Uploaded checkpoint-27500
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1793 -3
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 119975656
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:51ef8a0ba7bad5db9a03c998b7e0f0b9be8a11cac8ec229721c1db943b49949b
|
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 sha256:1a5eef17387e7ce5efb991a9df23b4cb28f5da1437f9349935c3f248a0519d26
|
3 |
size 240145026
|
rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14244
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:42ed1734c5823abfe806343a4de18dcccd1e9ad5af5349e08097c7bde2aa7437
|
3 |
size 14244
|
scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1064
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bae572518ab53ddc674f52a5ef01613875bea64a8d9c53d4b7d4a9aedc712f19
|
3 |
size 1064
|
trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": 0.9729424715042114,
|
3 |
"best_model_checkpoint": "runs/deepseek_lora_20240421-183352/checkpoint-25000",
|
4 |
-
"epoch": 0.
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -17907,6 +17907,1796 @@
|
|
17907 |
"eval_samples_per_second": 14.632,
|
17908 |
"eval_steps_per_second": 14.632,
|
17909 |
"step": 25000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17910 |
}
|
17911 |
],
|
17912 |
"logging_steps": 10,
|
@@ -17914,7 +19704,7 @@
|
|
17914 |
"num_input_tokens_seen": 0,
|
17915 |
"num_train_epochs": 1,
|
17916 |
"save_steps": 2500,
|
17917 |
-
"total_flos": 4.
|
17918 |
"train_batch_size": 1,
|
17919 |
"trial_name": null,
|
17920 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": 0.9729424715042114,
|
3 |
"best_model_checkpoint": "runs/deepseek_lora_20240421-183352/checkpoint-25000",
|
4 |
+
"epoch": 0.6366956652601481,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 27500,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
17907 |
"eval_samples_per_second": 14.632,
|
17908 |
"eval_steps_per_second": 14.632,
|
17909 |
"step": 25000
|
17910 |
+
},
|
17911 |
+
{
|
17912 |
+
"epoch": 0.58,
|
17913 |
+
"grad_norm": 5.260663032531738,
|
17914 |
+
"learning_rate": 3.3830508474576273e-06,
|
17915 |
+
"loss": 1.2702,
|
17916 |
+
"step": 25010
|
17917 |
+
},
|
17918 |
+
{
|
17919 |
+
"epoch": 0.58,
|
17920 |
+
"grad_norm": 5.188079357147217,
|
17921 |
+
"learning_rate": 3.3762711864406783e-06,
|
17922 |
+
"loss": 1.2995,
|
17923 |
+
"step": 25020
|
17924 |
+
},
|
17925 |
+
{
|
17926 |
+
"epoch": 0.58,
|
17927 |
+
"grad_norm": 7.982115745544434,
|
17928 |
+
"learning_rate": 3.3694915254237292e-06,
|
17929 |
+
"loss": 1.2687,
|
17930 |
+
"step": 25030
|
17931 |
+
},
|
17932 |
+
{
|
17933 |
+
"epoch": 0.58,
|
17934 |
+
"grad_norm": 3.1299421787261963,
|
17935 |
+
"learning_rate": 3.3627118644067802e-06,
|
17936 |
+
"loss": 1.3372,
|
17937 |
+
"step": 25040
|
17938 |
+
},
|
17939 |
+
{
|
17940 |
+
"epoch": 0.58,
|
17941 |
+
"grad_norm": 5.5412726402282715,
|
17942 |
+
"learning_rate": 3.3559322033898308e-06,
|
17943 |
+
"loss": 1.2002,
|
17944 |
+
"step": 25050
|
17945 |
+
},
|
17946 |
+
{
|
17947 |
+
"epoch": 0.58,
|
17948 |
+
"grad_norm": 6.219215393066406,
|
17949 |
+
"learning_rate": 3.3491525423728817e-06,
|
17950 |
+
"loss": 1.1117,
|
17951 |
+
"step": 25060
|
17952 |
+
},
|
17953 |
+
{
|
17954 |
+
"epoch": 0.58,
|
17955 |
+
"grad_norm": 3.6740598678588867,
|
17956 |
+
"learning_rate": 3.3423728813559327e-06,
|
17957 |
+
"loss": 1.2222,
|
17958 |
+
"step": 25070
|
17959 |
+
},
|
17960 |
+
{
|
17961 |
+
"epoch": 0.58,
|
17962 |
+
"grad_norm": 5.168256759643555,
|
17963 |
+
"learning_rate": 3.3355932203389833e-06,
|
17964 |
+
"loss": 1.2059,
|
17965 |
+
"step": 25080
|
17966 |
+
},
|
17967 |
+
{
|
17968 |
+
"epoch": 0.58,
|
17969 |
+
"grad_norm": 4.997997283935547,
|
17970 |
+
"learning_rate": 3.3288135593220343e-06,
|
17971 |
+
"loss": 1.3613,
|
17972 |
+
"step": 25090
|
17973 |
+
},
|
17974 |
+
{
|
17975 |
+
"epoch": 0.58,
|
17976 |
+
"grad_norm": 10.958406448364258,
|
17977 |
+
"learning_rate": 3.322033898305085e-06,
|
17978 |
+
"loss": 1.2508,
|
17979 |
+
"step": 25100
|
17980 |
+
},
|
17981 |
+
{
|
17982 |
+
"epoch": 0.58,
|
17983 |
+
"grad_norm": 7.793730735778809,
|
17984 |
+
"learning_rate": 3.3152542372881358e-06,
|
17985 |
+
"loss": 1.1418,
|
17986 |
+
"step": 25110
|
17987 |
+
},
|
17988 |
+
{
|
17989 |
+
"epoch": 0.58,
|
17990 |
+
"grad_norm": 6.706384181976318,
|
17991 |
+
"learning_rate": 3.3084745762711868e-06,
|
17992 |
+
"loss": 1.1661,
|
17993 |
+
"step": 25120
|
17994 |
+
},
|
17995 |
+
{
|
17996 |
+
"epoch": 0.58,
|
17997 |
+
"grad_norm": 9.17857551574707,
|
17998 |
+
"learning_rate": 3.3016949152542377e-06,
|
17999 |
+
"loss": 1.1569,
|
18000 |
+
"step": 25130
|
18001 |
+
},
|
18002 |
+
{
|
18003 |
+
"epoch": 0.58,
|
18004 |
+
"grad_norm": 1.5717544555664062,
|
18005 |
+
"learning_rate": 3.2949152542372887e-06,
|
18006 |
+
"loss": 1.0838,
|
18007 |
+
"step": 25140
|
18008 |
+
},
|
18009 |
+
{
|
18010 |
+
"epoch": 0.58,
|
18011 |
+
"grad_norm": 7.292511940002441,
|
18012 |
+
"learning_rate": 3.288135593220339e-06,
|
18013 |
+
"loss": 1.3231,
|
18014 |
+
"step": 25150
|
18015 |
+
},
|
18016 |
+
{
|
18017 |
+
"epoch": 0.58,
|
18018 |
+
"grad_norm": 6.030792236328125,
|
18019 |
+
"learning_rate": 3.28135593220339e-06,
|
18020 |
+
"loss": 1.1247,
|
18021 |
+
"step": 25160
|
18022 |
+
},
|
18023 |
+
{
|
18024 |
+
"epoch": 0.58,
|
18025 |
+
"grad_norm": 2.3071470260620117,
|
18026 |
+
"learning_rate": 3.274576271186441e-06,
|
18027 |
+
"loss": 1.113,
|
18028 |
+
"step": 25170
|
18029 |
+
},
|
18030 |
+
{
|
18031 |
+
"epoch": 0.58,
|
18032 |
+
"grad_norm": 5.088674545288086,
|
18033 |
+
"learning_rate": 3.2677966101694918e-06,
|
18034 |
+
"loss": 1.0984,
|
18035 |
+
"step": 25180
|
18036 |
+
},
|
18037 |
+
{
|
18038 |
+
"epoch": 0.58,
|
18039 |
+
"grad_norm": 13.216058731079102,
|
18040 |
+
"learning_rate": 3.2610169491525428e-06,
|
18041 |
+
"loss": 1.3596,
|
18042 |
+
"step": 25190
|
18043 |
+
},
|
18044 |
+
{
|
18045 |
+
"epoch": 0.58,
|
18046 |
+
"grad_norm": 5.061989784240723,
|
18047 |
+
"learning_rate": 3.2542372881355933e-06,
|
18048 |
+
"loss": 1.1028,
|
18049 |
+
"step": 25200
|
18050 |
+
},
|
18051 |
+
{
|
18052 |
+
"epoch": 0.58,
|
18053 |
+
"grad_norm": 5.740258693695068,
|
18054 |
+
"learning_rate": 3.2474576271186443e-06,
|
18055 |
+
"loss": 1.14,
|
18056 |
+
"step": 25210
|
18057 |
+
},
|
18058 |
+
{
|
18059 |
+
"epoch": 0.58,
|
18060 |
+
"grad_norm": 5.7218098640441895,
|
18061 |
+
"learning_rate": 3.2406779661016953e-06,
|
18062 |
+
"loss": 1.3268,
|
18063 |
+
"step": 25220
|
18064 |
+
},
|
18065 |
+
{
|
18066 |
+
"epoch": 0.58,
|
18067 |
+
"grad_norm": 10.579949378967285,
|
18068 |
+
"learning_rate": 3.2338983050847462e-06,
|
18069 |
+
"loss": 1.1409,
|
18070 |
+
"step": 25230
|
18071 |
+
},
|
18072 |
+
{
|
18073 |
+
"epoch": 0.58,
|
18074 |
+
"grad_norm": 10.984354019165039,
|
18075 |
+
"learning_rate": 3.2271186440677972e-06,
|
18076 |
+
"loss": 1.3797,
|
18077 |
+
"step": 25240
|
18078 |
+
},
|
18079 |
+
{
|
18080 |
+
"epoch": 0.58,
|
18081 |
+
"grad_norm": 8.028718948364258,
|
18082 |
+
"learning_rate": 3.2203389830508473e-06,
|
18083 |
+
"loss": 1.2275,
|
18084 |
+
"step": 25250
|
18085 |
+
},
|
18086 |
+
{
|
18087 |
+
"epoch": 0.58,
|
18088 |
+
"grad_norm": 2.102750301361084,
|
18089 |
+
"learning_rate": 3.2135593220338983e-06,
|
18090 |
+
"loss": 1.3638,
|
18091 |
+
"step": 25260
|
18092 |
+
},
|
18093 |
+
{
|
18094 |
+
"epoch": 0.59,
|
18095 |
+
"grad_norm": 12.67302417755127,
|
18096 |
+
"learning_rate": 3.2067796610169493e-06,
|
18097 |
+
"loss": 1.3955,
|
18098 |
+
"step": 25270
|
18099 |
+
},
|
18100 |
+
{
|
18101 |
+
"epoch": 0.59,
|
18102 |
+
"grad_norm": 9.690311431884766,
|
18103 |
+
"learning_rate": 3.2000000000000003e-06,
|
18104 |
+
"loss": 1.178,
|
18105 |
+
"step": 25280
|
18106 |
+
},
|
18107 |
+
{
|
18108 |
+
"epoch": 0.59,
|
18109 |
+
"grad_norm": 9.31050968170166,
|
18110 |
+
"learning_rate": 3.1932203389830513e-06,
|
18111 |
+
"loss": 1.3644,
|
18112 |
+
"step": 25290
|
18113 |
+
},
|
18114 |
+
{
|
18115 |
+
"epoch": 0.59,
|
18116 |
+
"grad_norm": 6.5038933753967285,
|
18117 |
+
"learning_rate": 3.186440677966102e-06,
|
18118 |
+
"loss": 1.4397,
|
18119 |
+
"step": 25300
|
18120 |
+
},
|
18121 |
+
{
|
18122 |
+
"epoch": 0.59,
|
18123 |
+
"grad_norm": 4.691766262054443,
|
18124 |
+
"learning_rate": 3.1796610169491528e-06,
|
18125 |
+
"loss": 1.3062,
|
18126 |
+
"step": 25310
|
18127 |
+
},
|
18128 |
+
{
|
18129 |
+
"epoch": 0.59,
|
18130 |
+
"grad_norm": 9.108213424682617,
|
18131 |
+
"learning_rate": 3.1728813559322038e-06,
|
18132 |
+
"loss": 1.1835,
|
18133 |
+
"step": 25320
|
18134 |
+
},
|
18135 |
+
{
|
18136 |
+
"epoch": 0.59,
|
18137 |
+
"grad_norm": 6.10239315032959,
|
18138 |
+
"learning_rate": 3.1661016949152547e-06,
|
18139 |
+
"loss": 1.3249,
|
18140 |
+
"step": 25330
|
18141 |
+
},
|
18142 |
+
{
|
18143 |
+
"epoch": 0.59,
|
18144 |
+
"grad_norm": 3.791691780090332,
|
18145 |
+
"learning_rate": 3.1593220338983053e-06,
|
18146 |
+
"loss": 1.2023,
|
18147 |
+
"step": 25340
|
18148 |
+
},
|
18149 |
+
{
|
18150 |
+
"epoch": 0.59,
|
18151 |
+
"grad_norm": 5.800852298736572,
|
18152 |
+
"learning_rate": 3.1525423728813563e-06,
|
18153 |
+
"loss": 1.352,
|
18154 |
+
"step": 25350
|
18155 |
+
},
|
18156 |
+
{
|
18157 |
+
"epoch": 0.59,
|
18158 |
+
"grad_norm": 6.795603275299072,
|
18159 |
+
"learning_rate": 3.145762711864407e-06,
|
18160 |
+
"loss": 1.3013,
|
18161 |
+
"step": 25360
|
18162 |
+
},
|
18163 |
+
{
|
18164 |
+
"epoch": 0.59,
|
18165 |
+
"grad_norm": 8.434121131896973,
|
18166 |
+
"learning_rate": 3.138983050847458e-06,
|
18167 |
+
"loss": 1.1872,
|
18168 |
+
"step": 25370
|
18169 |
+
},
|
18170 |
+
{
|
18171 |
+
"epoch": 0.59,
|
18172 |
+
"grad_norm": 4.576048851013184,
|
18173 |
+
"learning_rate": 3.1322033898305088e-06,
|
18174 |
+
"loss": 1.2377,
|
18175 |
+
"step": 25380
|
18176 |
+
},
|
18177 |
+
{
|
18178 |
+
"epoch": 0.59,
|
18179 |
+
"grad_norm": 11.319803237915039,
|
18180 |
+
"learning_rate": 3.1254237288135598e-06,
|
18181 |
+
"loss": 1.1306,
|
18182 |
+
"step": 25390
|
18183 |
+
},
|
18184 |
+
{
|
18185 |
+
"epoch": 0.59,
|
18186 |
+
"grad_norm": 11.858842849731445,
|
18187 |
+
"learning_rate": 3.1186440677966107e-06,
|
18188 |
+
"loss": 1.3123,
|
18189 |
+
"step": 25400
|
18190 |
+
},
|
18191 |
+
{
|
18192 |
+
"epoch": 0.59,
|
18193 |
+
"grad_norm": 10.17744255065918,
|
18194 |
+
"learning_rate": 3.111864406779661e-06,
|
18195 |
+
"loss": 1.2951,
|
18196 |
+
"step": 25410
|
18197 |
+
},
|
18198 |
+
{
|
18199 |
+
"epoch": 0.59,
|
18200 |
+
"grad_norm": 5.539567947387695,
|
18201 |
+
"learning_rate": 3.105084745762712e-06,
|
18202 |
+
"loss": 1.369,
|
18203 |
+
"step": 25420
|
18204 |
+
},
|
18205 |
+
{
|
18206 |
+
"epoch": 0.59,
|
18207 |
+
"grad_norm": 6.66847038269043,
|
18208 |
+
"learning_rate": 3.098305084745763e-06,
|
18209 |
+
"loss": 1.1254,
|
18210 |
+
"step": 25430
|
18211 |
+
},
|
18212 |
+
{
|
18213 |
+
"epoch": 0.59,
|
18214 |
+
"grad_norm": 2.0820562839508057,
|
18215 |
+
"learning_rate": 3.091525423728814e-06,
|
18216 |
+
"loss": 1.0718,
|
18217 |
+
"step": 25440
|
18218 |
+
},
|
18219 |
+
{
|
18220 |
+
"epoch": 0.59,
|
18221 |
+
"grad_norm": 1.1296765804290771,
|
18222 |
+
"learning_rate": 3.0847457627118648e-06,
|
18223 |
+
"loss": 1.3201,
|
18224 |
+
"step": 25450
|
18225 |
+
},
|
18226 |
+
{
|
18227 |
+
"epoch": 0.59,
|
18228 |
+
"grad_norm": 6.302907943725586,
|
18229 |
+
"learning_rate": 3.0779661016949153e-06,
|
18230 |
+
"loss": 1.3956,
|
18231 |
+
"step": 25460
|
18232 |
+
},
|
18233 |
+
{
|
18234 |
+
"epoch": 0.59,
|
18235 |
+
"grad_norm": 7.594703674316406,
|
18236 |
+
"learning_rate": 3.0711864406779663e-06,
|
18237 |
+
"loss": 1.109,
|
18238 |
+
"step": 25470
|
18239 |
+
},
|
18240 |
+
{
|
18241 |
+
"epoch": 0.59,
|
18242 |
+
"grad_norm": 4.60818338394165,
|
18243 |
+
"learning_rate": 3.0644067796610173e-06,
|
18244 |
+
"loss": 1.194,
|
18245 |
+
"step": 25480
|
18246 |
+
},
|
18247 |
+
{
|
18248 |
+
"epoch": 0.59,
|
18249 |
+
"grad_norm": 9.426482200622559,
|
18250 |
+
"learning_rate": 3.0576271186440683e-06,
|
18251 |
+
"loss": 1.1175,
|
18252 |
+
"step": 25490
|
18253 |
+
},
|
18254 |
+
{
|
18255 |
+
"epoch": 0.59,
|
18256 |
+
"grad_norm": 1.6626375913619995,
|
18257 |
+
"learning_rate": 3.0508474576271192e-06,
|
18258 |
+
"loss": 1.241,
|
18259 |
+
"step": 25500
|
18260 |
+
},
|
18261 |
+
{
|
18262 |
+
"epoch": 0.59,
|
18263 |
+
"eval_loss": 0.998432993888855,
|
18264 |
+
"eval_runtime": 66.8613,
|
18265 |
+
"eval_samples_per_second": 14.956,
|
18266 |
+
"eval_steps_per_second": 14.956,
|
18267 |
+
"step": 25500
|
18268 |
+
},
|
18269 |
+
{
|
18270 |
+
"epoch": 0.59,
|
18271 |
+
"grad_norm": 5.7481255531311035,
|
18272 |
+
"learning_rate": 3.0440677966101694e-06,
|
18273 |
+
"loss": 1.0914,
|
18274 |
+
"step": 25510
|
18275 |
+
},
|
18276 |
+
{
|
18277 |
+
"epoch": 0.59,
|
18278 |
+
"grad_norm": 6.295405864715576,
|
18279 |
+
"learning_rate": 3.0372881355932203e-06,
|
18280 |
+
"loss": 1.2265,
|
18281 |
+
"step": 25520
|
18282 |
+
},
|
18283 |
+
{
|
18284 |
+
"epoch": 0.59,
|
18285 |
+
"grad_norm": 7.302733898162842,
|
18286 |
+
"learning_rate": 3.0305084745762713e-06,
|
18287 |
+
"loss": 1.2384,
|
18288 |
+
"step": 25530
|
18289 |
+
},
|
18290 |
+
{
|
18291 |
+
"epoch": 0.59,
|
18292 |
+
"grad_norm": 6.079695224761963,
|
18293 |
+
"learning_rate": 3.0237288135593223e-06,
|
18294 |
+
"loss": 1.3285,
|
18295 |
+
"step": 25540
|
18296 |
+
},
|
18297 |
+
{
|
18298 |
+
"epoch": 0.59,
|
18299 |
+
"grad_norm": 3.6032302379608154,
|
18300 |
+
"learning_rate": 3.0169491525423733e-06,
|
18301 |
+
"loss": 1.3156,
|
18302 |
+
"step": 25550
|
18303 |
+
},
|
18304 |
+
{
|
18305 |
+
"epoch": 0.59,
|
18306 |
+
"grad_norm": 4.823906421661377,
|
18307 |
+
"learning_rate": 3.010169491525424e-06,
|
18308 |
+
"loss": 0.8886,
|
18309 |
+
"step": 25560
|
18310 |
+
},
|
18311 |
+
{
|
18312 |
+
"epoch": 0.59,
|
18313 |
+
"grad_norm": 3.6529293060302734,
|
18314 |
+
"learning_rate": 3.003389830508475e-06,
|
18315 |
+
"loss": 1.2771,
|
18316 |
+
"step": 25570
|
18317 |
+
},
|
18318 |
+
{
|
18319 |
+
"epoch": 0.59,
|
18320 |
+
"grad_norm": 8.288748741149902,
|
18321 |
+
"learning_rate": 2.9966101694915258e-06,
|
18322 |
+
"loss": 1.1765,
|
18323 |
+
"step": 25580
|
18324 |
+
},
|
18325 |
+
{
|
18326 |
+
"epoch": 0.59,
|
18327 |
+
"grad_norm": 17.39524269104004,
|
18328 |
+
"learning_rate": 2.9898305084745768e-06,
|
18329 |
+
"loss": 1.3095,
|
18330 |
+
"step": 25590
|
18331 |
+
},
|
18332 |
+
{
|
18333 |
+
"epoch": 0.59,
|
18334 |
+
"grad_norm": 6.149013042449951,
|
18335 |
+
"learning_rate": 2.9830508474576277e-06,
|
18336 |
+
"loss": 0.9663,
|
18337 |
+
"step": 25600
|
18338 |
+
},
|
18339 |
+
{
|
18340 |
+
"epoch": 0.59,
|
18341 |
+
"grad_norm": 9.429570198059082,
|
18342 |
+
"learning_rate": 2.9762711864406783e-06,
|
18343 |
+
"loss": 1.1291,
|
18344 |
+
"step": 25610
|
18345 |
+
},
|
18346 |
+
{
|
18347 |
+
"epoch": 0.59,
|
18348 |
+
"grad_norm": 7.571442127227783,
|
18349 |
+
"learning_rate": 2.969491525423729e-06,
|
18350 |
+
"loss": 1.0891,
|
18351 |
+
"step": 25620
|
18352 |
+
},
|
18353 |
+
{
|
18354 |
+
"epoch": 0.59,
|
18355 |
+
"grad_norm": 8.435641288757324,
|
18356 |
+
"learning_rate": 2.96271186440678e-06,
|
18357 |
+
"loss": 1.207,
|
18358 |
+
"step": 25630
|
18359 |
+
},
|
18360 |
+
{
|
18361 |
+
"epoch": 0.59,
|
18362 |
+
"grad_norm": 5.694164276123047,
|
18363 |
+
"learning_rate": 2.955932203389831e-06,
|
18364 |
+
"loss": 1.2288,
|
18365 |
+
"step": 25640
|
18366 |
+
},
|
18367 |
+
{
|
18368 |
+
"epoch": 0.59,
|
18369 |
+
"grad_norm": 7.887094974517822,
|
18370 |
+
"learning_rate": 2.9491525423728818e-06,
|
18371 |
+
"loss": 1.2683,
|
18372 |
+
"step": 25650
|
18373 |
+
},
|
18374 |
+
{
|
18375 |
+
"epoch": 0.59,
|
18376 |
+
"grad_norm": 1.4776153564453125,
|
18377 |
+
"learning_rate": 2.9423728813559327e-06,
|
18378 |
+
"loss": 1.2045,
|
18379 |
+
"step": 25660
|
18380 |
+
},
|
18381 |
+
{
|
18382 |
+
"epoch": 0.59,
|
18383 |
+
"grad_norm": 22.860342025756836,
|
18384 |
+
"learning_rate": 2.935593220338983e-06,
|
18385 |
+
"loss": 1.1633,
|
18386 |
+
"step": 25670
|
18387 |
+
},
|
18388 |
+
{
|
18389 |
+
"epoch": 0.59,
|
18390 |
+
"grad_norm": 5.612551689147949,
|
18391 |
+
"learning_rate": 2.928813559322034e-06,
|
18392 |
+
"loss": 1.2065,
|
18393 |
+
"step": 25680
|
18394 |
+
},
|
18395 |
+
{
|
18396 |
+
"epoch": 0.59,
|
18397 |
+
"grad_norm": 4.532423973083496,
|
18398 |
+
"learning_rate": 2.922033898305085e-06,
|
18399 |
+
"loss": 1.2672,
|
18400 |
+
"step": 25690
|
18401 |
+
},
|
18402 |
+
{
|
18403 |
+
"epoch": 0.6,
|
18404 |
+
"grad_norm": 8.843605041503906,
|
18405 |
+
"learning_rate": 2.915254237288136e-06,
|
18406 |
+
"loss": 1.2333,
|
18407 |
+
"step": 25700
|
18408 |
+
},
|
18409 |
+
{
|
18410 |
+
"epoch": 0.6,
|
18411 |
+
"grad_norm": 7.5249481201171875,
|
18412 |
+
"learning_rate": 2.9084745762711868e-06,
|
18413 |
+
"loss": 1.2351,
|
18414 |
+
"step": 25710
|
18415 |
+
},
|
18416 |
+
{
|
18417 |
+
"epoch": 0.6,
|
18418 |
+
"grad_norm": 4.021324634552002,
|
18419 |
+
"learning_rate": 2.9016949152542373e-06,
|
18420 |
+
"loss": 1.302,
|
18421 |
+
"step": 25720
|
18422 |
+
},
|
18423 |
+
{
|
18424 |
+
"epoch": 0.6,
|
18425 |
+
"grad_norm": 12.174166679382324,
|
18426 |
+
"learning_rate": 2.8949152542372883e-06,
|
18427 |
+
"loss": 1.1858,
|
18428 |
+
"step": 25730
|
18429 |
+
},
|
18430 |
+
{
|
18431 |
+
"epoch": 0.6,
|
18432 |
+
"grad_norm": 9.587007522583008,
|
18433 |
+
"learning_rate": 2.8881355932203393e-06,
|
18434 |
+
"loss": 1.2249,
|
18435 |
+
"step": 25740
|
18436 |
+
},
|
18437 |
+
{
|
18438 |
+
"epoch": 0.6,
|
18439 |
+
"grad_norm": 9.73293399810791,
|
18440 |
+
"learning_rate": 2.8813559322033903e-06,
|
18441 |
+
"loss": 1.1867,
|
18442 |
+
"step": 25750
|
18443 |
+
},
|
18444 |
+
{
|
18445 |
+
"epoch": 0.6,
|
18446 |
+
"grad_norm": 17.562885284423828,
|
18447 |
+
"learning_rate": 2.8745762711864412e-06,
|
18448 |
+
"loss": 1.2162,
|
18449 |
+
"step": 25760
|
18450 |
+
},
|
18451 |
+
{
|
18452 |
+
"epoch": 0.6,
|
18453 |
+
"grad_norm": 2.855794668197632,
|
18454 |
+
"learning_rate": 2.8677966101694914e-06,
|
18455 |
+
"loss": 1.2218,
|
18456 |
+
"step": 25770
|
18457 |
+
},
|
18458 |
+
{
|
18459 |
+
"epoch": 0.6,
|
18460 |
+
"grad_norm": 8.89600658416748,
|
18461 |
+
"learning_rate": 2.8610169491525424e-06,
|
18462 |
+
"loss": 1.288,
|
18463 |
+
"step": 25780
|
18464 |
+
},
|
18465 |
+
{
|
18466 |
+
"epoch": 0.6,
|
18467 |
+
"grad_norm": 9.732436180114746,
|
18468 |
+
"learning_rate": 2.8542372881355933e-06,
|
18469 |
+
"loss": 1.4694,
|
18470 |
+
"step": 25790
|
18471 |
+
},
|
18472 |
+
{
|
18473 |
+
"epoch": 0.6,
|
18474 |
+
"grad_norm": 7.497152805328369,
|
18475 |
+
"learning_rate": 2.8474576271186443e-06,
|
18476 |
+
"loss": 1.1525,
|
18477 |
+
"step": 25800
|
18478 |
+
},
|
18479 |
+
{
|
18480 |
+
"epoch": 0.6,
|
18481 |
+
"grad_norm": 1.6418439149856567,
|
18482 |
+
"learning_rate": 2.8406779661016953e-06,
|
18483 |
+
"loss": 1.1873,
|
18484 |
+
"step": 25810
|
18485 |
+
},
|
18486 |
+
{
|
18487 |
+
"epoch": 0.6,
|
18488 |
+
"grad_norm": 9.75865650177002,
|
18489 |
+
"learning_rate": 2.833898305084746e-06,
|
18490 |
+
"loss": 1.3353,
|
18491 |
+
"step": 25820
|
18492 |
+
},
|
18493 |
+
{
|
18494 |
+
"epoch": 0.6,
|
18495 |
+
"grad_norm": 9.915677070617676,
|
18496 |
+
"learning_rate": 2.827118644067797e-06,
|
18497 |
+
"loss": 1.2384,
|
18498 |
+
"step": 25830
|
18499 |
+
},
|
18500 |
+
{
|
18501 |
+
"epoch": 0.6,
|
18502 |
+
"grad_norm": 11.455119132995605,
|
18503 |
+
"learning_rate": 2.820338983050848e-06,
|
18504 |
+
"loss": 1.0484,
|
18505 |
+
"step": 25840
|
18506 |
+
},
|
18507 |
+
{
|
18508 |
+
"epoch": 0.6,
|
18509 |
+
"grad_norm": 4.330808162689209,
|
18510 |
+
"learning_rate": 2.8135593220338988e-06,
|
18511 |
+
"loss": 1.1905,
|
18512 |
+
"step": 25850
|
18513 |
+
},
|
18514 |
+
{
|
18515 |
+
"epoch": 0.6,
|
18516 |
+
"grad_norm": 7.7616376876831055,
|
18517 |
+
"learning_rate": 2.8067796610169497e-06,
|
18518 |
+
"loss": 1.2721,
|
18519 |
+
"step": 25860
|
18520 |
+
},
|
18521 |
+
{
|
18522 |
+
"epoch": 0.6,
|
18523 |
+
"grad_norm": 9.885458946228027,
|
18524 |
+
"learning_rate": 2.8000000000000003e-06,
|
18525 |
+
"loss": 0.9887,
|
18526 |
+
"step": 25870
|
18527 |
+
},
|
18528 |
+
{
|
18529 |
+
"epoch": 0.6,
|
18530 |
+
"grad_norm": 7.448554039001465,
|
18531 |
+
"learning_rate": 2.793220338983051e-06,
|
18532 |
+
"loss": 1.1294,
|
18533 |
+
"step": 25880
|
18534 |
+
},
|
18535 |
+
{
|
18536 |
+
"epoch": 0.6,
|
18537 |
+
"grad_norm": 3.008898973464966,
|
18538 |
+
"learning_rate": 2.786440677966102e-06,
|
18539 |
+
"loss": 1.2489,
|
18540 |
+
"step": 25890
|
18541 |
+
},
|
18542 |
+
{
|
18543 |
+
"epoch": 0.6,
|
18544 |
+
"grad_norm": 8.211142539978027,
|
18545 |
+
"learning_rate": 2.779661016949153e-06,
|
18546 |
+
"loss": 1.2185,
|
18547 |
+
"step": 25900
|
18548 |
+
},
|
18549 |
+
{
|
18550 |
+
"epoch": 0.6,
|
18551 |
+
"grad_norm": 7.966844081878662,
|
18552 |
+
"learning_rate": 2.7728813559322038e-06,
|
18553 |
+
"loss": 1.3666,
|
18554 |
+
"step": 25910
|
18555 |
+
},
|
18556 |
+
{
|
18557 |
+
"epoch": 0.6,
|
18558 |
+
"grad_norm": 7.830656051635742,
|
18559 |
+
"learning_rate": 2.7661016949152548e-06,
|
18560 |
+
"loss": 1.3537,
|
18561 |
+
"step": 25920
|
18562 |
+
},
|
18563 |
+
{
|
18564 |
+
"epoch": 0.6,
|
18565 |
+
"grad_norm": 4.714284420013428,
|
18566 |
+
"learning_rate": 2.7593220338983053e-06,
|
18567 |
+
"loss": 1.1768,
|
18568 |
+
"step": 25930
|
18569 |
+
},
|
18570 |
+
{
|
18571 |
+
"epoch": 0.6,
|
18572 |
+
"grad_norm": 11.090384483337402,
|
18573 |
+
"learning_rate": 2.752542372881356e-06,
|
18574 |
+
"loss": 1.3252,
|
18575 |
+
"step": 25940
|
18576 |
+
},
|
18577 |
+
{
|
18578 |
+
"epoch": 0.6,
|
18579 |
+
"grad_norm": 7.201011657714844,
|
18580 |
+
"learning_rate": 2.745762711864407e-06,
|
18581 |
+
"loss": 1.3252,
|
18582 |
+
"step": 25950
|
18583 |
+
},
|
18584 |
+
{
|
18585 |
+
"epoch": 0.6,
|
18586 |
+
"grad_norm": 3.7255921363830566,
|
18587 |
+
"learning_rate": 2.738983050847458e-06,
|
18588 |
+
"loss": 1.2302,
|
18589 |
+
"step": 25960
|
18590 |
+
},
|
18591 |
+
{
|
18592 |
+
"epoch": 0.6,
|
18593 |
+
"grad_norm": 9.323339462280273,
|
18594 |
+
"learning_rate": 2.732203389830509e-06,
|
18595 |
+
"loss": 1.2951,
|
18596 |
+
"step": 25970
|
18597 |
+
},
|
18598 |
+
{
|
18599 |
+
"epoch": 0.6,
|
18600 |
+
"grad_norm": 24.529966354370117,
|
18601 |
+
"learning_rate": 2.7254237288135593e-06,
|
18602 |
+
"loss": 1.2912,
|
18603 |
+
"step": 25980
|
18604 |
+
},
|
18605 |
+
{
|
18606 |
+
"epoch": 0.6,
|
18607 |
+
"grad_norm": 4.739007472991943,
|
18608 |
+
"learning_rate": 2.7186440677966103e-06,
|
18609 |
+
"loss": 0.9912,
|
18610 |
+
"step": 25990
|
18611 |
+
},
|
18612 |
+
{
|
18613 |
+
"epoch": 0.6,
|
18614 |
+
"grad_norm": 8.324442863464355,
|
18615 |
+
"learning_rate": 2.7118644067796613e-06,
|
18616 |
+
"loss": 1.1363,
|
18617 |
+
"step": 26000
|
18618 |
+
},
|
18619 |
+
{
|
18620 |
+
"epoch": 0.6,
|
18621 |
+
"eval_loss": 0.9869400858879089,
|
18622 |
+
"eval_runtime": 66.847,
|
18623 |
+
"eval_samples_per_second": 14.96,
|
18624 |
+
"eval_steps_per_second": 14.96,
|
18625 |
+
"step": 26000
|
18626 |
+
},
|
18627 |
+
{
|
18628 |
+
"epoch": 0.6,
|
18629 |
+
"grad_norm": 8.513010025024414,
|
18630 |
+
"learning_rate": 2.7050847457627123e-06,
|
18631 |
+
"loss": 1.1538,
|
18632 |
+
"step": 26010
|
18633 |
+
},
|
18634 |
+
{
|
18635 |
+
"epoch": 0.6,
|
18636 |
+
"grad_norm": 3.9168806076049805,
|
18637 |
+
"learning_rate": 2.6983050847457633e-06,
|
18638 |
+
"loss": 1.0728,
|
18639 |
+
"step": 26020
|
18640 |
+
},
|
18641 |
+
{
|
18642 |
+
"epoch": 0.6,
|
18643 |
+
"grad_norm": 5.090580463409424,
|
18644 |
+
"learning_rate": 2.6915254237288134e-06,
|
18645 |
+
"loss": 1.2442,
|
18646 |
+
"step": 26030
|
18647 |
+
},
|
18648 |
+
{
|
18649 |
+
"epoch": 0.6,
|
18650 |
+
"grad_norm": 5.569304466247559,
|
18651 |
+
"learning_rate": 2.6847457627118644e-06,
|
18652 |
+
"loss": 0.9942,
|
18653 |
+
"step": 26040
|
18654 |
+
},
|
18655 |
+
{
|
18656 |
+
"epoch": 0.6,
|
18657 |
+
"grad_norm": 6.856686115264893,
|
18658 |
+
"learning_rate": 2.6779661016949153e-06,
|
18659 |
+
"loss": 1.1926,
|
18660 |
+
"step": 26050
|
18661 |
+
},
|
18662 |
+
{
|
18663 |
+
"epoch": 0.6,
|
18664 |
+
"grad_norm": 4.85548210144043,
|
18665 |
+
"learning_rate": 2.6711864406779663e-06,
|
18666 |
+
"loss": 1.1911,
|
18667 |
+
"step": 26060
|
18668 |
+
},
|
18669 |
+
{
|
18670 |
+
"epoch": 0.6,
|
18671 |
+
"grad_norm": 5.3112335205078125,
|
18672 |
+
"learning_rate": 2.6644067796610173e-06,
|
18673 |
+
"loss": 1.1381,
|
18674 |
+
"step": 26070
|
18675 |
+
},
|
18676 |
+
{
|
18677 |
+
"epoch": 0.6,
|
18678 |
+
"grad_norm": 5.764985084533691,
|
18679 |
+
"learning_rate": 2.657627118644068e-06,
|
18680 |
+
"loss": 1.1531,
|
18681 |
+
"step": 26080
|
18682 |
+
},
|
18683 |
+
{
|
18684 |
+
"epoch": 0.6,
|
18685 |
+
"grad_norm": 6.8243184089660645,
|
18686 |
+
"learning_rate": 2.650847457627119e-06,
|
18687 |
+
"loss": 1.1892,
|
18688 |
+
"step": 26090
|
18689 |
+
},
|
18690 |
+
{
|
18691 |
+
"epoch": 0.6,
|
18692 |
+
"grad_norm": 6.5991692543029785,
|
18693 |
+
"learning_rate": 2.64406779661017e-06,
|
18694 |
+
"loss": 1.0998,
|
18695 |
+
"step": 26100
|
18696 |
+
},
|
18697 |
+
{
|
18698 |
+
"epoch": 0.6,
|
18699 |
+
"grad_norm": 9.651345252990723,
|
18700 |
+
"learning_rate": 2.6372881355932208e-06,
|
18701 |
+
"loss": 1.2483,
|
18702 |
+
"step": 26110
|
18703 |
+
},
|
18704 |
+
{
|
18705 |
+
"epoch": 0.6,
|
18706 |
+
"grad_norm": 3.886197328567505,
|
18707 |
+
"learning_rate": 2.6305084745762718e-06,
|
18708 |
+
"loss": 1.1655,
|
18709 |
+
"step": 26120
|
18710 |
+
},
|
18711 |
+
{
|
18712 |
+
"epoch": 0.6,
|
18713 |
+
"grad_norm": 4.277401924133301,
|
18714 |
+
"learning_rate": 2.6237288135593223e-06,
|
18715 |
+
"loss": 1.2908,
|
18716 |
+
"step": 26130
|
18717 |
+
},
|
18718 |
+
{
|
18719 |
+
"epoch": 0.61,
|
18720 |
+
"grad_norm": 4.072823524475098,
|
18721 |
+
"learning_rate": 2.616949152542373e-06,
|
18722 |
+
"loss": 1.0785,
|
18723 |
+
"step": 26140
|
18724 |
+
},
|
18725 |
+
{
|
18726 |
+
"epoch": 0.61,
|
18727 |
+
"grad_norm": 4.426355838775635,
|
18728 |
+
"learning_rate": 2.610169491525424e-06,
|
18729 |
+
"loss": 1.157,
|
18730 |
+
"step": 26150
|
18731 |
+
},
|
18732 |
+
{
|
18733 |
+
"epoch": 0.61,
|
18734 |
+
"grad_norm": 1.601050853729248,
|
18735 |
+
"learning_rate": 2.603389830508475e-06,
|
18736 |
+
"loss": 1.3465,
|
18737 |
+
"step": 26160
|
18738 |
+
},
|
18739 |
+
{
|
18740 |
+
"epoch": 0.61,
|
18741 |
+
"grad_norm": 22.35835075378418,
|
18742 |
+
"learning_rate": 2.596610169491526e-06,
|
18743 |
+
"loss": 1.3287,
|
18744 |
+
"step": 26170
|
18745 |
+
},
|
18746 |
+
{
|
18747 |
+
"epoch": 0.61,
|
18748 |
+
"grad_norm": 4.574925422668457,
|
18749 |
+
"learning_rate": 2.5898305084745768e-06,
|
18750 |
+
"loss": 1.1298,
|
18751 |
+
"step": 26180
|
18752 |
+
},
|
18753 |
+
{
|
18754 |
+
"epoch": 0.61,
|
18755 |
+
"grad_norm": 6.630609512329102,
|
18756 |
+
"learning_rate": 2.5830508474576273e-06,
|
18757 |
+
"loss": 1.1522,
|
18758 |
+
"step": 26190
|
18759 |
+
},
|
18760 |
+
{
|
18761 |
+
"epoch": 0.61,
|
18762 |
+
"grad_norm": 5.038731098175049,
|
18763 |
+
"learning_rate": 2.576271186440678e-06,
|
18764 |
+
"loss": 1.1709,
|
18765 |
+
"step": 26200
|
18766 |
+
},
|
18767 |
+
{
|
18768 |
+
"epoch": 0.61,
|
18769 |
+
"grad_norm": 9.418252944946289,
|
18770 |
+
"learning_rate": 2.569491525423729e-06,
|
18771 |
+
"loss": 1.2216,
|
18772 |
+
"step": 26210
|
18773 |
+
},
|
18774 |
+
{
|
18775 |
+
"epoch": 0.61,
|
18776 |
+
"grad_norm": 16.634244918823242,
|
18777 |
+
"learning_rate": 2.56271186440678e-06,
|
18778 |
+
"loss": 1.2476,
|
18779 |
+
"step": 26220
|
18780 |
+
},
|
18781 |
+
{
|
18782 |
+
"epoch": 0.61,
|
18783 |
+
"grad_norm": 6.329607009887695,
|
18784 |
+
"learning_rate": 2.555932203389831e-06,
|
18785 |
+
"loss": 1.3251,
|
18786 |
+
"step": 26230
|
18787 |
+
},
|
18788 |
+
{
|
18789 |
+
"epoch": 0.61,
|
18790 |
+
"grad_norm": 7.770134449005127,
|
18791 |
+
"learning_rate": 2.5491525423728814e-06,
|
18792 |
+
"loss": 1.3357,
|
18793 |
+
"step": 26240
|
18794 |
+
},
|
18795 |
+
{
|
18796 |
+
"epoch": 0.61,
|
18797 |
+
"grad_norm": 8.234681129455566,
|
18798 |
+
"learning_rate": 2.5423728813559323e-06,
|
18799 |
+
"loss": 1.3181,
|
18800 |
+
"step": 26250
|
18801 |
+
},
|
18802 |
+
{
|
18803 |
+
"epoch": 0.61,
|
18804 |
+
"grad_norm": 4.855250358581543,
|
18805 |
+
"learning_rate": 2.5355932203389833e-06,
|
18806 |
+
"loss": 1.3914,
|
18807 |
+
"step": 26260
|
18808 |
+
},
|
18809 |
+
{
|
18810 |
+
"epoch": 0.61,
|
18811 |
+
"grad_norm": 9.435198783874512,
|
18812 |
+
"learning_rate": 2.5288135593220343e-06,
|
18813 |
+
"loss": 1.225,
|
18814 |
+
"step": 26270
|
18815 |
+
},
|
18816 |
+
{
|
18817 |
+
"epoch": 0.61,
|
18818 |
+
"grad_norm": 2.1608779430389404,
|
18819 |
+
"learning_rate": 2.5220338983050853e-06,
|
18820 |
+
"loss": 1.4329,
|
18821 |
+
"step": 26280
|
18822 |
+
},
|
18823 |
+
{
|
18824 |
+
"epoch": 0.61,
|
18825 |
+
"grad_norm": 5.846624851226807,
|
18826 |
+
"learning_rate": 2.5152542372881354e-06,
|
18827 |
+
"loss": 1.2809,
|
18828 |
+
"step": 26290
|
18829 |
+
},
|
18830 |
+
{
|
18831 |
+
"epoch": 0.61,
|
18832 |
+
"grad_norm": 2.737983465194702,
|
18833 |
+
"learning_rate": 2.5084745762711864e-06,
|
18834 |
+
"loss": 1.1873,
|
18835 |
+
"step": 26300
|
18836 |
+
},
|
18837 |
+
{
|
18838 |
+
"epoch": 0.61,
|
18839 |
+
"grad_norm": 2.908550977706909,
|
18840 |
+
"learning_rate": 2.5016949152542374e-06,
|
18841 |
+
"loss": 1.1736,
|
18842 |
+
"step": 26310
|
18843 |
+
},
|
18844 |
+
{
|
18845 |
+
"epoch": 0.61,
|
18846 |
+
"grad_norm": 36.61166000366211,
|
18847 |
+
"learning_rate": 2.4949152542372883e-06,
|
18848 |
+
"loss": 1.1875,
|
18849 |
+
"step": 26320
|
18850 |
+
},
|
18851 |
+
{
|
18852 |
+
"epoch": 0.61,
|
18853 |
+
"grad_norm": 20.149581909179688,
|
18854 |
+
"learning_rate": 2.488135593220339e-06,
|
18855 |
+
"loss": 1.3924,
|
18856 |
+
"step": 26330
|
18857 |
+
},
|
18858 |
+
{
|
18859 |
+
"epoch": 0.61,
|
18860 |
+
"grad_norm": 8.97331428527832,
|
18861 |
+
"learning_rate": 2.48135593220339e-06,
|
18862 |
+
"loss": 1.0913,
|
18863 |
+
"step": 26340
|
18864 |
+
},
|
18865 |
+
{
|
18866 |
+
"epoch": 0.61,
|
18867 |
+
"grad_norm": 6.3798441886901855,
|
18868 |
+
"learning_rate": 2.474576271186441e-06,
|
18869 |
+
"loss": 1.0683,
|
18870 |
+
"step": 26350
|
18871 |
+
},
|
18872 |
+
{
|
18873 |
+
"epoch": 0.61,
|
18874 |
+
"grad_norm": 13.193538665771484,
|
18875 |
+
"learning_rate": 2.467796610169492e-06,
|
18876 |
+
"loss": 1.0535,
|
18877 |
+
"step": 26360
|
18878 |
+
},
|
18879 |
+
{
|
18880 |
+
"epoch": 0.61,
|
18881 |
+
"grad_norm": 14.773574829101562,
|
18882 |
+
"learning_rate": 2.461016949152543e-06,
|
18883 |
+
"loss": 1.0665,
|
18884 |
+
"step": 26370
|
18885 |
+
},
|
18886 |
+
{
|
18887 |
+
"epoch": 0.61,
|
18888 |
+
"grad_norm": 3.1088144779205322,
|
18889 |
+
"learning_rate": 2.4542372881355933e-06,
|
18890 |
+
"loss": 1.3309,
|
18891 |
+
"step": 26380
|
18892 |
+
},
|
18893 |
+
{
|
18894 |
+
"epoch": 0.61,
|
18895 |
+
"grad_norm": 3.1904797554016113,
|
18896 |
+
"learning_rate": 2.4474576271186443e-06,
|
18897 |
+
"loss": 1.2429,
|
18898 |
+
"step": 26390
|
18899 |
+
},
|
18900 |
+
{
|
18901 |
+
"epoch": 0.61,
|
18902 |
+
"grad_norm": 3.5016331672668457,
|
18903 |
+
"learning_rate": 2.4406779661016953e-06,
|
18904 |
+
"loss": 1.2189,
|
18905 |
+
"step": 26400
|
18906 |
+
},
|
18907 |
+
{
|
18908 |
+
"epoch": 0.61,
|
18909 |
+
"grad_norm": 6.315566539764404,
|
18910 |
+
"learning_rate": 2.433898305084746e-06,
|
18911 |
+
"loss": 1.2291,
|
18912 |
+
"step": 26410
|
18913 |
+
},
|
18914 |
+
{
|
18915 |
+
"epoch": 0.61,
|
18916 |
+
"grad_norm": 14.041393280029297,
|
18917 |
+
"learning_rate": 2.427118644067797e-06,
|
18918 |
+
"loss": 1.1644,
|
18919 |
+
"step": 26420
|
18920 |
+
},
|
18921 |
+
{
|
18922 |
+
"epoch": 0.61,
|
18923 |
+
"grad_norm": 20.225662231445312,
|
18924 |
+
"learning_rate": 2.4203389830508474e-06,
|
18925 |
+
"loss": 1.2319,
|
18926 |
+
"step": 26430
|
18927 |
+
},
|
18928 |
+
{
|
18929 |
+
"epoch": 0.61,
|
18930 |
+
"grad_norm": 3.7186009883880615,
|
18931 |
+
"learning_rate": 2.4135593220338984e-06,
|
18932 |
+
"loss": 1.2453,
|
18933 |
+
"step": 26440
|
18934 |
+
},
|
18935 |
+
{
|
18936 |
+
"epoch": 0.61,
|
18937 |
+
"grad_norm": 2.4279704093933105,
|
18938 |
+
"learning_rate": 2.4067796610169493e-06,
|
18939 |
+
"loss": 1.0585,
|
18940 |
+
"step": 26450
|
18941 |
+
},
|
18942 |
+
{
|
18943 |
+
"epoch": 0.61,
|
18944 |
+
"grad_norm": 4.853421211242676,
|
18945 |
+
"learning_rate": 2.4000000000000003e-06,
|
18946 |
+
"loss": 1.2464,
|
18947 |
+
"step": 26460
|
18948 |
+
},
|
18949 |
+
{
|
18950 |
+
"epoch": 0.61,
|
18951 |
+
"grad_norm": 4.436877250671387,
|
18952 |
+
"learning_rate": 2.393220338983051e-06,
|
18953 |
+
"loss": 1.3915,
|
18954 |
+
"step": 26470
|
18955 |
+
},
|
18956 |
+
{
|
18957 |
+
"epoch": 0.61,
|
18958 |
+
"grad_norm": 3.7520599365234375,
|
18959 |
+
"learning_rate": 2.386440677966102e-06,
|
18960 |
+
"loss": 1.3439,
|
18961 |
+
"step": 26480
|
18962 |
+
},
|
18963 |
+
{
|
18964 |
+
"epoch": 0.61,
|
18965 |
+
"grad_norm": 1.7208768129348755,
|
18966 |
+
"learning_rate": 2.379661016949153e-06,
|
18967 |
+
"loss": 1.2509,
|
18968 |
+
"step": 26490
|
18969 |
+
},
|
18970 |
+
{
|
18971 |
+
"epoch": 0.61,
|
18972 |
+
"grad_norm": 2.465243101119995,
|
18973 |
+
"learning_rate": 2.372881355932204e-06,
|
18974 |
+
"loss": 1.2785,
|
18975 |
+
"step": 26500
|
18976 |
+
},
|
18977 |
+
{
|
18978 |
+
"epoch": 0.61,
|
18979 |
+
"eval_loss": 0.9529283046722412,
|
18980 |
+
"eval_runtime": 66.4596,
|
18981 |
+
"eval_samples_per_second": 15.047,
|
18982 |
+
"eval_steps_per_second": 15.047,
|
18983 |
+
"step": 26500
|
18984 |
+
},
|
18985 |
+
{
|
18986 |
+
"epoch": 0.61,
|
18987 |
+
"grad_norm": 16.25847816467285,
|
18988 |
+
"learning_rate": 2.3661016949152544e-06,
|
18989 |
+
"loss": 1.2935,
|
18990 |
+
"step": 26510
|
18991 |
+
},
|
18992 |
+
{
|
18993 |
+
"epoch": 0.61,
|
18994 |
+
"grad_norm": 5.584811687469482,
|
18995 |
+
"learning_rate": 2.3593220338983053e-06,
|
18996 |
+
"loss": 1.1677,
|
18997 |
+
"step": 26520
|
18998 |
+
},
|
18999 |
+
{
|
19000 |
+
"epoch": 0.61,
|
19001 |
+
"grad_norm": 5.604238033294678,
|
19002 |
+
"learning_rate": 2.3525423728813563e-06,
|
19003 |
+
"loss": 1.1095,
|
19004 |
+
"step": 26530
|
19005 |
+
},
|
19006 |
+
{
|
19007 |
+
"epoch": 0.61,
|
19008 |
+
"grad_norm": 7.715309143066406,
|
19009 |
+
"learning_rate": 2.345762711864407e-06,
|
19010 |
+
"loss": 1.3044,
|
19011 |
+
"step": 26540
|
19012 |
+
},
|
19013 |
+
{
|
19014 |
+
"epoch": 0.61,
|
19015 |
+
"grad_norm": 1.3220744132995605,
|
19016 |
+
"learning_rate": 2.338983050847458e-06,
|
19017 |
+
"loss": 1.1298,
|
19018 |
+
"step": 26550
|
19019 |
+
},
|
19020 |
+
{
|
19021 |
+
"epoch": 0.61,
|
19022 |
+
"grad_norm": 2.8409392833709717,
|
19023 |
+
"learning_rate": 2.3322033898305084e-06,
|
19024 |
+
"loss": 1.2565,
|
19025 |
+
"step": 26560
|
19026 |
+
},
|
19027 |
+
{
|
19028 |
+
"epoch": 0.62,
|
19029 |
+
"grad_norm": 3.3840484619140625,
|
19030 |
+
"learning_rate": 2.3254237288135594e-06,
|
19031 |
+
"loss": 1.2529,
|
19032 |
+
"step": 26570
|
19033 |
+
},
|
19034 |
+
{
|
19035 |
+
"epoch": 0.62,
|
19036 |
+
"grad_norm": 2.952516555786133,
|
19037 |
+
"learning_rate": 2.3186440677966103e-06,
|
19038 |
+
"loss": 1.3135,
|
19039 |
+
"step": 26580
|
19040 |
+
},
|
19041 |
+
{
|
19042 |
+
"epoch": 0.62,
|
19043 |
+
"grad_norm": 6.298173904418945,
|
19044 |
+
"learning_rate": 2.3118644067796613e-06,
|
19045 |
+
"loss": 1.2517,
|
19046 |
+
"step": 26590
|
19047 |
+
},
|
19048 |
+
{
|
19049 |
+
"epoch": 0.62,
|
19050 |
+
"grad_norm": 17.91318702697754,
|
19051 |
+
"learning_rate": 2.305084745762712e-06,
|
19052 |
+
"loss": 1.1782,
|
19053 |
+
"step": 26600
|
19054 |
+
},
|
19055 |
+
{
|
19056 |
+
"epoch": 0.62,
|
19057 |
+
"grad_norm": 9.400506973266602,
|
19058 |
+
"learning_rate": 2.298305084745763e-06,
|
19059 |
+
"loss": 1.3047,
|
19060 |
+
"step": 26610
|
19061 |
+
},
|
19062 |
+
{
|
19063 |
+
"epoch": 0.62,
|
19064 |
+
"grad_norm": 8.402453422546387,
|
19065 |
+
"learning_rate": 2.291525423728814e-06,
|
19066 |
+
"loss": 1.3617,
|
19067 |
+
"step": 26620
|
19068 |
+
},
|
19069 |
+
{
|
19070 |
+
"epoch": 0.62,
|
19071 |
+
"grad_norm": 6.365287780761719,
|
19072 |
+
"learning_rate": 2.284745762711865e-06,
|
19073 |
+
"loss": 1.2257,
|
19074 |
+
"step": 26630
|
19075 |
+
},
|
19076 |
+
{
|
19077 |
+
"epoch": 0.62,
|
19078 |
+
"grad_norm": 2.8725926876068115,
|
19079 |
+
"learning_rate": 2.2779661016949154e-06,
|
19080 |
+
"loss": 1.0295,
|
19081 |
+
"step": 26640
|
19082 |
+
},
|
19083 |
+
{
|
19084 |
+
"epoch": 0.62,
|
19085 |
+
"grad_norm": 13.41898250579834,
|
19086 |
+
"learning_rate": 2.2711864406779663e-06,
|
19087 |
+
"loss": 1.1358,
|
19088 |
+
"step": 26650
|
19089 |
+
},
|
19090 |
+
{
|
19091 |
+
"epoch": 0.62,
|
19092 |
+
"grad_norm": 3.4462170600891113,
|
19093 |
+
"learning_rate": 2.2644067796610173e-06,
|
19094 |
+
"loss": 1.2295,
|
19095 |
+
"step": 26660
|
19096 |
+
},
|
19097 |
+
{
|
19098 |
+
"epoch": 0.62,
|
19099 |
+
"grad_norm": 1.85061514377594,
|
19100 |
+
"learning_rate": 2.257627118644068e-06,
|
19101 |
+
"loss": 1.2474,
|
19102 |
+
"step": 26670
|
19103 |
+
},
|
19104 |
+
{
|
19105 |
+
"epoch": 0.62,
|
19106 |
+
"grad_norm": 4.921010971069336,
|
19107 |
+
"learning_rate": 2.250847457627119e-06,
|
19108 |
+
"loss": 1.4324,
|
19109 |
+
"step": 26680
|
19110 |
+
},
|
19111 |
+
{
|
19112 |
+
"epoch": 0.62,
|
19113 |
+
"grad_norm": 5.649210453033447,
|
19114 |
+
"learning_rate": 2.2440677966101694e-06,
|
19115 |
+
"loss": 1.2167,
|
19116 |
+
"step": 26690
|
19117 |
+
},
|
19118 |
+
{
|
19119 |
+
"epoch": 0.62,
|
19120 |
+
"grad_norm": 13.460455894470215,
|
19121 |
+
"learning_rate": 2.2372881355932204e-06,
|
19122 |
+
"loss": 1.1124,
|
19123 |
+
"step": 26700
|
19124 |
+
},
|
19125 |
+
{
|
19126 |
+
"epoch": 0.62,
|
19127 |
+
"grad_norm": 4.873124599456787,
|
19128 |
+
"learning_rate": 2.2305084745762714e-06,
|
19129 |
+
"loss": 1.2533,
|
19130 |
+
"step": 26710
|
19131 |
+
},
|
19132 |
+
{
|
19133 |
+
"epoch": 0.62,
|
19134 |
+
"grad_norm": 4.260035514831543,
|
19135 |
+
"learning_rate": 2.2237288135593223e-06,
|
19136 |
+
"loss": 1.1214,
|
19137 |
+
"step": 26720
|
19138 |
+
},
|
19139 |
+
{
|
19140 |
+
"epoch": 0.62,
|
19141 |
+
"grad_norm": 9.646221160888672,
|
19142 |
+
"learning_rate": 2.216949152542373e-06,
|
19143 |
+
"loss": 1.3115,
|
19144 |
+
"step": 26730
|
19145 |
+
},
|
19146 |
+
{
|
19147 |
+
"epoch": 0.62,
|
19148 |
+
"grad_norm": 2.029663324356079,
|
19149 |
+
"learning_rate": 2.210169491525424e-06,
|
19150 |
+
"loss": 1.1277,
|
19151 |
+
"step": 26740
|
19152 |
+
},
|
19153 |
+
{
|
19154 |
+
"epoch": 0.62,
|
19155 |
+
"grad_norm": 2.2031261920928955,
|
19156 |
+
"learning_rate": 2.203389830508475e-06,
|
19157 |
+
"loss": 1.275,
|
19158 |
+
"step": 26750
|
19159 |
+
},
|
19160 |
+
{
|
19161 |
+
"epoch": 0.62,
|
19162 |
+
"grad_norm": 7.809500217437744,
|
19163 |
+
"learning_rate": 2.196610169491526e-06,
|
19164 |
+
"loss": 1.4312,
|
19165 |
+
"step": 26760
|
19166 |
+
},
|
19167 |
+
{
|
19168 |
+
"epoch": 0.62,
|
19169 |
+
"grad_norm": 3.175096273422241,
|
19170 |
+
"learning_rate": 2.1898305084745764e-06,
|
19171 |
+
"loss": 1.3533,
|
19172 |
+
"step": 26770
|
19173 |
+
},
|
19174 |
+
{
|
19175 |
+
"epoch": 0.62,
|
19176 |
+
"grad_norm": 7.439695835113525,
|
19177 |
+
"learning_rate": 2.1830508474576273e-06,
|
19178 |
+
"loss": 1.2436,
|
19179 |
+
"step": 26780
|
19180 |
+
},
|
19181 |
+
{
|
19182 |
+
"epoch": 0.62,
|
19183 |
+
"grad_norm": 6.698858737945557,
|
19184 |
+
"learning_rate": 2.1762711864406783e-06,
|
19185 |
+
"loss": 1.3145,
|
19186 |
+
"step": 26790
|
19187 |
+
},
|
19188 |
+
{
|
19189 |
+
"epoch": 0.62,
|
19190 |
+
"grad_norm": 4.412804126739502,
|
19191 |
+
"learning_rate": 2.169491525423729e-06,
|
19192 |
+
"loss": 1.185,
|
19193 |
+
"step": 26800
|
19194 |
+
},
|
19195 |
+
{
|
19196 |
+
"epoch": 0.62,
|
19197 |
+
"grad_norm": 9.170161247253418,
|
19198 |
+
"learning_rate": 2.16271186440678e-06,
|
19199 |
+
"loss": 1.1868,
|
19200 |
+
"step": 26810
|
19201 |
+
},
|
19202 |
+
{
|
19203 |
+
"epoch": 0.62,
|
19204 |
+
"grad_norm": 21.11345672607422,
|
19205 |
+
"learning_rate": 2.1559322033898304e-06,
|
19206 |
+
"loss": 1.2738,
|
19207 |
+
"step": 26820
|
19208 |
+
},
|
19209 |
+
{
|
19210 |
+
"epoch": 0.62,
|
19211 |
+
"grad_norm": 52.31527328491211,
|
19212 |
+
"learning_rate": 2.1491525423728814e-06,
|
19213 |
+
"loss": 1.1694,
|
19214 |
+
"step": 26830
|
19215 |
+
},
|
19216 |
+
{
|
19217 |
+
"epoch": 0.62,
|
19218 |
+
"grad_norm": 4.136544704437256,
|
19219 |
+
"learning_rate": 2.1423728813559324e-06,
|
19220 |
+
"loss": 1.0897,
|
19221 |
+
"step": 26840
|
19222 |
+
},
|
19223 |
+
{
|
19224 |
+
"epoch": 0.62,
|
19225 |
+
"grad_norm": 3.128509759902954,
|
19226 |
+
"learning_rate": 2.1355932203389833e-06,
|
19227 |
+
"loss": 1.1818,
|
19228 |
+
"step": 26850
|
19229 |
+
},
|
19230 |
+
{
|
19231 |
+
"epoch": 0.62,
|
19232 |
+
"grad_norm": 12.374292373657227,
|
19233 |
+
"learning_rate": 2.128813559322034e-06,
|
19234 |
+
"loss": 1.3163,
|
19235 |
+
"step": 26860
|
19236 |
+
},
|
19237 |
+
{
|
19238 |
+
"epoch": 0.62,
|
19239 |
+
"grad_norm": 8.347363471984863,
|
19240 |
+
"learning_rate": 2.122033898305085e-06,
|
19241 |
+
"loss": 1.1177,
|
19242 |
+
"step": 26870
|
19243 |
+
},
|
19244 |
+
{
|
19245 |
+
"epoch": 0.62,
|
19246 |
+
"grad_norm": 17.513408660888672,
|
19247 |
+
"learning_rate": 2.115254237288136e-06,
|
19248 |
+
"loss": 1.2128,
|
19249 |
+
"step": 26880
|
19250 |
+
},
|
19251 |
+
{
|
19252 |
+
"epoch": 0.62,
|
19253 |
+
"grad_norm": 5.053252220153809,
|
19254 |
+
"learning_rate": 2.108474576271187e-06,
|
19255 |
+
"loss": 1.4111,
|
19256 |
+
"step": 26890
|
19257 |
+
},
|
19258 |
+
{
|
19259 |
+
"epoch": 0.62,
|
19260 |
+
"grad_norm": 2.5421042442321777,
|
19261 |
+
"learning_rate": 2.1016949152542374e-06,
|
19262 |
+
"loss": 1.3909,
|
19263 |
+
"step": 26900
|
19264 |
+
},
|
19265 |
+
{
|
19266 |
+
"epoch": 0.62,
|
19267 |
+
"grad_norm": 14.27386474609375,
|
19268 |
+
"learning_rate": 2.0949152542372883e-06,
|
19269 |
+
"loss": 1.3056,
|
19270 |
+
"step": 26910
|
19271 |
+
},
|
19272 |
+
{
|
19273 |
+
"epoch": 0.62,
|
19274 |
+
"grad_norm": 3.5711681842803955,
|
19275 |
+
"learning_rate": 2.0881355932203393e-06,
|
19276 |
+
"loss": 1.1021,
|
19277 |
+
"step": 26920
|
19278 |
+
},
|
19279 |
+
{
|
19280 |
+
"epoch": 0.62,
|
19281 |
+
"grad_norm": 7.276656627655029,
|
19282 |
+
"learning_rate": 2.08135593220339e-06,
|
19283 |
+
"loss": 1.0432,
|
19284 |
+
"step": 26930
|
19285 |
+
},
|
19286 |
+
{
|
19287 |
+
"epoch": 0.62,
|
19288 |
+
"grad_norm": 2.4546101093292236,
|
19289 |
+
"learning_rate": 2.074576271186441e-06,
|
19290 |
+
"loss": 1.3391,
|
19291 |
+
"step": 26940
|
19292 |
+
},
|
19293 |
+
{
|
19294 |
+
"epoch": 0.62,
|
19295 |
+
"grad_norm": 6.01200008392334,
|
19296 |
+
"learning_rate": 2.0677966101694914e-06,
|
19297 |
+
"loss": 1.1382,
|
19298 |
+
"step": 26950
|
19299 |
+
},
|
19300 |
+
{
|
19301 |
+
"epoch": 0.62,
|
19302 |
+
"grad_norm": 6.168879508972168,
|
19303 |
+
"learning_rate": 2.0610169491525424e-06,
|
19304 |
+
"loss": 1.1858,
|
19305 |
+
"step": 26960
|
19306 |
+
},
|
19307 |
+
{
|
19308 |
+
"epoch": 0.62,
|
19309 |
+
"grad_norm": 8.381854057312012,
|
19310 |
+
"learning_rate": 2.0542372881355934e-06,
|
19311 |
+
"loss": 1.2046,
|
19312 |
+
"step": 26970
|
19313 |
+
},
|
19314 |
+
{
|
19315 |
+
"epoch": 0.62,
|
19316 |
+
"grad_norm": 3.6406259536743164,
|
19317 |
+
"learning_rate": 2.0474576271186443e-06,
|
19318 |
+
"loss": 1.2098,
|
19319 |
+
"step": 26980
|
19320 |
+
},
|
19321 |
+
{
|
19322 |
+
"epoch": 0.62,
|
19323 |
+
"grad_norm": 7.411496162414551,
|
19324 |
+
"learning_rate": 2.0406779661016953e-06,
|
19325 |
+
"loss": 1.2822,
|
19326 |
+
"step": 26990
|
19327 |
+
},
|
19328 |
+
{
|
19329 |
+
"epoch": 0.63,
|
19330 |
+
"grad_norm": 4.259427547454834,
|
19331 |
+
"learning_rate": 2.033898305084746e-06,
|
19332 |
+
"loss": 1.0367,
|
19333 |
+
"step": 27000
|
19334 |
+
},
|
19335 |
+
{
|
19336 |
+
"epoch": 0.63,
|
19337 |
+
"eval_loss": 1.0261023044586182,
|
19338 |
+
"eval_runtime": 66.6287,
|
19339 |
+
"eval_samples_per_second": 15.009,
|
19340 |
+
"eval_steps_per_second": 15.009,
|
19341 |
+
"step": 27000
|
19342 |
+
},
|
19343 |
+
{
|
19344 |
+
"epoch": 0.63,
|
19345 |
+
"grad_norm": 1.8957105875015259,
|
19346 |
+
"learning_rate": 2.027118644067797e-06,
|
19347 |
+
"loss": 1.1367,
|
19348 |
+
"step": 27010
|
19349 |
+
},
|
19350 |
+
{
|
19351 |
+
"epoch": 0.63,
|
19352 |
+
"grad_norm": 8.313997268676758,
|
19353 |
+
"learning_rate": 2.020338983050848e-06,
|
19354 |
+
"loss": 1.1379,
|
19355 |
+
"step": 27020
|
19356 |
+
},
|
19357 |
+
{
|
19358 |
+
"epoch": 0.63,
|
19359 |
+
"grad_norm": 5.363570690155029,
|
19360 |
+
"learning_rate": 2.0135593220338984e-06,
|
19361 |
+
"loss": 1.1295,
|
19362 |
+
"step": 27030
|
19363 |
+
},
|
19364 |
+
{
|
19365 |
+
"epoch": 0.63,
|
19366 |
+
"grad_norm": 15.122271537780762,
|
19367 |
+
"learning_rate": 2.0067796610169494e-06,
|
19368 |
+
"loss": 1.1451,
|
19369 |
+
"step": 27040
|
19370 |
+
},
|
19371 |
+
{
|
19372 |
+
"epoch": 0.63,
|
19373 |
+
"grad_norm": 4.8829755783081055,
|
19374 |
+
"learning_rate": 2.0000000000000003e-06,
|
19375 |
+
"loss": 1.411,
|
19376 |
+
"step": 27050
|
19377 |
+
},
|
19378 |
+
{
|
19379 |
+
"epoch": 0.63,
|
19380 |
+
"grad_norm": 9.953533172607422,
|
19381 |
+
"learning_rate": 1.993220338983051e-06,
|
19382 |
+
"loss": 1.0443,
|
19383 |
+
"step": 27060
|
19384 |
+
},
|
19385 |
+
{
|
19386 |
+
"epoch": 0.63,
|
19387 |
+
"grad_norm": 6.28991174697876,
|
19388 |
+
"learning_rate": 1.986440677966102e-06,
|
19389 |
+
"loss": 1.0607,
|
19390 |
+
"step": 27070
|
19391 |
+
},
|
19392 |
+
{
|
19393 |
+
"epoch": 0.63,
|
19394 |
+
"grad_norm": 8.279200553894043,
|
19395 |
+
"learning_rate": 1.9796610169491524e-06,
|
19396 |
+
"loss": 1.3226,
|
19397 |
+
"step": 27080
|
19398 |
+
},
|
19399 |
+
{
|
19400 |
+
"epoch": 0.63,
|
19401 |
+
"grad_norm": 6.716332912445068,
|
19402 |
+
"learning_rate": 1.9728813559322034e-06,
|
19403 |
+
"loss": 1.1337,
|
19404 |
+
"step": 27090
|
19405 |
+
},
|
19406 |
+
{
|
19407 |
+
"epoch": 0.63,
|
19408 |
+
"grad_norm": 8.114790916442871,
|
19409 |
+
"learning_rate": 1.9661016949152544e-06,
|
19410 |
+
"loss": 1.1927,
|
19411 |
+
"step": 27100
|
19412 |
+
},
|
19413 |
+
{
|
19414 |
+
"epoch": 0.63,
|
19415 |
+
"grad_norm": 8.28928279876709,
|
19416 |
+
"learning_rate": 1.9593220338983053e-06,
|
19417 |
+
"loss": 1.2793,
|
19418 |
+
"step": 27110
|
19419 |
+
},
|
19420 |
+
{
|
19421 |
+
"epoch": 0.63,
|
19422 |
+
"grad_norm": 5.299505710601807,
|
19423 |
+
"learning_rate": 1.9525423728813563e-06,
|
19424 |
+
"loss": 1.1913,
|
19425 |
+
"step": 27120
|
19426 |
+
},
|
19427 |
+
{
|
19428 |
+
"epoch": 0.63,
|
19429 |
+
"grad_norm": 5.9343061447143555,
|
19430 |
+
"learning_rate": 1.945762711864407e-06,
|
19431 |
+
"loss": 1.162,
|
19432 |
+
"step": 27130
|
19433 |
+
},
|
19434 |
+
{
|
19435 |
+
"epoch": 0.63,
|
19436 |
+
"grad_norm": 5.81511926651001,
|
19437 |
+
"learning_rate": 1.938983050847458e-06,
|
19438 |
+
"loss": 1.3911,
|
19439 |
+
"step": 27140
|
19440 |
+
},
|
19441 |
+
{
|
19442 |
+
"epoch": 0.63,
|
19443 |
+
"grad_norm": 2.0106678009033203,
|
19444 |
+
"learning_rate": 1.932203389830509e-06,
|
19445 |
+
"loss": 1.3768,
|
19446 |
+
"step": 27150
|
19447 |
+
},
|
19448 |
+
{
|
19449 |
+
"epoch": 0.63,
|
19450 |
+
"grad_norm": 5.970021724700928,
|
19451 |
+
"learning_rate": 1.9254237288135594e-06,
|
19452 |
+
"loss": 1.2548,
|
19453 |
+
"step": 27160
|
19454 |
+
},
|
19455 |
+
{
|
19456 |
+
"epoch": 0.63,
|
19457 |
+
"grad_norm": 12.892632484436035,
|
19458 |
+
"learning_rate": 1.9186440677966104e-06,
|
19459 |
+
"loss": 1.3527,
|
19460 |
+
"step": 27170
|
19461 |
+
},
|
19462 |
+
{
|
19463 |
+
"epoch": 0.63,
|
19464 |
+
"grad_norm": 12.3954496383667,
|
19465 |
+
"learning_rate": 1.9118644067796613e-06,
|
19466 |
+
"loss": 1.2633,
|
19467 |
+
"step": 27180
|
19468 |
+
},
|
19469 |
+
{
|
19470 |
+
"epoch": 0.63,
|
19471 |
+
"grad_norm": 9.317601203918457,
|
19472 |
+
"learning_rate": 1.9050847457627119e-06,
|
19473 |
+
"loss": 1.1907,
|
19474 |
+
"step": 27190
|
19475 |
+
},
|
19476 |
+
{
|
19477 |
+
"epoch": 0.63,
|
19478 |
+
"grad_norm": 2.2695984840393066,
|
19479 |
+
"learning_rate": 1.8983050847457629e-06,
|
19480 |
+
"loss": 1.2513,
|
19481 |
+
"step": 27200
|
19482 |
+
},
|
19483 |
+
{
|
19484 |
+
"epoch": 0.63,
|
19485 |
+
"grad_norm": 5.8285908699035645,
|
19486 |
+
"learning_rate": 1.8915254237288136e-06,
|
19487 |
+
"loss": 1.2707,
|
19488 |
+
"step": 27210
|
19489 |
+
},
|
19490 |
+
{
|
19491 |
+
"epoch": 0.63,
|
19492 |
+
"grad_norm": 1.861101508140564,
|
19493 |
+
"learning_rate": 1.8847457627118646e-06,
|
19494 |
+
"loss": 0.9353,
|
19495 |
+
"step": 27220
|
19496 |
+
},
|
19497 |
+
{
|
19498 |
+
"epoch": 0.63,
|
19499 |
+
"grad_norm": 5.153651714324951,
|
19500 |
+
"learning_rate": 1.8779661016949156e-06,
|
19501 |
+
"loss": 1.2232,
|
19502 |
+
"step": 27230
|
19503 |
+
},
|
19504 |
+
{
|
19505 |
+
"epoch": 0.63,
|
19506 |
+
"grad_norm": 4.55015230178833,
|
19507 |
+
"learning_rate": 1.8711864406779661e-06,
|
19508 |
+
"loss": 1.2731,
|
19509 |
+
"step": 27240
|
19510 |
+
},
|
19511 |
+
{
|
19512 |
+
"epoch": 0.63,
|
19513 |
+
"grad_norm": 2.4303572177886963,
|
19514 |
+
"learning_rate": 1.8644067796610171e-06,
|
19515 |
+
"loss": 1.3679,
|
19516 |
+
"step": 27250
|
19517 |
+
},
|
19518 |
+
{
|
19519 |
+
"epoch": 0.63,
|
19520 |
+
"grad_norm": 7.1049017906188965,
|
19521 |
+
"learning_rate": 1.857627118644068e-06,
|
19522 |
+
"loss": 1.2542,
|
19523 |
+
"step": 27260
|
19524 |
+
},
|
19525 |
+
{
|
19526 |
+
"epoch": 0.63,
|
19527 |
+
"grad_norm": 7.2281036376953125,
|
19528 |
+
"learning_rate": 1.8508474576271189e-06,
|
19529 |
+
"loss": 1.197,
|
19530 |
+
"step": 27270
|
19531 |
+
},
|
19532 |
+
{
|
19533 |
+
"epoch": 0.63,
|
19534 |
+
"grad_norm": 7.088107109069824,
|
19535 |
+
"learning_rate": 1.8440677966101696e-06,
|
19536 |
+
"loss": 1.3486,
|
19537 |
+
"step": 27280
|
19538 |
+
},
|
19539 |
+
{
|
19540 |
+
"epoch": 0.63,
|
19541 |
+
"grad_norm": 5.903868675231934,
|
19542 |
+
"learning_rate": 1.8372881355932204e-06,
|
19543 |
+
"loss": 1.2879,
|
19544 |
+
"step": 27290
|
19545 |
+
},
|
19546 |
+
{
|
19547 |
+
"epoch": 0.63,
|
19548 |
+
"grad_norm": 1.4684103727340698,
|
19549 |
+
"learning_rate": 1.8305084745762714e-06,
|
19550 |
+
"loss": 1.2303,
|
19551 |
+
"step": 27300
|
19552 |
+
},
|
19553 |
+
{
|
19554 |
+
"epoch": 0.63,
|
19555 |
+
"grad_norm": 10.257400512695312,
|
19556 |
+
"learning_rate": 1.8237288135593223e-06,
|
19557 |
+
"loss": 1.2068,
|
19558 |
+
"step": 27310
|
19559 |
+
},
|
19560 |
+
{
|
19561 |
+
"epoch": 0.63,
|
19562 |
+
"grad_norm": 4.519632816314697,
|
19563 |
+
"learning_rate": 1.816949152542373e-06,
|
19564 |
+
"loss": 1.1828,
|
19565 |
+
"step": 27320
|
19566 |
+
},
|
19567 |
+
{
|
19568 |
+
"epoch": 0.63,
|
19569 |
+
"grad_norm": 2.953141689300537,
|
19570 |
+
"learning_rate": 1.8101694915254239e-06,
|
19571 |
+
"loss": 1.2411,
|
19572 |
+
"step": 27330
|
19573 |
+
},
|
19574 |
+
{
|
19575 |
+
"epoch": 0.63,
|
19576 |
+
"grad_norm": 3.631997585296631,
|
19577 |
+
"learning_rate": 1.8033898305084746e-06,
|
19578 |
+
"loss": 1.267,
|
19579 |
+
"step": 27340
|
19580 |
+
},
|
19581 |
+
{
|
19582 |
+
"epoch": 0.63,
|
19583 |
+
"grad_norm": 4.540164947509766,
|
19584 |
+
"learning_rate": 1.7966101694915256e-06,
|
19585 |
+
"loss": 1.1731,
|
19586 |
+
"step": 27350
|
19587 |
+
},
|
19588 |
+
{
|
19589 |
+
"epoch": 0.63,
|
19590 |
+
"grad_norm": 5.641757965087891,
|
19591 |
+
"learning_rate": 1.7898305084745766e-06,
|
19592 |
+
"loss": 1.1819,
|
19593 |
+
"step": 27360
|
19594 |
+
},
|
19595 |
+
{
|
19596 |
+
"epoch": 0.63,
|
19597 |
+
"grad_norm": 10.886664390563965,
|
19598 |
+
"learning_rate": 1.7830508474576271e-06,
|
19599 |
+
"loss": 1.1085,
|
19600 |
+
"step": 27370
|
19601 |
+
},
|
19602 |
+
{
|
19603 |
+
"epoch": 0.63,
|
19604 |
+
"grad_norm": 8.04320240020752,
|
19605 |
+
"learning_rate": 1.7762711864406781e-06,
|
19606 |
+
"loss": 1.2514,
|
19607 |
+
"step": 27380
|
19608 |
+
},
|
19609 |
+
{
|
19610 |
+
"epoch": 0.63,
|
19611 |
+
"grad_norm": 8.10466480255127,
|
19612 |
+
"learning_rate": 1.769491525423729e-06,
|
19613 |
+
"loss": 1.3978,
|
19614 |
+
"step": 27390
|
19615 |
+
},
|
19616 |
+
{
|
19617 |
+
"epoch": 0.63,
|
19618 |
+
"grad_norm": 8.447578430175781,
|
19619 |
+
"learning_rate": 1.7627118644067799e-06,
|
19620 |
+
"loss": 1.1422,
|
19621 |
+
"step": 27400
|
19622 |
+
},
|
19623 |
+
{
|
19624 |
+
"epoch": 0.63,
|
19625 |
+
"grad_norm": 15.72360610961914,
|
19626 |
+
"learning_rate": 1.7559322033898306e-06,
|
19627 |
+
"loss": 1.2554,
|
19628 |
+
"step": 27410
|
19629 |
+
},
|
19630 |
+
{
|
19631 |
+
"epoch": 0.63,
|
19632 |
+
"grad_norm": 4.728666305541992,
|
19633 |
+
"learning_rate": 1.7491525423728814e-06,
|
19634 |
+
"loss": 1.0925,
|
19635 |
+
"step": 27420
|
19636 |
+
},
|
19637 |
+
{
|
19638 |
+
"epoch": 0.64,
|
19639 |
+
"grad_norm": 6.482224464416504,
|
19640 |
+
"learning_rate": 1.7423728813559324e-06,
|
19641 |
+
"loss": 1.2841,
|
19642 |
+
"step": 27430
|
19643 |
+
},
|
19644 |
+
{
|
19645 |
+
"epoch": 0.64,
|
19646 |
+
"grad_norm": 17.51344871520996,
|
19647 |
+
"learning_rate": 1.7355932203389834e-06,
|
19648 |
+
"loss": 1.2075,
|
19649 |
+
"step": 27440
|
19650 |
+
},
|
19651 |
+
{
|
19652 |
+
"epoch": 0.64,
|
19653 |
+
"grad_norm": 4.776230335235596,
|
19654 |
+
"learning_rate": 1.728813559322034e-06,
|
19655 |
+
"loss": 1.2672,
|
19656 |
+
"step": 27450
|
19657 |
+
},
|
19658 |
+
{
|
19659 |
+
"epoch": 0.64,
|
19660 |
+
"grad_norm": 11.152587890625,
|
19661 |
+
"learning_rate": 1.7220338983050849e-06,
|
19662 |
+
"loss": 1.3695,
|
19663 |
+
"step": 27460
|
19664 |
+
},
|
19665 |
+
{
|
19666 |
+
"epoch": 0.64,
|
19667 |
+
"grad_norm": 4.942652702331543,
|
19668 |
+
"learning_rate": 1.7152542372881356e-06,
|
19669 |
+
"loss": 1.4412,
|
19670 |
+
"step": 27470
|
19671 |
+
},
|
19672 |
+
{
|
19673 |
+
"epoch": 0.64,
|
19674 |
+
"grad_norm": 7.248074054718018,
|
19675 |
+
"learning_rate": 1.7084745762711866e-06,
|
19676 |
+
"loss": 1.2882,
|
19677 |
+
"step": 27480
|
19678 |
+
},
|
19679 |
+
{
|
19680 |
+
"epoch": 0.64,
|
19681 |
+
"grad_norm": 3.3746461868286133,
|
19682 |
+
"learning_rate": 1.7016949152542376e-06,
|
19683 |
+
"loss": 1.2192,
|
19684 |
+
"step": 27490
|
19685 |
+
},
|
19686 |
+
{
|
19687 |
+
"epoch": 0.64,
|
19688 |
+
"grad_norm": 0.75871741771698,
|
19689 |
+
"learning_rate": 1.6949152542372882e-06,
|
19690 |
+
"loss": 1.1612,
|
19691 |
+
"step": 27500
|
19692 |
+
},
|
19693 |
+
{
|
19694 |
+
"epoch": 0.64,
|
19695 |
+
"eval_loss": 0.9824415445327759,
|
19696 |
+
"eval_runtime": 66.6076,
|
19697 |
+
"eval_samples_per_second": 15.013,
|
19698 |
+
"eval_steps_per_second": 15.013,
|
19699 |
+
"step": 27500
|
19700 |
}
|
19701 |
],
|
19702 |
"logging_steps": 10,
|
|
|
19704 |
"num_input_tokens_seen": 0,
|
19705 |
"num_train_epochs": 1,
|
19706 |
"save_steps": 2500,
|
19707 |
+
"total_flos": 4.4280846483456e+17,
|
19708 |
"train_batch_size": 1,
|
19709 |
"trial_name": null,
|
19710 |
"trial_params": null
|