Commit
•
c37a6ac
1
Parent(s):
7e9a33f
Uploaded checkpoint-30000
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1763 -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:b525f711ea01af14a1a4109db33040d95091e1f5f4dccf0c07f874c01bfc2878
|
| 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:a5e41106e4c714a3b3dad5809e17e3c5394f938215aaa89f9f5521f67be46e41
|
| 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:f4922187d49de24e3ef28c2598a3510732f5f461011533915c69243f836690e1
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8b36093e06845c6146f3175c64f0e8bdb441d4f7fc67a6962ed0b80b6725daf1
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "runs/deepseek_lora_20240422-141601/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 2500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -19345,6 +19345,1764 @@
|
|
| 19345 |
"eval_samples_per_second": 8.171,
|
| 19346 |
"eval_steps_per_second": 8.171,
|
| 19347 |
"step": 27500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19348 |
}
|
| 19349 |
],
|
| 19350 |
"logging_steps": 10,
|
|
@@ -19352,7 +21110,7 @@
|
|
| 19352 |
"num_input_tokens_seen": 0,
|
| 19353 |
"num_train_epochs": 1,
|
| 19354 |
"save_steps": 2500,
|
| 19355 |
-
"total_flos": 4.
|
| 19356 |
"train_batch_size": 1,
|
| 19357 |
"trial_name": null,
|
| 19358 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.2914531230926514,
|
| 3 |
+
"best_model_checkpoint": "runs/deepseek_lora_20240422-141601/checkpoint-30000",
|
| 4 |
+
"epoch": 0.75,
|
| 5 |
"eval_steps": 2500,
|
| 6 |
+
"global_step": 30000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 19345 |
"eval_samples_per_second": 8.171,
|
| 19346 |
"eval_steps_per_second": 8.171,
|
| 19347 |
"step": 27500
|
| 19348 |
+
},
|
| 19349 |
+
{
|
| 19350 |
+
"epoch": 0.69,
|
| 19351 |
+
"grad_norm": 4.339414119720459,
|
| 19352 |
+
"learning_rate": 1.6881355932203391e-06,
|
| 19353 |
+
"loss": 1.3429,
|
| 19354 |
+
"step": 27510
|
| 19355 |
+
},
|
| 19356 |
+
{
|
| 19357 |
+
"epoch": 0.69,
|
| 19358 |
+
"grad_norm": 2.123643159866333,
|
| 19359 |
+
"learning_rate": 1.6813559322033901e-06,
|
| 19360 |
+
"loss": 1.2949,
|
| 19361 |
+
"step": 27520
|
| 19362 |
+
},
|
| 19363 |
+
{
|
| 19364 |
+
"epoch": 0.69,
|
| 19365 |
+
"grad_norm": 6.727291584014893,
|
| 19366 |
+
"learning_rate": 1.6745762711864409e-06,
|
| 19367 |
+
"loss": 1.4269,
|
| 19368 |
+
"step": 27530
|
| 19369 |
+
},
|
| 19370 |
+
{
|
| 19371 |
+
"epoch": 0.69,
|
| 19372 |
+
"grad_norm": 5.464277744293213,
|
| 19373 |
+
"learning_rate": 1.6677966101694916e-06,
|
| 19374 |
+
"loss": 1.3967,
|
| 19375 |
+
"step": 27540
|
| 19376 |
+
},
|
| 19377 |
+
{
|
| 19378 |
+
"epoch": 0.69,
|
| 19379 |
+
"grad_norm": 3.3927230834960938,
|
| 19380 |
+
"learning_rate": 1.6610169491525424e-06,
|
| 19381 |
+
"loss": 1.2701,
|
| 19382 |
+
"step": 27550
|
| 19383 |
+
},
|
| 19384 |
+
{
|
| 19385 |
+
"epoch": 0.69,
|
| 19386 |
+
"grad_norm": 6.092131614685059,
|
| 19387 |
+
"learning_rate": 1.6542372881355934e-06,
|
| 19388 |
+
"loss": 1.1989,
|
| 19389 |
+
"step": 27560
|
| 19390 |
+
},
|
| 19391 |
+
{
|
| 19392 |
+
"epoch": 0.69,
|
| 19393 |
+
"grad_norm": 4.336033821105957,
|
| 19394 |
+
"learning_rate": 1.6474576271186444e-06,
|
| 19395 |
+
"loss": 1.2679,
|
| 19396 |
+
"step": 27570
|
| 19397 |
+
},
|
| 19398 |
+
{
|
| 19399 |
+
"epoch": 0.69,
|
| 19400 |
+
"grad_norm": 3.4040894508361816,
|
| 19401 |
+
"learning_rate": 1.640677966101695e-06,
|
| 19402 |
+
"loss": 1.4791,
|
| 19403 |
+
"step": 27580
|
| 19404 |
+
},
|
| 19405 |
+
{
|
| 19406 |
+
"epoch": 0.69,
|
| 19407 |
+
"grad_norm": 5.1817626953125,
|
| 19408 |
+
"learning_rate": 1.6338983050847459e-06,
|
| 19409 |
+
"loss": 1.3513,
|
| 19410 |
+
"step": 27590
|
| 19411 |
+
},
|
| 19412 |
+
{
|
| 19413 |
+
"epoch": 0.69,
|
| 19414 |
+
"grad_norm": 14.48134708404541,
|
| 19415 |
+
"learning_rate": 1.6271186440677967e-06,
|
| 19416 |
+
"loss": 1.3058,
|
| 19417 |
+
"step": 27600
|
| 19418 |
+
},
|
| 19419 |
+
{
|
| 19420 |
+
"epoch": 0.69,
|
| 19421 |
+
"grad_norm": 6.538887977600098,
|
| 19422 |
+
"learning_rate": 1.6203389830508476e-06,
|
| 19423 |
+
"loss": 1.3431,
|
| 19424 |
+
"step": 27610
|
| 19425 |
+
},
|
| 19426 |
+
{
|
| 19427 |
+
"epoch": 0.69,
|
| 19428 |
+
"grad_norm": 7.871098518371582,
|
| 19429 |
+
"learning_rate": 1.6135593220338986e-06,
|
| 19430 |
+
"loss": 1.3213,
|
| 19431 |
+
"step": 27620
|
| 19432 |
+
},
|
| 19433 |
+
{
|
| 19434 |
+
"epoch": 0.69,
|
| 19435 |
+
"grad_norm": 6.496534824371338,
|
| 19436 |
+
"learning_rate": 1.6067796610169492e-06,
|
| 19437 |
+
"loss": 1.4389,
|
| 19438 |
+
"step": 27630
|
| 19439 |
+
},
|
| 19440 |
+
{
|
| 19441 |
+
"epoch": 0.69,
|
| 19442 |
+
"grad_norm": 6.736118793487549,
|
| 19443 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 19444 |
+
"loss": 1.3323,
|
| 19445 |
+
"step": 27640
|
| 19446 |
+
},
|
| 19447 |
+
{
|
| 19448 |
+
"epoch": 0.69,
|
| 19449 |
+
"grad_norm": 13.042952537536621,
|
| 19450 |
+
"learning_rate": 1.593220338983051e-06,
|
| 19451 |
+
"loss": 1.1813,
|
| 19452 |
+
"step": 27650
|
| 19453 |
+
},
|
| 19454 |
+
{
|
| 19455 |
+
"epoch": 0.69,
|
| 19456 |
+
"grad_norm": 2.258842706680298,
|
| 19457 |
+
"learning_rate": 1.5864406779661019e-06,
|
| 19458 |
+
"loss": 1.3224,
|
| 19459 |
+
"step": 27660
|
| 19460 |
+
},
|
| 19461 |
+
{
|
| 19462 |
+
"epoch": 0.69,
|
| 19463 |
+
"grad_norm": 6.339553356170654,
|
| 19464 |
+
"learning_rate": 1.5796610169491526e-06,
|
| 19465 |
+
"loss": 1.3433,
|
| 19466 |
+
"step": 27670
|
| 19467 |
+
},
|
| 19468 |
+
{
|
| 19469 |
+
"epoch": 0.69,
|
| 19470 |
+
"grad_norm": 7.741724491119385,
|
| 19471 |
+
"learning_rate": 1.5728813559322034e-06,
|
| 19472 |
+
"loss": 1.3168,
|
| 19473 |
+
"step": 27680
|
| 19474 |
+
},
|
| 19475 |
+
{
|
| 19476 |
+
"epoch": 0.69,
|
| 19477 |
+
"grad_norm": 16.320072174072266,
|
| 19478 |
+
"learning_rate": 1.5661016949152544e-06,
|
| 19479 |
+
"loss": 1.2347,
|
| 19480 |
+
"step": 27690
|
| 19481 |
+
},
|
| 19482 |
+
{
|
| 19483 |
+
"epoch": 0.69,
|
| 19484 |
+
"grad_norm": 5.457180023193359,
|
| 19485 |
+
"learning_rate": 1.5593220338983054e-06,
|
| 19486 |
+
"loss": 1.2835,
|
| 19487 |
+
"step": 27700
|
| 19488 |
+
},
|
| 19489 |
+
{
|
| 19490 |
+
"epoch": 0.69,
|
| 19491 |
+
"grad_norm": 5.750472068786621,
|
| 19492 |
+
"learning_rate": 1.552542372881356e-06,
|
| 19493 |
+
"loss": 1.2346,
|
| 19494 |
+
"step": 27710
|
| 19495 |
+
},
|
| 19496 |
+
{
|
| 19497 |
+
"epoch": 0.69,
|
| 19498 |
+
"grad_norm": 1.067400574684143,
|
| 19499 |
+
"learning_rate": 1.545762711864407e-06,
|
| 19500 |
+
"loss": 1.26,
|
| 19501 |
+
"step": 27720
|
| 19502 |
+
},
|
| 19503 |
+
{
|
| 19504 |
+
"epoch": 0.69,
|
| 19505 |
+
"grad_norm": 14.453791618347168,
|
| 19506 |
+
"learning_rate": 1.5389830508474577e-06,
|
| 19507 |
+
"loss": 1.1674,
|
| 19508 |
+
"step": 27730
|
| 19509 |
+
},
|
| 19510 |
+
{
|
| 19511 |
+
"epoch": 0.69,
|
| 19512 |
+
"grad_norm": 9.273608207702637,
|
| 19513 |
+
"learning_rate": 1.5322033898305086e-06,
|
| 19514 |
+
"loss": 1.2715,
|
| 19515 |
+
"step": 27740
|
| 19516 |
+
},
|
| 19517 |
+
{
|
| 19518 |
+
"epoch": 0.69,
|
| 19519 |
+
"grad_norm": 2.4199931621551514,
|
| 19520 |
+
"learning_rate": 1.5254237288135596e-06,
|
| 19521 |
+
"loss": 1.3738,
|
| 19522 |
+
"step": 27750
|
| 19523 |
+
},
|
| 19524 |
+
{
|
| 19525 |
+
"epoch": 0.69,
|
| 19526 |
+
"grad_norm": 11.302939414978027,
|
| 19527 |
+
"learning_rate": 1.5186440677966102e-06,
|
| 19528 |
+
"loss": 1.2809,
|
| 19529 |
+
"step": 27760
|
| 19530 |
+
},
|
| 19531 |
+
{
|
| 19532 |
+
"epoch": 0.69,
|
| 19533 |
+
"grad_norm": 9.895012855529785,
|
| 19534 |
+
"learning_rate": 1.5118644067796611e-06,
|
| 19535 |
+
"loss": 1.2577,
|
| 19536 |
+
"step": 27770
|
| 19537 |
+
},
|
| 19538 |
+
{
|
| 19539 |
+
"epoch": 0.69,
|
| 19540 |
+
"grad_norm": 9.7351655960083,
|
| 19541 |
+
"learning_rate": 1.505084745762712e-06,
|
| 19542 |
+
"loss": 1.2743,
|
| 19543 |
+
"step": 27780
|
| 19544 |
+
},
|
| 19545 |
+
{
|
| 19546 |
+
"epoch": 0.69,
|
| 19547 |
+
"grad_norm": 3.2647464275360107,
|
| 19548 |
+
"learning_rate": 1.4983050847457629e-06,
|
| 19549 |
+
"loss": 1.32,
|
| 19550 |
+
"step": 27790
|
| 19551 |
+
},
|
| 19552 |
+
{
|
| 19553 |
+
"epoch": 0.69,
|
| 19554 |
+
"grad_norm": 9.726398468017578,
|
| 19555 |
+
"learning_rate": 1.4915254237288139e-06,
|
| 19556 |
+
"loss": 1.3836,
|
| 19557 |
+
"step": 27800
|
| 19558 |
+
},
|
| 19559 |
+
{
|
| 19560 |
+
"epoch": 0.7,
|
| 19561 |
+
"grad_norm": 7.257278919219971,
|
| 19562 |
+
"learning_rate": 1.4847457627118644e-06,
|
| 19563 |
+
"loss": 1.4125,
|
| 19564 |
+
"step": 27810
|
| 19565 |
+
},
|
| 19566 |
+
{
|
| 19567 |
+
"epoch": 0.7,
|
| 19568 |
+
"grad_norm": 10.737898826599121,
|
| 19569 |
+
"learning_rate": 1.4779661016949154e-06,
|
| 19570 |
+
"loss": 1.3774,
|
| 19571 |
+
"step": 27820
|
| 19572 |
+
},
|
| 19573 |
+
{
|
| 19574 |
+
"epoch": 0.7,
|
| 19575 |
+
"grad_norm": 10.912455558776855,
|
| 19576 |
+
"learning_rate": 1.4711864406779664e-06,
|
| 19577 |
+
"loss": 1.302,
|
| 19578 |
+
"step": 27830
|
| 19579 |
+
},
|
| 19580 |
+
{
|
| 19581 |
+
"epoch": 0.7,
|
| 19582 |
+
"grad_norm": 4.399783611297607,
|
| 19583 |
+
"learning_rate": 1.464406779661017e-06,
|
| 19584 |
+
"loss": 1.2622,
|
| 19585 |
+
"step": 27840
|
| 19586 |
+
},
|
| 19587 |
+
{
|
| 19588 |
+
"epoch": 0.7,
|
| 19589 |
+
"grad_norm": 14.4951171875,
|
| 19590 |
+
"learning_rate": 1.457627118644068e-06,
|
| 19591 |
+
"loss": 1.415,
|
| 19592 |
+
"step": 27850
|
| 19593 |
+
},
|
| 19594 |
+
{
|
| 19595 |
+
"epoch": 0.7,
|
| 19596 |
+
"grad_norm": 5.326625347137451,
|
| 19597 |
+
"learning_rate": 1.4508474576271187e-06,
|
| 19598 |
+
"loss": 1.2882,
|
| 19599 |
+
"step": 27860
|
| 19600 |
+
},
|
| 19601 |
+
{
|
| 19602 |
+
"epoch": 0.7,
|
| 19603 |
+
"grad_norm": 6.698885917663574,
|
| 19604 |
+
"learning_rate": 1.4440677966101696e-06,
|
| 19605 |
+
"loss": 1.3461,
|
| 19606 |
+
"step": 27870
|
| 19607 |
+
},
|
| 19608 |
+
{
|
| 19609 |
+
"epoch": 0.7,
|
| 19610 |
+
"grad_norm": 2.7243850231170654,
|
| 19611 |
+
"learning_rate": 1.4372881355932206e-06,
|
| 19612 |
+
"loss": 1.4233,
|
| 19613 |
+
"step": 27880
|
| 19614 |
+
},
|
| 19615 |
+
{
|
| 19616 |
+
"epoch": 0.7,
|
| 19617 |
+
"grad_norm": 2.273730516433716,
|
| 19618 |
+
"learning_rate": 1.4305084745762712e-06,
|
| 19619 |
+
"loss": 1.2245,
|
| 19620 |
+
"step": 27890
|
| 19621 |
+
},
|
| 19622 |
+
{
|
| 19623 |
+
"epoch": 0.7,
|
| 19624 |
+
"grad_norm": 3.1223866939544678,
|
| 19625 |
+
"learning_rate": 1.4237288135593222e-06,
|
| 19626 |
+
"loss": 1.3181,
|
| 19627 |
+
"step": 27900
|
| 19628 |
+
},
|
| 19629 |
+
{
|
| 19630 |
+
"epoch": 0.7,
|
| 19631 |
+
"grad_norm": 8.419058799743652,
|
| 19632 |
+
"learning_rate": 1.416949152542373e-06,
|
| 19633 |
+
"loss": 1.2782,
|
| 19634 |
+
"step": 27910
|
| 19635 |
+
},
|
| 19636 |
+
{
|
| 19637 |
+
"epoch": 0.7,
|
| 19638 |
+
"grad_norm": 4.810690879821777,
|
| 19639 |
+
"learning_rate": 1.410169491525424e-06,
|
| 19640 |
+
"loss": 1.2527,
|
| 19641 |
+
"step": 27920
|
| 19642 |
+
},
|
| 19643 |
+
{
|
| 19644 |
+
"epoch": 0.7,
|
| 19645 |
+
"grad_norm": 2.7679851055145264,
|
| 19646 |
+
"learning_rate": 1.4033898305084749e-06,
|
| 19647 |
+
"loss": 1.1995,
|
| 19648 |
+
"step": 27930
|
| 19649 |
+
},
|
| 19650 |
+
{
|
| 19651 |
+
"epoch": 0.7,
|
| 19652 |
+
"grad_norm": 1.7833703756332397,
|
| 19653 |
+
"learning_rate": 1.3966101694915254e-06,
|
| 19654 |
+
"loss": 1.3988,
|
| 19655 |
+
"step": 27940
|
| 19656 |
+
},
|
| 19657 |
+
{
|
| 19658 |
+
"epoch": 0.7,
|
| 19659 |
+
"grad_norm": 8.348339080810547,
|
| 19660 |
+
"learning_rate": 1.3898305084745764e-06,
|
| 19661 |
+
"loss": 1.2983,
|
| 19662 |
+
"step": 27950
|
| 19663 |
+
},
|
| 19664 |
+
{
|
| 19665 |
+
"epoch": 0.7,
|
| 19666 |
+
"grad_norm": 8.304106712341309,
|
| 19667 |
+
"learning_rate": 1.3830508474576274e-06,
|
| 19668 |
+
"loss": 1.2926,
|
| 19669 |
+
"step": 27960
|
| 19670 |
+
},
|
| 19671 |
+
{
|
| 19672 |
+
"epoch": 0.7,
|
| 19673 |
+
"grad_norm": 5.610420227050781,
|
| 19674 |
+
"learning_rate": 1.376271186440678e-06,
|
| 19675 |
+
"loss": 1.2064,
|
| 19676 |
+
"step": 27970
|
| 19677 |
+
},
|
| 19678 |
+
{
|
| 19679 |
+
"epoch": 0.7,
|
| 19680 |
+
"grad_norm": 13.0822114944458,
|
| 19681 |
+
"learning_rate": 1.369491525423729e-06,
|
| 19682 |
+
"loss": 1.3171,
|
| 19683 |
+
"step": 27980
|
| 19684 |
+
},
|
| 19685 |
+
{
|
| 19686 |
+
"epoch": 0.7,
|
| 19687 |
+
"grad_norm": 3.2502501010894775,
|
| 19688 |
+
"learning_rate": 1.3627118644067797e-06,
|
| 19689 |
+
"loss": 1.2354,
|
| 19690 |
+
"step": 27990
|
| 19691 |
+
},
|
| 19692 |
+
{
|
| 19693 |
+
"epoch": 0.7,
|
| 19694 |
+
"grad_norm": 19.23220443725586,
|
| 19695 |
+
"learning_rate": 1.3559322033898307e-06,
|
| 19696 |
+
"loss": 1.3304,
|
| 19697 |
+
"step": 28000
|
| 19698 |
+
},
|
| 19699 |
+
{
|
| 19700 |
+
"epoch": 0.7,
|
| 19701 |
+
"grad_norm": 5.642111778259277,
|
| 19702 |
+
"learning_rate": 1.3491525423728816e-06,
|
| 19703 |
+
"loss": 1.2085,
|
| 19704 |
+
"step": 28010
|
| 19705 |
+
},
|
| 19706 |
+
{
|
| 19707 |
+
"epoch": 0.7,
|
| 19708 |
+
"grad_norm": 5.464583396911621,
|
| 19709 |
+
"learning_rate": 1.3423728813559322e-06,
|
| 19710 |
+
"loss": 1.3588,
|
| 19711 |
+
"step": 28020
|
| 19712 |
+
},
|
| 19713 |
+
{
|
| 19714 |
+
"epoch": 0.7,
|
| 19715 |
+
"grad_norm": 2.51157808303833,
|
| 19716 |
+
"learning_rate": 1.3355932203389832e-06,
|
| 19717 |
+
"loss": 1.1803,
|
| 19718 |
+
"step": 28030
|
| 19719 |
+
},
|
| 19720 |
+
{
|
| 19721 |
+
"epoch": 0.7,
|
| 19722 |
+
"grad_norm": 10.629277229309082,
|
| 19723 |
+
"learning_rate": 1.328813559322034e-06,
|
| 19724 |
+
"loss": 1.3943,
|
| 19725 |
+
"step": 28040
|
| 19726 |
+
},
|
| 19727 |
+
{
|
| 19728 |
+
"epoch": 0.7,
|
| 19729 |
+
"grad_norm": 10.221532821655273,
|
| 19730 |
+
"learning_rate": 1.322033898305085e-06,
|
| 19731 |
+
"loss": 1.4424,
|
| 19732 |
+
"step": 28050
|
| 19733 |
+
},
|
| 19734 |
+
{
|
| 19735 |
+
"epoch": 0.7,
|
| 19736 |
+
"grad_norm": 1.1695493459701538,
|
| 19737 |
+
"learning_rate": 1.3152542372881359e-06,
|
| 19738 |
+
"loss": 1.3071,
|
| 19739 |
+
"step": 28060
|
| 19740 |
+
},
|
| 19741 |
+
{
|
| 19742 |
+
"epoch": 0.7,
|
| 19743 |
+
"grad_norm": 4.1398539543151855,
|
| 19744 |
+
"learning_rate": 1.3084745762711864e-06,
|
| 19745 |
+
"loss": 1.2239,
|
| 19746 |
+
"step": 28070
|
| 19747 |
+
},
|
| 19748 |
+
{
|
| 19749 |
+
"epoch": 0.7,
|
| 19750 |
+
"grad_norm": 20.721601486206055,
|
| 19751 |
+
"learning_rate": 1.3016949152542374e-06,
|
| 19752 |
+
"loss": 1.3761,
|
| 19753 |
+
"step": 28080
|
| 19754 |
+
},
|
| 19755 |
+
{
|
| 19756 |
+
"epoch": 0.7,
|
| 19757 |
+
"grad_norm": 8.006892204284668,
|
| 19758 |
+
"learning_rate": 1.2949152542372884e-06,
|
| 19759 |
+
"loss": 1.2867,
|
| 19760 |
+
"step": 28090
|
| 19761 |
+
},
|
| 19762 |
+
{
|
| 19763 |
+
"epoch": 0.7,
|
| 19764 |
+
"grad_norm": 4.007428169250488,
|
| 19765 |
+
"learning_rate": 1.288135593220339e-06,
|
| 19766 |
+
"loss": 1.2511,
|
| 19767 |
+
"step": 28100
|
| 19768 |
+
},
|
| 19769 |
+
{
|
| 19770 |
+
"epoch": 0.7,
|
| 19771 |
+
"grad_norm": 8.041426658630371,
|
| 19772 |
+
"learning_rate": 1.28135593220339e-06,
|
| 19773 |
+
"loss": 1.3597,
|
| 19774 |
+
"step": 28110
|
| 19775 |
+
},
|
| 19776 |
+
{
|
| 19777 |
+
"epoch": 0.7,
|
| 19778 |
+
"grad_norm": 4.553910732269287,
|
| 19779 |
+
"learning_rate": 1.2745762711864407e-06,
|
| 19780 |
+
"loss": 1.4066,
|
| 19781 |
+
"step": 28120
|
| 19782 |
+
},
|
| 19783 |
+
{
|
| 19784 |
+
"epoch": 0.7,
|
| 19785 |
+
"grad_norm": 8.02763843536377,
|
| 19786 |
+
"learning_rate": 1.2677966101694917e-06,
|
| 19787 |
+
"loss": 1.1639,
|
| 19788 |
+
"step": 28130
|
| 19789 |
+
},
|
| 19790 |
+
{
|
| 19791 |
+
"epoch": 0.7,
|
| 19792 |
+
"grad_norm": 9.010587692260742,
|
| 19793 |
+
"learning_rate": 1.2610169491525426e-06,
|
| 19794 |
+
"loss": 1.5038,
|
| 19795 |
+
"step": 28140
|
| 19796 |
+
},
|
| 19797 |
+
{
|
| 19798 |
+
"epoch": 0.7,
|
| 19799 |
+
"grad_norm": 4.593667984008789,
|
| 19800 |
+
"learning_rate": 1.2542372881355932e-06,
|
| 19801 |
+
"loss": 1.4871,
|
| 19802 |
+
"step": 28150
|
| 19803 |
+
},
|
| 19804 |
+
{
|
| 19805 |
+
"epoch": 0.7,
|
| 19806 |
+
"grad_norm": 6.7055158615112305,
|
| 19807 |
+
"learning_rate": 1.2474576271186442e-06,
|
| 19808 |
+
"loss": 1.2533,
|
| 19809 |
+
"step": 28160
|
| 19810 |
+
},
|
| 19811 |
+
{
|
| 19812 |
+
"epoch": 0.7,
|
| 19813 |
+
"grad_norm": 13.491644859313965,
|
| 19814 |
+
"learning_rate": 1.240677966101695e-06,
|
| 19815 |
+
"loss": 1.2957,
|
| 19816 |
+
"step": 28170
|
| 19817 |
+
},
|
| 19818 |
+
{
|
| 19819 |
+
"epoch": 0.7,
|
| 19820 |
+
"grad_norm": 5.513734340667725,
|
| 19821 |
+
"learning_rate": 1.233898305084746e-06,
|
| 19822 |
+
"loss": 1.2898,
|
| 19823 |
+
"step": 28180
|
| 19824 |
+
},
|
| 19825 |
+
{
|
| 19826 |
+
"epoch": 0.7,
|
| 19827 |
+
"grad_norm": 11.136821746826172,
|
| 19828 |
+
"learning_rate": 1.2271186440677967e-06,
|
| 19829 |
+
"loss": 1.4257,
|
| 19830 |
+
"step": 28190
|
| 19831 |
+
},
|
| 19832 |
+
{
|
| 19833 |
+
"epoch": 0.7,
|
| 19834 |
+
"grad_norm": 4.229417324066162,
|
| 19835 |
+
"learning_rate": 1.2203389830508477e-06,
|
| 19836 |
+
"loss": 1.322,
|
| 19837 |
+
"step": 28200
|
| 19838 |
+
},
|
| 19839 |
+
{
|
| 19840 |
+
"epoch": 0.71,
|
| 19841 |
+
"grad_norm": 4.852339267730713,
|
| 19842 |
+
"learning_rate": 1.2135593220338984e-06,
|
| 19843 |
+
"loss": 1.3371,
|
| 19844 |
+
"step": 28210
|
| 19845 |
+
},
|
| 19846 |
+
{
|
| 19847 |
+
"epoch": 0.71,
|
| 19848 |
+
"grad_norm": 9.924147605895996,
|
| 19849 |
+
"learning_rate": 1.2067796610169492e-06,
|
| 19850 |
+
"loss": 1.1373,
|
| 19851 |
+
"step": 28220
|
| 19852 |
+
},
|
| 19853 |
+
{
|
| 19854 |
+
"epoch": 0.71,
|
| 19855 |
+
"grad_norm": 3.660102128982544,
|
| 19856 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 19857 |
+
"loss": 1.253,
|
| 19858 |
+
"step": 28230
|
| 19859 |
+
},
|
| 19860 |
+
{
|
| 19861 |
+
"epoch": 0.71,
|
| 19862 |
+
"grad_norm": 5.641490936279297,
|
| 19863 |
+
"learning_rate": 1.193220338983051e-06,
|
| 19864 |
+
"loss": 1.3156,
|
| 19865 |
+
"step": 28240
|
| 19866 |
+
},
|
| 19867 |
+
{
|
| 19868 |
+
"epoch": 0.71,
|
| 19869 |
+
"grad_norm": 7.371779441833496,
|
| 19870 |
+
"learning_rate": 1.186440677966102e-06,
|
| 19871 |
+
"loss": 1.3273,
|
| 19872 |
+
"step": 28250
|
| 19873 |
+
},
|
| 19874 |
+
{
|
| 19875 |
+
"epoch": 0.71,
|
| 19876 |
+
"grad_norm": 11.060622215270996,
|
| 19877 |
+
"learning_rate": 1.1796610169491527e-06,
|
| 19878 |
+
"loss": 1.2729,
|
| 19879 |
+
"step": 28260
|
| 19880 |
+
},
|
| 19881 |
+
{
|
| 19882 |
+
"epoch": 0.71,
|
| 19883 |
+
"grad_norm": 2.1023409366607666,
|
| 19884 |
+
"learning_rate": 1.1728813559322034e-06,
|
| 19885 |
+
"loss": 1.4814,
|
| 19886 |
+
"step": 28270
|
| 19887 |
+
},
|
| 19888 |
+
{
|
| 19889 |
+
"epoch": 0.71,
|
| 19890 |
+
"grad_norm": 2.1587941646575928,
|
| 19891 |
+
"learning_rate": 1.1661016949152542e-06,
|
| 19892 |
+
"loss": 1.4983,
|
| 19893 |
+
"step": 28280
|
| 19894 |
+
},
|
| 19895 |
+
{
|
| 19896 |
+
"epoch": 0.71,
|
| 19897 |
+
"grad_norm": 10.272698402404785,
|
| 19898 |
+
"learning_rate": 1.1593220338983052e-06,
|
| 19899 |
+
"loss": 1.2093,
|
| 19900 |
+
"step": 28290
|
| 19901 |
+
},
|
| 19902 |
+
{
|
| 19903 |
+
"epoch": 0.71,
|
| 19904 |
+
"grad_norm": 3.677959442138672,
|
| 19905 |
+
"learning_rate": 1.152542372881356e-06,
|
| 19906 |
+
"loss": 1.1745,
|
| 19907 |
+
"step": 28300
|
| 19908 |
+
},
|
| 19909 |
+
{
|
| 19910 |
+
"epoch": 0.71,
|
| 19911 |
+
"grad_norm": 2.187302350997925,
|
| 19912 |
+
"learning_rate": 1.145762711864407e-06,
|
| 19913 |
+
"loss": 1.3257,
|
| 19914 |
+
"step": 28310
|
| 19915 |
+
},
|
| 19916 |
+
{
|
| 19917 |
+
"epoch": 0.71,
|
| 19918 |
+
"grad_norm": 5.065845489501953,
|
| 19919 |
+
"learning_rate": 1.1389830508474577e-06,
|
| 19920 |
+
"loss": 1.3038,
|
| 19921 |
+
"step": 28320
|
| 19922 |
+
},
|
| 19923 |
+
{
|
| 19924 |
+
"epoch": 0.71,
|
| 19925 |
+
"grad_norm": 6.725832462310791,
|
| 19926 |
+
"learning_rate": 1.1322033898305087e-06,
|
| 19927 |
+
"loss": 1.2525,
|
| 19928 |
+
"step": 28330
|
| 19929 |
+
},
|
| 19930 |
+
{
|
| 19931 |
+
"epoch": 0.71,
|
| 19932 |
+
"grad_norm": 13.034323692321777,
|
| 19933 |
+
"learning_rate": 1.1254237288135594e-06,
|
| 19934 |
+
"loss": 1.3096,
|
| 19935 |
+
"step": 28340
|
| 19936 |
+
},
|
| 19937 |
+
{
|
| 19938 |
+
"epoch": 0.71,
|
| 19939 |
+
"grad_norm": 12.125313758850098,
|
| 19940 |
+
"learning_rate": 1.1186440677966102e-06,
|
| 19941 |
+
"loss": 1.1823,
|
| 19942 |
+
"step": 28350
|
| 19943 |
+
},
|
| 19944 |
+
{
|
| 19945 |
+
"epoch": 0.71,
|
| 19946 |
+
"grad_norm": 12.07535457611084,
|
| 19947 |
+
"learning_rate": 1.1118644067796612e-06,
|
| 19948 |
+
"loss": 1.2526,
|
| 19949 |
+
"step": 28360
|
| 19950 |
+
},
|
| 19951 |
+
{
|
| 19952 |
+
"epoch": 0.71,
|
| 19953 |
+
"grad_norm": 6.179876804351807,
|
| 19954 |
+
"learning_rate": 1.105084745762712e-06,
|
| 19955 |
+
"loss": 1.3062,
|
| 19956 |
+
"step": 28370
|
| 19957 |
+
},
|
| 19958 |
+
{
|
| 19959 |
+
"epoch": 0.71,
|
| 19960 |
+
"grad_norm": 2.814276933670044,
|
| 19961 |
+
"learning_rate": 1.098305084745763e-06,
|
| 19962 |
+
"loss": 1.2911,
|
| 19963 |
+
"step": 28380
|
| 19964 |
+
},
|
| 19965 |
+
{
|
| 19966 |
+
"epoch": 0.71,
|
| 19967 |
+
"grad_norm": 14.710918426513672,
|
| 19968 |
+
"learning_rate": 1.0915254237288137e-06,
|
| 19969 |
+
"loss": 1.1594,
|
| 19970 |
+
"step": 28390
|
| 19971 |
+
},
|
| 19972 |
+
{
|
| 19973 |
+
"epoch": 0.71,
|
| 19974 |
+
"grad_norm": 12.056600570678711,
|
| 19975 |
+
"learning_rate": 1.0847457627118644e-06,
|
| 19976 |
+
"loss": 1.2203,
|
| 19977 |
+
"step": 28400
|
| 19978 |
+
},
|
| 19979 |
+
{
|
| 19980 |
+
"epoch": 0.71,
|
| 19981 |
+
"grad_norm": 9.935985565185547,
|
| 19982 |
+
"learning_rate": 1.0779661016949152e-06,
|
| 19983 |
+
"loss": 1.4101,
|
| 19984 |
+
"step": 28410
|
| 19985 |
+
},
|
| 19986 |
+
{
|
| 19987 |
+
"epoch": 0.71,
|
| 19988 |
+
"grad_norm": 15.330471992492676,
|
| 19989 |
+
"learning_rate": 1.0711864406779662e-06,
|
| 19990 |
+
"loss": 1.1429,
|
| 19991 |
+
"step": 28420
|
| 19992 |
+
},
|
| 19993 |
+
{
|
| 19994 |
+
"epoch": 0.71,
|
| 19995 |
+
"grad_norm": 2.6031837463378906,
|
| 19996 |
+
"learning_rate": 1.064406779661017e-06,
|
| 19997 |
+
"loss": 1.336,
|
| 19998 |
+
"step": 28430
|
| 19999 |
+
},
|
| 20000 |
+
{
|
| 20001 |
+
"epoch": 0.71,
|
| 20002 |
+
"grad_norm": 10.921528816223145,
|
| 20003 |
+
"learning_rate": 1.057627118644068e-06,
|
| 20004 |
+
"loss": 1.3067,
|
| 20005 |
+
"step": 28440
|
| 20006 |
+
},
|
| 20007 |
+
{
|
| 20008 |
+
"epoch": 0.71,
|
| 20009 |
+
"grad_norm": 10.396923065185547,
|
| 20010 |
+
"learning_rate": 1.0508474576271187e-06,
|
| 20011 |
+
"loss": 1.3811,
|
| 20012 |
+
"step": 28450
|
| 20013 |
+
},
|
| 20014 |
+
{
|
| 20015 |
+
"epoch": 0.71,
|
| 20016 |
+
"grad_norm": 10.552454948425293,
|
| 20017 |
+
"learning_rate": 1.0440677966101697e-06,
|
| 20018 |
+
"loss": 1.3656,
|
| 20019 |
+
"step": 28460
|
| 20020 |
+
},
|
| 20021 |
+
{
|
| 20022 |
+
"epoch": 0.71,
|
| 20023 |
+
"grad_norm": 11.751737594604492,
|
| 20024 |
+
"learning_rate": 1.0372881355932204e-06,
|
| 20025 |
+
"loss": 1.2437,
|
| 20026 |
+
"step": 28470
|
| 20027 |
+
},
|
| 20028 |
+
{
|
| 20029 |
+
"epoch": 0.71,
|
| 20030 |
+
"grad_norm": 3.7091989517211914,
|
| 20031 |
+
"learning_rate": 1.0305084745762712e-06,
|
| 20032 |
+
"loss": 1.3765,
|
| 20033 |
+
"step": 28480
|
| 20034 |
+
},
|
| 20035 |
+
{
|
| 20036 |
+
"epoch": 0.71,
|
| 20037 |
+
"grad_norm": 5.946925163269043,
|
| 20038 |
+
"learning_rate": 1.0237288135593222e-06,
|
| 20039 |
+
"loss": 1.4501,
|
| 20040 |
+
"step": 28490
|
| 20041 |
+
},
|
| 20042 |
+
{
|
| 20043 |
+
"epoch": 0.71,
|
| 20044 |
+
"grad_norm": 10.043877601623535,
|
| 20045 |
+
"learning_rate": 1.016949152542373e-06,
|
| 20046 |
+
"loss": 1.3166,
|
| 20047 |
+
"step": 28500
|
| 20048 |
+
},
|
| 20049 |
+
{
|
| 20050 |
+
"epoch": 0.71,
|
| 20051 |
+
"grad_norm": 12.921855926513672,
|
| 20052 |
+
"learning_rate": 1.010169491525424e-06,
|
| 20053 |
+
"loss": 1.2181,
|
| 20054 |
+
"step": 28510
|
| 20055 |
+
},
|
| 20056 |
+
{
|
| 20057 |
+
"epoch": 0.71,
|
| 20058 |
+
"grad_norm": 8.31167984008789,
|
| 20059 |
+
"learning_rate": 1.0033898305084747e-06,
|
| 20060 |
+
"loss": 1.1729,
|
| 20061 |
+
"step": 28520
|
| 20062 |
+
},
|
| 20063 |
+
{
|
| 20064 |
+
"epoch": 0.71,
|
| 20065 |
+
"grad_norm": 13.391186714172363,
|
| 20066 |
+
"learning_rate": 9.966101694915254e-07,
|
| 20067 |
+
"loss": 1.4122,
|
| 20068 |
+
"step": 28530
|
| 20069 |
+
},
|
| 20070 |
+
{
|
| 20071 |
+
"epoch": 0.71,
|
| 20072 |
+
"grad_norm": 3.428011178970337,
|
| 20073 |
+
"learning_rate": 9.898305084745762e-07,
|
| 20074 |
+
"loss": 1.4873,
|
| 20075 |
+
"step": 28540
|
| 20076 |
+
},
|
| 20077 |
+
{
|
| 20078 |
+
"epoch": 0.71,
|
| 20079 |
+
"grad_norm": 5.871237754821777,
|
| 20080 |
+
"learning_rate": 9.830508474576272e-07,
|
| 20081 |
+
"loss": 1.5205,
|
| 20082 |
+
"step": 28550
|
| 20083 |
+
},
|
| 20084 |
+
{
|
| 20085 |
+
"epoch": 0.71,
|
| 20086 |
+
"grad_norm": 11.823165893554688,
|
| 20087 |
+
"learning_rate": 9.762711864406782e-07,
|
| 20088 |
+
"loss": 1.1785,
|
| 20089 |
+
"step": 28560
|
| 20090 |
+
},
|
| 20091 |
+
{
|
| 20092 |
+
"epoch": 0.71,
|
| 20093 |
+
"grad_norm": 13.24584674835205,
|
| 20094 |
+
"learning_rate": 9.69491525423729e-07,
|
| 20095 |
+
"loss": 1.3375,
|
| 20096 |
+
"step": 28570
|
| 20097 |
+
},
|
| 20098 |
+
{
|
| 20099 |
+
"epoch": 0.71,
|
| 20100 |
+
"grad_norm": 4.6865386962890625,
|
| 20101 |
+
"learning_rate": 9.627118644067797e-07,
|
| 20102 |
+
"loss": 1.3289,
|
| 20103 |
+
"step": 28580
|
| 20104 |
+
},
|
| 20105 |
+
{
|
| 20106 |
+
"epoch": 0.71,
|
| 20107 |
+
"grad_norm": 4.408345699310303,
|
| 20108 |
+
"learning_rate": 9.559322033898307e-07,
|
| 20109 |
+
"loss": 1.3748,
|
| 20110 |
+
"step": 28590
|
| 20111 |
+
},
|
| 20112 |
+
{
|
| 20113 |
+
"epoch": 0.71,
|
| 20114 |
+
"grad_norm": 6.063961982727051,
|
| 20115 |
+
"learning_rate": 9.491525423728814e-07,
|
| 20116 |
+
"loss": 1.2393,
|
| 20117 |
+
"step": 28600
|
| 20118 |
+
},
|
| 20119 |
+
{
|
| 20120 |
+
"epoch": 0.72,
|
| 20121 |
+
"grad_norm": 12.433189392089844,
|
| 20122 |
+
"learning_rate": 9.423728813559323e-07,
|
| 20123 |
+
"loss": 1.295,
|
| 20124 |
+
"step": 28610
|
| 20125 |
+
},
|
| 20126 |
+
{
|
| 20127 |
+
"epoch": 0.72,
|
| 20128 |
+
"grad_norm": 4.160829067230225,
|
| 20129 |
+
"learning_rate": 9.355932203389831e-07,
|
| 20130 |
+
"loss": 1.1346,
|
| 20131 |
+
"step": 28620
|
| 20132 |
+
},
|
| 20133 |
+
{
|
| 20134 |
+
"epoch": 0.72,
|
| 20135 |
+
"grad_norm": 4.698163032531738,
|
| 20136 |
+
"learning_rate": 9.28813559322034e-07,
|
| 20137 |
+
"loss": 1.2795,
|
| 20138 |
+
"step": 28630
|
| 20139 |
+
},
|
| 20140 |
+
{
|
| 20141 |
+
"epoch": 0.72,
|
| 20142 |
+
"grad_norm": 9.586724281311035,
|
| 20143 |
+
"learning_rate": 9.220338983050848e-07,
|
| 20144 |
+
"loss": 1.3546,
|
| 20145 |
+
"step": 28640
|
| 20146 |
+
},
|
| 20147 |
+
{
|
| 20148 |
+
"epoch": 0.72,
|
| 20149 |
+
"grad_norm": 5.39924955368042,
|
| 20150 |
+
"learning_rate": 9.152542372881357e-07,
|
| 20151 |
+
"loss": 1.3835,
|
| 20152 |
+
"step": 28650
|
| 20153 |
+
},
|
| 20154 |
+
{
|
| 20155 |
+
"epoch": 0.72,
|
| 20156 |
+
"grad_norm": 9.498052597045898,
|
| 20157 |
+
"learning_rate": 9.084745762711864e-07,
|
| 20158 |
+
"loss": 1.3516,
|
| 20159 |
+
"step": 28660
|
| 20160 |
+
},
|
| 20161 |
+
{
|
| 20162 |
+
"epoch": 0.72,
|
| 20163 |
+
"grad_norm": 7.17230749130249,
|
| 20164 |
+
"learning_rate": 9.016949152542373e-07,
|
| 20165 |
+
"loss": 1.2309,
|
| 20166 |
+
"step": 28670
|
| 20167 |
+
},
|
| 20168 |
+
{
|
| 20169 |
+
"epoch": 0.72,
|
| 20170 |
+
"grad_norm": 5.45039176940918,
|
| 20171 |
+
"learning_rate": 8.949152542372883e-07,
|
| 20172 |
+
"loss": 1.2632,
|
| 20173 |
+
"step": 28680
|
| 20174 |
+
},
|
| 20175 |
+
{
|
| 20176 |
+
"epoch": 0.72,
|
| 20177 |
+
"grad_norm": 5.3587260246276855,
|
| 20178 |
+
"learning_rate": 8.881355932203391e-07,
|
| 20179 |
+
"loss": 1.4843,
|
| 20180 |
+
"step": 28690
|
| 20181 |
+
},
|
| 20182 |
+
{
|
| 20183 |
+
"epoch": 0.72,
|
| 20184 |
+
"grad_norm": 1.8675847053527832,
|
| 20185 |
+
"learning_rate": 8.813559322033899e-07,
|
| 20186 |
+
"loss": 1.4973,
|
| 20187 |
+
"step": 28700
|
| 20188 |
+
},
|
| 20189 |
+
{
|
| 20190 |
+
"epoch": 0.72,
|
| 20191 |
+
"grad_norm": 3.713160753250122,
|
| 20192 |
+
"learning_rate": 8.745762711864407e-07,
|
| 20193 |
+
"loss": 1.309,
|
| 20194 |
+
"step": 28710
|
| 20195 |
+
},
|
| 20196 |
+
{
|
| 20197 |
+
"epoch": 0.72,
|
| 20198 |
+
"grad_norm": 10.668930053710938,
|
| 20199 |
+
"learning_rate": 8.677966101694917e-07,
|
| 20200 |
+
"loss": 1.3676,
|
| 20201 |
+
"step": 28720
|
| 20202 |
+
},
|
| 20203 |
+
{
|
| 20204 |
+
"epoch": 0.72,
|
| 20205 |
+
"grad_norm": 2.1564583778381348,
|
| 20206 |
+
"learning_rate": 8.610169491525424e-07,
|
| 20207 |
+
"loss": 1.2518,
|
| 20208 |
+
"step": 28730
|
| 20209 |
+
},
|
| 20210 |
+
{
|
| 20211 |
+
"epoch": 0.72,
|
| 20212 |
+
"grad_norm": 4.37905216217041,
|
| 20213 |
+
"learning_rate": 8.542372881355933e-07,
|
| 20214 |
+
"loss": 1.3369,
|
| 20215 |
+
"step": 28740
|
| 20216 |
+
},
|
| 20217 |
+
{
|
| 20218 |
+
"epoch": 0.72,
|
| 20219 |
+
"grad_norm": 15.10600471496582,
|
| 20220 |
+
"learning_rate": 8.474576271186441e-07,
|
| 20221 |
+
"loss": 1.1867,
|
| 20222 |
+
"step": 28750
|
| 20223 |
+
},
|
| 20224 |
+
{
|
| 20225 |
+
"epoch": 0.72,
|
| 20226 |
+
"grad_norm": 6.618597030639648,
|
| 20227 |
+
"learning_rate": 8.406779661016951e-07,
|
| 20228 |
+
"loss": 1.2391,
|
| 20229 |
+
"step": 28760
|
| 20230 |
+
},
|
| 20231 |
+
{
|
| 20232 |
+
"epoch": 0.72,
|
| 20233 |
+
"grad_norm": 16.292184829711914,
|
| 20234 |
+
"learning_rate": 8.338983050847458e-07,
|
| 20235 |
+
"loss": 1.4165,
|
| 20236 |
+
"step": 28770
|
| 20237 |
+
},
|
| 20238 |
+
{
|
| 20239 |
+
"epoch": 0.72,
|
| 20240 |
+
"grad_norm": 8.334046363830566,
|
| 20241 |
+
"learning_rate": 8.271186440677967e-07,
|
| 20242 |
+
"loss": 1.344,
|
| 20243 |
+
"step": 28780
|
| 20244 |
+
},
|
| 20245 |
+
{
|
| 20246 |
+
"epoch": 0.72,
|
| 20247 |
+
"grad_norm": 10.543550491333008,
|
| 20248 |
+
"learning_rate": 8.203389830508475e-07,
|
| 20249 |
+
"loss": 1.3328,
|
| 20250 |
+
"step": 28790
|
| 20251 |
+
},
|
| 20252 |
+
{
|
| 20253 |
+
"epoch": 0.72,
|
| 20254 |
+
"grad_norm": 2.6791608333587646,
|
| 20255 |
+
"learning_rate": 8.135593220338983e-07,
|
| 20256 |
+
"loss": 1.4221,
|
| 20257 |
+
"step": 28800
|
| 20258 |
+
},
|
| 20259 |
+
{
|
| 20260 |
+
"epoch": 0.72,
|
| 20261 |
+
"grad_norm": 3.6508536338806152,
|
| 20262 |
+
"learning_rate": 8.067796610169493e-07,
|
| 20263 |
+
"loss": 1.3269,
|
| 20264 |
+
"step": 28810
|
| 20265 |
+
},
|
| 20266 |
+
{
|
| 20267 |
+
"epoch": 0.72,
|
| 20268 |
+
"grad_norm": 7.089112758636475,
|
| 20269 |
+
"learning_rate": 8.000000000000001e-07,
|
| 20270 |
+
"loss": 1.3038,
|
| 20271 |
+
"step": 28820
|
| 20272 |
+
},
|
| 20273 |
+
{
|
| 20274 |
+
"epoch": 0.72,
|
| 20275 |
+
"grad_norm": 6.554429531097412,
|
| 20276 |
+
"learning_rate": 7.932203389830509e-07,
|
| 20277 |
+
"loss": 1.2598,
|
| 20278 |
+
"step": 28830
|
| 20279 |
+
},
|
| 20280 |
+
{
|
| 20281 |
+
"epoch": 0.72,
|
| 20282 |
+
"grad_norm": 5.9939799308776855,
|
| 20283 |
+
"learning_rate": 7.864406779661017e-07,
|
| 20284 |
+
"loss": 1.419,
|
| 20285 |
+
"step": 28840
|
| 20286 |
+
},
|
| 20287 |
+
{
|
| 20288 |
+
"epoch": 0.72,
|
| 20289 |
+
"grad_norm": 10.088254928588867,
|
| 20290 |
+
"learning_rate": 7.796610169491527e-07,
|
| 20291 |
+
"loss": 1.2858,
|
| 20292 |
+
"step": 28850
|
| 20293 |
+
},
|
| 20294 |
+
{
|
| 20295 |
+
"epoch": 0.72,
|
| 20296 |
+
"grad_norm": 16.476806640625,
|
| 20297 |
+
"learning_rate": 7.728813559322034e-07,
|
| 20298 |
+
"loss": 1.3498,
|
| 20299 |
+
"step": 28860
|
| 20300 |
+
},
|
| 20301 |
+
{
|
| 20302 |
+
"epoch": 0.72,
|
| 20303 |
+
"grad_norm": 11.453723907470703,
|
| 20304 |
+
"learning_rate": 7.661016949152543e-07,
|
| 20305 |
+
"loss": 1.2346,
|
| 20306 |
+
"step": 28870
|
| 20307 |
+
},
|
| 20308 |
+
{
|
| 20309 |
+
"epoch": 0.72,
|
| 20310 |
+
"grad_norm": 4.832727909088135,
|
| 20311 |
+
"learning_rate": 7.593220338983051e-07,
|
| 20312 |
+
"loss": 1.252,
|
| 20313 |
+
"step": 28880
|
| 20314 |
+
},
|
| 20315 |
+
{
|
| 20316 |
+
"epoch": 0.72,
|
| 20317 |
+
"grad_norm": 8.281596183776855,
|
| 20318 |
+
"learning_rate": 7.52542372881356e-07,
|
| 20319 |
+
"loss": 1.3162,
|
| 20320 |
+
"step": 28890
|
| 20321 |
+
},
|
| 20322 |
+
{
|
| 20323 |
+
"epoch": 0.72,
|
| 20324 |
+
"grad_norm": 8.322393417358398,
|
| 20325 |
+
"learning_rate": 7.457627118644069e-07,
|
| 20326 |
+
"loss": 1.1593,
|
| 20327 |
+
"step": 28900
|
| 20328 |
+
},
|
| 20329 |
+
{
|
| 20330 |
+
"epoch": 0.72,
|
| 20331 |
+
"grad_norm": 10.499345779418945,
|
| 20332 |
+
"learning_rate": 7.389830508474577e-07,
|
| 20333 |
+
"loss": 1.243,
|
| 20334 |
+
"step": 28910
|
| 20335 |
+
},
|
| 20336 |
+
{
|
| 20337 |
+
"epoch": 0.72,
|
| 20338 |
+
"grad_norm": 14.577847480773926,
|
| 20339 |
+
"learning_rate": 7.322033898305085e-07,
|
| 20340 |
+
"loss": 1.3396,
|
| 20341 |
+
"step": 28920
|
| 20342 |
+
},
|
| 20343 |
+
{
|
| 20344 |
+
"epoch": 0.72,
|
| 20345 |
+
"grad_norm": 3.6276347637176514,
|
| 20346 |
+
"learning_rate": 7.254237288135593e-07,
|
| 20347 |
+
"loss": 1.4325,
|
| 20348 |
+
"step": 28930
|
| 20349 |
+
},
|
| 20350 |
+
{
|
| 20351 |
+
"epoch": 0.72,
|
| 20352 |
+
"grad_norm": 10.75643539428711,
|
| 20353 |
+
"learning_rate": 7.186440677966103e-07,
|
| 20354 |
+
"loss": 1.3319,
|
| 20355 |
+
"step": 28940
|
| 20356 |
+
},
|
| 20357 |
+
{
|
| 20358 |
+
"epoch": 0.72,
|
| 20359 |
+
"grad_norm": 8.437240600585938,
|
| 20360 |
+
"learning_rate": 7.118644067796611e-07,
|
| 20361 |
+
"loss": 1.2917,
|
| 20362 |
+
"step": 28950
|
| 20363 |
+
},
|
| 20364 |
+
{
|
| 20365 |
+
"epoch": 0.72,
|
| 20366 |
+
"grad_norm": 7.0036139488220215,
|
| 20367 |
+
"learning_rate": 7.05084745762712e-07,
|
| 20368 |
+
"loss": 1.306,
|
| 20369 |
+
"step": 28960
|
| 20370 |
+
},
|
| 20371 |
+
{
|
| 20372 |
+
"epoch": 0.72,
|
| 20373 |
+
"grad_norm": 6.939202308654785,
|
| 20374 |
+
"learning_rate": 6.983050847457627e-07,
|
| 20375 |
+
"loss": 1.1919,
|
| 20376 |
+
"step": 28970
|
| 20377 |
+
},
|
| 20378 |
+
{
|
| 20379 |
+
"epoch": 0.72,
|
| 20380 |
+
"grad_norm": 5.897158145904541,
|
| 20381 |
+
"learning_rate": 6.915254237288137e-07,
|
| 20382 |
+
"loss": 1.3183,
|
| 20383 |
+
"step": 28980
|
| 20384 |
+
},
|
| 20385 |
+
{
|
| 20386 |
+
"epoch": 0.72,
|
| 20387 |
+
"grad_norm": 8.489838600158691,
|
| 20388 |
+
"learning_rate": 6.847457627118645e-07,
|
| 20389 |
+
"loss": 1.3525,
|
| 20390 |
+
"step": 28990
|
| 20391 |
+
},
|
| 20392 |
+
{
|
| 20393 |
+
"epoch": 0.72,
|
| 20394 |
+
"grad_norm": 11.85141658782959,
|
| 20395 |
+
"learning_rate": 6.779661016949153e-07,
|
| 20396 |
+
"loss": 1.3708,
|
| 20397 |
+
"step": 29000
|
| 20398 |
+
},
|
| 20399 |
+
{
|
| 20400 |
+
"epoch": 0.73,
|
| 20401 |
+
"grad_norm": 12.649805068969727,
|
| 20402 |
+
"learning_rate": 6.711864406779661e-07,
|
| 20403 |
+
"loss": 1.2521,
|
| 20404 |
+
"step": 29010
|
| 20405 |
+
},
|
| 20406 |
+
{
|
| 20407 |
+
"epoch": 0.73,
|
| 20408 |
+
"grad_norm": 4.55898380279541,
|
| 20409 |
+
"learning_rate": 6.64406779661017e-07,
|
| 20410 |
+
"loss": 1.3484,
|
| 20411 |
+
"step": 29020
|
| 20412 |
+
},
|
| 20413 |
+
{
|
| 20414 |
+
"epoch": 0.73,
|
| 20415 |
+
"grad_norm": 7.323060989379883,
|
| 20416 |
+
"learning_rate": 6.576271186440679e-07,
|
| 20417 |
+
"loss": 1.4211,
|
| 20418 |
+
"step": 29030
|
| 20419 |
+
},
|
| 20420 |
+
{
|
| 20421 |
+
"epoch": 0.73,
|
| 20422 |
+
"grad_norm": 4.497279167175293,
|
| 20423 |
+
"learning_rate": 6.508474576271187e-07,
|
| 20424 |
+
"loss": 1.3893,
|
| 20425 |
+
"step": 29040
|
| 20426 |
+
},
|
| 20427 |
+
{
|
| 20428 |
+
"epoch": 0.73,
|
| 20429 |
+
"grad_norm": 10.126681327819824,
|
| 20430 |
+
"learning_rate": 6.440677966101695e-07,
|
| 20431 |
+
"loss": 1.2969,
|
| 20432 |
+
"step": 29050
|
| 20433 |
+
},
|
| 20434 |
+
{
|
| 20435 |
+
"epoch": 0.73,
|
| 20436 |
+
"grad_norm": 4.058706760406494,
|
| 20437 |
+
"learning_rate": 6.372881355932203e-07,
|
| 20438 |
+
"loss": 1.297,
|
| 20439 |
+
"step": 29060
|
| 20440 |
+
},
|
| 20441 |
+
{
|
| 20442 |
+
"epoch": 0.73,
|
| 20443 |
+
"grad_norm": 3.2129762172698975,
|
| 20444 |
+
"learning_rate": 6.305084745762713e-07,
|
| 20445 |
+
"loss": 1.336,
|
| 20446 |
+
"step": 29070
|
| 20447 |
+
},
|
| 20448 |
+
{
|
| 20449 |
+
"epoch": 0.73,
|
| 20450 |
+
"grad_norm": 2.8219425678253174,
|
| 20451 |
+
"learning_rate": 6.237288135593221e-07,
|
| 20452 |
+
"loss": 1.2462,
|
| 20453 |
+
"step": 29080
|
| 20454 |
+
},
|
| 20455 |
+
{
|
| 20456 |
+
"epoch": 0.73,
|
| 20457 |
+
"grad_norm": 3.939793109893799,
|
| 20458 |
+
"learning_rate": 6.16949152542373e-07,
|
| 20459 |
+
"loss": 1.417,
|
| 20460 |
+
"step": 29090
|
| 20461 |
+
},
|
| 20462 |
+
{
|
| 20463 |
+
"epoch": 0.73,
|
| 20464 |
+
"grad_norm": 5.102091312408447,
|
| 20465 |
+
"learning_rate": 6.101694915254238e-07,
|
| 20466 |
+
"loss": 1.467,
|
| 20467 |
+
"step": 29100
|
| 20468 |
+
},
|
| 20469 |
+
{
|
| 20470 |
+
"epoch": 0.73,
|
| 20471 |
+
"grad_norm": 24.394020080566406,
|
| 20472 |
+
"learning_rate": 6.033898305084746e-07,
|
| 20473 |
+
"loss": 1.0779,
|
| 20474 |
+
"step": 29110
|
| 20475 |
+
},
|
| 20476 |
+
{
|
| 20477 |
+
"epoch": 0.73,
|
| 20478 |
+
"grad_norm": 9.672082901000977,
|
| 20479 |
+
"learning_rate": 5.966101694915255e-07,
|
| 20480 |
+
"loss": 1.2263,
|
| 20481 |
+
"step": 29120
|
| 20482 |
+
},
|
| 20483 |
+
{
|
| 20484 |
+
"epoch": 0.73,
|
| 20485 |
+
"grad_norm": 3.5495758056640625,
|
| 20486 |
+
"learning_rate": 5.898305084745763e-07,
|
| 20487 |
+
"loss": 1.3195,
|
| 20488 |
+
"step": 29130
|
| 20489 |
+
},
|
| 20490 |
+
{
|
| 20491 |
+
"epoch": 0.73,
|
| 20492 |
+
"grad_norm": 3.160689115524292,
|
| 20493 |
+
"learning_rate": 5.830508474576271e-07,
|
| 20494 |
+
"loss": 1.4946,
|
| 20495 |
+
"step": 29140
|
| 20496 |
+
},
|
| 20497 |
+
{
|
| 20498 |
+
"epoch": 0.73,
|
| 20499 |
+
"grad_norm": 7.806612014770508,
|
| 20500 |
+
"learning_rate": 5.76271186440678e-07,
|
| 20501 |
+
"loss": 1.2732,
|
| 20502 |
+
"step": 29150
|
| 20503 |
+
},
|
| 20504 |
+
{
|
| 20505 |
+
"epoch": 0.73,
|
| 20506 |
+
"grad_norm": 16.695268630981445,
|
| 20507 |
+
"learning_rate": 5.694915254237288e-07,
|
| 20508 |
+
"loss": 1.4132,
|
| 20509 |
+
"step": 29160
|
| 20510 |
+
},
|
| 20511 |
+
{
|
| 20512 |
+
"epoch": 0.73,
|
| 20513 |
+
"grad_norm": 9.606965065002441,
|
| 20514 |
+
"learning_rate": 5.627118644067797e-07,
|
| 20515 |
+
"loss": 1.2185,
|
| 20516 |
+
"step": 29170
|
| 20517 |
+
},
|
| 20518 |
+
{
|
| 20519 |
+
"epoch": 0.73,
|
| 20520 |
+
"grad_norm": 3.9440929889678955,
|
| 20521 |
+
"learning_rate": 5.559322033898306e-07,
|
| 20522 |
+
"loss": 1.3204,
|
| 20523 |
+
"step": 29180
|
| 20524 |
+
},
|
| 20525 |
+
{
|
| 20526 |
+
"epoch": 0.73,
|
| 20527 |
+
"grad_norm": 11.45638370513916,
|
| 20528 |
+
"learning_rate": 5.491525423728815e-07,
|
| 20529 |
+
"loss": 1.3867,
|
| 20530 |
+
"step": 29190
|
| 20531 |
+
},
|
| 20532 |
+
{
|
| 20533 |
+
"epoch": 0.73,
|
| 20534 |
+
"grad_norm": 5.335674285888672,
|
| 20535 |
+
"learning_rate": 5.423728813559322e-07,
|
| 20536 |
+
"loss": 1.2608,
|
| 20537 |
+
"step": 29200
|
| 20538 |
+
},
|
| 20539 |
+
{
|
| 20540 |
+
"epoch": 0.73,
|
| 20541 |
+
"grad_norm": 5.219457149505615,
|
| 20542 |
+
"learning_rate": 5.355932203389831e-07,
|
| 20543 |
+
"loss": 1.3998,
|
| 20544 |
+
"step": 29210
|
| 20545 |
+
},
|
| 20546 |
+
{
|
| 20547 |
+
"epoch": 0.73,
|
| 20548 |
+
"grad_norm": 8.143280029296875,
|
| 20549 |
+
"learning_rate": 5.28813559322034e-07,
|
| 20550 |
+
"loss": 1.2506,
|
| 20551 |
+
"step": 29220
|
| 20552 |
+
},
|
| 20553 |
+
{
|
| 20554 |
+
"epoch": 0.73,
|
| 20555 |
+
"grad_norm": 21.4418888092041,
|
| 20556 |
+
"learning_rate": 5.220338983050848e-07,
|
| 20557 |
+
"loss": 1.3894,
|
| 20558 |
+
"step": 29230
|
| 20559 |
+
},
|
| 20560 |
+
{
|
| 20561 |
+
"epoch": 0.73,
|
| 20562 |
+
"grad_norm": 9.489405632019043,
|
| 20563 |
+
"learning_rate": 5.152542372881356e-07,
|
| 20564 |
+
"loss": 1.3329,
|
| 20565 |
+
"step": 29240
|
| 20566 |
+
},
|
| 20567 |
+
{
|
| 20568 |
+
"epoch": 0.73,
|
| 20569 |
+
"grad_norm": 4.9300031661987305,
|
| 20570 |
+
"learning_rate": 5.084745762711865e-07,
|
| 20571 |
+
"loss": 1.2334,
|
| 20572 |
+
"step": 29250
|
| 20573 |
+
},
|
| 20574 |
+
{
|
| 20575 |
+
"epoch": 0.73,
|
| 20576 |
+
"grad_norm": 4.382762908935547,
|
| 20577 |
+
"learning_rate": 5.016949152542373e-07,
|
| 20578 |
+
"loss": 1.2921,
|
| 20579 |
+
"step": 29260
|
| 20580 |
+
},
|
| 20581 |
+
{
|
| 20582 |
+
"epoch": 0.73,
|
| 20583 |
+
"grad_norm": 9.448204040527344,
|
| 20584 |
+
"learning_rate": 4.949152542372881e-07,
|
| 20585 |
+
"loss": 1.2535,
|
| 20586 |
+
"step": 29270
|
| 20587 |
+
},
|
| 20588 |
+
{
|
| 20589 |
+
"epoch": 0.73,
|
| 20590 |
+
"grad_norm": 12.2984619140625,
|
| 20591 |
+
"learning_rate": 4.881355932203391e-07,
|
| 20592 |
+
"loss": 1.1073,
|
| 20593 |
+
"step": 29280
|
| 20594 |
+
},
|
| 20595 |
+
{
|
| 20596 |
+
"epoch": 0.73,
|
| 20597 |
+
"grad_norm": 8.206221580505371,
|
| 20598 |
+
"learning_rate": 4.813559322033898e-07,
|
| 20599 |
+
"loss": 1.0739,
|
| 20600 |
+
"step": 29290
|
| 20601 |
+
},
|
| 20602 |
+
{
|
| 20603 |
+
"epoch": 0.73,
|
| 20604 |
+
"grad_norm": 16.299449920654297,
|
| 20605 |
+
"learning_rate": 4.745762711864407e-07,
|
| 20606 |
+
"loss": 1.2505,
|
| 20607 |
+
"step": 29300
|
| 20608 |
+
},
|
| 20609 |
+
{
|
| 20610 |
+
"epoch": 0.73,
|
| 20611 |
+
"grad_norm": 5.297379493713379,
|
| 20612 |
+
"learning_rate": 4.6779661016949154e-07,
|
| 20613 |
+
"loss": 1.2957,
|
| 20614 |
+
"step": 29310
|
| 20615 |
+
},
|
| 20616 |
+
{
|
| 20617 |
+
"epoch": 0.73,
|
| 20618 |
+
"grad_norm": 11.336212158203125,
|
| 20619 |
+
"learning_rate": 4.610169491525424e-07,
|
| 20620 |
+
"loss": 1.2837,
|
| 20621 |
+
"step": 29320
|
| 20622 |
+
},
|
| 20623 |
+
{
|
| 20624 |
+
"epoch": 0.73,
|
| 20625 |
+
"grad_norm": 8.407357215881348,
|
| 20626 |
+
"learning_rate": 4.542372881355932e-07,
|
| 20627 |
+
"loss": 1.2719,
|
| 20628 |
+
"step": 29330
|
| 20629 |
+
},
|
| 20630 |
+
{
|
| 20631 |
+
"epoch": 0.73,
|
| 20632 |
+
"grad_norm": 4.997983932495117,
|
| 20633 |
+
"learning_rate": 4.4745762711864415e-07,
|
| 20634 |
+
"loss": 1.3578,
|
| 20635 |
+
"step": 29340
|
| 20636 |
+
},
|
| 20637 |
+
{
|
| 20638 |
+
"epoch": 0.73,
|
| 20639 |
+
"grad_norm": 1.8061422109603882,
|
| 20640 |
+
"learning_rate": 4.4067796610169497e-07,
|
| 20641 |
+
"loss": 1.2524,
|
| 20642 |
+
"step": 29350
|
| 20643 |
+
},
|
| 20644 |
+
{
|
| 20645 |
+
"epoch": 0.73,
|
| 20646 |
+
"grad_norm": 7.534559726715088,
|
| 20647 |
+
"learning_rate": 4.3389830508474584e-07,
|
| 20648 |
+
"loss": 1.2727,
|
| 20649 |
+
"step": 29360
|
| 20650 |
+
},
|
| 20651 |
+
{
|
| 20652 |
+
"epoch": 0.73,
|
| 20653 |
+
"grad_norm": 2.897261619567871,
|
| 20654 |
+
"learning_rate": 4.2711864406779666e-07,
|
| 20655 |
+
"loss": 1.35,
|
| 20656 |
+
"step": 29370
|
| 20657 |
+
},
|
| 20658 |
+
{
|
| 20659 |
+
"epoch": 0.73,
|
| 20660 |
+
"grad_norm": 11.500360488891602,
|
| 20661 |
+
"learning_rate": 4.2033898305084753e-07,
|
| 20662 |
+
"loss": 1.3056,
|
| 20663 |
+
"step": 29380
|
| 20664 |
+
},
|
| 20665 |
+
{
|
| 20666 |
+
"epoch": 0.73,
|
| 20667 |
+
"grad_norm": 11.723714828491211,
|
| 20668 |
+
"learning_rate": 4.1355932203389835e-07,
|
| 20669 |
+
"loss": 1.3156,
|
| 20670 |
+
"step": 29390
|
| 20671 |
+
},
|
| 20672 |
+
{
|
| 20673 |
+
"epoch": 0.73,
|
| 20674 |
+
"grad_norm": 3.963747978210449,
|
| 20675 |
+
"learning_rate": 4.0677966101694916e-07,
|
| 20676 |
+
"loss": 1.3509,
|
| 20677 |
+
"step": 29400
|
| 20678 |
+
},
|
| 20679 |
+
{
|
| 20680 |
+
"epoch": 0.74,
|
| 20681 |
+
"grad_norm": 1.9269698858261108,
|
| 20682 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 20683 |
+
"loss": 1.3019,
|
| 20684 |
+
"step": 29410
|
| 20685 |
+
},
|
| 20686 |
+
{
|
| 20687 |
+
"epoch": 0.74,
|
| 20688 |
+
"grad_norm": 6.010022163391113,
|
| 20689 |
+
"learning_rate": 3.9322033898305085e-07,
|
| 20690 |
+
"loss": 1.2579,
|
| 20691 |
+
"step": 29420
|
| 20692 |
+
},
|
| 20693 |
+
{
|
| 20694 |
+
"epoch": 0.74,
|
| 20695 |
+
"grad_norm": 11.123760223388672,
|
| 20696 |
+
"learning_rate": 3.864406779661017e-07,
|
| 20697 |
+
"loss": 1.3221,
|
| 20698 |
+
"step": 29430
|
| 20699 |
+
},
|
| 20700 |
+
{
|
| 20701 |
+
"epoch": 0.74,
|
| 20702 |
+
"grad_norm": 11.927657127380371,
|
| 20703 |
+
"learning_rate": 3.7966101694915254e-07,
|
| 20704 |
+
"loss": 1.2219,
|
| 20705 |
+
"step": 29440
|
| 20706 |
+
},
|
| 20707 |
+
{
|
| 20708 |
+
"epoch": 0.74,
|
| 20709 |
+
"grad_norm": 8.796635627746582,
|
| 20710 |
+
"learning_rate": 3.7288135593220347e-07,
|
| 20711 |
+
"loss": 1.2437,
|
| 20712 |
+
"step": 29450
|
| 20713 |
+
},
|
| 20714 |
+
{
|
| 20715 |
+
"epoch": 0.74,
|
| 20716 |
+
"grad_norm": 15.227128028869629,
|
| 20717 |
+
"learning_rate": 3.6610169491525423e-07,
|
| 20718 |
+
"loss": 1.2653,
|
| 20719 |
+
"step": 29460
|
| 20720 |
+
},
|
| 20721 |
+
{
|
| 20722 |
+
"epoch": 0.74,
|
| 20723 |
+
"grad_norm": 3.172653913497925,
|
| 20724 |
+
"learning_rate": 3.5932203389830516e-07,
|
| 20725 |
+
"loss": 1.4635,
|
| 20726 |
+
"step": 29470
|
| 20727 |
+
},
|
| 20728 |
+
{
|
| 20729 |
+
"epoch": 0.74,
|
| 20730 |
+
"grad_norm": 5.841287612915039,
|
| 20731 |
+
"learning_rate": 3.52542372881356e-07,
|
| 20732 |
+
"loss": 1.2766,
|
| 20733 |
+
"step": 29480
|
| 20734 |
+
},
|
| 20735 |
+
{
|
| 20736 |
+
"epoch": 0.74,
|
| 20737 |
+
"grad_norm": 9.996382713317871,
|
| 20738 |
+
"learning_rate": 3.4576271186440684e-07,
|
| 20739 |
+
"loss": 1.3705,
|
| 20740 |
+
"step": 29490
|
| 20741 |
+
},
|
| 20742 |
+
{
|
| 20743 |
+
"epoch": 0.74,
|
| 20744 |
+
"grad_norm": 9.059891700744629,
|
| 20745 |
+
"learning_rate": 3.3898305084745766e-07,
|
| 20746 |
+
"loss": 1.282,
|
| 20747 |
+
"step": 29500
|
| 20748 |
+
},
|
| 20749 |
+
{
|
| 20750 |
+
"epoch": 0.74,
|
| 20751 |
+
"grad_norm": 6.326866626739502,
|
| 20752 |
+
"learning_rate": 3.322033898305085e-07,
|
| 20753 |
+
"loss": 1.1478,
|
| 20754 |
+
"step": 29510
|
| 20755 |
+
},
|
| 20756 |
+
{
|
| 20757 |
+
"epoch": 0.74,
|
| 20758 |
+
"grad_norm": 4.546393394470215,
|
| 20759 |
+
"learning_rate": 3.2542372881355935e-07,
|
| 20760 |
+
"loss": 1.3345,
|
| 20761 |
+
"step": 29520
|
| 20762 |
+
},
|
| 20763 |
+
{
|
| 20764 |
+
"epoch": 0.74,
|
| 20765 |
+
"grad_norm": 3.654453992843628,
|
| 20766 |
+
"learning_rate": 3.1864406779661017e-07,
|
| 20767 |
+
"loss": 1.5465,
|
| 20768 |
+
"step": 29530
|
| 20769 |
+
},
|
| 20770 |
+
{
|
| 20771 |
+
"epoch": 0.74,
|
| 20772 |
+
"grad_norm": 8.529802322387695,
|
| 20773 |
+
"learning_rate": 3.1186440677966104e-07,
|
| 20774 |
+
"loss": 1.5145,
|
| 20775 |
+
"step": 29540
|
| 20776 |
+
},
|
| 20777 |
+
{
|
| 20778 |
+
"epoch": 0.74,
|
| 20779 |
+
"grad_norm": 4.871727466583252,
|
| 20780 |
+
"learning_rate": 3.050847457627119e-07,
|
| 20781 |
+
"loss": 1.3838,
|
| 20782 |
+
"step": 29550
|
| 20783 |
+
},
|
| 20784 |
+
{
|
| 20785 |
+
"epoch": 0.74,
|
| 20786 |
+
"grad_norm": 6.172389507293701,
|
| 20787 |
+
"learning_rate": 2.9830508474576273e-07,
|
| 20788 |
+
"loss": 1.3202,
|
| 20789 |
+
"step": 29560
|
| 20790 |
+
},
|
| 20791 |
+
{
|
| 20792 |
+
"epoch": 0.74,
|
| 20793 |
+
"grad_norm": 9.824962615966797,
|
| 20794 |
+
"learning_rate": 2.9152542372881355e-07,
|
| 20795 |
+
"loss": 1.2167,
|
| 20796 |
+
"step": 29570
|
| 20797 |
+
},
|
| 20798 |
+
{
|
| 20799 |
+
"epoch": 0.74,
|
| 20800 |
+
"grad_norm": 10.725250244140625,
|
| 20801 |
+
"learning_rate": 2.847457627118644e-07,
|
| 20802 |
+
"loss": 1.2796,
|
| 20803 |
+
"step": 29580
|
| 20804 |
+
},
|
| 20805 |
+
{
|
| 20806 |
+
"epoch": 0.74,
|
| 20807 |
+
"grad_norm": 14.172257423400879,
|
| 20808 |
+
"learning_rate": 2.779661016949153e-07,
|
| 20809 |
+
"loss": 1.1175,
|
| 20810 |
+
"step": 29590
|
| 20811 |
+
},
|
| 20812 |
+
{
|
| 20813 |
+
"epoch": 0.74,
|
| 20814 |
+
"grad_norm": 6.030600547790527,
|
| 20815 |
+
"learning_rate": 2.711864406779661e-07,
|
| 20816 |
+
"loss": 1.0479,
|
| 20817 |
+
"step": 29600
|
| 20818 |
+
},
|
| 20819 |
+
{
|
| 20820 |
+
"epoch": 0.74,
|
| 20821 |
+
"grad_norm": 3.860551357269287,
|
| 20822 |
+
"learning_rate": 2.64406779661017e-07,
|
| 20823 |
+
"loss": 1.1551,
|
| 20824 |
+
"step": 29610
|
| 20825 |
+
},
|
| 20826 |
+
{
|
| 20827 |
+
"epoch": 0.74,
|
| 20828 |
+
"grad_norm": 8.990274429321289,
|
| 20829 |
+
"learning_rate": 2.576271186440678e-07,
|
| 20830 |
+
"loss": 1.2577,
|
| 20831 |
+
"step": 29620
|
| 20832 |
+
},
|
| 20833 |
+
{
|
| 20834 |
+
"epoch": 0.74,
|
| 20835 |
+
"grad_norm": 9.873722076416016,
|
| 20836 |
+
"learning_rate": 2.5084745762711867e-07,
|
| 20837 |
+
"loss": 1.333,
|
| 20838 |
+
"step": 29630
|
| 20839 |
+
},
|
| 20840 |
+
{
|
| 20841 |
+
"epoch": 0.74,
|
| 20842 |
+
"grad_norm": 10.069340705871582,
|
| 20843 |
+
"learning_rate": 2.4406779661016954e-07,
|
| 20844 |
+
"loss": 1.4405,
|
| 20845 |
+
"step": 29640
|
| 20846 |
+
},
|
| 20847 |
+
{
|
| 20848 |
+
"epoch": 0.74,
|
| 20849 |
+
"grad_norm": 4.778994083404541,
|
| 20850 |
+
"learning_rate": 2.3728813559322036e-07,
|
| 20851 |
+
"loss": 1.5492,
|
| 20852 |
+
"step": 29650
|
| 20853 |
+
},
|
| 20854 |
+
{
|
| 20855 |
+
"epoch": 0.74,
|
| 20856 |
+
"grad_norm": 14.848403930664062,
|
| 20857 |
+
"learning_rate": 2.305084745762712e-07,
|
| 20858 |
+
"loss": 1.136,
|
| 20859 |
+
"step": 29660
|
| 20860 |
+
},
|
| 20861 |
+
{
|
| 20862 |
+
"epoch": 0.74,
|
| 20863 |
+
"grad_norm": 17.621976852416992,
|
| 20864 |
+
"learning_rate": 2.2372881355932207e-07,
|
| 20865 |
+
"loss": 1.3884,
|
| 20866 |
+
"step": 29670
|
| 20867 |
+
},
|
| 20868 |
+
{
|
| 20869 |
+
"epoch": 0.74,
|
| 20870 |
+
"grad_norm": 5.0682597160339355,
|
| 20871 |
+
"learning_rate": 2.1694915254237292e-07,
|
| 20872 |
+
"loss": 1.2395,
|
| 20873 |
+
"step": 29680
|
| 20874 |
+
},
|
| 20875 |
+
{
|
| 20876 |
+
"epoch": 0.74,
|
| 20877 |
+
"grad_norm": 10.465557098388672,
|
| 20878 |
+
"learning_rate": 2.1016949152542376e-07,
|
| 20879 |
+
"loss": 1.3657,
|
| 20880 |
+
"step": 29690
|
| 20881 |
+
},
|
| 20882 |
+
{
|
| 20883 |
+
"epoch": 0.74,
|
| 20884 |
+
"grad_norm": 16.993274688720703,
|
| 20885 |
+
"learning_rate": 2.0338983050847458e-07,
|
| 20886 |
+
"loss": 1.2805,
|
| 20887 |
+
"step": 29700
|
| 20888 |
+
},
|
| 20889 |
+
{
|
| 20890 |
+
"epoch": 0.74,
|
| 20891 |
+
"grad_norm": 8.59652042388916,
|
| 20892 |
+
"learning_rate": 1.9661016949152543e-07,
|
| 20893 |
+
"loss": 1.0588,
|
| 20894 |
+
"step": 29710
|
| 20895 |
+
},
|
| 20896 |
+
{
|
| 20897 |
+
"epoch": 0.74,
|
| 20898 |
+
"grad_norm": 3.580928087234497,
|
| 20899 |
+
"learning_rate": 1.8983050847457627e-07,
|
| 20900 |
+
"loss": 1.3986,
|
| 20901 |
+
"step": 29720
|
| 20902 |
+
},
|
| 20903 |
+
{
|
| 20904 |
+
"epoch": 0.74,
|
| 20905 |
+
"grad_norm": 5.2861552238464355,
|
| 20906 |
+
"learning_rate": 1.8305084745762712e-07,
|
| 20907 |
+
"loss": 1.1782,
|
| 20908 |
+
"step": 29730
|
| 20909 |
+
},
|
| 20910 |
+
{
|
| 20911 |
+
"epoch": 0.74,
|
| 20912 |
+
"grad_norm": 16.308183670043945,
|
| 20913 |
+
"learning_rate": 1.76271186440678e-07,
|
| 20914 |
+
"loss": 1.153,
|
| 20915 |
+
"step": 29740
|
| 20916 |
+
},
|
| 20917 |
+
{
|
| 20918 |
+
"epoch": 0.74,
|
| 20919 |
+
"grad_norm": 4.2665863037109375,
|
| 20920 |
+
"learning_rate": 1.6949152542372883e-07,
|
| 20921 |
+
"loss": 1.3667,
|
| 20922 |
+
"step": 29750
|
| 20923 |
+
},
|
| 20924 |
+
{
|
| 20925 |
+
"epoch": 0.74,
|
| 20926 |
+
"grad_norm": 7.868191719055176,
|
| 20927 |
+
"learning_rate": 1.6271186440677968e-07,
|
| 20928 |
+
"loss": 1.2511,
|
| 20929 |
+
"step": 29760
|
| 20930 |
+
},
|
| 20931 |
+
{
|
| 20932 |
+
"epoch": 0.74,
|
| 20933 |
+
"grad_norm": 6.3579936027526855,
|
| 20934 |
+
"learning_rate": 1.5593220338983052e-07,
|
| 20935 |
+
"loss": 1.4062,
|
| 20936 |
+
"step": 29770
|
| 20937 |
+
},
|
| 20938 |
+
{
|
| 20939 |
+
"epoch": 0.74,
|
| 20940 |
+
"grad_norm": 9.477784156799316,
|
| 20941 |
+
"learning_rate": 1.4915254237288137e-07,
|
| 20942 |
+
"loss": 1.2801,
|
| 20943 |
+
"step": 29780
|
| 20944 |
+
},
|
| 20945 |
+
{
|
| 20946 |
+
"epoch": 0.74,
|
| 20947 |
+
"grad_norm": 5.881575584411621,
|
| 20948 |
+
"learning_rate": 1.423728813559322e-07,
|
| 20949 |
+
"loss": 1.343,
|
| 20950 |
+
"step": 29790
|
| 20951 |
+
},
|
| 20952 |
+
{
|
| 20953 |
+
"epoch": 0.74,
|
| 20954 |
+
"grad_norm": 6.489737510681152,
|
| 20955 |
+
"learning_rate": 1.3559322033898305e-07,
|
| 20956 |
+
"loss": 1.302,
|
| 20957 |
+
"step": 29800
|
| 20958 |
+
},
|
| 20959 |
+
{
|
| 20960 |
+
"epoch": 0.75,
|
| 20961 |
+
"grad_norm": 14.81570816040039,
|
| 20962 |
+
"learning_rate": 1.288135593220339e-07,
|
| 20963 |
+
"loss": 1.251,
|
| 20964 |
+
"step": 29810
|
| 20965 |
+
},
|
| 20966 |
+
{
|
| 20967 |
+
"epoch": 0.75,
|
| 20968 |
+
"grad_norm": 15.79887866973877,
|
| 20969 |
+
"learning_rate": 1.2203389830508477e-07,
|
| 20970 |
+
"loss": 1.3992,
|
| 20971 |
+
"step": 29820
|
| 20972 |
+
},
|
| 20973 |
+
{
|
| 20974 |
+
"epoch": 0.75,
|
| 20975 |
+
"grad_norm": 5.4701828956604,
|
| 20976 |
+
"learning_rate": 1.152542372881356e-07,
|
| 20977 |
+
"loss": 1.3728,
|
| 20978 |
+
"step": 29830
|
| 20979 |
+
},
|
| 20980 |
+
{
|
| 20981 |
+
"epoch": 0.75,
|
| 20982 |
+
"grad_norm": 5.326042175292969,
|
| 20983 |
+
"learning_rate": 1.0847457627118646e-07,
|
| 20984 |
+
"loss": 1.3645,
|
| 20985 |
+
"step": 29840
|
| 20986 |
+
},
|
| 20987 |
+
{
|
| 20988 |
+
"epoch": 0.75,
|
| 20989 |
+
"grad_norm": 3.9920434951782227,
|
| 20990 |
+
"learning_rate": 1.0169491525423729e-07,
|
| 20991 |
+
"loss": 1.4768,
|
| 20992 |
+
"step": 29850
|
| 20993 |
+
},
|
| 20994 |
+
{
|
| 20995 |
+
"epoch": 0.75,
|
| 20996 |
+
"grad_norm": 10.122292518615723,
|
| 20997 |
+
"learning_rate": 9.491525423728814e-08,
|
| 20998 |
+
"loss": 1.4596,
|
| 20999 |
+
"step": 29860
|
| 21000 |
+
},
|
| 21001 |
+
{
|
| 21002 |
+
"epoch": 0.75,
|
| 21003 |
+
"grad_norm": 6.636433124542236,
|
| 21004 |
+
"learning_rate": 8.8135593220339e-08,
|
| 21005 |
+
"loss": 1.3858,
|
| 21006 |
+
"step": 29870
|
| 21007 |
+
},
|
| 21008 |
+
{
|
| 21009 |
+
"epoch": 0.75,
|
| 21010 |
+
"grad_norm": 6.18862247467041,
|
| 21011 |
+
"learning_rate": 8.135593220338984e-08,
|
| 21012 |
+
"loss": 1.3912,
|
| 21013 |
+
"step": 29880
|
| 21014 |
+
},
|
| 21015 |
+
{
|
| 21016 |
+
"epoch": 0.75,
|
| 21017 |
+
"grad_norm": 10.676321983337402,
|
| 21018 |
+
"learning_rate": 7.457627118644068e-08,
|
| 21019 |
+
"loss": 1.3445,
|
| 21020 |
+
"step": 29890
|
| 21021 |
+
},
|
| 21022 |
+
{
|
| 21023 |
+
"epoch": 0.75,
|
| 21024 |
+
"grad_norm": 14.703264236450195,
|
| 21025 |
+
"learning_rate": 6.779661016949153e-08,
|
| 21026 |
+
"loss": 1.3391,
|
| 21027 |
+
"step": 29900
|
| 21028 |
+
},
|
| 21029 |
+
{
|
| 21030 |
+
"epoch": 0.75,
|
| 21031 |
+
"grad_norm": 11.101775169372559,
|
| 21032 |
+
"learning_rate": 6.101694915254239e-08,
|
| 21033 |
+
"loss": 1.2044,
|
| 21034 |
+
"step": 29910
|
| 21035 |
+
},
|
| 21036 |
+
{
|
| 21037 |
+
"epoch": 0.75,
|
| 21038 |
+
"grad_norm": 5.793562889099121,
|
| 21039 |
+
"learning_rate": 5.423728813559323e-08,
|
| 21040 |
+
"loss": 1.3205,
|
| 21041 |
+
"step": 29920
|
| 21042 |
+
},
|
| 21043 |
+
{
|
| 21044 |
+
"epoch": 0.75,
|
| 21045 |
+
"grad_norm": 4.186032295227051,
|
| 21046 |
+
"learning_rate": 4.745762711864407e-08,
|
| 21047 |
+
"loss": 1.3155,
|
| 21048 |
+
"step": 29930
|
| 21049 |
+
},
|
| 21050 |
+
{
|
| 21051 |
+
"epoch": 0.75,
|
| 21052 |
+
"grad_norm": 5.155704021453857,
|
| 21053 |
+
"learning_rate": 4.067796610169492e-08,
|
| 21054 |
+
"loss": 1.3661,
|
| 21055 |
+
"step": 29940
|
| 21056 |
+
},
|
| 21057 |
+
{
|
| 21058 |
+
"epoch": 0.75,
|
| 21059 |
+
"grad_norm": 10.625030517578125,
|
| 21060 |
+
"learning_rate": 3.3898305084745764e-08,
|
| 21061 |
+
"loss": 1.2094,
|
| 21062 |
+
"step": 29950
|
| 21063 |
+
},
|
| 21064 |
+
{
|
| 21065 |
+
"epoch": 0.75,
|
| 21066 |
+
"grad_norm": 9.991485595703125,
|
| 21067 |
+
"learning_rate": 2.7118644067796615e-08,
|
| 21068 |
+
"loss": 1.2198,
|
| 21069 |
+
"step": 29960
|
| 21070 |
+
},
|
| 21071 |
+
{
|
| 21072 |
+
"epoch": 0.75,
|
| 21073 |
+
"grad_norm": 3.559561014175415,
|
| 21074 |
+
"learning_rate": 2.033898305084746e-08,
|
| 21075 |
+
"loss": 1.33,
|
| 21076 |
+
"step": 29970
|
| 21077 |
+
},
|
| 21078 |
+
{
|
| 21079 |
+
"epoch": 0.75,
|
| 21080 |
+
"grad_norm": 1.8645702600479126,
|
| 21081 |
+
"learning_rate": 1.3559322033898307e-08,
|
| 21082 |
+
"loss": 1.3138,
|
| 21083 |
+
"step": 29980
|
| 21084 |
+
},
|
| 21085 |
+
{
|
| 21086 |
+
"epoch": 0.75,
|
| 21087 |
+
"grad_norm": 4.156149387359619,
|
| 21088 |
+
"learning_rate": 6.779661016949154e-09,
|
| 21089 |
+
"loss": 1.2674,
|
| 21090 |
+
"step": 29990
|
| 21091 |
+
},
|
| 21092 |
+
{
|
| 21093 |
+
"epoch": 0.75,
|
| 21094 |
+
"grad_norm": 4.104331016540527,
|
| 21095 |
+
"learning_rate": 0.0,
|
| 21096 |
+
"loss": 1.2401,
|
| 21097 |
+
"step": 30000
|
| 21098 |
+
},
|
| 21099 |
+
{
|
| 21100 |
+
"epoch": 0.75,
|
| 21101 |
+
"eval_loss": 1.2914531230926514,
|
| 21102 |
+
"eval_runtime": 122.439,
|
| 21103 |
+
"eval_samples_per_second": 8.167,
|
| 21104 |
+
"eval_steps_per_second": 8.167,
|
| 21105 |
+
"step": 30000
|
| 21106 |
}
|
| 21107 |
],
|
| 21108 |
"logging_steps": 10,
|
|
|
|
| 21110 |
"num_input_tokens_seen": 0,
|
| 21111 |
"num_train_epochs": 1,
|
| 21112 |
"save_steps": 2500,
|
| 21113 |
+
"total_flos": 4.8306377981952e+17,
|
| 21114 |
"train_batch_size": 1,
|
| 21115 |
"trial_name": null,
|
| 21116 |
"trial_params": null
|