Upload checkpoint-258060
Browse files- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1891 -5
- training_args.bin +1 -1
optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 866895354
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cf389c91b851e668d5e63f6f7c5b36c73e641d8e4b093e49b672a0d77b0c5abe
|
3 |
size 866895354
|
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:fc23f9a9f7aa172955396035c69940d79883c5359b47c42257084c82f32f20ed
|
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 sha256:52acb6d543ff5fdb99c731bea89a5d2499fee1cdf9577497042d302270267fa2
|
3 |
size 1064
|
trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch":
|
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,
|
@@ -1893,12 +1893,1898 @@
|
|
1893 |
"eval_samples_per_second": 797.168,
|
1894 |
"eval_steps_per_second": 12.46,
|
1895 |
"step": 129030
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1896 |
}
|
1897 |
],
|
1898 |
"logging_steps": 500,
|
1899 |
-
"max_steps":
|
1900 |
"num_input_tokens_seen": 0,
|
1901 |
-
"num_train_epochs":
|
1902 |
"save_steps": 500,
|
1903 |
"stateful_callbacks": {
|
1904 |
"TrainerControl": {
|
@@ -1912,7 +3798,7 @@
|
|
1912 |
"attributes": {}
|
1913 |
}
|
1914 |
},
|
1915 |
-
"total_flos":
|
1916 |
"train_batch_size": 64,
|
1917 |
"trial_name": null,
|
1918 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 20.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 258060,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
1893 |
"eval_samples_per_second": 797.168,
|
1894 |
"eval_steps_per_second": 12.46,
|
1895 |
"step": 129030
|
1896 |
+
},
|
1897 |
+
{
|
1898 |
+
"epoch": 10.036425637448655,
|
1899 |
+
"grad_norm": 6.090233325958252,
|
1900 |
+
"learning_rate": 9.963574362551346e-06,
|
1901 |
+
"loss": 2.4784,
|
1902 |
+
"step": 129500
|
1903 |
+
},
|
1904 |
+
{
|
1905 |
+
"epoch": 10.075176315585523,
|
1906 |
+
"grad_norm": 6.285606384277344,
|
1907 |
+
"learning_rate": 9.924823684414478e-06,
|
1908 |
+
"loss": 2.4657,
|
1909 |
+
"step": 130000
|
1910 |
+
},
|
1911 |
+
{
|
1912 |
+
"epoch": 10.11392699372239,
|
1913 |
+
"grad_norm": 5.937399864196777,
|
1914 |
+
"learning_rate": 9.886073006277611e-06,
|
1915 |
+
"loss": 2.4869,
|
1916 |
+
"step": 130500
|
1917 |
+
},
|
1918 |
+
{
|
1919 |
+
"epoch": 10.152677671859257,
|
1920 |
+
"grad_norm": 7.235742568969727,
|
1921 |
+
"learning_rate": 9.847322328140743e-06,
|
1922 |
+
"loss": 2.4726,
|
1923 |
+
"step": 131000
|
1924 |
+
},
|
1925 |
+
{
|
1926 |
+
"epoch": 10.191428349996125,
|
1927 |
+
"grad_norm": 6.6334028244018555,
|
1928 |
+
"learning_rate": 9.808571650003877e-06,
|
1929 |
+
"loss": 2.472,
|
1930 |
+
"step": 131500
|
1931 |
+
},
|
1932 |
+
{
|
1933 |
+
"epoch": 10.230179028132993,
|
1934 |
+
"grad_norm": 7.366402626037598,
|
1935 |
+
"learning_rate": 9.769820971867009e-06,
|
1936 |
+
"loss": 2.4887,
|
1937 |
+
"step": 132000
|
1938 |
+
},
|
1939 |
+
{
|
1940 |
+
"epoch": 10.26892970626986,
|
1941 |
+
"grad_norm": 6.17592716217041,
|
1942 |
+
"learning_rate": 9.731070293730142e-06,
|
1943 |
+
"loss": 2.4854,
|
1944 |
+
"step": 132500
|
1945 |
+
},
|
1946 |
+
{
|
1947 |
+
"epoch": 10.307680384406726,
|
1948 |
+
"grad_norm": 6.376716613769531,
|
1949 |
+
"learning_rate": 9.692319615593274e-06,
|
1950 |
+
"loss": 2.486,
|
1951 |
+
"step": 133000
|
1952 |
+
},
|
1953 |
+
{
|
1954 |
+
"epoch": 10.346431062543594,
|
1955 |
+
"grad_norm": 6.293849945068359,
|
1956 |
+
"learning_rate": 9.653568937456407e-06,
|
1957 |
+
"loss": 2.4707,
|
1958 |
+
"step": 133500
|
1959 |
+
},
|
1960 |
+
{
|
1961 |
+
"epoch": 10.385181740680462,
|
1962 |
+
"grad_norm": 6.606166839599609,
|
1963 |
+
"learning_rate": 9.614818259319539e-06,
|
1964 |
+
"loss": 2.4704,
|
1965 |
+
"step": 134000
|
1966 |
+
},
|
1967 |
+
{
|
1968 |
+
"epoch": 10.42393241881733,
|
1969 |
+
"grad_norm": 6.805929660797119,
|
1970 |
+
"learning_rate": 9.576067581182673e-06,
|
1971 |
+
"loss": 2.4727,
|
1972 |
+
"step": 134500
|
1973 |
+
},
|
1974 |
+
{
|
1975 |
+
"epoch": 10.462683096954196,
|
1976 |
+
"grad_norm": 6.598349571228027,
|
1977 |
+
"learning_rate": 9.537316903045804e-06,
|
1978 |
+
"loss": 2.4825,
|
1979 |
+
"step": 135000
|
1980 |
+
},
|
1981 |
+
{
|
1982 |
+
"epoch": 10.501433775091064,
|
1983 |
+
"grad_norm": 5.807904243469238,
|
1984 |
+
"learning_rate": 9.498566224908938e-06,
|
1985 |
+
"loss": 2.4721,
|
1986 |
+
"step": 135500
|
1987 |
+
},
|
1988 |
+
{
|
1989 |
+
"epoch": 10.540184453227932,
|
1990 |
+
"grad_norm": 6.681980609893799,
|
1991 |
+
"learning_rate": 9.45981554677207e-06,
|
1992 |
+
"loss": 2.4764,
|
1993 |
+
"step": 136000
|
1994 |
+
},
|
1995 |
+
{
|
1996 |
+
"epoch": 10.5789351313648,
|
1997 |
+
"grad_norm": 6.540719032287598,
|
1998 |
+
"learning_rate": 9.421064868635203e-06,
|
1999 |
+
"loss": 2.4545,
|
2000 |
+
"step": 136500
|
2001 |
+
},
|
2002 |
+
{
|
2003 |
+
"epoch": 10.617685809501666,
|
2004 |
+
"grad_norm": 6.627035140991211,
|
2005 |
+
"learning_rate": 9.382314190498335e-06,
|
2006 |
+
"loss": 2.4778,
|
2007 |
+
"step": 137000
|
2008 |
+
},
|
2009 |
+
{
|
2010 |
+
"epoch": 10.656436487638533,
|
2011 |
+
"grad_norm": 6.348284721374512,
|
2012 |
+
"learning_rate": 9.343563512361468e-06,
|
2013 |
+
"loss": 2.4597,
|
2014 |
+
"step": 137500
|
2015 |
+
},
|
2016 |
+
{
|
2017 |
+
"epoch": 10.695187165775401,
|
2018 |
+
"grad_norm": 6.790314197540283,
|
2019 |
+
"learning_rate": 9.3048128342246e-06,
|
2020 |
+
"loss": 2.471,
|
2021 |
+
"step": 138000
|
2022 |
+
},
|
2023 |
+
{
|
2024 |
+
"epoch": 10.733937843912269,
|
2025 |
+
"grad_norm": 6.8181233406066895,
|
2026 |
+
"learning_rate": 9.266062156087732e-06,
|
2027 |
+
"loss": 2.4571,
|
2028 |
+
"step": 138500
|
2029 |
+
},
|
2030 |
+
{
|
2031 |
+
"epoch": 10.772688522049135,
|
2032 |
+
"grad_norm": 6.593683242797852,
|
2033 |
+
"learning_rate": 9.227311477950864e-06,
|
2034 |
+
"loss": 2.4843,
|
2035 |
+
"step": 139000
|
2036 |
+
},
|
2037 |
+
{
|
2038 |
+
"epoch": 10.811439200186003,
|
2039 |
+
"grad_norm": 6.600128650665283,
|
2040 |
+
"learning_rate": 9.188560799813997e-06,
|
2041 |
+
"loss": 2.464,
|
2042 |
+
"step": 139500
|
2043 |
+
},
|
2044 |
+
{
|
2045 |
+
"epoch": 10.85018987832287,
|
2046 |
+
"grad_norm": 6.368162631988525,
|
2047 |
+
"learning_rate": 9.149810121677129e-06,
|
2048 |
+
"loss": 2.4598,
|
2049 |
+
"step": 140000
|
2050 |
+
},
|
2051 |
+
{
|
2052 |
+
"epoch": 10.888940556459739,
|
2053 |
+
"grad_norm": 6.5435943603515625,
|
2054 |
+
"learning_rate": 9.111059443540262e-06,
|
2055 |
+
"loss": 2.4704,
|
2056 |
+
"step": 140500
|
2057 |
+
},
|
2058 |
+
{
|
2059 |
+
"epoch": 10.927691234596605,
|
2060 |
+
"grad_norm": 6.06011962890625,
|
2061 |
+
"learning_rate": 9.072308765403394e-06,
|
2062 |
+
"loss": 2.4514,
|
2063 |
+
"step": 141000
|
2064 |
+
},
|
2065 |
+
{
|
2066 |
+
"epoch": 10.966441912733472,
|
2067 |
+
"grad_norm": 7.2288689613342285,
|
2068 |
+
"learning_rate": 9.033558087266528e-06,
|
2069 |
+
"loss": 2.4521,
|
2070 |
+
"step": 141500
|
2071 |
+
},
|
2072 |
+
{
|
2073 |
+
"epoch": 11.0,
|
2074 |
+
"eval_loss": 2.3912322521209717,
|
2075 |
+
"eval_runtime": 258.9953,
|
2076 |
+
"eval_samples_per_second": 797.176,
|
2077 |
+
"eval_steps_per_second": 12.46,
|
2078 |
+
"step": 141933
|
2079 |
+
},
|
2080 |
+
{
|
2081 |
+
"epoch": 11.00519259087034,
|
2082 |
+
"grad_norm": 6.698403358459473,
|
2083 |
+
"learning_rate": 8.99480740912966e-06,
|
2084 |
+
"loss": 2.4457,
|
2085 |
+
"step": 142000
|
2086 |
+
},
|
2087 |
+
{
|
2088 |
+
"epoch": 11.043943269007208,
|
2089 |
+
"grad_norm": 6.455236911773682,
|
2090 |
+
"learning_rate": 8.956056730992793e-06,
|
2091 |
+
"loss": 2.4507,
|
2092 |
+
"step": 142500
|
2093 |
+
},
|
2094 |
+
{
|
2095 |
+
"epoch": 11.082693947144074,
|
2096 |
+
"grad_norm": 6.590576648712158,
|
2097 |
+
"learning_rate": 8.917306052855925e-06,
|
2098 |
+
"loss": 2.4256,
|
2099 |
+
"step": 143000
|
2100 |
+
},
|
2101 |
+
{
|
2102 |
+
"epoch": 11.121444625280942,
|
2103 |
+
"grad_norm": 6.957404136657715,
|
2104 |
+
"learning_rate": 8.878555374719058e-06,
|
2105 |
+
"loss": 2.4549,
|
2106 |
+
"step": 143500
|
2107 |
+
},
|
2108 |
+
{
|
2109 |
+
"epoch": 11.16019530341781,
|
2110 |
+
"grad_norm": 6.926699161529541,
|
2111 |
+
"learning_rate": 8.83980469658219e-06,
|
2112 |
+
"loss": 2.4499,
|
2113 |
+
"step": 144000
|
2114 |
+
},
|
2115 |
+
{
|
2116 |
+
"epoch": 11.198945981554678,
|
2117 |
+
"grad_norm": 6.484086036682129,
|
2118 |
+
"learning_rate": 8.801054018445324e-06,
|
2119 |
+
"loss": 2.4443,
|
2120 |
+
"step": 144500
|
2121 |
+
},
|
2122 |
+
{
|
2123 |
+
"epoch": 11.237696659691544,
|
2124 |
+
"grad_norm": 6.107706069946289,
|
2125 |
+
"learning_rate": 8.762303340308455e-06,
|
2126 |
+
"loss": 2.4459,
|
2127 |
+
"step": 145000
|
2128 |
+
},
|
2129 |
+
{
|
2130 |
+
"epoch": 11.276447337828412,
|
2131 |
+
"grad_norm": 7.301278591156006,
|
2132 |
+
"learning_rate": 8.723552662171589e-06,
|
2133 |
+
"loss": 2.4463,
|
2134 |
+
"step": 145500
|
2135 |
+
},
|
2136 |
+
{
|
2137 |
+
"epoch": 11.31519801596528,
|
2138 |
+
"grad_norm": 6.378045082092285,
|
2139 |
+
"learning_rate": 8.68480198403472e-06,
|
2140 |
+
"loss": 2.4494,
|
2141 |
+
"step": 146000
|
2142 |
+
},
|
2143 |
+
{
|
2144 |
+
"epoch": 11.353948694102147,
|
2145 |
+
"grad_norm": 6.803300857543945,
|
2146 |
+
"learning_rate": 8.646051305897854e-06,
|
2147 |
+
"loss": 2.4235,
|
2148 |
+
"step": 146500
|
2149 |
+
},
|
2150 |
+
{
|
2151 |
+
"epoch": 11.392699372239015,
|
2152 |
+
"grad_norm": 6.401794910430908,
|
2153 |
+
"learning_rate": 8.607300627760986e-06,
|
2154 |
+
"loss": 2.4353,
|
2155 |
+
"step": 147000
|
2156 |
+
},
|
2157 |
+
{
|
2158 |
+
"epoch": 11.431450050375881,
|
2159 |
+
"grad_norm": 6.455550193786621,
|
2160 |
+
"learning_rate": 8.56854994962412e-06,
|
2161 |
+
"loss": 2.4306,
|
2162 |
+
"step": 147500
|
2163 |
+
},
|
2164 |
+
{
|
2165 |
+
"epoch": 11.470200728512749,
|
2166 |
+
"grad_norm": 6.416442394256592,
|
2167 |
+
"learning_rate": 8.529799271487251e-06,
|
2168 |
+
"loss": 2.4143,
|
2169 |
+
"step": 148000
|
2170 |
+
},
|
2171 |
+
{
|
2172 |
+
"epoch": 11.508951406649617,
|
2173 |
+
"grad_norm": 6.768812656402588,
|
2174 |
+
"learning_rate": 8.491048593350385e-06,
|
2175 |
+
"loss": 2.4184,
|
2176 |
+
"step": 148500
|
2177 |
+
},
|
2178 |
+
{
|
2179 |
+
"epoch": 11.547702084786483,
|
2180 |
+
"grad_norm": 6.085323810577393,
|
2181 |
+
"learning_rate": 8.452297915213516e-06,
|
2182 |
+
"loss": 2.4318,
|
2183 |
+
"step": 149000
|
2184 |
+
},
|
2185 |
+
{
|
2186 |
+
"epoch": 11.58645276292335,
|
2187 |
+
"grad_norm": 6.181857585906982,
|
2188 |
+
"learning_rate": 8.41354723707665e-06,
|
2189 |
+
"loss": 2.4348,
|
2190 |
+
"step": 149500
|
2191 |
+
},
|
2192 |
+
{
|
2193 |
+
"epoch": 11.625203441060219,
|
2194 |
+
"grad_norm": 6.558756351470947,
|
2195 |
+
"learning_rate": 8.374796558939782e-06,
|
2196 |
+
"loss": 2.413,
|
2197 |
+
"step": 150000
|
2198 |
+
},
|
2199 |
+
{
|
2200 |
+
"epoch": 11.663954119197086,
|
2201 |
+
"grad_norm": 6.249685287475586,
|
2202 |
+
"learning_rate": 8.336045880802915e-06,
|
2203 |
+
"loss": 2.4271,
|
2204 |
+
"step": 150500
|
2205 |
+
},
|
2206 |
+
{
|
2207 |
+
"epoch": 11.702704797333954,
|
2208 |
+
"grad_norm": 6.789103984832764,
|
2209 |
+
"learning_rate": 8.297295202666047e-06,
|
2210 |
+
"loss": 2.4226,
|
2211 |
+
"step": 151000
|
2212 |
+
},
|
2213 |
+
{
|
2214 |
+
"epoch": 11.74145547547082,
|
2215 |
+
"grad_norm": 6.4289140701293945,
|
2216 |
+
"learning_rate": 8.25854452452918e-06,
|
2217 |
+
"loss": 2.4184,
|
2218 |
+
"step": 151500
|
2219 |
+
},
|
2220 |
+
{
|
2221 |
+
"epoch": 11.780206153607688,
|
2222 |
+
"grad_norm": 6.098612308502197,
|
2223 |
+
"learning_rate": 8.219793846392312e-06,
|
2224 |
+
"loss": 2.4132,
|
2225 |
+
"step": 152000
|
2226 |
+
},
|
2227 |
+
{
|
2228 |
+
"epoch": 11.818956831744556,
|
2229 |
+
"grad_norm": 6.500378608703613,
|
2230 |
+
"learning_rate": 8.181043168255444e-06,
|
2231 |
+
"loss": 2.4184,
|
2232 |
+
"step": 152500
|
2233 |
+
},
|
2234 |
+
{
|
2235 |
+
"epoch": 11.857707509881424,
|
2236 |
+
"grad_norm": 6.583259105682373,
|
2237 |
+
"learning_rate": 8.142292490118577e-06,
|
2238 |
+
"loss": 2.4259,
|
2239 |
+
"step": 153000
|
2240 |
+
},
|
2241 |
+
{
|
2242 |
+
"epoch": 11.89645818801829,
|
2243 |
+
"grad_norm": 6.7018303871154785,
|
2244 |
+
"learning_rate": 8.10354181198171e-06,
|
2245 |
+
"loss": 2.4185,
|
2246 |
+
"step": 153500
|
2247 |
+
},
|
2248 |
+
{
|
2249 |
+
"epoch": 11.935208866155158,
|
2250 |
+
"grad_norm": 6.679374694824219,
|
2251 |
+
"learning_rate": 8.064791133844843e-06,
|
2252 |
+
"loss": 2.4078,
|
2253 |
+
"step": 154000
|
2254 |
+
},
|
2255 |
+
{
|
2256 |
+
"epoch": 11.973959544292025,
|
2257 |
+
"grad_norm": 6.576003551483154,
|
2258 |
+
"learning_rate": 8.026040455707974e-06,
|
2259 |
+
"loss": 2.4212,
|
2260 |
+
"step": 154500
|
2261 |
+
},
|
2262 |
+
{
|
2263 |
+
"epoch": 12.0,
|
2264 |
+
"eval_loss": 2.3491039276123047,
|
2265 |
+
"eval_runtime": 260.4232,
|
2266 |
+
"eval_samples_per_second": 792.806,
|
2267 |
+
"eval_steps_per_second": 12.391,
|
2268 |
+
"step": 154836
|
2269 |
+
},
|
2270 |
+
{
|
2271 |
+
"epoch": 12.012710222428893,
|
2272 |
+
"grad_norm": 6.768045902252197,
|
2273 |
+
"learning_rate": 7.987289777571108e-06,
|
2274 |
+
"loss": 2.399,
|
2275 |
+
"step": 155000
|
2276 |
+
},
|
2277 |
+
{
|
2278 |
+
"epoch": 12.05146090056576,
|
2279 |
+
"grad_norm": 6.445169925689697,
|
2280 |
+
"learning_rate": 7.94853909943424e-06,
|
2281 |
+
"loss": 2.4055,
|
2282 |
+
"step": 155500
|
2283 |
+
},
|
2284 |
+
{
|
2285 |
+
"epoch": 12.090211578702627,
|
2286 |
+
"grad_norm": 6.684764385223389,
|
2287 |
+
"learning_rate": 7.909788421297373e-06,
|
2288 |
+
"loss": 2.3979,
|
2289 |
+
"step": 156000
|
2290 |
+
},
|
2291 |
+
{
|
2292 |
+
"epoch": 12.128962256839495,
|
2293 |
+
"grad_norm": 7.150822162628174,
|
2294 |
+
"learning_rate": 7.871037743160505e-06,
|
2295 |
+
"loss": 2.4091,
|
2296 |
+
"step": 156500
|
2297 |
+
},
|
2298 |
+
{
|
2299 |
+
"epoch": 12.167712934976363,
|
2300 |
+
"grad_norm": 6.7067131996154785,
|
2301 |
+
"learning_rate": 7.832287065023639e-06,
|
2302 |
+
"loss": 2.4057,
|
2303 |
+
"step": 157000
|
2304 |
+
},
|
2305 |
+
{
|
2306 |
+
"epoch": 12.206463613113229,
|
2307 |
+
"grad_norm": 6.288236141204834,
|
2308 |
+
"learning_rate": 7.79353638688677e-06,
|
2309 |
+
"loss": 2.4024,
|
2310 |
+
"step": 157500
|
2311 |
+
},
|
2312 |
+
{
|
2313 |
+
"epoch": 12.245214291250097,
|
2314 |
+
"grad_norm": 6.532754898071289,
|
2315 |
+
"learning_rate": 7.754785708749904e-06,
|
2316 |
+
"loss": 2.4119,
|
2317 |
+
"step": 158000
|
2318 |
+
},
|
2319 |
+
{
|
2320 |
+
"epoch": 12.283964969386965,
|
2321 |
+
"grad_norm": 6.437507629394531,
|
2322 |
+
"learning_rate": 7.716035030613036e-06,
|
2323 |
+
"loss": 2.4048,
|
2324 |
+
"step": 158500
|
2325 |
+
},
|
2326 |
+
{
|
2327 |
+
"epoch": 12.322715647523832,
|
2328 |
+
"grad_norm": 6.648064136505127,
|
2329 |
+
"learning_rate": 7.677284352476169e-06,
|
2330 |
+
"loss": 2.3954,
|
2331 |
+
"step": 159000
|
2332 |
+
},
|
2333 |
+
{
|
2334 |
+
"epoch": 12.361466325660698,
|
2335 |
+
"grad_norm": 6.406070232391357,
|
2336 |
+
"learning_rate": 7.6385336743393e-06,
|
2337 |
+
"loss": 2.4069,
|
2338 |
+
"step": 159500
|
2339 |
+
},
|
2340 |
+
{
|
2341 |
+
"epoch": 12.400217003797566,
|
2342 |
+
"grad_norm": 6.75925350189209,
|
2343 |
+
"learning_rate": 7.5997829962024335e-06,
|
2344 |
+
"loss": 2.3803,
|
2345 |
+
"step": 160000
|
2346 |
+
},
|
2347 |
+
{
|
2348 |
+
"epoch": 12.438967681934434,
|
2349 |
+
"grad_norm": 7.390876770019531,
|
2350 |
+
"learning_rate": 7.561032318065566e-06,
|
2351 |
+
"loss": 2.3952,
|
2352 |
+
"step": 160500
|
2353 |
+
},
|
2354 |
+
{
|
2355 |
+
"epoch": 12.477718360071302,
|
2356 |
+
"grad_norm": 6.584438800811768,
|
2357 |
+
"learning_rate": 7.522281639928699e-06,
|
2358 |
+
"loss": 2.3921,
|
2359 |
+
"step": 161000
|
2360 |
+
},
|
2361 |
+
{
|
2362 |
+
"epoch": 12.516469038208168,
|
2363 |
+
"grad_norm": 6.7814040184021,
|
2364 |
+
"learning_rate": 7.483530961791831e-06,
|
2365 |
+
"loss": 2.4035,
|
2366 |
+
"step": 161500
|
2367 |
+
},
|
2368 |
+
{
|
2369 |
+
"epoch": 12.555219716345036,
|
2370 |
+
"grad_norm": 6.544926166534424,
|
2371 |
+
"learning_rate": 7.444780283654964e-06,
|
2372 |
+
"loss": 2.3855,
|
2373 |
+
"step": 162000
|
2374 |
+
},
|
2375 |
+
{
|
2376 |
+
"epoch": 12.593970394481904,
|
2377 |
+
"grad_norm": 6.649155139923096,
|
2378 |
+
"learning_rate": 7.406029605518097e-06,
|
2379 |
+
"loss": 2.3884,
|
2380 |
+
"step": 162500
|
2381 |
+
},
|
2382 |
+
{
|
2383 |
+
"epoch": 12.632721072618772,
|
2384 |
+
"grad_norm": 6.128752708435059,
|
2385 |
+
"learning_rate": 7.367278927381229e-06,
|
2386 |
+
"loss": 2.3915,
|
2387 |
+
"step": 163000
|
2388 |
+
},
|
2389 |
+
{
|
2390 |
+
"epoch": 12.671471750755638,
|
2391 |
+
"grad_norm": 6.694360733032227,
|
2392 |
+
"learning_rate": 7.328528249244362e-06,
|
2393 |
+
"loss": 2.4065,
|
2394 |
+
"step": 163500
|
2395 |
+
},
|
2396 |
+
{
|
2397 |
+
"epoch": 12.710222428892505,
|
2398 |
+
"grad_norm": 6.9979963302612305,
|
2399 |
+
"learning_rate": 7.2897775711074945e-06,
|
2400 |
+
"loss": 2.3816,
|
2401 |
+
"step": 164000
|
2402 |
+
},
|
2403 |
+
{
|
2404 |
+
"epoch": 12.748973107029373,
|
2405 |
+
"grad_norm": 6.7657294273376465,
|
2406 |
+
"learning_rate": 7.251026892970627e-06,
|
2407 |
+
"loss": 2.385,
|
2408 |
+
"step": 164500
|
2409 |
+
},
|
2410 |
+
{
|
2411 |
+
"epoch": 12.787723785166241,
|
2412 |
+
"grad_norm": 7.142265796661377,
|
2413 |
+
"learning_rate": 7.21227621483376e-06,
|
2414 |
+
"loss": 2.3809,
|
2415 |
+
"step": 165000
|
2416 |
+
},
|
2417 |
+
{
|
2418 |
+
"epoch": 12.826474463303107,
|
2419 |
+
"grad_norm": 6.2213134765625,
|
2420 |
+
"learning_rate": 7.1735255366968924e-06,
|
2421 |
+
"loss": 2.3883,
|
2422 |
+
"step": 165500
|
2423 |
+
},
|
2424 |
+
{
|
2425 |
+
"epoch": 12.865225141439975,
|
2426 |
+
"grad_norm": 6.274342060089111,
|
2427 |
+
"learning_rate": 7.134774858560025e-06,
|
2428 |
+
"loss": 2.3838,
|
2429 |
+
"step": 166000
|
2430 |
+
},
|
2431 |
+
{
|
2432 |
+
"epoch": 12.903975819576843,
|
2433 |
+
"grad_norm": 6.5893049240112305,
|
2434 |
+
"learning_rate": 7.096024180423158e-06,
|
2435 |
+
"loss": 2.3832,
|
2436 |
+
"step": 166500
|
2437 |
+
},
|
2438 |
+
{
|
2439 |
+
"epoch": 12.94272649771371,
|
2440 |
+
"grad_norm": 6.229060173034668,
|
2441 |
+
"learning_rate": 7.05727350228629e-06,
|
2442 |
+
"loss": 2.3839,
|
2443 |
+
"step": 167000
|
2444 |
+
},
|
2445 |
+
{
|
2446 |
+
"epoch": 12.981477175850577,
|
2447 |
+
"grad_norm": 7.251420497894287,
|
2448 |
+
"learning_rate": 7.018522824149423e-06,
|
2449 |
+
"loss": 2.3838,
|
2450 |
+
"step": 167500
|
2451 |
+
},
|
2452 |
+
{
|
2453 |
+
"epoch": 13.0,
|
2454 |
+
"eval_loss": 2.3215689659118652,
|
2455 |
+
"eval_runtime": 259.7568,
|
2456 |
+
"eval_samples_per_second": 794.84,
|
2457 |
+
"eval_steps_per_second": 12.423,
|
2458 |
+
"step": 167739
|
2459 |
+
},
|
2460 |
+
{
|
2461 |
+
"epoch": 13.020227853987445,
|
2462 |
+
"grad_norm": 5.944735050201416,
|
2463 |
+
"learning_rate": 6.979772146012556e-06,
|
2464 |
+
"loss": 2.3687,
|
2465 |
+
"step": 168000
|
2466 |
+
},
|
2467 |
+
{
|
2468 |
+
"epoch": 13.058978532124312,
|
2469 |
+
"grad_norm": 6.25685977935791,
|
2470 |
+
"learning_rate": 6.941021467875688e-06,
|
2471 |
+
"loss": 2.3761,
|
2472 |
+
"step": 168500
|
2473 |
+
},
|
2474 |
+
{
|
2475 |
+
"epoch": 13.09772921026118,
|
2476 |
+
"grad_norm": 6.244680881500244,
|
2477 |
+
"learning_rate": 6.902270789738821e-06,
|
2478 |
+
"loss": 2.3463,
|
2479 |
+
"step": 169000
|
2480 |
+
},
|
2481 |
+
{
|
2482 |
+
"epoch": 13.136479888398046,
|
2483 |
+
"grad_norm": 6.370804309844971,
|
2484 |
+
"learning_rate": 6.8635201116019535e-06,
|
2485 |
+
"loss": 2.3597,
|
2486 |
+
"step": 169500
|
2487 |
+
},
|
2488 |
+
{
|
2489 |
+
"epoch": 13.175230566534914,
|
2490 |
+
"grad_norm": 6.249234676361084,
|
2491 |
+
"learning_rate": 6.824769433465086e-06,
|
2492 |
+
"loss": 2.3679,
|
2493 |
+
"step": 170000
|
2494 |
+
},
|
2495 |
+
{
|
2496 |
+
"epoch": 13.213981244671782,
|
2497 |
+
"grad_norm": 6.973300933837891,
|
2498 |
+
"learning_rate": 6.786018755328219e-06,
|
2499 |
+
"loss": 2.3669,
|
2500 |
+
"step": 170500
|
2501 |
+
},
|
2502 |
+
{
|
2503 |
+
"epoch": 13.25273192280865,
|
2504 |
+
"grad_norm": 7.319492816925049,
|
2505 |
+
"learning_rate": 6.747268077191351e-06,
|
2506 |
+
"loss": 2.3528,
|
2507 |
+
"step": 171000
|
2508 |
+
},
|
2509 |
+
{
|
2510 |
+
"epoch": 13.291482600945516,
|
2511 |
+
"grad_norm": 6.924526214599609,
|
2512 |
+
"learning_rate": 6.708517399054484e-06,
|
2513 |
+
"loss": 2.3662,
|
2514 |
+
"step": 171500
|
2515 |
+
},
|
2516 |
+
{
|
2517 |
+
"epoch": 13.330233279082384,
|
2518 |
+
"grad_norm": 6.761091709136963,
|
2519 |
+
"learning_rate": 6.669766720917617e-06,
|
2520 |
+
"loss": 2.3608,
|
2521 |
+
"step": 172000
|
2522 |
+
},
|
2523 |
+
{
|
2524 |
+
"epoch": 13.368983957219251,
|
2525 |
+
"grad_norm": 6.105197429656982,
|
2526 |
+
"learning_rate": 6.631016042780749e-06,
|
2527 |
+
"loss": 2.3536,
|
2528 |
+
"step": 172500
|
2529 |
+
},
|
2530 |
+
{
|
2531 |
+
"epoch": 13.40773463535612,
|
2532 |
+
"grad_norm": 6.724457740783691,
|
2533 |
+
"learning_rate": 6.592265364643882e-06,
|
2534 |
+
"loss": 2.3682,
|
2535 |
+
"step": 173000
|
2536 |
+
},
|
2537 |
+
{
|
2538 |
+
"epoch": 13.446485313492985,
|
2539 |
+
"grad_norm": 6.62090539932251,
|
2540 |
+
"learning_rate": 6.553514686507015e-06,
|
2541 |
+
"loss": 2.3549,
|
2542 |
+
"step": 173500
|
2543 |
+
},
|
2544 |
+
{
|
2545 |
+
"epoch": 13.485235991629853,
|
2546 |
+
"grad_norm": 6.862425327301025,
|
2547 |
+
"learning_rate": 6.514764008370147e-06,
|
2548 |
+
"loss": 2.3475,
|
2549 |
+
"step": 174000
|
2550 |
+
},
|
2551 |
+
{
|
2552 |
+
"epoch": 13.523986669766721,
|
2553 |
+
"grad_norm": 6.164032936096191,
|
2554 |
+
"learning_rate": 6.47601333023328e-06,
|
2555 |
+
"loss": 2.3625,
|
2556 |
+
"step": 174500
|
2557 |
+
},
|
2558 |
+
{
|
2559 |
+
"epoch": 13.562737347903589,
|
2560 |
+
"grad_norm": 7.522220134735107,
|
2561 |
+
"learning_rate": 6.4372626520964125e-06,
|
2562 |
+
"loss": 2.3676,
|
2563 |
+
"step": 175000
|
2564 |
+
},
|
2565 |
+
{
|
2566 |
+
"epoch": 13.601488026040455,
|
2567 |
+
"grad_norm": 6.564206600189209,
|
2568 |
+
"learning_rate": 6.398511973959545e-06,
|
2569 |
+
"loss": 2.3606,
|
2570 |
+
"step": 175500
|
2571 |
+
},
|
2572 |
+
{
|
2573 |
+
"epoch": 13.640238704177323,
|
2574 |
+
"grad_norm": 6.069074630737305,
|
2575 |
+
"learning_rate": 6.359761295822677e-06,
|
2576 |
+
"loss": 2.3644,
|
2577 |
+
"step": 176000
|
2578 |
+
},
|
2579 |
+
{
|
2580 |
+
"epoch": 13.67898938231419,
|
2581 |
+
"grad_norm": 6.570771217346191,
|
2582 |
+
"learning_rate": 6.3210106176858095e-06,
|
2583 |
+
"loss": 2.3711,
|
2584 |
+
"step": 176500
|
2585 |
+
},
|
2586 |
+
{
|
2587 |
+
"epoch": 13.717740060451058,
|
2588 |
+
"grad_norm": 6.1281609535217285,
|
2589 |
+
"learning_rate": 6.282259939548942e-06,
|
2590 |
+
"loss": 2.348,
|
2591 |
+
"step": 177000
|
2592 |
+
},
|
2593 |
+
{
|
2594 |
+
"epoch": 13.756490738587924,
|
2595 |
+
"grad_norm": 6.176905632019043,
|
2596 |
+
"learning_rate": 6.243509261412075e-06,
|
2597 |
+
"loss": 2.379,
|
2598 |
+
"step": 177500
|
2599 |
+
},
|
2600 |
+
{
|
2601 |
+
"epoch": 13.795241416724792,
|
2602 |
+
"grad_norm": 7.890781402587891,
|
2603 |
+
"learning_rate": 6.2047585832752074e-06,
|
2604 |
+
"loss": 2.365,
|
2605 |
+
"step": 178000
|
2606 |
+
},
|
2607 |
+
{
|
2608 |
+
"epoch": 13.83399209486166,
|
2609 |
+
"grad_norm": 6.160940647125244,
|
2610 |
+
"learning_rate": 6.16600790513834e-06,
|
2611 |
+
"loss": 2.3391,
|
2612 |
+
"step": 178500
|
2613 |
+
},
|
2614 |
+
{
|
2615 |
+
"epoch": 13.872742772998528,
|
2616 |
+
"grad_norm": 6.732828617095947,
|
2617 |
+
"learning_rate": 6.127257227001473e-06,
|
2618 |
+
"loss": 2.355,
|
2619 |
+
"step": 179000
|
2620 |
+
},
|
2621 |
+
{
|
2622 |
+
"epoch": 13.911493451135394,
|
2623 |
+
"grad_norm": 6.500529766082764,
|
2624 |
+
"learning_rate": 6.088506548864605e-06,
|
2625 |
+
"loss": 2.3512,
|
2626 |
+
"step": 179500
|
2627 |
+
},
|
2628 |
+
{
|
2629 |
+
"epoch": 13.950244129272262,
|
2630 |
+
"grad_norm": 7.362790584564209,
|
2631 |
+
"learning_rate": 6.049755870727738e-06,
|
2632 |
+
"loss": 2.3654,
|
2633 |
+
"step": 180000
|
2634 |
+
},
|
2635 |
+
{
|
2636 |
+
"epoch": 13.98899480740913,
|
2637 |
+
"grad_norm": 7.070291519165039,
|
2638 |
+
"learning_rate": 6.011005192590871e-06,
|
2639 |
+
"loss": 2.3444,
|
2640 |
+
"step": 180500
|
2641 |
+
},
|
2642 |
+
{
|
2643 |
+
"epoch": 14.0,
|
2644 |
+
"eval_loss": 2.2924630641937256,
|
2645 |
+
"eval_runtime": 259.3076,
|
2646 |
+
"eval_samples_per_second": 796.217,
|
2647 |
+
"eval_steps_per_second": 12.445,
|
2648 |
+
"step": 180642
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 14.027745485545998,
|
2652 |
+
"grad_norm": 7.284486293792725,
|
2653 |
+
"learning_rate": 5.972254514454003e-06,
|
2654 |
+
"loss": 2.3296,
|
2655 |
+
"step": 181000
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 14.066496163682864,
|
2659 |
+
"grad_norm": 7.636621952056885,
|
2660 |
+
"learning_rate": 5.933503836317136e-06,
|
2661 |
+
"loss": 2.3314,
|
2662 |
+
"step": 181500
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 14.105246841819731,
|
2666 |
+
"grad_norm": 6.692602634429932,
|
2667 |
+
"learning_rate": 5.8947531581802685e-06,
|
2668 |
+
"loss": 2.3363,
|
2669 |
+
"step": 182000
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 14.1439975199566,
|
2673 |
+
"grad_norm": 6.751750469207764,
|
2674 |
+
"learning_rate": 5.856002480043401e-06,
|
2675 |
+
"loss": 2.3174,
|
2676 |
+
"step": 182500
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 14.182748198093467,
|
2680 |
+
"grad_norm": 7.041817665100098,
|
2681 |
+
"learning_rate": 5.817251801906534e-06,
|
2682 |
+
"loss": 2.3295,
|
2683 |
+
"step": 183000
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 14.221498876230335,
|
2687 |
+
"grad_norm": 7.414912700653076,
|
2688 |
+
"learning_rate": 5.778501123769666e-06,
|
2689 |
+
"loss": 2.3386,
|
2690 |
+
"step": 183500
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 14.260249554367201,
|
2694 |
+
"grad_norm": 7.009491920471191,
|
2695 |
+
"learning_rate": 5.739750445632799e-06,
|
2696 |
+
"loss": 2.3282,
|
2697 |
+
"step": 184000
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 14.299000232504069,
|
2701 |
+
"grad_norm": 6.77699089050293,
|
2702 |
+
"learning_rate": 5.700999767495932e-06,
|
2703 |
+
"loss": 2.3323,
|
2704 |
+
"step": 184500
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 14.337750910640937,
|
2708 |
+
"grad_norm": 6.922458171844482,
|
2709 |
+
"learning_rate": 5.662249089359064e-06,
|
2710 |
+
"loss": 2.3545,
|
2711 |
+
"step": 185000
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 14.376501588777803,
|
2715 |
+
"grad_norm": 7.635495185852051,
|
2716 |
+
"learning_rate": 5.623498411222197e-06,
|
2717 |
+
"loss": 2.3429,
|
2718 |
+
"step": 185500
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 14.41525226691467,
|
2722 |
+
"grad_norm": 6.657200813293457,
|
2723 |
+
"learning_rate": 5.58474773308533e-06,
|
2724 |
+
"loss": 2.3371,
|
2725 |
+
"step": 186000
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 14.454002945051538,
|
2729 |
+
"grad_norm": 6.328368663787842,
|
2730 |
+
"learning_rate": 5.545997054948462e-06,
|
2731 |
+
"loss": 2.3225,
|
2732 |
+
"step": 186500
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 14.492753623188406,
|
2736 |
+
"grad_norm": 6.7084503173828125,
|
2737 |
+
"learning_rate": 5.507246376811595e-06,
|
2738 |
+
"loss": 2.3141,
|
2739 |
+
"step": 187000
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 14.531504301325274,
|
2743 |
+
"grad_norm": 6.23046875,
|
2744 |
+
"learning_rate": 5.4684956986747275e-06,
|
2745 |
+
"loss": 2.3387,
|
2746 |
+
"step": 187500
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 14.57025497946214,
|
2750 |
+
"grad_norm": 6.53918981552124,
|
2751 |
+
"learning_rate": 5.42974502053786e-06,
|
2752 |
+
"loss": 2.3355,
|
2753 |
+
"step": 188000
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 14.609005657599008,
|
2757 |
+
"grad_norm": 6.816432952880859,
|
2758 |
+
"learning_rate": 5.390994342400993e-06,
|
2759 |
+
"loss": 2.3409,
|
2760 |
+
"step": 188500
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 14.647756335735876,
|
2764 |
+
"grad_norm": 6.9504475593566895,
|
2765 |
+
"learning_rate": 5.352243664264125e-06,
|
2766 |
+
"loss": 2.3274,
|
2767 |
+
"step": 189000
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 14.686507013872744,
|
2771 |
+
"grad_norm": 7.058226585388184,
|
2772 |
+
"learning_rate": 5.313492986127258e-06,
|
2773 |
+
"loss": 2.3295,
|
2774 |
+
"step": 189500
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 14.72525769200961,
|
2778 |
+
"grad_norm": 6.337547302246094,
|
2779 |
+
"learning_rate": 5.274742307990391e-06,
|
2780 |
+
"loss": 2.316,
|
2781 |
+
"step": 190000
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 14.764008370146477,
|
2785 |
+
"grad_norm": 7.420670032501221,
|
2786 |
+
"learning_rate": 5.235991629853523e-06,
|
2787 |
+
"loss": 2.3313,
|
2788 |
+
"step": 190500
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 14.802759048283345,
|
2792 |
+
"grad_norm": 6.559388637542725,
|
2793 |
+
"learning_rate": 5.197240951716656e-06,
|
2794 |
+
"loss": 2.3368,
|
2795 |
+
"step": 191000
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 14.841509726420213,
|
2799 |
+
"grad_norm": 6.416265487670898,
|
2800 |
+
"learning_rate": 5.1584902735797886e-06,
|
2801 |
+
"loss": 2.3139,
|
2802 |
+
"step": 191500
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 14.88026040455708,
|
2806 |
+
"grad_norm": 6.204991817474365,
|
2807 |
+
"learning_rate": 5.11973959544292e-06,
|
2808 |
+
"loss": 2.3209,
|
2809 |
+
"step": 192000
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 14.919011082693947,
|
2813 |
+
"grad_norm": 7.657558441162109,
|
2814 |
+
"learning_rate": 5.080988917306053e-06,
|
2815 |
+
"loss": 2.3346,
|
2816 |
+
"step": 192500
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 14.957761760830815,
|
2820 |
+
"grad_norm": 6.812448024749756,
|
2821 |
+
"learning_rate": 5.042238239169186e-06,
|
2822 |
+
"loss": 2.3226,
|
2823 |
+
"step": 193000
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 14.996512438967683,
|
2827 |
+
"grad_norm": 5.866453170776367,
|
2828 |
+
"learning_rate": 5.003487561032318e-06,
|
2829 |
+
"loss": 2.3034,
|
2830 |
+
"step": 193500
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 15.0,
|
2834 |
+
"eval_loss": 2.2758021354675293,
|
2835 |
+
"eval_runtime": 268.9287,
|
2836 |
+
"eval_samples_per_second": 767.731,
|
2837 |
+
"eval_steps_per_second": 11.999,
|
2838 |
+
"step": 193545
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 15.035263117104549,
|
2842 |
+
"grad_norm": 6.998913288116455,
|
2843 |
+
"learning_rate": 4.964736882895451e-06,
|
2844 |
+
"loss": 2.3103,
|
2845 |
+
"step": 194000
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 15.074013795241417,
|
2849 |
+
"grad_norm": 7.022980213165283,
|
2850 |
+
"learning_rate": 4.9259862047585835e-06,
|
2851 |
+
"loss": 2.3121,
|
2852 |
+
"step": 194500
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 15.112764473378284,
|
2856 |
+
"grad_norm": 6.3553056716918945,
|
2857 |
+
"learning_rate": 4.887235526621716e-06,
|
2858 |
+
"loss": 2.325,
|
2859 |
+
"step": 195000
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 15.151515151515152,
|
2863 |
+
"grad_norm": 7.574887752532959,
|
2864 |
+
"learning_rate": 4.848484848484849e-06,
|
2865 |
+
"loss": 2.3128,
|
2866 |
+
"step": 195500
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 15.190265829652018,
|
2870 |
+
"grad_norm": 6.3977556228637695,
|
2871 |
+
"learning_rate": 4.809734170347981e-06,
|
2872 |
+
"loss": 2.3058,
|
2873 |
+
"step": 196000
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 15.229016507788886,
|
2877 |
+
"grad_norm": 6.198862552642822,
|
2878 |
+
"learning_rate": 4.770983492211114e-06,
|
2879 |
+
"loss": 2.3111,
|
2880 |
+
"step": 196500
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 15.267767185925754,
|
2884 |
+
"grad_norm": 7.1892499923706055,
|
2885 |
+
"learning_rate": 4.732232814074247e-06,
|
2886 |
+
"loss": 2.3181,
|
2887 |
+
"step": 197000
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 15.306517864062622,
|
2891 |
+
"grad_norm": 6.773824214935303,
|
2892 |
+
"learning_rate": 4.693482135937379e-06,
|
2893 |
+
"loss": 2.3158,
|
2894 |
+
"step": 197500
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 15.345268542199488,
|
2898 |
+
"grad_norm": 6.595972537994385,
|
2899 |
+
"learning_rate": 4.654731457800512e-06,
|
2900 |
+
"loss": 2.2989,
|
2901 |
+
"step": 198000
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 15.384019220336356,
|
2905 |
+
"grad_norm": 7.397641658782959,
|
2906 |
+
"learning_rate": 4.615980779663645e-06,
|
2907 |
+
"loss": 2.3143,
|
2908 |
+
"step": 198500
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 15.422769898473224,
|
2912 |
+
"grad_norm": 7.2511820793151855,
|
2913 |
+
"learning_rate": 4.577230101526777e-06,
|
2914 |
+
"loss": 2.3077,
|
2915 |
+
"step": 199000
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 15.461520576610091,
|
2919 |
+
"grad_norm": 6.52310848236084,
|
2920 |
+
"learning_rate": 4.53847942338991e-06,
|
2921 |
+
"loss": 2.3062,
|
2922 |
+
"step": 199500
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 15.500271254746957,
|
2926 |
+
"grad_norm": 6.681788921356201,
|
2927 |
+
"learning_rate": 4.4997287452530425e-06,
|
2928 |
+
"loss": 2.3078,
|
2929 |
+
"step": 200000
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 15.539021932883825,
|
2933 |
+
"grad_norm": 7.010565280914307,
|
2934 |
+
"learning_rate": 4.460978067116175e-06,
|
2935 |
+
"loss": 2.3031,
|
2936 |
+
"step": 200500
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 15.577772611020693,
|
2940 |
+
"grad_norm": 7.412187576293945,
|
2941 |
+
"learning_rate": 4.422227388979308e-06,
|
2942 |
+
"loss": 2.3029,
|
2943 |
+
"step": 201000
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 15.616523289157561,
|
2947 |
+
"grad_norm": 6.586581707000732,
|
2948 |
+
"learning_rate": 4.38347671084244e-06,
|
2949 |
+
"loss": 2.3092,
|
2950 |
+
"step": 201500
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 15.655273967294427,
|
2954 |
+
"grad_norm": 6.430338382720947,
|
2955 |
+
"learning_rate": 4.344726032705573e-06,
|
2956 |
+
"loss": 2.2972,
|
2957 |
+
"step": 202000
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 15.694024645431295,
|
2961 |
+
"grad_norm": 6.151809215545654,
|
2962 |
+
"learning_rate": 4.305975354568706e-06,
|
2963 |
+
"loss": 2.2972,
|
2964 |
+
"step": 202500
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 15.732775323568163,
|
2968 |
+
"grad_norm": 7.195096492767334,
|
2969 |
+
"learning_rate": 4.267224676431838e-06,
|
2970 |
+
"loss": 2.3045,
|
2971 |
+
"step": 203000
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 15.77152600170503,
|
2975 |
+
"grad_norm": 6.76158332824707,
|
2976 |
+
"learning_rate": 4.228473998294971e-06,
|
2977 |
+
"loss": 2.2995,
|
2978 |
+
"step": 203500
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 15.810276679841897,
|
2982 |
+
"grad_norm": 6.710601329803467,
|
2983 |
+
"learning_rate": 4.1897233201581036e-06,
|
2984 |
+
"loss": 2.3045,
|
2985 |
+
"step": 204000
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 15.849027357978764,
|
2989 |
+
"grad_norm": 6.813743591308594,
|
2990 |
+
"learning_rate": 4.150972642021236e-06,
|
2991 |
+
"loss": 2.3114,
|
2992 |
+
"step": 204500
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 15.887778036115632,
|
2996 |
+
"grad_norm": 7.168315410614014,
|
2997 |
+
"learning_rate": 4.112221963884369e-06,
|
2998 |
+
"loss": 2.2995,
|
2999 |
+
"step": 205000
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 15.9265287142525,
|
3003 |
+
"grad_norm": 6.606774806976318,
|
3004 |
+
"learning_rate": 4.0734712857475015e-06,
|
3005 |
+
"loss": 2.3023,
|
3006 |
+
"step": 205500
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 15.965279392389366,
|
3010 |
+
"grad_norm": 6.852230548858643,
|
3011 |
+
"learning_rate": 4.034720607610634e-06,
|
3012 |
+
"loss": 2.311,
|
3013 |
+
"step": 206000
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 16.0,
|
3017 |
+
"eval_loss": 2.252058982849121,
|
3018 |
+
"eval_runtime": 272.2097,
|
3019 |
+
"eval_samples_per_second": 758.478,
|
3020 |
+
"eval_steps_per_second": 11.855,
|
3021 |
+
"step": 206448
|
3022 |
+
},
|
3023 |
+
{
|
3024 |
+
"epoch": 16.004030070526234,
|
3025 |
+
"grad_norm": 7.245954990386963,
|
3026 |
+
"learning_rate": 3.995969929473767e-06,
|
3027 |
+
"loss": 2.2896,
|
3028 |
+
"step": 206500
|
3029 |
+
},
|
3030 |
+
{
|
3031 |
+
"epoch": 16.0427807486631,
|
3032 |
+
"grad_norm": 6.094116687774658,
|
3033 |
+
"learning_rate": 3.957219251336899e-06,
|
3034 |
+
"loss": 2.2999,
|
3035 |
+
"step": 207000
|
3036 |
+
},
|
3037 |
+
{
|
3038 |
+
"epoch": 16.08153142679997,
|
3039 |
+
"grad_norm": 6.302695274353027,
|
3040 |
+
"learning_rate": 3.918468573200031e-06,
|
3041 |
+
"loss": 2.3017,
|
3042 |
+
"step": 207500
|
3043 |
+
},
|
3044 |
+
{
|
3045 |
+
"epoch": 16.120282104936837,
|
3046 |
+
"grad_norm": 6.800222873687744,
|
3047 |
+
"learning_rate": 3.879717895063164e-06,
|
3048 |
+
"loss": 2.2876,
|
3049 |
+
"step": 208000
|
3050 |
+
},
|
3051 |
+
{
|
3052 |
+
"epoch": 16.159032783073705,
|
3053 |
+
"grad_norm": 7.139950752258301,
|
3054 |
+
"learning_rate": 3.840967216926296e-06,
|
3055 |
+
"loss": 2.2975,
|
3056 |
+
"step": 208500
|
3057 |
+
},
|
3058 |
+
{
|
3059 |
+
"epoch": 16.19778346121057,
|
3060 |
+
"grad_norm": 6.805322170257568,
|
3061 |
+
"learning_rate": 3.802216538789429e-06,
|
3062 |
+
"loss": 2.2994,
|
3063 |
+
"step": 209000
|
3064 |
+
},
|
3065 |
+
{
|
3066 |
+
"epoch": 16.236534139347437,
|
3067 |
+
"grad_norm": 6.6877336502075195,
|
3068 |
+
"learning_rate": 3.7634658606525617e-06,
|
3069 |
+
"loss": 2.277,
|
3070 |
+
"step": 209500
|
3071 |
+
},
|
3072 |
+
{
|
3073 |
+
"epoch": 16.275284817484305,
|
3074 |
+
"grad_norm": 6.0831193923950195,
|
3075 |
+
"learning_rate": 3.724715182515694e-06,
|
3076 |
+
"loss": 2.3029,
|
3077 |
+
"step": 210000
|
3078 |
+
},
|
3079 |
+
{
|
3080 |
+
"epoch": 16.314035495621173,
|
3081 |
+
"grad_norm": 6.021631240844727,
|
3082 |
+
"learning_rate": 3.6859645043788265e-06,
|
3083 |
+
"loss": 2.2959,
|
3084 |
+
"step": 210500
|
3085 |
+
},
|
3086 |
+
{
|
3087 |
+
"epoch": 16.35278617375804,
|
3088 |
+
"grad_norm": 7.072383403778076,
|
3089 |
+
"learning_rate": 3.647213826241959e-06,
|
3090 |
+
"loss": 2.2709,
|
3091 |
+
"step": 211000
|
3092 |
+
},
|
3093 |
+
{
|
3094 |
+
"epoch": 16.39153685189491,
|
3095 |
+
"grad_norm": 6.0719404220581055,
|
3096 |
+
"learning_rate": 3.608463148105092e-06,
|
3097 |
+
"loss": 2.2952,
|
3098 |
+
"step": 211500
|
3099 |
+
},
|
3100 |
+
{
|
3101 |
+
"epoch": 16.430287530031777,
|
3102 |
+
"grad_norm": 6.733717441558838,
|
3103 |
+
"learning_rate": 3.5697124699682244e-06,
|
3104 |
+
"loss": 2.2984,
|
3105 |
+
"step": 212000
|
3106 |
+
},
|
3107 |
+
{
|
3108 |
+
"epoch": 16.469038208168644,
|
3109 |
+
"grad_norm": 7.269532203674316,
|
3110 |
+
"learning_rate": 3.530961791831357e-06,
|
3111 |
+
"loss": 2.2855,
|
3112 |
+
"step": 212500
|
3113 |
+
},
|
3114 |
+
{
|
3115 |
+
"epoch": 16.50778888630551,
|
3116 |
+
"grad_norm": 7.440357208251953,
|
3117 |
+
"learning_rate": 3.4922111136944897e-06,
|
3118 |
+
"loss": 2.2747,
|
3119 |
+
"step": 213000
|
3120 |
+
},
|
3121 |
+
{
|
3122 |
+
"epoch": 16.546539564442377,
|
3123 |
+
"grad_norm": 7.448116302490234,
|
3124 |
+
"learning_rate": 3.4534604355576223e-06,
|
3125 |
+
"loss": 2.2933,
|
3126 |
+
"step": 213500
|
3127 |
+
},
|
3128 |
+
{
|
3129 |
+
"epoch": 16.585290242579244,
|
3130 |
+
"grad_norm": 6.202878475189209,
|
3131 |
+
"learning_rate": 3.414709757420755e-06,
|
3132 |
+
"loss": 2.2963,
|
3133 |
+
"step": 214000
|
3134 |
+
},
|
3135 |
+
{
|
3136 |
+
"epoch": 16.624040920716112,
|
3137 |
+
"grad_norm": 7.019168376922607,
|
3138 |
+
"learning_rate": 3.3759590792838876e-06,
|
3139 |
+
"loss": 2.2667,
|
3140 |
+
"step": 214500
|
3141 |
+
},
|
3142 |
+
{
|
3143 |
+
"epoch": 16.66279159885298,
|
3144 |
+
"grad_norm": 6.448665142059326,
|
3145 |
+
"learning_rate": 3.3372084011470202e-06,
|
3146 |
+
"loss": 2.2905,
|
3147 |
+
"step": 215000
|
3148 |
+
},
|
3149 |
+
{
|
3150 |
+
"epoch": 16.701542276989848,
|
3151 |
+
"grad_norm": 6.160965442657471,
|
3152 |
+
"learning_rate": 3.298457723010153e-06,
|
3153 |
+
"loss": 2.2854,
|
3154 |
+
"step": 215500
|
3155 |
+
},
|
3156 |
+
{
|
3157 |
+
"epoch": 16.740292955126716,
|
3158 |
+
"grad_norm": 6.956637859344482,
|
3159 |
+
"learning_rate": 3.2597070448732855e-06,
|
3160 |
+
"loss": 2.2944,
|
3161 |
+
"step": 216000
|
3162 |
+
},
|
3163 |
+
{
|
3164 |
+
"epoch": 16.779043633263584,
|
3165 |
+
"grad_norm": 6.935131549835205,
|
3166 |
+
"learning_rate": 3.220956366736418e-06,
|
3167 |
+
"loss": 2.2795,
|
3168 |
+
"step": 216500
|
3169 |
+
},
|
3170 |
+
{
|
3171 |
+
"epoch": 16.817794311400448,
|
3172 |
+
"grad_norm": 6.656859397888184,
|
3173 |
+
"learning_rate": 3.1822056885995508e-06,
|
3174 |
+
"loss": 2.2872,
|
3175 |
+
"step": 217000
|
3176 |
+
},
|
3177 |
+
{
|
3178 |
+
"epoch": 16.856544989537316,
|
3179 |
+
"grad_norm": 6.204549312591553,
|
3180 |
+
"learning_rate": 3.1434550104626834e-06,
|
3181 |
+
"loss": 2.2832,
|
3182 |
+
"step": 217500
|
3183 |
+
},
|
3184 |
+
{
|
3185 |
+
"epoch": 16.895295667674183,
|
3186 |
+
"grad_norm": 6.77413272857666,
|
3187 |
+
"learning_rate": 3.1047043323258156e-06,
|
3188 |
+
"loss": 2.2719,
|
3189 |
+
"step": 218000
|
3190 |
+
},
|
3191 |
+
{
|
3192 |
+
"epoch": 16.93404634581105,
|
3193 |
+
"grad_norm": 6.447382926940918,
|
3194 |
+
"learning_rate": 3.0659536541889482e-06,
|
3195 |
+
"loss": 2.2702,
|
3196 |
+
"step": 218500
|
3197 |
+
},
|
3198 |
+
{
|
3199 |
+
"epoch": 16.97279702394792,
|
3200 |
+
"grad_norm": 7.396761894226074,
|
3201 |
+
"learning_rate": 3.027202976052081e-06,
|
3202 |
+
"loss": 2.2813,
|
3203 |
+
"step": 219000
|
3204 |
+
},
|
3205 |
+
{
|
3206 |
+
"epoch": 17.0,
|
3207 |
+
"eval_loss": 2.2362165451049805,
|
3208 |
+
"eval_runtime": 266.8391,
|
3209 |
+
"eval_samples_per_second": 773.743,
|
3210 |
+
"eval_steps_per_second": 12.093,
|
3211 |
+
"step": 219351
|
3212 |
+
},
|
3213 |
+
{
|
3214 |
+
"epoch": 17.011547702084787,
|
3215 |
+
"grad_norm": 6.575385093688965,
|
3216 |
+
"learning_rate": 2.9884522979152135e-06,
|
3217 |
+
"loss": 2.2747,
|
3218 |
+
"step": 219500
|
3219 |
+
},
|
3220 |
+
{
|
3221 |
+
"epoch": 17.050298380221655,
|
3222 |
+
"grad_norm": 7.168444633483887,
|
3223 |
+
"learning_rate": 2.949701619778346e-06,
|
3224 |
+
"loss": 2.2868,
|
3225 |
+
"step": 220000
|
3226 |
+
},
|
3227 |
+
{
|
3228 |
+
"epoch": 17.089049058358523,
|
3229 |
+
"grad_norm": 7.069167613983154,
|
3230 |
+
"learning_rate": 2.9109509416414788e-06,
|
3231 |
+
"loss": 2.2838,
|
3232 |
+
"step": 220500
|
3233 |
+
},
|
3234 |
+
{
|
3235 |
+
"epoch": 17.12779973649539,
|
3236 |
+
"grad_norm": 6.792834758758545,
|
3237 |
+
"learning_rate": 2.8722002635046114e-06,
|
3238 |
+
"loss": 2.2836,
|
3239 |
+
"step": 221000
|
3240 |
+
},
|
3241 |
+
{
|
3242 |
+
"epoch": 17.166550414632255,
|
3243 |
+
"grad_norm": 6.546488285064697,
|
3244 |
+
"learning_rate": 2.833449585367744e-06,
|
3245 |
+
"loss": 2.2733,
|
3246 |
+
"step": 221500
|
3247 |
+
},
|
3248 |
+
{
|
3249 |
+
"epoch": 17.205301092769123,
|
3250 |
+
"grad_norm": 6.293231010437012,
|
3251 |
+
"learning_rate": 2.7946989072308767e-06,
|
3252 |
+
"loss": 2.2688,
|
3253 |
+
"step": 222000
|
3254 |
+
},
|
3255 |
+
{
|
3256 |
+
"epoch": 17.24405177090599,
|
3257 |
+
"grad_norm": 6.560914039611816,
|
3258 |
+
"learning_rate": 2.7559482290940093e-06,
|
3259 |
+
"loss": 2.2787,
|
3260 |
+
"step": 222500
|
3261 |
+
},
|
3262 |
+
{
|
3263 |
+
"epoch": 17.28280244904286,
|
3264 |
+
"grad_norm": 6.571765422821045,
|
3265 |
+
"learning_rate": 2.717197550957142e-06,
|
3266 |
+
"loss": 2.2801,
|
3267 |
+
"step": 223000
|
3268 |
+
},
|
3269 |
+
{
|
3270 |
+
"epoch": 17.321553127179726,
|
3271 |
+
"grad_norm": 7.396661281585693,
|
3272 |
+
"learning_rate": 2.6784468728202746e-06,
|
3273 |
+
"loss": 2.28,
|
3274 |
+
"step": 223500
|
3275 |
+
},
|
3276 |
+
{
|
3277 |
+
"epoch": 17.360303805316594,
|
3278 |
+
"grad_norm": 6.239862442016602,
|
3279 |
+
"learning_rate": 2.6396961946834072e-06,
|
3280 |
+
"loss": 2.2743,
|
3281 |
+
"step": 224000
|
3282 |
+
},
|
3283 |
+
{
|
3284 |
+
"epoch": 17.39905448345346,
|
3285 |
+
"grad_norm": 6.766594886779785,
|
3286 |
+
"learning_rate": 2.60094551654654e-06,
|
3287 |
+
"loss": 2.2456,
|
3288 |
+
"step": 224500
|
3289 |
+
},
|
3290 |
+
{
|
3291 |
+
"epoch": 17.43780516159033,
|
3292 |
+
"grad_norm": 6.488914966583252,
|
3293 |
+
"learning_rate": 2.5621948384096725e-06,
|
3294 |
+
"loss": 2.2666,
|
3295 |
+
"step": 225000
|
3296 |
+
},
|
3297 |
+
{
|
3298 |
+
"epoch": 17.476555839727194,
|
3299 |
+
"grad_norm": 6.036900043487549,
|
3300 |
+
"learning_rate": 2.523444160272805e-06,
|
3301 |
+
"loss": 2.2577,
|
3302 |
+
"step": 225500
|
3303 |
+
},
|
3304 |
+
{
|
3305 |
+
"epoch": 17.51530651786406,
|
3306 |
+
"grad_norm": 6.977652549743652,
|
3307 |
+
"learning_rate": 2.4846934821359373e-06,
|
3308 |
+
"loss": 2.2657,
|
3309 |
+
"step": 226000
|
3310 |
+
},
|
3311 |
+
{
|
3312 |
+
"epoch": 17.55405719600093,
|
3313 |
+
"grad_norm": 6.468418121337891,
|
3314 |
+
"learning_rate": 2.44594280399907e-06,
|
3315 |
+
"loss": 2.2737,
|
3316 |
+
"step": 226500
|
3317 |
+
},
|
3318 |
+
{
|
3319 |
+
"epoch": 17.592807874137797,
|
3320 |
+
"grad_norm": 6.7042646408081055,
|
3321 |
+
"learning_rate": 2.4071921258622026e-06,
|
3322 |
+
"loss": 2.2685,
|
3323 |
+
"step": 227000
|
3324 |
+
},
|
3325 |
+
{
|
3326 |
+
"epoch": 17.631558552274665,
|
3327 |
+
"grad_norm": 6.591056823730469,
|
3328 |
+
"learning_rate": 2.3684414477253352e-06,
|
3329 |
+
"loss": 2.2836,
|
3330 |
+
"step": 227500
|
3331 |
+
},
|
3332 |
+
{
|
3333 |
+
"epoch": 17.670309230411533,
|
3334 |
+
"grad_norm": 7.078721523284912,
|
3335 |
+
"learning_rate": 2.329690769588468e-06,
|
3336 |
+
"loss": 2.2754,
|
3337 |
+
"step": 228000
|
3338 |
+
},
|
3339 |
+
{
|
3340 |
+
"epoch": 17.7090599085484,
|
3341 |
+
"grad_norm": 6.701901435852051,
|
3342 |
+
"learning_rate": 2.2909400914516005e-06,
|
3343 |
+
"loss": 2.2494,
|
3344 |
+
"step": 228500
|
3345 |
+
},
|
3346 |
+
{
|
3347 |
+
"epoch": 17.74781058668527,
|
3348 |
+
"grad_norm": 6.622567176818848,
|
3349 |
+
"learning_rate": 2.252189413314733e-06,
|
3350 |
+
"loss": 2.2689,
|
3351 |
+
"step": 229000
|
3352 |
+
},
|
3353 |
+
{
|
3354 |
+
"epoch": 17.786561264822133,
|
3355 |
+
"grad_norm": 6.573280334472656,
|
3356 |
+
"learning_rate": 2.2134387351778658e-06,
|
3357 |
+
"loss": 2.271,
|
3358 |
+
"step": 229500
|
3359 |
+
},
|
3360 |
+
{
|
3361 |
+
"epoch": 17.825311942959,
|
3362 |
+
"grad_norm": 6.9067206382751465,
|
3363 |
+
"learning_rate": 2.1746880570409984e-06,
|
3364 |
+
"loss": 2.2573,
|
3365 |
+
"step": 230000
|
3366 |
+
},
|
3367 |
+
{
|
3368 |
+
"epoch": 17.86406262109587,
|
3369 |
+
"grad_norm": 6.601592063903809,
|
3370 |
+
"learning_rate": 2.135937378904131e-06,
|
3371 |
+
"loss": 2.2743,
|
3372 |
+
"step": 230500
|
3373 |
+
},
|
3374 |
+
{
|
3375 |
+
"epoch": 17.902813299232736,
|
3376 |
+
"grad_norm": 6.949497699737549,
|
3377 |
+
"learning_rate": 2.0971867007672637e-06,
|
3378 |
+
"loss": 2.2644,
|
3379 |
+
"step": 231000
|
3380 |
+
},
|
3381 |
+
{
|
3382 |
+
"epoch": 17.941563977369604,
|
3383 |
+
"grad_norm": 5.614126205444336,
|
3384 |
+
"learning_rate": 2.0584360226303963e-06,
|
3385 |
+
"loss": 2.2608,
|
3386 |
+
"step": 231500
|
3387 |
+
},
|
3388 |
+
{
|
3389 |
+
"epoch": 17.980314655506472,
|
3390 |
+
"grad_norm": 6.880855560302734,
|
3391 |
+
"learning_rate": 2.019685344493529e-06,
|
3392 |
+
"loss": 2.2862,
|
3393 |
+
"step": 232000
|
3394 |
+
},
|
3395 |
+
{
|
3396 |
+
"epoch": 18.0,
|
3397 |
+
"eval_loss": 2.229489326477051,
|
3398 |
+
"eval_runtime": 270.0811,
|
3399 |
+
"eval_samples_per_second": 764.456,
|
3400 |
+
"eval_steps_per_second": 11.948,
|
3401 |
+
"step": 232254
|
3402 |
+
},
|
3403 |
+
{
|
3404 |
+
"epoch": 18.01906533364334,
|
3405 |
+
"grad_norm": 6.630836486816406,
|
3406 |
+
"learning_rate": 1.9809346663566616e-06,
|
3407 |
+
"loss": 2.2632,
|
3408 |
+
"step": 232500
|
3409 |
+
},
|
3410 |
+
{
|
3411 |
+
"epoch": 18.057816011780208,
|
3412 |
+
"grad_norm": 6.50869607925415,
|
3413 |
+
"learning_rate": 1.942183988219794e-06,
|
3414 |
+
"loss": 2.2584,
|
3415 |
+
"step": 233000
|
3416 |
+
},
|
3417 |
+
{
|
3418 |
+
"epoch": 18.096566689917072,
|
3419 |
+
"grad_norm": 6.81369161605835,
|
3420 |
+
"learning_rate": 1.9034333100829266e-06,
|
3421 |
+
"loss": 2.2599,
|
3422 |
+
"step": 233500
|
3423 |
+
},
|
3424 |
+
{
|
3425 |
+
"epoch": 18.13531736805394,
|
3426 |
+
"grad_norm": 6.202197074890137,
|
3427 |
+
"learning_rate": 1.8646826319460593e-06,
|
3428 |
+
"loss": 2.2532,
|
3429 |
+
"step": 234000
|
3430 |
+
},
|
3431 |
+
{
|
3432 |
+
"epoch": 18.174068046190808,
|
3433 |
+
"grad_norm": 6.907183647155762,
|
3434 |
+
"learning_rate": 1.8259319538091919e-06,
|
3435 |
+
"loss": 2.2553,
|
3436 |
+
"step": 234500
|
3437 |
+
},
|
3438 |
+
{
|
3439 |
+
"epoch": 18.212818724327676,
|
3440 |
+
"grad_norm": 7.445714473724365,
|
3441 |
+
"learning_rate": 1.7871812756723245e-06,
|
3442 |
+
"loss": 2.2586,
|
3443 |
+
"step": 235000
|
3444 |
+
},
|
3445 |
+
{
|
3446 |
+
"epoch": 18.251569402464543,
|
3447 |
+
"grad_norm": 6.844184398651123,
|
3448 |
+
"learning_rate": 1.7484305975354572e-06,
|
3449 |
+
"loss": 2.2502,
|
3450 |
+
"step": 235500
|
3451 |
+
},
|
3452 |
+
{
|
3453 |
+
"epoch": 18.29032008060141,
|
3454 |
+
"grad_norm": 6.495091438293457,
|
3455 |
+
"learning_rate": 1.7096799193985896e-06,
|
3456 |
+
"loss": 2.2703,
|
3457 |
+
"step": 236000
|
3458 |
+
},
|
3459 |
+
{
|
3460 |
+
"epoch": 18.32907075873828,
|
3461 |
+
"grad_norm": 6.848631858825684,
|
3462 |
+
"learning_rate": 1.6709292412617222e-06,
|
3463 |
+
"loss": 2.2494,
|
3464 |
+
"step": 236500
|
3465 |
+
},
|
3466 |
+
{
|
3467 |
+
"epoch": 18.367821436875147,
|
3468 |
+
"grad_norm": 6.527080535888672,
|
3469 |
+
"learning_rate": 1.6321785631248548e-06,
|
3470 |
+
"loss": 2.2676,
|
3471 |
+
"step": 237000
|
3472 |
+
},
|
3473 |
+
{
|
3474 |
+
"epoch": 18.40657211501201,
|
3475 |
+
"grad_norm": 6.402927875518799,
|
3476 |
+
"learning_rate": 1.5934278849879875e-06,
|
3477 |
+
"loss": 2.256,
|
3478 |
+
"step": 237500
|
3479 |
+
},
|
3480 |
+
{
|
3481 |
+
"epoch": 18.44532279314888,
|
3482 |
+
"grad_norm": 6.720060348510742,
|
3483 |
+
"learning_rate": 1.5546772068511201e-06,
|
3484 |
+
"loss": 2.256,
|
3485 |
+
"step": 238000
|
3486 |
+
},
|
3487 |
+
{
|
3488 |
+
"epoch": 18.484073471285747,
|
3489 |
+
"grad_norm": 6.392049312591553,
|
3490 |
+
"learning_rate": 1.5159265287142528e-06,
|
3491 |
+
"loss": 2.272,
|
3492 |
+
"step": 238500
|
3493 |
+
},
|
3494 |
+
{
|
3495 |
+
"epoch": 18.522824149422615,
|
3496 |
+
"grad_norm": 6.625200271606445,
|
3497 |
+
"learning_rate": 1.4771758505773854e-06,
|
3498 |
+
"loss": 2.2561,
|
3499 |
+
"step": 239000
|
3500 |
+
},
|
3501 |
+
{
|
3502 |
+
"epoch": 18.561574827559483,
|
3503 |
+
"grad_norm": 6.451653003692627,
|
3504 |
+
"learning_rate": 1.438425172440518e-06,
|
3505 |
+
"loss": 2.2518,
|
3506 |
+
"step": 239500
|
3507 |
+
},
|
3508 |
+
{
|
3509 |
+
"epoch": 18.60032550569635,
|
3510 |
+
"grad_norm": 6.246822357177734,
|
3511 |
+
"learning_rate": 1.3996744943036504e-06,
|
3512 |
+
"loss": 2.2541,
|
3513 |
+
"step": 240000
|
3514 |
+
},
|
3515 |
+
{
|
3516 |
+
"epoch": 18.639076183833218,
|
3517 |
+
"grad_norm": 6.265354156494141,
|
3518 |
+
"learning_rate": 1.360923816166783e-06,
|
3519 |
+
"loss": 2.2546,
|
3520 |
+
"step": 240500
|
3521 |
+
},
|
3522 |
+
{
|
3523 |
+
"epoch": 18.677826861970086,
|
3524 |
+
"grad_norm": 6.439133644104004,
|
3525 |
+
"learning_rate": 1.3221731380299157e-06,
|
3526 |
+
"loss": 2.2583,
|
3527 |
+
"step": 241000
|
3528 |
+
},
|
3529 |
+
{
|
3530 |
+
"epoch": 18.71657754010695,
|
3531 |
+
"grad_norm": 6.528525352478027,
|
3532 |
+
"learning_rate": 1.2834224598930483e-06,
|
3533 |
+
"loss": 2.2467,
|
3534 |
+
"step": 241500
|
3535 |
+
},
|
3536 |
+
{
|
3537 |
+
"epoch": 18.755328218243818,
|
3538 |
+
"grad_norm": 7.4315900802612305,
|
3539 |
+
"learning_rate": 1.2446717817561808e-06,
|
3540 |
+
"loss": 2.2585,
|
3541 |
+
"step": 242000
|
3542 |
+
},
|
3543 |
+
{
|
3544 |
+
"epoch": 18.794078896380686,
|
3545 |
+
"grad_norm": 7.4202141761779785,
|
3546 |
+
"learning_rate": 1.2059211036193134e-06,
|
3547 |
+
"loss": 2.2637,
|
3548 |
+
"step": 242500
|
3549 |
+
},
|
3550 |
+
{
|
3551 |
+
"epoch": 18.832829574517554,
|
3552 |
+
"grad_norm": 6.3204145431518555,
|
3553 |
+
"learning_rate": 1.167170425482446e-06,
|
3554 |
+
"loss": 2.264,
|
3555 |
+
"step": 243000
|
3556 |
+
},
|
3557 |
+
{
|
3558 |
+
"epoch": 18.87158025265442,
|
3559 |
+
"grad_norm": 6.220766067504883,
|
3560 |
+
"learning_rate": 1.1284197473455787e-06,
|
3561 |
+
"loss": 2.2705,
|
3562 |
+
"step": 243500
|
3563 |
+
},
|
3564 |
+
{
|
3565 |
+
"epoch": 18.91033093079129,
|
3566 |
+
"grad_norm": 6.558001518249512,
|
3567 |
+
"learning_rate": 1.0896690692087113e-06,
|
3568 |
+
"loss": 2.2632,
|
3569 |
+
"step": 244000
|
3570 |
+
},
|
3571 |
+
{
|
3572 |
+
"epoch": 18.949081608928157,
|
3573 |
+
"grad_norm": 6.786870956420898,
|
3574 |
+
"learning_rate": 1.050918391071844e-06,
|
3575 |
+
"loss": 2.2441,
|
3576 |
+
"step": 244500
|
3577 |
+
},
|
3578 |
+
{
|
3579 |
+
"epoch": 18.987832287065025,
|
3580 |
+
"grad_norm": 6.955057621002197,
|
3581 |
+
"learning_rate": 1.0121677129349766e-06,
|
3582 |
+
"loss": 2.2503,
|
3583 |
+
"step": 245000
|
3584 |
+
},
|
3585 |
+
{
|
3586 |
+
"epoch": 19.0,
|
3587 |
+
"eval_loss": 2.2231059074401855,
|
3588 |
+
"eval_runtime": 272.187,
|
3589 |
+
"eval_samples_per_second": 758.541,
|
3590 |
+
"eval_steps_per_second": 11.856,
|
3591 |
+
"step": 245157
|
3592 |
+
},
|
3593 |
+
{
|
3594 |
+
"epoch": 19.02658296520189,
|
3595 |
+
"grad_norm": 6.136529922485352,
|
3596 |
+
"learning_rate": 9.73417034798109e-07,
|
3597 |
+
"loss": 2.2528,
|
3598 |
+
"step": 245500
|
3599 |
+
},
|
3600 |
+
{
|
3601 |
+
"epoch": 19.065333643338757,
|
3602 |
+
"grad_norm": 7.144802093505859,
|
3603 |
+
"learning_rate": 9.346663566612417e-07,
|
3604 |
+
"loss": 2.248,
|
3605 |
+
"step": 246000
|
3606 |
+
},
|
3607 |
+
{
|
3608 |
+
"epoch": 19.104084321475625,
|
3609 |
+
"grad_norm": 5.582034111022949,
|
3610 |
+
"learning_rate": 8.959156785243743e-07,
|
3611 |
+
"loss": 2.2513,
|
3612 |
+
"step": 246500
|
3613 |
+
},
|
3614 |
+
{
|
3615 |
+
"epoch": 19.142834999612493,
|
3616 |
+
"grad_norm": 6.747804164886475,
|
3617 |
+
"learning_rate": 8.571650003875069e-07,
|
3618 |
+
"loss": 2.2647,
|
3619 |
+
"step": 247000
|
3620 |
+
},
|
3621 |
+
{
|
3622 |
+
"epoch": 19.18158567774936,
|
3623 |
+
"grad_norm": 6.1470417976379395,
|
3624 |
+
"learning_rate": 8.184143222506395e-07,
|
3625 |
+
"loss": 2.2548,
|
3626 |
+
"step": 247500
|
3627 |
+
},
|
3628 |
+
{
|
3629 |
+
"epoch": 19.22033635588623,
|
3630 |
+
"grad_norm": 6.574125289916992,
|
3631 |
+
"learning_rate": 7.79663644113772e-07,
|
3632 |
+
"loss": 2.2714,
|
3633 |
+
"step": 248000
|
3634 |
+
},
|
3635 |
+
{
|
3636 |
+
"epoch": 19.259087034023096,
|
3637 |
+
"grad_norm": 6.6587982177734375,
|
3638 |
+
"learning_rate": 7.409129659769046e-07,
|
3639 |
+
"loss": 2.2491,
|
3640 |
+
"step": 248500
|
3641 |
+
},
|
3642 |
+
{
|
3643 |
+
"epoch": 19.297837712159964,
|
3644 |
+
"grad_norm": 6.578282356262207,
|
3645 |
+
"learning_rate": 7.021622878400372e-07,
|
3646 |
+
"loss": 2.2483,
|
3647 |
+
"step": 249000
|
3648 |
+
},
|
3649 |
+
{
|
3650 |
+
"epoch": 19.33658839029683,
|
3651 |
+
"grad_norm": 6.449355602264404,
|
3652 |
+
"learning_rate": 6.634116097031699e-07,
|
3653 |
+
"loss": 2.2558,
|
3654 |
+
"step": 249500
|
3655 |
+
},
|
3656 |
+
{
|
3657 |
+
"epoch": 19.375339068433696,
|
3658 |
+
"grad_norm": 5.921240329742432,
|
3659 |
+
"learning_rate": 6.246609315663025e-07,
|
3660 |
+
"loss": 2.2428,
|
3661 |
+
"step": 250000
|
3662 |
+
},
|
3663 |
+
{
|
3664 |
+
"epoch": 19.414089746570564,
|
3665 |
+
"grad_norm": 6.655218124389648,
|
3666 |
+
"learning_rate": 5.859102534294351e-07,
|
3667 |
+
"loss": 2.2616,
|
3668 |
+
"step": 250500
|
3669 |
+
},
|
3670 |
+
{
|
3671 |
+
"epoch": 19.452840424707432,
|
3672 |
+
"grad_norm": 6.733659744262695,
|
3673 |
+
"learning_rate": 5.471595752925676e-07,
|
3674 |
+
"loss": 2.2481,
|
3675 |
+
"step": 251000
|
3676 |
+
},
|
3677 |
+
{
|
3678 |
+
"epoch": 19.4915911028443,
|
3679 |
+
"grad_norm": 6.9586968421936035,
|
3680 |
+
"learning_rate": 5.084088971557003e-07,
|
3681 |
+
"loss": 2.2495,
|
3682 |
+
"step": 251500
|
3683 |
+
},
|
3684 |
+
{
|
3685 |
+
"epoch": 19.530341780981168,
|
3686 |
+
"grad_norm": 6.441699028015137,
|
3687 |
+
"learning_rate": 4.6965821901883286e-07,
|
3688 |
+
"loss": 2.2456,
|
3689 |
+
"step": 252000
|
3690 |
+
},
|
3691 |
+
{
|
3692 |
+
"epoch": 19.569092459118036,
|
3693 |
+
"grad_norm": 6.126708984375,
|
3694 |
+
"learning_rate": 4.3090754088196544e-07,
|
3695 |
+
"loss": 2.2561,
|
3696 |
+
"step": 252500
|
3697 |
+
},
|
3698 |
+
{
|
3699 |
+
"epoch": 19.607843137254903,
|
3700 |
+
"grad_norm": 6.69553279876709,
|
3701 |
+
"learning_rate": 3.921568627450981e-07,
|
3702 |
+
"loss": 2.2435,
|
3703 |
+
"step": 253000
|
3704 |
+
},
|
3705 |
+
{
|
3706 |
+
"epoch": 19.646593815391768,
|
3707 |
+
"grad_norm": 7.468321800231934,
|
3708 |
+
"learning_rate": 3.5340618460823066e-07,
|
3709 |
+
"loss": 2.2517,
|
3710 |
+
"step": 253500
|
3711 |
+
},
|
3712 |
+
{
|
3713 |
+
"epoch": 19.685344493528635,
|
3714 |
+
"grad_norm": 6.724938869476318,
|
3715 |
+
"learning_rate": 3.146555064713633e-07,
|
3716 |
+
"loss": 2.2307,
|
3717 |
+
"step": 254000
|
3718 |
+
},
|
3719 |
+
{
|
3720 |
+
"epoch": 19.724095171665503,
|
3721 |
+
"grad_norm": 6.407966613769531,
|
3722 |
+
"learning_rate": 2.7590482833449587e-07,
|
3723 |
+
"loss": 2.2469,
|
3724 |
+
"step": 254500
|
3725 |
+
},
|
3726 |
+
{
|
3727 |
+
"epoch": 19.76284584980237,
|
3728 |
+
"grad_norm": 6.521556377410889,
|
3729 |
+
"learning_rate": 2.3715415019762845e-07,
|
3730 |
+
"loss": 2.2266,
|
3731 |
+
"step": 255000
|
3732 |
+
},
|
3733 |
+
{
|
3734 |
+
"epoch": 19.80159652793924,
|
3735 |
+
"grad_norm": 6.066943645477295,
|
3736 |
+
"learning_rate": 1.9840347206076106e-07,
|
3737 |
+
"loss": 2.2448,
|
3738 |
+
"step": 255500
|
3739 |
+
},
|
3740 |
+
{
|
3741 |
+
"epoch": 19.840347206076107,
|
3742 |
+
"grad_norm": 6.806808948516846,
|
3743 |
+
"learning_rate": 1.5965279392389367e-07,
|
3744 |
+
"loss": 2.2355,
|
3745 |
+
"step": 256000
|
3746 |
+
},
|
3747 |
+
{
|
3748 |
+
"epoch": 19.879097884212975,
|
3749 |
+
"grad_norm": 6.567816734313965,
|
3750 |
+
"learning_rate": 1.2090211578702627e-07,
|
3751 |
+
"loss": 2.2581,
|
3752 |
+
"step": 256500
|
3753 |
+
},
|
3754 |
+
{
|
3755 |
+
"epoch": 19.917848562349842,
|
3756 |
+
"grad_norm": 7.037693977355957,
|
3757 |
+
"learning_rate": 8.21514376501589e-08,
|
3758 |
+
"loss": 2.2558,
|
3759 |
+
"step": 257000
|
3760 |
+
},
|
3761 |
+
{
|
3762 |
+
"epoch": 19.95659924048671,
|
3763 |
+
"grad_norm": 7.10353422164917,
|
3764 |
+
"learning_rate": 4.340075951329148e-08,
|
3765 |
+
"loss": 2.2513,
|
3766 |
+
"step": 257500
|
3767 |
+
},
|
3768 |
+
{
|
3769 |
+
"epoch": 19.995349918623575,
|
3770 |
+
"grad_norm": 7.407341003417969,
|
3771 |
+
"learning_rate": 4.6500813764240875e-09,
|
3772 |
+
"loss": 2.2448,
|
3773 |
+
"step": 258000
|
3774 |
+
},
|
3775 |
+
{
|
3776 |
+
"epoch": 20.0,
|
3777 |
+
"eval_loss": 2.211674213409424,
|
3778 |
+
"eval_runtime": 270.2294,
|
3779 |
+
"eval_samples_per_second": 764.036,
|
3780 |
+
"eval_steps_per_second": 11.942,
|
3781 |
+
"step": 258060
|
3782 |
}
|
3783 |
],
|
3784 |
"logging_steps": 500,
|
3785 |
+
"max_steps": 258060,
|
3786 |
"num_input_tokens_seen": 0,
|
3787 |
+
"num_train_epochs": 20,
|
3788 |
"save_steps": 500,
|
3789 |
"stateful_callbacks": {
|
3790 |
"TrainerControl": {
|
|
|
3798 |
"attributes": {}
|
3799 |
}
|
3800 |
},
|
3801 |
+
"total_flos": 6.469828059827256e+17,
|
3802 |
"train_batch_size": 64,
|
3803 |
"trial_name": null,
|
3804 |
"trial_params": null
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5048
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:74ba329b03609527987c8126060f1c2c7e67ac7fad6da2fce4fad8d3324d3a3b
|
3 |
size 5048
|