 |
+
},
|
| 17868 |
+
{
|
| 17869 |
+
"epoch": 0.62,
|
| 17870 |
+
"grad_norm": 4.1047515869140625,
|
| 17871 |
+
"learning_rate": 3.4169491525423733e-06,
|
| 17872 |
+
"loss": 1.3449,
|
| 17873 |
+
"step": 24960
|
| 17874 |
+
},
|
| 17875 |
+
{
|
| 17876 |
+
"epoch": 0.62,
|
| 17877 |
+
"grad_norm": 19.238649368286133,
|
| 17878 |
+
"learning_rate": 3.4101694915254242e-06,
|
| 17879 |
+
"loss": 1.3224,
|
| 17880 |
+
"step": 24970
|
| 17881 |
+
},
|
| 17882 |
+
{
|
| 17883 |
+
"epoch": 0.62,
|
| 17884 |
+
"grad_norm": 11.792010307312012,
|
| 17885 |
+
"learning_rate": 3.403389830508475e-06,
|
| 17886 |
+
"loss": 1.1061,
|
| 17887 |
+
"step": 24980
|
| 17888 |
+
},
|
| 17889 |
+
{
|
| 17890 |
+
"epoch": 0.62,
|
| 17891 |
+
"grad_norm": 10.934020042419434,
|
| 17892 |
+
"learning_rate": 3.3966101694915253e-06,
|
| 17893 |
+
"loss": 1.4179,
|
| 17894 |
+
"step": 24990
|
| 17895 |
+
},
|
| 17896 |
+
{
|
| 17897 |
+
"epoch": 0.62,
|
| 17898 |
+
"grad_norm": 1.5776662826538086,
|
| 17899 |
+
"learning_rate": 3.3898305084745763e-06,
|
| 17900 |
+
"loss": 1.3254,
|
| 17901 |
+
"step": 25000
|
| 17902 |
+
},
|
| 17903 |
+
{
|
| 17904 |
+
"epoch": 0.62,
|
| 17905 |
+
"eval_loss": 1.3409814834594727,
|
| 17906 |
+
"eval_runtime": 66.1173,
|
| 17907 |
+
"eval_samples_per_second": 15.125,
|
| 17908 |
+
"eval_steps_per_second": 15.125,
|
| 17909 |
+
"step": 25000
|
| 17910 |
}
|
| 17911 |
],
|
| 17912 |
"logging_steps": 10,
|
|
|
|
| 17914 |
"num_input_tokens_seen": 0,
|
| 17915 |
"num_train_epochs": 1,
|
| 17916 |
"save_steps": 2500,
|
| 17917 |
+
"total_flos": 4.025531498496e+17,
|
| 17918 |
"train_batch_size": 1,
|
| 17919 |
"trial_name": null,
|
| 17920 |
"trial_params": null
|