Upload folder using huggingface_hub
Browse files- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1319 -6
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 990185320
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a362ad74ecdd3d29a102c8117642174c93ee465df5da1b4595fa94a234de7234
|
| 3 |
size 990185320
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b82db22b6891448c5c31d179151f5cbd660554523ea3ec6c76f2eb97b1de18ef
|
| 3 |
+
size 1980540922
|
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:f78a674aa2b4fdd82caff54d14a91a9039ca5220c343c30e97377494701aa69a
|
| 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:e8ed5b5855a820a266d1940d2bebfabd66ce7527bdfc2ac754460ff7467631d4
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1674,12 +1674,1325 @@
|
|
| 1674 |
"eval_samples_per_second": 44.247,
|
| 1675 |
"eval_steps_per_second": 5.542,
|
| 1676 |
"step": 22274
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1677 |
}
|
| 1678 |
],
|
| 1679 |
"logging_steps": 100,
|
| 1680 |
-
"max_steps":
|
| 1681 |
"num_input_tokens_seen": 0,
|
| 1682 |
-
"num_train_epochs":
|
| 1683 |
"save_steps": 500,
|
| 1684 |
"stateful_callbacks": {
|
| 1685 |
"TrainerControl": {
|
|
@@ -1688,12 +3001,12 @@
|
|
| 1688 |
"should_evaluate": false,
|
| 1689 |
"should_log": false,
|
| 1690 |
"should_save": true,
|
| 1691 |
-
"should_training_stop":
|
| 1692 |
},
|
| 1693 |
"attributes": {}
|
| 1694 |
}
|
| 1695 |
},
|
| 1696 |
-
"total_flos":
|
| 1697 |
"train_batch_size": 8,
|
| 1698 |
"trial_name": null,
|
| 1699 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 25.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 39775,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1674 |
"eval_samples_per_second": 44.247,
|
| 1675 |
"eval_steps_per_second": 5.542,
|
| 1676 |
"step": 22274
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 14.016341923318668,
|
| 1680 |
+
"grad_norm": 4.320845127105713,
|
| 1681 |
+
"learning_rate": 2.1968573224387177e-05,
|
| 1682 |
+
"loss": 1.7866,
|
| 1683 |
+
"step": 22300
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 14.079195474544312,
|
| 1687 |
+
"grad_norm": 5.575278282165527,
|
| 1688 |
+
"learning_rate": 2.184286612193589e-05,
|
| 1689 |
+
"loss": 1.7572,
|
| 1690 |
+
"step": 22400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 14.142049025769955,
|
| 1694 |
+
"grad_norm": 5.764155387878418,
|
| 1695 |
+
"learning_rate": 2.17171590194846e-05,
|
| 1696 |
+
"loss": 1.7566,
|
| 1697 |
+
"step": 22500
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 14.2049025769956,
|
| 1701 |
+
"grad_norm": 4.854477882385254,
|
| 1702 |
+
"learning_rate": 2.1591451917033316e-05,
|
| 1703 |
+
"loss": 1.7517,
|
| 1704 |
+
"step": 22600
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 14.267756128221244,
|
| 1708 |
+
"grad_norm": 4.7141618728637695,
|
| 1709 |
+
"learning_rate": 2.1465744814582025e-05,
|
| 1710 |
+
"loss": 1.713,
|
| 1711 |
+
"step": 22700
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 14.330609679446889,
|
| 1715 |
+
"grad_norm": 4.3324785232543945,
|
| 1716 |
+
"learning_rate": 2.1340037712130736e-05,
|
| 1717 |
+
"loss": 1.7511,
|
| 1718 |
+
"step": 22800
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 14.393463230672532,
|
| 1722 |
+
"grad_norm": 3.4204530715942383,
|
| 1723 |
+
"learning_rate": 2.1214330609679448e-05,
|
| 1724 |
+
"loss": 1.7451,
|
| 1725 |
+
"step": 22900
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 14.456316781898177,
|
| 1729 |
+
"grad_norm": 4.925296783447266,
|
| 1730 |
+
"learning_rate": 2.108862350722816e-05,
|
| 1731 |
+
"loss": 1.6868,
|
| 1732 |
+
"step": 23000
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 14.51917033312382,
|
| 1736 |
+
"grad_norm": 4.997200965881348,
|
| 1737 |
+
"learning_rate": 2.0962916404776872e-05,
|
| 1738 |
+
"loss": 1.7259,
|
| 1739 |
+
"step": 23100
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 14.582023884349466,
|
| 1743 |
+
"grad_norm": 4.816483497619629,
|
| 1744 |
+
"learning_rate": 2.0837209302325584e-05,
|
| 1745 |
+
"loss": 1.7716,
|
| 1746 |
+
"step": 23200
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 14.64487743557511,
|
| 1750 |
+
"grad_norm": 5.224360466003418,
|
| 1751 |
+
"learning_rate": 2.0711502199874295e-05,
|
| 1752 |
+
"loss": 1.7039,
|
| 1753 |
+
"step": 23300
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 14.707730986800755,
|
| 1757 |
+
"grad_norm": 7.450541019439697,
|
| 1758 |
+
"learning_rate": 2.0585795097423004e-05,
|
| 1759 |
+
"loss": 1.6634,
|
| 1760 |
+
"step": 23400
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 14.770584538026398,
|
| 1764 |
+
"grad_norm": 5.811767101287842,
|
| 1765 |
+
"learning_rate": 2.0460087994971716e-05,
|
| 1766 |
+
"loss": 1.7526,
|
| 1767 |
+
"step": 23500
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 14.833438089252043,
|
| 1771 |
+
"grad_norm": 4.1061272621154785,
|
| 1772 |
+
"learning_rate": 2.0334380892520427e-05,
|
| 1773 |
+
"loss": 1.7612,
|
| 1774 |
+
"step": 23600
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 14.896291640477687,
|
| 1778 |
+
"grad_norm": 4.599556922912598,
|
| 1779 |
+
"learning_rate": 2.020867379006914e-05,
|
| 1780 |
+
"loss": 1.776,
|
| 1781 |
+
"step": 23700
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 14.959145191703332,
|
| 1785 |
+
"grad_norm": 4.085700988769531,
|
| 1786 |
+
"learning_rate": 2.008296668761785e-05,
|
| 1787 |
+
"loss": 1.7143,
|
| 1788 |
+
"step": 23800
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 15.0,
|
| 1792 |
+
"eval_loss": 1.6270309686660767,
|
| 1793 |
+
"eval_runtime": 20.346,
|
| 1794 |
+
"eval_samples_per_second": 47.085,
|
| 1795 |
+
"eval_steps_per_second": 5.898,
|
| 1796 |
+
"step": 23865
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 15.021998742928975,
|
| 1800 |
+
"grad_norm": 8.476902961730957,
|
| 1801 |
+
"learning_rate": 1.9957259585166563e-05,
|
| 1802 |
+
"loss": 1.6504,
|
| 1803 |
+
"step": 23900
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 15.08485229415462,
|
| 1807 |
+
"grad_norm": 4.84979772567749,
|
| 1808 |
+
"learning_rate": 1.9831552482715275e-05,
|
| 1809 |
+
"loss": 1.7259,
|
| 1810 |
+
"step": 24000
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 15.147705845380264,
|
| 1814 |
+
"grad_norm": 4.314637184143066,
|
| 1815 |
+
"learning_rate": 1.9705845380263983e-05,
|
| 1816 |
+
"loss": 1.6254,
|
| 1817 |
+
"step": 24100
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 15.210559396605909,
|
| 1821 |
+
"grad_norm": 4.656597137451172,
|
| 1822 |
+
"learning_rate": 1.9580138277812698e-05,
|
| 1823 |
+
"loss": 1.7493,
|
| 1824 |
+
"step": 24200
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 15.273412947831552,
|
| 1828 |
+
"grad_norm": 4.276788711547852,
|
| 1829 |
+
"learning_rate": 1.945443117536141e-05,
|
| 1830 |
+
"loss": 1.6797,
|
| 1831 |
+
"step": 24300
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 15.336266499057198,
|
| 1835 |
+
"grad_norm": 3.9574031829833984,
|
| 1836 |
+
"learning_rate": 1.9328724072910122e-05,
|
| 1837 |
+
"loss": 1.716,
|
| 1838 |
+
"step": 24400
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 15.399120050282841,
|
| 1842 |
+
"grad_norm": 8.148831367492676,
|
| 1843 |
+
"learning_rate": 1.920301697045883e-05,
|
| 1844 |
+
"loss": 1.6737,
|
| 1845 |
+
"step": 24500
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 15.461973601508486,
|
| 1849 |
+
"grad_norm": 3.8734018802642822,
|
| 1850 |
+
"learning_rate": 1.9077309868007542e-05,
|
| 1851 |
+
"loss": 1.6452,
|
| 1852 |
+
"step": 24600
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 15.52482715273413,
|
| 1856 |
+
"grad_norm": 4.928835391998291,
|
| 1857 |
+
"learning_rate": 1.8951602765556257e-05,
|
| 1858 |
+
"loss": 1.7134,
|
| 1859 |
+
"step": 24700
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 15.587680703959773,
|
| 1863 |
+
"grad_norm": 4.991033554077148,
|
| 1864 |
+
"learning_rate": 1.8825895663104966e-05,
|
| 1865 |
+
"loss": 1.7327,
|
| 1866 |
+
"step": 24800
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 15.650534255185418,
|
| 1870 |
+
"grad_norm": 4.160732269287109,
|
| 1871 |
+
"learning_rate": 1.8700188560653677e-05,
|
| 1872 |
+
"loss": 1.6678,
|
| 1873 |
+
"step": 24900
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 15.713387806411061,
|
| 1877 |
+
"grad_norm": 6.523078441619873,
|
| 1878 |
+
"learning_rate": 1.857448145820239e-05,
|
| 1879 |
+
"loss": 1.6856,
|
| 1880 |
+
"step": 25000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 15.776241357636707,
|
| 1884 |
+
"grad_norm": 6.306403636932373,
|
| 1885 |
+
"learning_rate": 1.84487743557511e-05,
|
| 1886 |
+
"loss": 1.6699,
|
| 1887 |
+
"step": 25100
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 15.83909490886235,
|
| 1891 |
+
"grad_norm": 4.479640483856201,
|
| 1892 |
+
"learning_rate": 1.832306725329981e-05,
|
| 1893 |
+
"loss": 1.676,
|
| 1894 |
+
"step": 25200
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 15.901948460087995,
|
| 1898 |
+
"grad_norm": 4.6891279220581055,
|
| 1899 |
+
"learning_rate": 1.8197360150848525e-05,
|
| 1900 |
+
"loss": 1.667,
|
| 1901 |
+
"step": 25300
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 15.964802011313639,
|
| 1905 |
+
"grad_norm": 5.908668518066406,
|
| 1906 |
+
"learning_rate": 1.8071653048397236e-05,
|
| 1907 |
+
"loss": 1.6267,
|
| 1908 |
+
"step": 25400
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 16.0,
|
| 1912 |
+
"eval_loss": 1.608726143836975,
|
| 1913 |
+
"eval_runtime": 20.3571,
|
| 1914 |
+
"eval_samples_per_second": 47.06,
|
| 1915 |
+
"eval_steps_per_second": 5.895,
|
| 1916 |
+
"step": 25456
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 16.027655562539284,
|
| 1920 |
+
"grad_norm": 4.081086158752441,
|
| 1921 |
+
"learning_rate": 1.7945945945945948e-05,
|
| 1922 |
+
"loss": 1.5625,
|
| 1923 |
+
"step": 25500
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 16.090509113764927,
|
| 1927 |
+
"grad_norm": 3.7648415565490723,
|
| 1928 |
+
"learning_rate": 1.7820238843494657e-05,
|
| 1929 |
+
"loss": 1.6818,
|
| 1930 |
+
"step": 25600
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 16.15336266499057,
|
| 1934 |
+
"grad_norm": 5.430357456207275,
|
| 1935 |
+
"learning_rate": 1.769453174104337e-05,
|
| 1936 |
+
"loss": 1.6125,
|
| 1937 |
+
"step": 25700
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 16.216216216216218,
|
| 1941 |
+
"grad_norm": 5.235119819641113,
|
| 1942 |
+
"learning_rate": 1.7568824638592084e-05,
|
| 1943 |
+
"loss": 1.6985,
|
| 1944 |
+
"step": 25800
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 16.27906976744186,
|
| 1948 |
+
"grad_norm": 5.521476745605469,
|
| 1949 |
+
"learning_rate": 1.7443117536140792e-05,
|
| 1950 |
+
"loss": 1.6291,
|
| 1951 |
+
"step": 25900
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 16.341923318667504,
|
| 1955 |
+
"grad_norm": 5.7086873054504395,
|
| 1956 |
+
"learning_rate": 1.7317410433689504e-05,
|
| 1957 |
+
"loss": 1.6523,
|
| 1958 |
+
"step": 26000
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 16.404776869893148,
|
| 1962 |
+
"grad_norm": 5.697257041931152,
|
| 1963 |
+
"learning_rate": 1.7191703331238216e-05,
|
| 1964 |
+
"loss": 1.6518,
|
| 1965 |
+
"step": 26100
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 16.467630421118795,
|
| 1969 |
+
"grad_norm": 8.258442878723145,
|
| 1970 |
+
"learning_rate": 1.7065996228786928e-05,
|
| 1971 |
+
"loss": 1.6314,
|
| 1972 |
+
"step": 26200
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 16.530483972344438,
|
| 1976 |
+
"grad_norm": 4.087442874908447,
|
| 1977 |
+
"learning_rate": 1.694028912633564e-05,
|
| 1978 |
+
"loss": 1.7048,
|
| 1979 |
+
"step": 26300
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 16.59333752357008,
|
| 1983 |
+
"grad_norm": 4.184548377990723,
|
| 1984 |
+
"learning_rate": 1.681458202388435e-05,
|
| 1985 |
+
"loss": 1.6062,
|
| 1986 |
+
"step": 26400
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 16.656191074795725,
|
| 1990 |
+
"grad_norm": 5.8042707443237305,
|
| 1991 |
+
"learning_rate": 1.6688874921433063e-05,
|
| 1992 |
+
"loss": 1.6239,
|
| 1993 |
+
"step": 26500
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 16.719044626021372,
|
| 1997 |
+
"grad_norm": 4.104475498199463,
|
| 1998 |
+
"learning_rate": 1.656316781898177e-05,
|
| 1999 |
+
"loss": 1.5742,
|
| 2000 |
+
"step": 26600
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 16.781898177247015,
|
| 2004 |
+
"grad_norm": 4.2934722900390625,
|
| 2005 |
+
"learning_rate": 1.6437460716530483e-05,
|
| 2006 |
+
"loss": 1.6069,
|
| 2007 |
+
"step": 26700
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 16.84475172847266,
|
| 2011 |
+
"grad_norm": 4.601330757141113,
|
| 2012 |
+
"learning_rate": 1.6311753614079195e-05,
|
| 2013 |
+
"loss": 1.5827,
|
| 2014 |
+
"step": 26800
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 16.907605279698302,
|
| 2018 |
+
"grad_norm": 4.304816246032715,
|
| 2019 |
+
"learning_rate": 1.618604651162791e-05,
|
| 2020 |
+
"loss": 1.6461,
|
| 2021 |
+
"step": 26900
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 16.970458830923945,
|
| 2025 |
+
"grad_norm": 6.80120325088501,
|
| 2026 |
+
"learning_rate": 1.606033940917662e-05,
|
| 2027 |
+
"loss": 1.6143,
|
| 2028 |
+
"step": 27000
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 17.0,
|
| 2032 |
+
"eval_loss": 1.5869935750961304,
|
| 2033 |
+
"eval_runtime": 20.3162,
|
| 2034 |
+
"eval_samples_per_second": 47.154,
|
| 2035 |
+
"eval_steps_per_second": 5.907,
|
| 2036 |
+
"step": 27047
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 17.033312382149592,
|
| 2040 |
+
"grad_norm": 4.368440628051758,
|
| 2041 |
+
"learning_rate": 1.593463230672533e-05,
|
| 2042 |
+
"loss": 1.6352,
|
| 2043 |
+
"step": 27100
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 17.096165933375236,
|
| 2047 |
+
"grad_norm": 4.066120624542236,
|
| 2048 |
+
"learning_rate": 1.5808925204274042e-05,
|
| 2049 |
+
"loss": 1.5052,
|
| 2050 |
+
"step": 27200
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 17.15901948460088,
|
| 2054 |
+
"grad_norm": 6.150811672210693,
|
| 2055 |
+
"learning_rate": 1.5683218101822754e-05,
|
| 2056 |
+
"loss": 1.5449,
|
| 2057 |
+
"step": 27300
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 17.221873035826523,
|
| 2061 |
+
"grad_norm": 7.994663715362549,
|
| 2062 |
+
"learning_rate": 1.5557510999371466e-05,
|
| 2063 |
+
"loss": 1.7157,
|
| 2064 |
+
"step": 27400
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 17.28472658705217,
|
| 2068 |
+
"grad_norm": 3.554856061935425,
|
| 2069 |
+
"learning_rate": 1.5431803896920178e-05,
|
| 2070 |
+
"loss": 1.5878,
|
| 2071 |
+
"step": 27500
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 17.347580138277813,
|
| 2075 |
+
"grad_norm": 4.025883674621582,
|
| 2076 |
+
"learning_rate": 1.530609679446889e-05,
|
| 2077 |
+
"loss": 1.6454,
|
| 2078 |
+
"step": 27600
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 17.410433689503456,
|
| 2082 |
+
"grad_norm": 2.9825448989868164,
|
| 2083 |
+
"learning_rate": 1.51803896920176e-05,
|
| 2084 |
+
"loss": 1.5605,
|
| 2085 |
+
"step": 27700
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 17.4732872407291,
|
| 2089 |
+
"grad_norm": 4.528345584869385,
|
| 2090 |
+
"learning_rate": 1.505468258956631e-05,
|
| 2091 |
+
"loss": 1.626,
|
| 2092 |
+
"step": 27800
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 17.536140791954747,
|
| 2096 |
+
"grad_norm": 4.549004554748535,
|
| 2097 |
+
"learning_rate": 1.4928975487115023e-05,
|
| 2098 |
+
"loss": 1.5508,
|
| 2099 |
+
"step": 27900
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 17.59899434318039,
|
| 2103 |
+
"grad_norm": 4.830588340759277,
|
| 2104 |
+
"learning_rate": 1.4803268384663735e-05,
|
| 2105 |
+
"loss": 1.5394,
|
| 2106 |
+
"step": 28000
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 17.661847894406034,
|
| 2110 |
+
"grad_norm": 4.127079486846924,
|
| 2111 |
+
"learning_rate": 1.4677561282212447e-05,
|
| 2112 |
+
"loss": 1.5548,
|
| 2113 |
+
"step": 28100
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 17.724701445631677,
|
| 2117 |
+
"grad_norm": 3.208592414855957,
|
| 2118 |
+
"learning_rate": 1.4551854179761157e-05,
|
| 2119 |
+
"loss": 1.5595,
|
| 2120 |
+
"step": 28200
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 17.787554996857324,
|
| 2124 |
+
"grad_norm": 4.784154891967773,
|
| 2125 |
+
"learning_rate": 1.4426147077309869e-05,
|
| 2126 |
+
"loss": 1.6029,
|
| 2127 |
+
"step": 28300
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 17.850408548082967,
|
| 2131 |
+
"grad_norm": 5.0941481590271,
|
| 2132 |
+
"learning_rate": 1.4300439974858582e-05,
|
| 2133 |
+
"loss": 1.634,
|
| 2134 |
+
"step": 28400
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 17.91326209930861,
|
| 2138 |
+
"grad_norm": 6.4498982429504395,
|
| 2139 |
+
"learning_rate": 1.4174732872407292e-05,
|
| 2140 |
+
"loss": 1.6685,
|
| 2141 |
+
"step": 28500
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 17.976115650534254,
|
| 2145 |
+
"grad_norm": 5.136322021484375,
|
| 2146 |
+
"learning_rate": 1.4049025769956004e-05,
|
| 2147 |
+
"loss": 1.5587,
|
| 2148 |
+
"step": 28600
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 18.0,
|
| 2152 |
+
"eval_loss": 1.565408706665039,
|
| 2153 |
+
"eval_runtime": 20.3165,
|
| 2154 |
+
"eval_samples_per_second": 47.154,
|
| 2155 |
+
"eval_steps_per_second": 5.907,
|
| 2156 |
+
"step": 28638
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 18.0389692017599,
|
| 2160 |
+
"grad_norm": 7.265219211578369,
|
| 2161 |
+
"learning_rate": 1.3923318667504714e-05,
|
| 2162 |
+
"loss": 1.534,
|
| 2163 |
+
"step": 28700
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 18.101822752985544,
|
| 2167 |
+
"grad_norm": 5.552704334259033,
|
| 2168 |
+
"learning_rate": 1.3797611565053426e-05,
|
| 2169 |
+
"loss": 1.5396,
|
| 2170 |
+
"step": 28800
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 18.164676304211188,
|
| 2174 |
+
"grad_norm": 7.356419086456299,
|
| 2175 |
+
"learning_rate": 1.3671904462602136e-05,
|
| 2176 |
+
"loss": 1.5851,
|
| 2177 |
+
"step": 28900
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 18.22752985543683,
|
| 2181 |
+
"grad_norm": 5.519120693206787,
|
| 2182 |
+
"learning_rate": 1.354619736015085e-05,
|
| 2183 |
+
"loss": 1.6331,
|
| 2184 |
+
"step": 29000
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 18.290383406662478,
|
| 2188 |
+
"grad_norm": 4.4178242683410645,
|
| 2189 |
+
"learning_rate": 1.3420490257699561e-05,
|
| 2190 |
+
"loss": 1.508,
|
| 2191 |
+
"step": 29100
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 18.35323695788812,
|
| 2195 |
+
"grad_norm": 4.479162216186523,
|
| 2196 |
+
"learning_rate": 1.3294783155248271e-05,
|
| 2197 |
+
"loss": 1.5201,
|
| 2198 |
+
"step": 29200
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 18.416090509113765,
|
| 2202 |
+
"grad_norm": 4.4193806648254395,
|
| 2203 |
+
"learning_rate": 1.3169076052796983e-05,
|
| 2204 |
+
"loss": 1.5393,
|
| 2205 |
+
"step": 29300
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 18.47894406033941,
|
| 2209 |
+
"grad_norm": 6.695824146270752,
|
| 2210 |
+
"learning_rate": 1.3043368950345693e-05,
|
| 2211 |
+
"loss": 1.6264,
|
| 2212 |
+
"step": 29400
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 18.541797611565052,
|
| 2216 |
+
"grad_norm": 4.760421276092529,
|
| 2217 |
+
"learning_rate": 1.2917661847894409e-05,
|
| 2218 |
+
"loss": 1.5465,
|
| 2219 |
+
"step": 29500
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 18.6046511627907,
|
| 2223 |
+
"grad_norm": 4.158078193664551,
|
| 2224 |
+
"learning_rate": 1.2791954745443119e-05,
|
| 2225 |
+
"loss": 1.5533,
|
| 2226 |
+
"step": 29600
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 18.667504714016342,
|
| 2230 |
+
"grad_norm": 6.8502092361450195,
|
| 2231 |
+
"learning_rate": 1.266624764299183e-05,
|
| 2232 |
+
"loss": 1.6525,
|
| 2233 |
+
"step": 29700
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 18.730358265241986,
|
| 2237 |
+
"grad_norm": 4.013594150543213,
|
| 2238 |
+
"learning_rate": 1.254054054054054e-05,
|
| 2239 |
+
"loss": 1.5357,
|
| 2240 |
+
"step": 29800
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 18.79321181646763,
|
| 2244 |
+
"grad_norm": 6.064908981323242,
|
| 2245 |
+
"learning_rate": 1.2414833438089252e-05,
|
| 2246 |
+
"loss": 1.5659,
|
| 2247 |
+
"step": 29900
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 18.856065367693276,
|
| 2251 |
+
"grad_norm": 5.281710624694824,
|
| 2252 |
+
"learning_rate": 1.2289126335637964e-05,
|
| 2253 |
+
"loss": 1.4692,
|
| 2254 |
+
"step": 30000
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 18.91891891891892,
|
| 2258 |
+
"grad_norm": 4.661835193634033,
|
| 2259 |
+
"learning_rate": 1.2163419233186674e-05,
|
| 2260 |
+
"loss": 1.5126,
|
| 2261 |
+
"step": 30100
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 18.981772470144563,
|
| 2265 |
+
"grad_norm": 3.9490227699279785,
|
| 2266 |
+
"learning_rate": 1.2037712130735388e-05,
|
| 2267 |
+
"loss": 1.5389,
|
| 2268 |
+
"step": 30200
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 19.0,
|
| 2272 |
+
"eval_loss": 1.5563335418701172,
|
| 2273 |
+
"eval_runtime": 20.3631,
|
| 2274 |
+
"eval_samples_per_second": 47.046,
|
| 2275 |
+
"eval_steps_per_second": 5.893,
|
| 2276 |
+
"step": 30229
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 19.044626021370206,
|
| 2280 |
+
"grad_norm": 4.6667866706848145,
|
| 2281 |
+
"learning_rate": 1.1912005028284098e-05,
|
| 2282 |
+
"loss": 1.5508,
|
| 2283 |
+
"step": 30300
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 19.107479572595853,
|
| 2287 |
+
"grad_norm": 4.471792697906494,
|
| 2288 |
+
"learning_rate": 1.1786297925832811e-05,
|
| 2289 |
+
"loss": 1.5253,
|
| 2290 |
+
"step": 30400
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 19.170333123821496,
|
| 2294 |
+
"grad_norm": 4.01970100402832,
|
| 2295 |
+
"learning_rate": 1.1660590823381521e-05,
|
| 2296 |
+
"loss": 1.5047,
|
| 2297 |
+
"step": 30500
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 19.23318667504714,
|
| 2301 |
+
"grad_norm": 5.021801471710205,
|
| 2302 |
+
"learning_rate": 1.1534883720930233e-05,
|
| 2303 |
+
"loss": 1.5459,
|
| 2304 |
+
"step": 30600
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 19.296040226272783,
|
| 2308 |
+
"grad_norm": 4.681889533996582,
|
| 2309 |
+
"learning_rate": 1.1409176618478945e-05,
|
| 2310 |
+
"loss": 1.561,
|
| 2311 |
+
"step": 30700
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 19.35889377749843,
|
| 2315 |
+
"grad_norm": 4.114772319793701,
|
| 2316 |
+
"learning_rate": 1.1283469516027655e-05,
|
| 2317 |
+
"loss": 1.532,
|
| 2318 |
+
"step": 30800
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 19.421747328724074,
|
| 2322 |
+
"grad_norm": 3.9337844848632812,
|
| 2323 |
+
"learning_rate": 1.1157762413576367e-05,
|
| 2324 |
+
"loss": 1.5512,
|
| 2325 |
+
"step": 30900
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 19.484600879949717,
|
| 2329 |
+
"grad_norm": 4.935436725616455,
|
| 2330 |
+
"learning_rate": 1.1032055311125079e-05,
|
| 2331 |
+
"loss": 1.5328,
|
| 2332 |
+
"step": 31000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 19.54745443117536,
|
| 2336 |
+
"grad_norm": 5.703494071960449,
|
| 2337 |
+
"learning_rate": 1.090634820867379e-05,
|
| 2338 |
+
"loss": 1.5889,
|
| 2339 |
+
"step": 31100
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 19.610307982401007,
|
| 2343 |
+
"grad_norm": 6.010659217834473,
|
| 2344 |
+
"learning_rate": 1.0780641106222502e-05,
|
| 2345 |
+
"loss": 1.5166,
|
| 2346 |
+
"step": 31200
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 19.67316153362665,
|
| 2350 |
+
"grad_norm": 5.14444637298584,
|
| 2351 |
+
"learning_rate": 1.0654934003771214e-05,
|
| 2352 |
+
"loss": 1.5096,
|
| 2353 |
+
"step": 31300
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 19.736015084852294,
|
| 2357 |
+
"grad_norm": 7.321188449859619,
|
| 2358 |
+
"learning_rate": 1.0529226901319924e-05,
|
| 2359 |
+
"loss": 1.4865,
|
| 2360 |
+
"step": 31400
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 19.798868636077938,
|
| 2364 |
+
"grad_norm": 3.7702994346618652,
|
| 2365 |
+
"learning_rate": 1.0403519798868636e-05,
|
| 2366 |
+
"loss": 1.5122,
|
| 2367 |
+
"step": 31500
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 19.86172218730358,
|
| 2371 |
+
"grad_norm": 5.493444442749023,
|
| 2372 |
+
"learning_rate": 1.0277812696417348e-05,
|
| 2373 |
+
"loss": 1.4974,
|
| 2374 |
+
"step": 31600
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 19.924575738529228,
|
| 2378 |
+
"grad_norm": 5.273486137390137,
|
| 2379 |
+
"learning_rate": 1.015210559396606e-05,
|
| 2380 |
+
"loss": 1.5619,
|
| 2381 |
+
"step": 31700
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 19.98742928975487,
|
| 2385 |
+
"grad_norm": 4.340183734893799,
|
| 2386 |
+
"learning_rate": 1.0026398491514772e-05,
|
| 2387 |
+
"loss": 1.4476,
|
| 2388 |
+
"step": 31800
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 20.0,
|
| 2392 |
+
"eval_loss": 1.5459223985671997,
|
| 2393 |
+
"eval_runtime": 20.3264,
|
| 2394 |
+
"eval_samples_per_second": 47.131,
|
| 2395 |
+
"eval_steps_per_second": 5.904,
|
| 2396 |
+
"step": 31820
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 20.050282840980515,
|
| 2400 |
+
"grad_norm": 3.8120639324188232,
|
| 2401 |
+
"learning_rate": 9.900691389063482e-06,
|
| 2402 |
+
"loss": 1.4837,
|
| 2403 |
+
"step": 31900
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 20.113136392206158,
|
| 2407 |
+
"grad_norm": 4.154244899749756,
|
| 2408 |
+
"learning_rate": 9.774984286612195e-06,
|
| 2409 |
+
"loss": 1.4684,
|
| 2410 |
+
"step": 32000
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 20.175989943431805,
|
| 2414 |
+
"grad_norm": 3.925746202468872,
|
| 2415 |
+
"learning_rate": 9.649277184160905e-06,
|
| 2416 |
+
"loss": 1.4685,
|
| 2417 |
+
"step": 32100
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 20.23884349465745,
|
| 2421 |
+
"grad_norm": 5.944131374359131,
|
| 2422 |
+
"learning_rate": 9.523570081709617e-06,
|
| 2423 |
+
"loss": 1.5097,
|
| 2424 |
+
"step": 32200
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 20.301697045883092,
|
| 2428 |
+
"grad_norm": 4.755185127258301,
|
| 2429 |
+
"learning_rate": 9.397862979258329e-06,
|
| 2430 |
+
"loss": 1.4334,
|
| 2431 |
+
"step": 32300
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 20.364550597108735,
|
| 2435 |
+
"grad_norm": 4.627038478851318,
|
| 2436 |
+
"learning_rate": 9.27215587680704e-06,
|
| 2437 |
+
"loss": 1.503,
|
| 2438 |
+
"step": 32400
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 20.427404148334382,
|
| 2442 |
+
"grad_norm": 9.863165855407715,
|
| 2443 |
+
"learning_rate": 9.14644877435575e-06,
|
| 2444 |
+
"loss": 1.4607,
|
| 2445 |
+
"step": 32500
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 20.490257699560026,
|
| 2449 |
+
"grad_norm": 4.401854991912842,
|
| 2450 |
+
"learning_rate": 9.020741671904463e-06,
|
| 2451 |
+
"loss": 1.4653,
|
| 2452 |
+
"step": 32600
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 20.55311125078567,
|
| 2456 |
+
"grad_norm": 6.041737079620361,
|
| 2457 |
+
"learning_rate": 8.895034569453174e-06,
|
| 2458 |
+
"loss": 1.504,
|
| 2459 |
+
"step": 32700
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 20.615964802011312,
|
| 2463 |
+
"grad_norm": 6.523427963256836,
|
| 2464 |
+
"learning_rate": 8.769327467001886e-06,
|
| 2465 |
+
"loss": 1.6205,
|
| 2466 |
+
"step": 32800
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 20.67881835323696,
|
| 2470 |
+
"grad_norm": 5.47548246383667,
|
| 2471 |
+
"learning_rate": 8.643620364550598e-06,
|
| 2472 |
+
"loss": 1.4491,
|
| 2473 |
+
"step": 32900
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 20.741671904462603,
|
| 2477 |
+
"grad_norm": 5.3726959228515625,
|
| 2478 |
+
"learning_rate": 8.517913262099308e-06,
|
| 2479 |
+
"loss": 1.5817,
|
| 2480 |
+
"step": 33000
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 20.804525455688246,
|
| 2484 |
+
"grad_norm": 3.872283935546875,
|
| 2485 |
+
"learning_rate": 8.392206159648022e-06,
|
| 2486 |
+
"loss": 1.5482,
|
| 2487 |
+
"step": 33100
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 20.86737900691389,
|
| 2491 |
+
"grad_norm": 4.935946464538574,
|
| 2492 |
+
"learning_rate": 8.266499057196732e-06,
|
| 2493 |
+
"loss": 1.5006,
|
| 2494 |
+
"step": 33200
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 20.930232558139537,
|
| 2498 |
+
"grad_norm": 6.805904388427734,
|
| 2499 |
+
"learning_rate": 8.140791954745444e-06,
|
| 2500 |
+
"loss": 1.5314,
|
| 2501 |
+
"step": 33300
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 20.99308610936518,
|
| 2505 |
+
"grad_norm": 4.420083522796631,
|
| 2506 |
+
"learning_rate": 8.015084852294155e-06,
|
| 2507 |
+
"loss": 1.5417,
|
| 2508 |
+
"step": 33400
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 21.0,
|
| 2512 |
+
"eval_loss": 1.5356966257095337,
|
| 2513 |
+
"eval_runtime": 20.4137,
|
| 2514 |
+
"eval_samples_per_second": 46.929,
|
| 2515 |
+
"eval_steps_per_second": 5.878,
|
| 2516 |
+
"step": 33411
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 21.055939660590823,
|
| 2520 |
+
"grad_norm": 3.697171688079834,
|
| 2521 |
+
"learning_rate": 7.889377749842865e-06,
|
| 2522 |
+
"loss": 1.4994,
|
| 2523 |
+
"step": 33500
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 21.118793211816467,
|
| 2527 |
+
"grad_norm": 5.232399940490723,
|
| 2528 |
+
"learning_rate": 7.763670647391579e-06,
|
| 2529 |
+
"loss": 1.5351,
|
| 2530 |
+
"step": 33600
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 21.18164676304211,
|
| 2534 |
+
"grad_norm": 4.508577823638916,
|
| 2535 |
+
"learning_rate": 7.637963544940289e-06,
|
| 2536 |
+
"loss": 1.4301,
|
| 2537 |
+
"step": 33700
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 21.244500314267757,
|
| 2541 |
+
"grad_norm": 5.425107479095459,
|
| 2542 |
+
"learning_rate": 7.512256442489001e-06,
|
| 2543 |
+
"loss": 1.4739,
|
| 2544 |
+
"step": 33800
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 21.3073538654934,
|
| 2548 |
+
"grad_norm": 6.195432186126709,
|
| 2549 |
+
"learning_rate": 7.386549340037713e-06,
|
| 2550 |
+
"loss": 1.5458,
|
| 2551 |
+
"step": 33900
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 21.370207416719044,
|
| 2555 |
+
"grad_norm": 5.850045204162598,
|
| 2556 |
+
"learning_rate": 7.260842237586424e-06,
|
| 2557 |
+
"loss": 1.5189,
|
| 2558 |
+
"step": 34000
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 21.433060967944687,
|
| 2562 |
+
"grad_norm": 7.121579170227051,
|
| 2563 |
+
"learning_rate": 7.135135135135136e-06,
|
| 2564 |
+
"loss": 1.5273,
|
| 2565 |
+
"step": 34100
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 21.495914519170334,
|
| 2569 |
+
"grad_norm": 4.316208362579346,
|
| 2570 |
+
"learning_rate": 7.009428032683847e-06,
|
| 2571 |
+
"loss": 1.4437,
|
| 2572 |
+
"step": 34200
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 21.558768070395978,
|
| 2576 |
+
"grad_norm": 4.3052873611450195,
|
| 2577 |
+
"learning_rate": 6.883720930232558e-06,
|
| 2578 |
+
"loss": 1.4266,
|
| 2579 |
+
"step": 34300
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 21.62162162162162,
|
| 2583 |
+
"grad_norm": 4.691330432891846,
|
| 2584 |
+
"learning_rate": 6.758013827781271e-06,
|
| 2585 |
+
"loss": 1.422,
|
| 2586 |
+
"step": 34400
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 21.684475172847264,
|
| 2590 |
+
"grad_norm": 4.346444129943848,
|
| 2591 |
+
"learning_rate": 6.632306725329982e-06,
|
| 2592 |
+
"loss": 1.5511,
|
| 2593 |
+
"step": 34500
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 21.74732872407291,
|
| 2597 |
+
"grad_norm": 5.304843902587891,
|
| 2598 |
+
"learning_rate": 6.506599622878693e-06,
|
| 2599 |
+
"loss": 1.4961,
|
| 2600 |
+
"step": 34600
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 21.810182275298555,
|
| 2604 |
+
"grad_norm": 4.877419948577881,
|
| 2605 |
+
"learning_rate": 6.3808925204274045e-06,
|
| 2606 |
+
"loss": 1.4837,
|
| 2607 |
+
"step": 34700
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 21.873035826524198,
|
| 2611 |
+
"grad_norm": 4.086881637573242,
|
| 2612 |
+
"learning_rate": 6.2551854179761155e-06,
|
| 2613 |
+
"loss": 1.5164,
|
| 2614 |
+
"step": 34800
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 21.93588937774984,
|
| 2618 |
+
"grad_norm": 4.570976734161377,
|
| 2619 |
+
"learning_rate": 6.129478315524827e-06,
|
| 2620 |
+
"loss": 1.4681,
|
| 2621 |
+
"step": 34900
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 21.99874292897549,
|
| 2625 |
+
"grad_norm": 25.407676696777344,
|
| 2626 |
+
"learning_rate": 6.003771213073539e-06,
|
| 2627 |
+
"loss": 1.4062,
|
| 2628 |
+
"step": 35000
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 22.0,
|
| 2632 |
+
"eval_loss": 1.5373815298080444,
|
| 2633 |
+
"eval_runtime": 20.3495,
|
| 2634 |
+
"eval_samples_per_second": 47.077,
|
| 2635 |
+
"eval_steps_per_second": 5.897,
|
| 2636 |
+
"step": 35002
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 22.061596480201132,
|
| 2640 |
+
"grad_norm": 4.965208053588867,
|
| 2641 |
+
"learning_rate": 5.878064110622251e-06,
|
| 2642 |
+
"loss": 1.446,
|
| 2643 |
+
"step": 35100
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 22.124450031426775,
|
| 2647 |
+
"grad_norm": 5.620969772338867,
|
| 2648 |
+
"learning_rate": 5.752357008170962e-06,
|
| 2649 |
+
"loss": 1.475,
|
| 2650 |
+
"step": 35200
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 22.18730358265242,
|
| 2654 |
+
"grad_norm": 4.315845489501953,
|
| 2655 |
+
"learning_rate": 5.626649905719674e-06,
|
| 2656 |
+
"loss": 1.4866,
|
| 2657 |
+
"step": 35300
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 22.250157133878066,
|
| 2661 |
+
"grad_norm": 4.076879501342773,
|
| 2662 |
+
"learning_rate": 5.5009428032683854e-06,
|
| 2663 |
+
"loss": 1.5079,
|
| 2664 |
+
"step": 35400
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 22.31301068510371,
|
| 2668 |
+
"grad_norm": 9.52351188659668,
|
| 2669 |
+
"learning_rate": 5.375235700817096e-06,
|
| 2670 |
+
"loss": 1.5637,
|
| 2671 |
+
"step": 35500
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 22.375864236329353,
|
| 2675 |
+
"grad_norm": 5.529058933258057,
|
| 2676 |
+
"learning_rate": 5.249528598365807e-06,
|
| 2677 |
+
"loss": 1.4702,
|
| 2678 |
+
"step": 35600
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 22.438717787554996,
|
| 2682 |
+
"grad_norm": 4.761877536773682,
|
| 2683 |
+
"learning_rate": 5.123821495914519e-06,
|
| 2684 |
+
"loss": 1.4367,
|
| 2685 |
+
"step": 35700
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 22.501571338780643,
|
| 2689 |
+
"grad_norm": 6.587429046630859,
|
| 2690 |
+
"learning_rate": 4.998114393463231e-06,
|
| 2691 |
+
"loss": 1.4052,
|
| 2692 |
+
"step": 35800
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 22.564424890006286,
|
| 2696 |
+
"grad_norm": 5.834304332733154,
|
| 2697 |
+
"learning_rate": 4.872407291011943e-06,
|
| 2698 |
+
"loss": 1.4186,
|
| 2699 |
+
"step": 35900
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 22.62727844123193,
|
| 2703 |
+
"grad_norm": 3.871225595474243,
|
| 2704 |
+
"learning_rate": 4.746700188560654e-06,
|
| 2705 |
+
"loss": 1.51,
|
| 2706 |
+
"step": 36000
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 22.690131992457573,
|
| 2710 |
+
"grad_norm": 3.876692771911621,
|
| 2711 |
+
"learning_rate": 4.6209930861093655e-06,
|
| 2712 |
+
"loss": 1.5022,
|
| 2713 |
+
"step": 36100
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 22.752985543683216,
|
| 2717 |
+
"grad_norm": 4.569952964782715,
|
| 2718 |
+
"learning_rate": 4.495285983658077e-06,
|
| 2719 |
+
"loss": 1.454,
|
| 2720 |
+
"step": 36200
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 22.815839094908863,
|
| 2724 |
+
"grad_norm": 5.837776184082031,
|
| 2725 |
+
"learning_rate": 4.369578881206788e-06,
|
| 2726 |
+
"loss": 1.4472,
|
| 2727 |
+
"step": 36300
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 22.878692646134507,
|
| 2731 |
+
"grad_norm": 5.9942426681518555,
|
| 2732 |
+
"learning_rate": 4.243871778755499e-06,
|
| 2733 |
+
"loss": 1.4198,
|
| 2734 |
+
"step": 36400
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 22.94154619736015,
|
| 2738 |
+
"grad_norm": 4.1033220291137695,
|
| 2739 |
+
"learning_rate": 4.118164676304211e-06,
|
| 2740 |
+
"loss": 1.4658,
|
| 2741 |
+
"step": 36500
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 23.0,
|
| 2745 |
+
"eval_loss": 1.5307875871658325,
|
| 2746 |
+
"eval_runtime": 20.3299,
|
| 2747 |
+
"eval_samples_per_second": 47.123,
|
| 2748 |
+
"eval_steps_per_second": 5.903,
|
| 2749 |
+
"step": 36593
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 23.004399748585794,
|
| 2753 |
+
"grad_norm": 4.649007320404053,
|
| 2754 |
+
"learning_rate": 3.992457573852923e-06,
|
| 2755 |
+
"loss": 1.4064,
|
| 2756 |
+
"step": 36600
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 23.06725329981144,
|
| 2760 |
+
"grad_norm": 4.318711757659912,
|
| 2761 |
+
"learning_rate": 3.866750471401635e-06,
|
| 2762 |
+
"loss": 1.4249,
|
| 2763 |
+
"step": 36700
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 23.130106851037084,
|
| 2767 |
+
"grad_norm": 6.213062286376953,
|
| 2768 |
+
"learning_rate": 3.7410433689503456e-06,
|
| 2769 |
+
"loss": 1.4317,
|
| 2770 |
+
"step": 36800
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 23.192960402262727,
|
| 2774 |
+
"grad_norm": 4.529442310333252,
|
| 2775 |
+
"learning_rate": 3.6153362664990574e-06,
|
| 2776 |
+
"loss": 1.5102,
|
| 2777 |
+
"step": 36900
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 23.25581395348837,
|
| 2781 |
+
"grad_norm": 4.912539005279541,
|
| 2782 |
+
"learning_rate": 3.4896291640477688e-06,
|
| 2783 |
+
"loss": 1.4684,
|
| 2784 |
+
"step": 37000
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 23.318667504714018,
|
| 2788 |
+
"grad_norm": 4.593921661376953,
|
| 2789 |
+
"learning_rate": 3.3639220615964806e-06,
|
| 2790 |
+
"loss": 1.4181,
|
| 2791 |
+
"step": 37100
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 23.38152105593966,
|
| 2795 |
+
"grad_norm": 5.35049295425415,
|
| 2796 |
+
"learning_rate": 3.2382149591451915e-06,
|
| 2797 |
+
"loss": 1.4813,
|
| 2798 |
+
"step": 37200
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 23.444374607165305,
|
| 2802 |
+
"grad_norm": 4.00051212310791,
|
| 2803 |
+
"learning_rate": 3.1125078566939033e-06,
|
| 2804 |
+
"loss": 1.4392,
|
| 2805 |
+
"step": 37300
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 23.507228158390948,
|
| 2809 |
+
"grad_norm": 5.91484260559082,
|
| 2810 |
+
"learning_rate": 2.9868007542426147e-06,
|
| 2811 |
+
"loss": 1.4386,
|
| 2812 |
+
"step": 37400
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 23.570081709616595,
|
| 2816 |
+
"grad_norm": 7.114585876464844,
|
| 2817 |
+
"learning_rate": 2.861093651791326e-06,
|
| 2818 |
+
"loss": 1.4115,
|
| 2819 |
+
"step": 37500
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 23.63293526084224,
|
| 2823 |
+
"grad_norm": 2.977877378463745,
|
| 2824 |
+
"learning_rate": 2.735386549340038e-06,
|
| 2825 |
+
"loss": 1.4211,
|
| 2826 |
+
"step": 37600
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 23.69578881206788,
|
| 2830 |
+
"grad_norm": 3.83953857421875,
|
| 2831 |
+
"learning_rate": 2.6096794468887493e-06,
|
| 2832 |
+
"loss": 1.4601,
|
| 2833 |
+
"step": 37700
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 23.758642363293525,
|
| 2837 |
+
"grad_norm": 4.377187728881836,
|
| 2838 |
+
"learning_rate": 2.483972344437461e-06,
|
| 2839 |
+
"loss": 1.4281,
|
| 2840 |
+
"step": 37800
|
| 2841 |
+
},
|
| 2842 |
+
{
|
| 2843 |
+
"epoch": 23.821495914519172,
|
| 2844 |
+
"grad_norm": 3.9868085384368896,
|
| 2845 |
+
"learning_rate": 2.358265241986172e-06,
|
| 2846 |
+
"loss": 1.4585,
|
| 2847 |
+
"step": 37900
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"epoch": 23.884349465744815,
|
| 2851 |
+
"grad_norm": 3.989767551422119,
|
| 2852 |
+
"learning_rate": 2.232558139534884e-06,
|
| 2853 |
+
"loss": 1.5302,
|
| 2854 |
+
"step": 38000
|
| 2855 |
+
},
|
| 2856 |
+
{
|
| 2857 |
+
"epoch": 23.94720301697046,
|
| 2858 |
+
"grad_norm": 4.481296062469482,
|
| 2859 |
+
"learning_rate": 2.1068510370835952e-06,
|
| 2860 |
+
"loss": 1.4366,
|
| 2861 |
+
"step": 38100
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 24.0,
|
| 2865 |
+
"eval_loss": 1.5289642810821533,
|
| 2866 |
+
"eval_runtime": 20.3269,
|
| 2867 |
+
"eval_samples_per_second": 47.13,
|
| 2868 |
+
"eval_steps_per_second": 5.904,
|
| 2869 |
+
"step": 38184
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"epoch": 24.010056568196102,
|
| 2873 |
+
"grad_norm": 4.909224033355713,
|
| 2874 |
+
"learning_rate": 1.981143934632307e-06,
|
| 2875 |
+
"loss": 1.4956,
|
| 2876 |
+
"step": 38200
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 24.072910119421746,
|
| 2880 |
+
"grad_norm": 4.9214372634887695,
|
| 2881 |
+
"learning_rate": 1.8554368321810182e-06,
|
| 2882 |
+
"loss": 1.4725,
|
| 2883 |
+
"step": 38300
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 24.135763670647393,
|
| 2887 |
+
"grad_norm": 4.345515251159668,
|
| 2888 |
+
"learning_rate": 1.7297297297297298e-06,
|
| 2889 |
+
"loss": 1.4407,
|
| 2890 |
+
"step": 38400
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 24.198617221873036,
|
| 2894 |
+
"grad_norm": 4.926340579986572,
|
| 2895 |
+
"learning_rate": 1.6040226272784412e-06,
|
| 2896 |
+
"loss": 1.5008,
|
| 2897 |
+
"step": 38500
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 24.26147077309868,
|
| 2901 |
+
"grad_norm": 4.5064263343811035,
|
| 2902 |
+
"learning_rate": 1.4783155248271527e-06,
|
| 2903 |
+
"loss": 1.4868,
|
| 2904 |
+
"step": 38600
|
| 2905 |
+
},
|
| 2906 |
+
{
|
| 2907 |
+
"epoch": 24.324324324324323,
|
| 2908 |
+
"grad_norm": 5.347716808319092,
|
| 2909 |
+
"learning_rate": 1.3526084223758643e-06,
|
| 2910 |
+
"loss": 1.45,
|
| 2911 |
+
"step": 38700
|
| 2912 |
+
},
|
| 2913 |
+
{
|
| 2914 |
+
"epoch": 24.38717787554997,
|
| 2915 |
+
"grad_norm": 5.024169921875,
|
| 2916 |
+
"learning_rate": 1.2269013199245757e-06,
|
| 2917 |
+
"loss": 1.3905,
|
| 2918 |
+
"step": 38800
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 24.450031426775613,
|
| 2922 |
+
"grad_norm": 4.319692611694336,
|
| 2923 |
+
"learning_rate": 1.1011942174732873e-06,
|
| 2924 |
+
"loss": 1.4671,
|
| 2925 |
+
"step": 38900
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 24.512884978001257,
|
| 2929 |
+
"grad_norm": 2.880321979522705,
|
| 2930 |
+
"learning_rate": 9.75487115021999e-07,
|
| 2931 |
+
"loss": 1.4211,
|
| 2932 |
+
"step": 39000
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 24.5757385292269,
|
| 2936 |
+
"grad_norm": 4.416039943695068,
|
| 2937 |
+
"learning_rate": 8.497800125707103e-07,
|
| 2938 |
+
"loss": 1.4176,
|
| 2939 |
+
"step": 39100
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 24.638592080452547,
|
| 2943 |
+
"grad_norm": 4.598896503448486,
|
| 2944 |
+
"learning_rate": 7.240729101194218e-07,
|
| 2945 |
+
"loss": 1.4194,
|
| 2946 |
+
"step": 39200
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 24.70144563167819,
|
| 2950 |
+
"grad_norm": 4.256235599517822,
|
| 2951 |
+
"learning_rate": 5.983658076681333e-07,
|
| 2952 |
+
"loss": 1.4331,
|
| 2953 |
+
"step": 39300
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 24.764299182903834,
|
| 2957 |
+
"grad_norm": 4.7764811515808105,
|
| 2958 |
+
"learning_rate": 4.726587052168448e-07,
|
| 2959 |
+
"loss": 1.4491,
|
| 2960 |
+
"step": 39400
|
| 2961 |
+
},
|
| 2962 |
+
{
|
| 2963 |
+
"epoch": 24.827152734129477,
|
| 2964 |
+
"grad_norm": 4.296844005584717,
|
| 2965 |
+
"learning_rate": 3.4695160276555627e-07,
|
| 2966 |
+
"loss": 1.4443,
|
| 2967 |
+
"step": 39500
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 24.890006285355124,
|
| 2971 |
+
"grad_norm": 3.9589693546295166,
|
| 2972 |
+
"learning_rate": 2.2124450031426776e-07,
|
| 2973 |
+
"loss": 1.4612,
|
| 2974 |
+
"step": 39600
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 24.952859836580767,
|
| 2978 |
+
"grad_norm": 4.165828227996826,
|
| 2979 |
+
"learning_rate": 9.553739786297926e-08,
|
| 2980 |
+
"loss": 1.48,
|
| 2981 |
+
"step": 39700
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 25.0,
|
| 2985 |
+
"eval_loss": 1.528791069984436,
|
| 2986 |
+
"eval_runtime": 20.2887,
|
| 2987 |
+
"eval_samples_per_second": 47.218,
|
| 2988 |
+
"eval_steps_per_second": 5.915,
|
| 2989 |
+
"step": 39775
|
| 2990 |
}
|
| 2991 |
],
|
| 2992 |
"logging_steps": 100,
|
| 2993 |
+
"max_steps": 39775,
|
| 2994 |
"num_input_tokens_seen": 0,
|
| 2995 |
+
"num_train_epochs": 25,
|
| 2996 |
"save_steps": 500,
|
| 2997 |
"stateful_callbacks": {
|
| 2998 |
"TrainerControl": {
|
|
|
|
| 3001 |
"should_evaluate": false,
|
| 3002 |
"should_log": false,
|
| 3003 |
"should_save": true,
|
| 3004 |
+
"should_training_stop": true
|
| 3005 |
},
|
| 3006 |
"attributes": {}
|
| 3007 |
}
|
| 3008 |
},
|
| 3009 |
+
"total_flos": 5.44418915549184e+16,
|
| 3010 |
"train_batch_size": 8,
|
| 3011 |
"trial_name": null,
|
| 3012 |
"trial_params": null
|