Upload checkpoint-387090
Browse files- optimizer.pt +1 -1
- rng_state.pth +2 -2
- 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:da49aa28746ed8097072a44404d04cda00ada0bf98ca46ae77a286f9d797ae58
|
| 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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0ff4e93208fe1783462c5ae204a4cc834ea3fc02e3256b2aae94f996b93a622
|
| 3 |
+
size 14180
|
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:c53cdb4a080fe565430cda0918d011de3751759b213ac69153c506f7adfadce6
|
| 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,
|
|
@@ -3779,12 +3779,1898 @@
|
|
| 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":
|
| 3786 |
"num_input_tokens_seen": 0,
|
| 3787 |
-
"num_train_epochs":
|
| 3788 |
"save_steps": 500,
|
| 3789 |
"stateful_callbacks": {
|
| 3790 |
"TrainerControl": {
|
|
@@ -3798,7 +5684,7 @@
|
|
| 3798 |
"attributes": {}
|
| 3799 |
}
|
| 3800 |
},
|
| 3801 |
-
"total_flos":
|
| 3802 |
"train_batch_size": 64,
|
| 3803 |
"trial_name": null,
|
| 3804 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 387090,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3779 |
"eval_samples_per_second": 764.036,
|
| 3780 |
"eval_steps_per_second": 11.942,
|
| 3781 |
"step": 258060
|
| 3782 |
+
},
|
| 3783 |
+
{
|
| 3784 |
+
"epoch": 20.034100596760442,
|
| 3785 |
+
"grad_norm": 6.319613456726074,
|
| 3786 |
+
"learning_rate": 6.643932935493038e-06,
|
| 3787 |
+
"loss": 2.2464,
|
| 3788 |
+
"step": 258500
|
| 3789 |
+
},
|
| 3790 |
+
{
|
| 3791 |
+
"epoch": 20.07285127489731,
|
| 3792 |
+
"grad_norm": 10.947772026062012,
|
| 3793 |
+
"learning_rate": 6.61809915006846e-06,
|
| 3794 |
+
"loss": 2.2717,
|
| 3795 |
+
"step": 259000
|
| 3796 |
+
},
|
| 3797 |
+
{
|
| 3798 |
+
"epoch": 20.111601953034178,
|
| 3799 |
+
"grad_norm": 6.688451290130615,
|
| 3800 |
+
"learning_rate": 6.592265364643882e-06,
|
| 3801 |
+
"loss": 2.246,
|
| 3802 |
+
"step": 259500
|
| 3803 |
+
},
|
| 3804 |
+
{
|
| 3805 |
+
"epoch": 20.150352631171046,
|
| 3806 |
+
"grad_norm": 7.084783554077148,
|
| 3807 |
+
"learning_rate": 6.566431579219303e-06,
|
| 3808 |
+
"loss": 2.2547,
|
| 3809 |
+
"step": 260000
|
| 3810 |
+
},
|
| 3811 |
+
{
|
| 3812 |
+
"epoch": 20.189103309307914,
|
| 3813 |
+
"grad_norm": 7.182523250579834,
|
| 3814 |
+
"learning_rate": 6.540597793794725e-06,
|
| 3815 |
+
"loss": 2.2673,
|
| 3816 |
+
"step": 260500
|
| 3817 |
+
},
|
| 3818 |
+
{
|
| 3819 |
+
"epoch": 20.22785398744478,
|
| 3820 |
+
"grad_norm": 6.572226524353027,
|
| 3821 |
+
"learning_rate": 6.514764008370147e-06,
|
| 3822 |
+
"loss": 2.2696,
|
| 3823 |
+
"step": 261000
|
| 3824 |
+
},
|
| 3825 |
+
{
|
| 3826 |
+
"epoch": 20.26660466558165,
|
| 3827 |
+
"grad_norm": 6.861509323120117,
|
| 3828 |
+
"learning_rate": 6.488930222945569e-06,
|
| 3829 |
+
"loss": 2.2602,
|
| 3830 |
+
"step": 261500
|
| 3831 |
+
},
|
| 3832 |
+
{
|
| 3833 |
+
"epoch": 20.305355343718514,
|
| 3834 |
+
"grad_norm": 7.068969249725342,
|
| 3835 |
+
"learning_rate": 6.46309643752099e-06,
|
| 3836 |
+
"loss": 2.2736,
|
| 3837 |
+
"step": 262000
|
| 3838 |
+
},
|
| 3839 |
+
{
|
| 3840 |
+
"epoch": 20.34410602185538,
|
| 3841 |
+
"grad_norm": 6.5293660163879395,
|
| 3842 |
+
"learning_rate": 6.4372626520964125e-06,
|
| 3843 |
+
"loss": 2.2698,
|
| 3844 |
+
"step": 262500
|
| 3845 |
+
},
|
| 3846 |
+
{
|
| 3847 |
+
"epoch": 20.38285669999225,
|
| 3848 |
+
"grad_norm": 6.285311698913574,
|
| 3849 |
+
"learning_rate": 6.411428866671834e-06,
|
| 3850 |
+
"loss": 2.2671,
|
| 3851 |
+
"step": 263000
|
| 3852 |
+
},
|
| 3853 |
+
{
|
| 3854 |
+
"epoch": 20.421607378129117,
|
| 3855 |
+
"grad_norm": 6.466723918914795,
|
| 3856 |
+
"learning_rate": 6.3855950812472554e-06,
|
| 3857 |
+
"loss": 2.267,
|
| 3858 |
+
"step": 263500
|
| 3859 |
+
},
|
| 3860 |
+
{
|
| 3861 |
+
"epoch": 20.460358056265985,
|
| 3862 |
+
"grad_norm": 7.045479774475098,
|
| 3863 |
+
"learning_rate": 6.359761295822677e-06,
|
| 3864 |
+
"loss": 2.2499,
|
| 3865 |
+
"step": 264000
|
| 3866 |
+
},
|
| 3867 |
+
{
|
| 3868 |
+
"epoch": 20.499108734402853,
|
| 3869 |
+
"grad_norm": 7.05580472946167,
|
| 3870 |
+
"learning_rate": 6.333927510398099e-06,
|
| 3871 |
+
"loss": 2.2793,
|
| 3872 |
+
"step": 264500
|
| 3873 |
+
},
|
| 3874 |
+
{
|
| 3875 |
+
"epoch": 20.53785941253972,
|
| 3876 |
+
"grad_norm": 7.213685035705566,
|
| 3877 |
+
"learning_rate": 6.308093724973521e-06,
|
| 3878 |
+
"loss": 2.2519,
|
| 3879 |
+
"step": 265000
|
| 3880 |
+
},
|
| 3881 |
+
{
|
| 3882 |
+
"epoch": 20.57661009067659,
|
| 3883 |
+
"grad_norm": 6.6378984451293945,
|
| 3884 |
+
"learning_rate": 6.282259939548942e-06,
|
| 3885 |
+
"loss": 2.2699,
|
| 3886 |
+
"step": 265500
|
| 3887 |
+
},
|
| 3888 |
+
{
|
| 3889 |
+
"epoch": 20.615360768813453,
|
| 3890 |
+
"grad_norm": 6.8442463874816895,
|
| 3891 |
+
"learning_rate": 6.2564261541243645e-06,
|
| 3892 |
+
"loss": 2.2697,
|
| 3893 |
+
"step": 266000
|
| 3894 |
+
},
|
| 3895 |
+
{
|
| 3896 |
+
"epoch": 20.65411144695032,
|
| 3897 |
+
"grad_norm": 7.099138259887695,
|
| 3898 |
+
"learning_rate": 6.230592368699786e-06,
|
| 3899 |
+
"loss": 2.2622,
|
| 3900 |
+
"step": 266500
|
| 3901 |
+
},
|
| 3902 |
+
{
|
| 3903 |
+
"epoch": 20.69286212508719,
|
| 3904 |
+
"grad_norm": 6.572378635406494,
|
| 3905 |
+
"learning_rate": 6.2047585832752074e-06,
|
| 3906 |
+
"loss": 2.2709,
|
| 3907 |
+
"step": 267000
|
| 3908 |
+
},
|
| 3909 |
+
{
|
| 3910 |
+
"epoch": 20.731612803224056,
|
| 3911 |
+
"grad_norm": 6.410079479217529,
|
| 3912 |
+
"learning_rate": 6.17892479785063e-06,
|
| 3913 |
+
"loss": 2.2599,
|
| 3914 |
+
"step": 267500
|
| 3915 |
+
},
|
| 3916 |
+
{
|
| 3917 |
+
"epoch": 20.770363481360924,
|
| 3918 |
+
"grad_norm": 7.154236316680908,
|
| 3919 |
+
"learning_rate": 6.153091012426051e-06,
|
| 3920 |
+
"loss": 2.2654,
|
| 3921 |
+
"step": 268000
|
| 3922 |
+
},
|
| 3923 |
+
{
|
| 3924 |
+
"epoch": 20.809114159497792,
|
| 3925 |
+
"grad_norm": 7.05757999420166,
|
| 3926 |
+
"learning_rate": 6.127257227001473e-06,
|
| 3927 |
+
"loss": 2.2673,
|
| 3928 |
+
"step": 268500
|
| 3929 |
+
},
|
| 3930 |
+
{
|
| 3931 |
+
"epoch": 20.84786483763466,
|
| 3932 |
+
"grad_norm": 7.457660675048828,
|
| 3933 |
+
"learning_rate": 6.101423441576895e-06,
|
| 3934 |
+
"loss": 2.2534,
|
| 3935 |
+
"step": 269000
|
| 3936 |
+
},
|
| 3937 |
+
{
|
| 3938 |
+
"epoch": 20.886615515771528,
|
| 3939 |
+
"grad_norm": 6.697342872619629,
|
| 3940 |
+
"learning_rate": 6.0755896561523165e-06,
|
| 3941 |
+
"loss": 2.2721,
|
| 3942 |
+
"step": 269500
|
| 3943 |
+
},
|
| 3944 |
+
{
|
| 3945 |
+
"epoch": 20.925366193908392,
|
| 3946 |
+
"grad_norm": 6.83280611038208,
|
| 3947 |
+
"learning_rate": 6.049755870727738e-06,
|
| 3948 |
+
"loss": 2.276,
|
| 3949 |
+
"step": 270000
|
| 3950 |
+
},
|
| 3951 |
+
{
|
| 3952 |
+
"epoch": 20.96411687204526,
|
| 3953 |
+
"grad_norm": 6.298649311065674,
|
| 3954 |
+
"learning_rate": 6.02392208530316e-06,
|
| 3955 |
+
"loss": 2.265,
|
| 3956 |
+
"step": 270500
|
| 3957 |
+
},
|
| 3958 |
+
{
|
| 3959 |
+
"epoch": 21.0,
|
| 3960 |
+
"eval_loss": 2.2148919105529785,
|
| 3961 |
+
"eval_runtime": 275.6952,
|
| 3962 |
+
"eval_samples_per_second": 748.889,
|
| 3963 |
+
"eval_steps_per_second": 11.705,
|
| 3964 |
+
"step": 270963
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 21.002867550182128,
|
| 3968 |
+
"grad_norm": 6.698548316955566,
|
| 3969 |
+
"learning_rate": 5.998088299878582e-06,
|
| 3970 |
+
"loss": 2.2576,
|
| 3971 |
+
"step": 271000
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 21.041618228318995,
|
| 3975 |
+
"grad_norm": 6.784346103668213,
|
| 3976 |
+
"learning_rate": 5.972254514454003e-06,
|
| 3977 |
+
"loss": 2.2398,
|
| 3978 |
+
"step": 271500
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 21.080368906455863,
|
| 3982 |
+
"grad_norm": 7.072300910949707,
|
| 3983 |
+
"learning_rate": 5.946420729029425e-06,
|
| 3984 |
+
"loss": 2.2557,
|
| 3985 |
+
"step": 272000
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 21.11911958459273,
|
| 3989 |
+
"grad_norm": 6.624369144439697,
|
| 3990 |
+
"learning_rate": 5.920586943604847e-06,
|
| 3991 |
+
"loss": 2.2337,
|
| 3992 |
+
"step": 272500
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 21.1578702627296,
|
| 3996 |
+
"grad_norm": 6.317164897918701,
|
| 3997 |
+
"learning_rate": 5.8947531581802685e-06,
|
| 3998 |
+
"loss": 2.2534,
|
| 3999 |
+
"step": 273000
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 21.196620940866467,
|
| 4003 |
+
"grad_norm": 6.728669166564941,
|
| 4004 |
+
"learning_rate": 5.86891937275569e-06,
|
| 4005 |
+
"loss": 2.2505,
|
| 4006 |
+
"step": 273500
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 21.23537161900333,
|
| 4010 |
+
"grad_norm": 6.596154689788818,
|
| 4011 |
+
"learning_rate": 5.843085587331112e-06,
|
| 4012 |
+
"loss": 2.253,
|
| 4013 |
+
"step": 274000
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 21.2741222971402,
|
| 4017 |
+
"grad_norm": 6.471163749694824,
|
| 4018 |
+
"learning_rate": 5.817251801906534e-06,
|
| 4019 |
+
"loss": 2.2556,
|
| 4020 |
+
"step": 274500
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 21.312872975277067,
|
| 4024 |
+
"grad_norm": 6.29288911819458,
|
| 4025 |
+
"learning_rate": 5.791418016481955e-06,
|
| 4026 |
+
"loss": 2.2567,
|
| 4027 |
+
"step": 275000
|
| 4028 |
+
},
|
| 4029 |
+
{
|
| 4030 |
+
"epoch": 21.351623653413935,
|
| 4031 |
+
"grad_norm": 7.078927040100098,
|
| 4032 |
+
"learning_rate": 5.7655842310573776e-06,
|
| 4033 |
+
"loss": 2.2294,
|
| 4034 |
+
"step": 275500
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 21.390374331550802,
|
| 4038 |
+
"grad_norm": 6.867557525634766,
|
| 4039 |
+
"learning_rate": 5.739750445632799e-06,
|
| 4040 |
+
"loss": 2.2574,
|
| 4041 |
+
"step": 276000
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 21.42912500968767,
|
| 4045 |
+
"grad_norm": 6.830238342285156,
|
| 4046 |
+
"learning_rate": 5.7139166602082205e-06,
|
| 4047 |
+
"loss": 2.2794,
|
| 4048 |
+
"step": 276500
|
| 4049 |
+
},
|
| 4050 |
+
{
|
| 4051 |
+
"epoch": 21.467875687824538,
|
| 4052 |
+
"grad_norm": 6.694831371307373,
|
| 4053 |
+
"learning_rate": 5.688082874783643e-06,
|
| 4054 |
+
"loss": 2.253,
|
| 4055 |
+
"step": 277000
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 21.506626365961406,
|
| 4059 |
+
"grad_norm": 7.064994812011719,
|
| 4060 |
+
"learning_rate": 5.662249089359064e-06,
|
| 4061 |
+
"loss": 2.2435,
|
| 4062 |
+
"step": 277500
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 21.54537704409827,
|
| 4066 |
+
"grad_norm": 6.832572937011719,
|
| 4067 |
+
"learning_rate": 5.636415303934486e-06,
|
| 4068 |
+
"loss": 2.2478,
|
| 4069 |
+
"step": 278000
|
| 4070 |
+
},
|
| 4071 |
+
{
|
| 4072 |
+
"epoch": 21.584127722235138,
|
| 4073 |
+
"grad_norm": 7.045238494873047,
|
| 4074 |
+
"learning_rate": 5.610581518509908e-06,
|
| 4075 |
+
"loss": 2.2434,
|
| 4076 |
+
"step": 278500
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 21.622878400372006,
|
| 4080 |
+
"grad_norm": 6.720279216766357,
|
| 4081 |
+
"learning_rate": 5.58474773308533e-06,
|
| 4082 |
+
"loss": 2.238,
|
| 4083 |
+
"step": 279000
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 21.661629078508874,
|
| 4087 |
+
"grad_norm": 7.401440143585205,
|
| 4088 |
+
"learning_rate": 5.558913947660751e-06,
|
| 4089 |
+
"loss": 2.2461,
|
| 4090 |
+
"step": 279500
|
| 4091 |
+
},
|
| 4092 |
+
{
|
| 4093 |
+
"epoch": 21.70037975664574,
|
| 4094 |
+
"grad_norm": 6.497147560119629,
|
| 4095 |
+
"learning_rate": 5.5330801622361725e-06,
|
| 4096 |
+
"loss": 2.2339,
|
| 4097 |
+
"step": 280000
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 21.73913043478261,
|
| 4101 |
+
"grad_norm": 6.529776096343994,
|
| 4102 |
+
"learning_rate": 5.507246376811595e-06,
|
| 4103 |
+
"loss": 2.2501,
|
| 4104 |
+
"step": 280500
|
| 4105 |
+
},
|
| 4106 |
+
{
|
| 4107 |
+
"epoch": 21.777881112919477,
|
| 4108 |
+
"grad_norm": 6.42600679397583,
|
| 4109 |
+
"learning_rate": 5.481412591387016e-06,
|
| 4110 |
+
"loss": 2.235,
|
| 4111 |
+
"step": 281000
|
| 4112 |
+
},
|
| 4113 |
+
{
|
| 4114 |
+
"epoch": 21.816631791056345,
|
| 4115 |
+
"grad_norm": 6.715229034423828,
|
| 4116 |
+
"learning_rate": 5.455578805962438e-06,
|
| 4117 |
+
"loss": 2.2401,
|
| 4118 |
+
"step": 281500
|
| 4119 |
+
},
|
| 4120 |
+
{
|
| 4121 |
+
"epoch": 21.85538246919321,
|
| 4122 |
+
"grad_norm": 6.575899124145508,
|
| 4123 |
+
"learning_rate": 5.42974502053786e-06,
|
| 4124 |
+
"loss": 2.2576,
|
| 4125 |
+
"step": 282000
|
| 4126 |
+
},
|
| 4127 |
+
{
|
| 4128 |
+
"epoch": 21.894133147330077,
|
| 4129 |
+
"grad_norm": 5.999971866607666,
|
| 4130 |
+
"learning_rate": 5.403911235113282e-06,
|
| 4131 |
+
"loss": 2.2379,
|
| 4132 |
+
"step": 282500
|
| 4133 |
+
},
|
| 4134 |
+
{
|
| 4135 |
+
"epoch": 21.932883825466945,
|
| 4136 |
+
"grad_norm": 6.936278343200684,
|
| 4137 |
+
"learning_rate": 5.378077449688703e-06,
|
| 4138 |
+
"loss": 2.2534,
|
| 4139 |
+
"step": 283000
|
| 4140 |
+
},
|
| 4141 |
+
{
|
| 4142 |
+
"epoch": 21.971634503603813,
|
| 4143 |
+
"grad_norm": 6.040930271148682,
|
| 4144 |
+
"learning_rate": 5.352243664264125e-06,
|
| 4145 |
+
"loss": 2.2391,
|
| 4146 |
+
"step": 283500
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 22.0,
|
| 4150 |
+
"eval_loss": 2.1943371295928955,
|
| 4151 |
+
"eval_runtime": 268.1318,
|
| 4152 |
+
"eval_samples_per_second": 770.013,
|
| 4153 |
+
"eval_steps_per_second": 12.035,
|
| 4154 |
+
"step": 283866
|
| 4155 |
+
},
|
| 4156 |
+
{
|
| 4157 |
+
"epoch": 22.01038518174068,
|
| 4158 |
+
"grad_norm": 6.7548747062683105,
|
| 4159 |
+
"learning_rate": 5.326409878839547e-06,
|
| 4160 |
+
"loss": 2.2428,
|
| 4161 |
+
"step": 284000
|
| 4162 |
+
},
|
| 4163 |
+
{
|
| 4164 |
+
"epoch": 22.04913585987755,
|
| 4165 |
+
"grad_norm": 7.0850749015808105,
|
| 4166 |
+
"learning_rate": 5.300576093414968e-06,
|
| 4167 |
+
"loss": 2.2273,
|
| 4168 |
+
"step": 284500
|
| 4169 |
+
},
|
| 4170 |
+
{
|
| 4171 |
+
"epoch": 22.087886538014416,
|
| 4172 |
+
"grad_norm": 6.658077239990234,
|
| 4173 |
+
"learning_rate": 5.274742307990391e-06,
|
| 4174 |
+
"loss": 2.2214,
|
| 4175 |
+
"step": 285000
|
| 4176 |
+
},
|
| 4177 |
+
{
|
| 4178 |
+
"epoch": 22.126637216151284,
|
| 4179 |
+
"grad_norm": 7.19653844833374,
|
| 4180 |
+
"learning_rate": 5.248908522565812e-06,
|
| 4181 |
+
"loss": 2.2273,
|
| 4182 |
+
"step": 285500
|
| 4183 |
+
},
|
| 4184 |
+
{
|
| 4185 |
+
"epoch": 22.16538789428815,
|
| 4186 |
+
"grad_norm": 7.094461441040039,
|
| 4187 |
+
"learning_rate": 5.223074737141234e-06,
|
| 4188 |
+
"loss": 2.2359,
|
| 4189 |
+
"step": 286000
|
| 4190 |
+
},
|
| 4191 |
+
{
|
| 4192 |
+
"epoch": 22.204138572425016,
|
| 4193 |
+
"grad_norm": 7.156402587890625,
|
| 4194 |
+
"learning_rate": 5.197240951716656e-06,
|
| 4195 |
+
"loss": 2.1969,
|
| 4196 |
+
"step": 286500
|
| 4197 |
+
},
|
| 4198 |
+
{
|
| 4199 |
+
"epoch": 22.242889250561884,
|
| 4200 |
+
"grad_norm": 6.595995903015137,
|
| 4201 |
+
"learning_rate": 5.171407166292077e-06,
|
| 4202 |
+
"loss": 2.2223,
|
| 4203 |
+
"step": 287000
|
| 4204 |
+
},
|
| 4205 |
+
{
|
| 4206 |
+
"epoch": 22.281639928698752,
|
| 4207 |
+
"grad_norm": 7.04496955871582,
|
| 4208 |
+
"learning_rate": 5.145573380867499e-06,
|
| 4209 |
+
"loss": 2.2343,
|
| 4210 |
+
"step": 287500
|
| 4211 |
+
},
|
| 4212 |
+
{
|
| 4213 |
+
"epoch": 22.32039060683562,
|
| 4214 |
+
"grad_norm": 7.146208763122559,
|
| 4215 |
+
"learning_rate": 5.11973959544292e-06,
|
| 4216 |
+
"loss": 2.2338,
|
| 4217 |
+
"step": 288000
|
| 4218 |
+
},
|
| 4219 |
+
{
|
| 4220 |
+
"epoch": 22.359141284972488,
|
| 4221 |
+
"grad_norm": 6.4659576416015625,
|
| 4222 |
+
"learning_rate": 5.093905810018343e-06,
|
| 4223 |
+
"loss": 2.2273,
|
| 4224 |
+
"step": 288500
|
| 4225 |
+
},
|
| 4226 |
+
{
|
| 4227 |
+
"epoch": 22.397891963109355,
|
| 4228 |
+
"grad_norm": 6.372287750244141,
|
| 4229 |
+
"learning_rate": 5.068072024593764e-06,
|
| 4230 |
+
"loss": 2.2247,
|
| 4231 |
+
"step": 289000
|
| 4232 |
+
},
|
| 4233 |
+
{
|
| 4234 |
+
"epoch": 22.436642641246223,
|
| 4235 |
+
"grad_norm": 7.088085174560547,
|
| 4236 |
+
"learning_rate": 5.042238239169186e-06,
|
| 4237 |
+
"loss": 2.2474,
|
| 4238 |
+
"step": 289500
|
| 4239 |
+
},
|
| 4240 |
+
{
|
| 4241 |
+
"epoch": 22.475393319383087,
|
| 4242 |
+
"grad_norm": 6.911520004272461,
|
| 4243 |
+
"learning_rate": 5.016404453744608e-06,
|
| 4244 |
+
"loss": 2.2356,
|
| 4245 |
+
"step": 290000
|
| 4246 |
+
},
|
| 4247 |
+
{
|
| 4248 |
+
"epoch": 22.514143997519955,
|
| 4249 |
+
"grad_norm": 7.5756611824035645,
|
| 4250 |
+
"learning_rate": 4.990570668320029e-06,
|
| 4251 |
+
"loss": 2.2297,
|
| 4252 |
+
"step": 290500
|
| 4253 |
+
},
|
| 4254 |
+
{
|
| 4255 |
+
"epoch": 22.552894675656823,
|
| 4256 |
+
"grad_norm": 6.587701320648193,
|
| 4257 |
+
"learning_rate": 4.964736882895451e-06,
|
| 4258 |
+
"loss": 2.2245,
|
| 4259 |
+
"step": 291000
|
| 4260 |
+
},
|
| 4261 |
+
{
|
| 4262 |
+
"epoch": 22.59164535379369,
|
| 4263 |
+
"grad_norm": 5.8870849609375,
|
| 4264 |
+
"learning_rate": 4.938903097470873e-06,
|
| 4265 |
+
"loss": 2.229,
|
| 4266 |
+
"step": 291500
|
| 4267 |
+
},
|
| 4268 |
+
{
|
| 4269 |
+
"epoch": 22.63039603193056,
|
| 4270 |
+
"grad_norm": 6.882173538208008,
|
| 4271 |
+
"learning_rate": 4.913069312046295e-06,
|
| 4272 |
+
"loss": 2.2254,
|
| 4273 |
+
"step": 292000
|
| 4274 |
+
},
|
| 4275 |
+
{
|
| 4276 |
+
"epoch": 22.669146710067427,
|
| 4277 |
+
"grad_norm": 6.710127830505371,
|
| 4278 |
+
"learning_rate": 4.887235526621716e-06,
|
| 4279 |
+
"loss": 2.223,
|
| 4280 |
+
"step": 292500
|
| 4281 |
+
},
|
| 4282 |
+
{
|
| 4283 |
+
"epoch": 22.707897388204294,
|
| 4284 |
+
"grad_norm": 6.753304481506348,
|
| 4285 |
+
"learning_rate": 4.8614017411971385e-06,
|
| 4286 |
+
"loss": 2.2311,
|
| 4287 |
+
"step": 293000
|
| 4288 |
+
},
|
| 4289 |
+
{
|
| 4290 |
+
"epoch": 22.746648066341162,
|
| 4291 |
+
"grad_norm": 6.02184534072876,
|
| 4292 |
+
"learning_rate": 4.83556795577256e-06,
|
| 4293 |
+
"loss": 2.2198,
|
| 4294 |
+
"step": 293500
|
| 4295 |
+
},
|
| 4296 |
+
{
|
| 4297 |
+
"epoch": 22.78539874447803,
|
| 4298 |
+
"grad_norm": 7.022054195404053,
|
| 4299 |
+
"learning_rate": 4.809734170347981e-06,
|
| 4300 |
+
"loss": 2.2275,
|
| 4301 |
+
"step": 294000
|
| 4302 |
+
},
|
| 4303 |
+
{
|
| 4304 |
+
"epoch": 22.824149422614894,
|
| 4305 |
+
"grad_norm": 7.951735019683838,
|
| 4306 |
+
"learning_rate": 4.783900384923404e-06,
|
| 4307 |
+
"loss": 2.2314,
|
| 4308 |
+
"step": 294500
|
| 4309 |
+
},
|
| 4310 |
+
{
|
| 4311 |
+
"epoch": 22.862900100751762,
|
| 4312 |
+
"grad_norm": 5.854333877563477,
|
| 4313 |
+
"learning_rate": 4.758066599498825e-06,
|
| 4314 |
+
"loss": 2.2172,
|
| 4315 |
+
"step": 295000
|
| 4316 |
+
},
|
| 4317 |
+
{
|
| 4318 |
+
"epoch": 22.90165077888863,
|
| 4319 |
+
"grad_norm": 6.547132968902588,
|
| 4320 |
+
"learning_rate": 4.732232814074247e-06,
|
| 4321 |
+
"loss": 2.226,
|
| 4322 |
+
"step": 295500
|
| 4323 |
+
},
|
| 4324 |
+
{
|
| 4325 |
+
"epoch": 22.940401457025498,
|
| 4326 |
+
"grad_norm": 6.535789966583252,
|
| 4327 |
+
"learning_rate": 4.706399028649668e-06,
|
| 4328 |
+
"loss": 2.2299,
|
| 4329 |
+
"step": 296000
|
| 4330 |
+
},
|
| 4331 |
+
{
|
| 4332 |
+
"epoch": 22.979152135162366,
|
| 4333 |
+
"grad_norm": 7.285912036895752,
|
| 4334 |
+
"learning_rate": 4.6805652432250905e-06,
|
| 4335 |
+
"loss": 2.2239,
|
| 4336 |
+
"step": 296500
|
| 4337 |
+
},
|
| 4338 |
+
{
|
| 4339 |
+
"epoch": 23.0,
|
| 4340 |
+
"eval_loss": 2.185373067855835,
|
| 4341 |
+
"eval_runtime": 265.4806,
|
| 4342 |
+
"eval_samples_per_second": 777.703,
|
| 4343 |
+
"eval_steps_per_second": 12.155,
|
| 4344 |
+
"step": 296769
|
| 4345 |
+
},
|
| 4346 |
+
{
|
| 4347 |
+
"epoch": 23.017902813299234,
|
| 4348 |
+
"grad_norm": 6.972716808319092,
|
| 4349 |
+
"learning_rate": 4.654731457800512e-06,
|
| 4350 |
+
"loss": 2.2193,
|
| 4351 |
+
"step": 297000
|
| 4352 |
+
},
|
| 4353 |
+
{
|
| 4354 |
+
"epoch": 23.0566534914361,
|
| 4355 |
+
"grad_norm": 6.841848373413086,
|
| 4356 |
+
"learning_rate": 4.628897672375933e-06,
|
| 4357 |
+
"loss": 2.2139,
|
| 4358 |
+
"step": 297500
|
| 4359 |
+
},
|
| 4360 |
+
{
|
| 4361 |
+
"epoch": 23.09540416957297,
|
| 4362 |
+
"grad_norm": 6.285813331604004,
|
| 4363 |
+
"learning_rate": 4.603063886951356e-06,
|
| 4364 |
+
"loss": 2.2092,
|
| 4365 |
+
"step": 298000
|
| 4366 |
+
},
|
| 4367 |
+
{
|
| 4368 |
+
"epoch": 23.134154847709834,
|
| 4369 |
+
"grad_norm": 6.615530967712402,
|
| 4370 |
+
"learning_rate": 4.577230101526777e-06,
|
| 4371 |
+
"loss": 2.2141,
|
| 4372 |
+
"step": 298500
|
| 4373 |
+
},
|
| 4374 |
+
{
|
| 4375 |
+
"epoch": 23.1729055258467,
|
| 4376 |
+
"grad_norm": 6.762087821960449,
|
| 4377 |
+
"learning_rate": 4.551396316102199e-06,
|
| 4378 |
+
"loss": 2.1944,
|
| 4379 |
+
"step": 299000
|
| 4380 |
+
},
|
| 4381 |
+
{
|
| 4382 |
+
"epoch": 23.21165620398357,
|
| 4383 |
+
"grad_norm": 7.053805351257324,
|
| 4384 |
+
"learning_rate": 4.525562530677621e-06,
|
| 4385 |
+
"loss": 2.2129,
|
| 4386 |
+
"step": 299500
|
| 4387 |
+
},
|
| 4388 |
+
{
|
| 4389 |
+
"epoch": 23.250406882120437,
|
| 4390 |
+
"grad_norm": 7.14516544342041,
|
| 4391 |
+
"learning_rate": 4.4997287452530425e-06,
|
| 4392 |
+
"loss": 2.2038,
|
| 4393 |
+
"step": 300000
|
| 4394 |
+
},
|
| 4395 |
+
{
|
| 4396 |
+
"epoch": 23.289157560257305,
|
| 4397 |
+
"grad_norm": 6.8478803634643555,
|
| 4398 |
+
"learning_rate": 4.473894959828464e-06,
|
| 4399 |
+
"loss": 2.2166,
|
| 4400 |
+
"step": 300500
|
| 4401 |
+
},
|
| 4402 |
+
{
|
| 4403 |
+
"epoch": 23.327908238394173,
|
| 4404 |
+
"grad_norm": 6.808053970336914,
|
| 4405 |
+
"learning_rate": 4.448061174403886e-06,
|
| 4406 |
+
"loss": 2.224,
|
| 4407 |
+
"step": 301000
|
| 4408 |
+
},
|
| 4409 |
+
{
|
| 4410 |
+
"epoch": 23.36665891653104,
|
| 4411 |
+
"grad_norm": 7.149857521057129,
|
| 4412 |
+
"learning_rate": 4.422227388979308e-06,
|
| 4413 |
+
"loss": 2.2081,
|
| 4414 |
+
"step": 301500
|
| 4415 |
+
},
|
| 4416 |
+
{
|
| 4417 |
+
"epoch": 23.40540959466791,
|
| 4418 |
+
"grad_norm": 6.334920406341553,
|
| 4419 |
+
"learning_rate": 4.396393603554729e-06,
|
| 4420 |
+
"loss": 2.2217,
|
| 4421 |
+
"step": 302000
|
| 4422 |
+
},
|
| 4423 |
+
{
|
| 4424 |
+
"epoch": 23.444160272804773,
|
| 4425 |
+
"grad_norm": 7.154323577880859,
|
| 4426 |
+
"learning_rate": 4.3705598181301515e-06,
|
| 4427 |
+
"loss": 2.2129,
|
| 4428 |
+
"step": 302500
|
| 4429 |
+
},
|
| 4430 |
+
{
|
| 4431 |
+
"epoch": 23.48291095094164,
|
| 4432 |
+
"grad_norm": 7.202456474304199,
|
| 4433 |
+
"learning_rate": 4.344726032705573e-06,
|
| 4434 |
+
"loss": 2.2019,
|
| 4435 |
+
"step": 303000
|
| 4436 |
+
},
|
| 4437 |
+
{
|
| 4438 |
+
"epoch": 23.52166162907851,
|
| 4439 |
+
"grad_norm": 6.832441806793213,
|
| 4440 |
+
"learning_rate": 4.3188922472809945e-06,
|
| 4441 |
+
"loss": 2.214,
|
| 4442 |
+
"step": 303500
|
| 4443 |
+
},
|
| 4444 |
+
{
|
| 4445 |
+
"epoch": 23.560412307215376,
|
| 4446 |
+
"grad_norm": 6.258272647857666,
|
| 4447 |
+
"learning_rate": 4.293058461856417e-06,
|
| 4448 |
+
"loss": 2.21,
|
| 4449 |
+
"step": 304000
|
| 4450 |
+
},
|
| 4451 |
+
{
|
| 4452 |
+
"epoch": 23.599162985352244,
|
| 4453 |
+
"grad_norm": 6.8391194343566895,
|
| 4454 |
+
"learning_rate": 4.267224676431838e-06,
|
| 4455 |
+
"loss": 2.2106,
|
| 4456 |
+
"step": 304500
|
| 4457 |
+
},
|
| 4458 |
+
{
|
| 4459 |
+
"epoch": 23.637913663489112,
|
| 4460 |
+
"grad_norm": 6.621433734893799,
|
| 4461 |
+
"learning_rate": 4.24139089100726e-06,
|
| 4462 |
+
"loss": 2.2219,
|
| 4463 |
+
"step": 305000
|
| 4464 |
+
},
|
| 4465 |
+
{
|
| 4466 |
+
"epoch": 23.67666434162598,
|
| 4467 |
+
"grad_norm": 6.718801498413086,
|
| 4468 |
+
"learning_rate": 4.215557105582681e-06,
|
| 4469 |
+
"loss": 2.2215,
|
| 4470 |
+
"step": 305500
|
| 4471 |
+
},
|
| 4472 |
+
{
|
| 4473 |
+
"epoch": 23.715415019762847,
|
| 4474 |
+
"grad_norm": 7.0543622970581055,
|
| 4475 |
+
"learning_rate": 4.1897233201581036e-06,
|
| 4476 |
+
"loss": 2.2182,
|
| 4477 |
+
"step": 306000
|
| 4478 |
+
},
|
| 4479 |
+
{
|
| 4480 |
+
"epoch": 23.75416569789971,
|
| 4481 |
+
"grad_norm": 7.598169326782227,
|
| 4482 |
+
"learning_rate": 4.163889534733525e-06,
|
| 4483 |
+
"loss": 2.2218,
|
| 4484 |
+
"step": 306500
|
| 4485 |
+
},
|
| 4486 |
+
{
|
| 4487 |
+
"epoch": 23.79291637603658,
|
| 4488 |
+
"grad_norm": 6.874271392822266,
|
| 4489 |
+
"learning_rate": 4.1380557493089465e-06,
|
| 4490 |
+
"loss": 2.2061,
|
| 4491 |
+
"step": 307000
|
| 4492 |
+
},
|
| 4493 |
+
{
|
| 4494 |
+
"epoch": 23.831667054173447,
|
| 4495 |
+
"grad_norm": 6.820863723754883,
|
| 4496 |
+
"learning_rate": 4.112221963884369e-06,
|
| 4497 |
+
"loss": 2.2166,
|
| 4498 |
+
"step": 307500
|
| 4499 |
+
},
|
| 4500 |
+
{
|
| 4501 |
+
"epoch": 23.870417732310315,
|
| 4502 |
+
"grad_norm": 7.149729251861572,
|
| 4503 |
+
"learning_rate": 4.08638817845979e-06,
|
| 4504 |
+
"loss": 2.2089,
|
| 4505 |
+
"step": 308000
|
| 4506 |
+
},
|
| 4507 |
+
{
|
| 4508 |
+
"epoch": 23.909168410447183,
|
| 4509 |
+
"grad_norm": 6.278995990753174,
|
| 4510 |
+
"learning_rate": 4.060554393035212e-06,
|
| 4511 |
+
"loss": 2.2163,
|
| 4512 |
+
"step": 308500
|
| 4513 |
+
},
|
| 4514 |
+
{
|
| 4515 |
+
"epoch": 23.94791908858405,
|
| 4516 |
+
"grad_norm": 7.162642002105713,
|
| 4517 |
+
"learning_rate": 4.034720607610634e-06,
|
| 4518 |
+
"loss": 2.2222,
|
| 4519 |
+
"step": 309000
|
| 4520 |
+
},
|
| 4521 |
+
{
|
| 4522 |
+
"epoch": 23.98666976672092,
|
| 4523 |
+
"grad_norm": 6.67965841293335,
|
| 4524 |
+
"learning_rate": 4.0088868221860556e-06,
|
| 4525 |
+
"loss": 2.1965,
|
| 4526 |
+
"step": 309500
|
| 4527 |
+
},
|
| 4528 |
+
{
|
| 4529 |
+
"epoch": 24.0,
|
| 4530 |
+
"eval_loss": 2.1702771186828613,
|
| 4531 |
+
"eval_runtime": 264.5615,
|
| 4532 |
+
"eval_samples_per_second": 780.405,
|
| 4533 |
+
"eval_steps_per_second": 12.198,
|
| 4534 |
+
"step": 309672
|
| 4535 |
+
},
|
| 4536 |
+
{
|
| 4537 |
+
"epoch": 24.025420444857787,
|
| 4538 |
+
"grad_norm": 6.355005264282227,
|
| 4539 |
+
"learning_rate": 3.983053036761477e-06,
|
| 4540 |
+
"loss": 2.1935,
|
| 4541 |
+
"step": 310000
|
| 4542 |
+
},
|
| 4543 |
+
{
|
| 4544 |
+
"epoch": 24.06417112299465,
|
| 4545 |
+
"grad_norm": 6.339087963104248,
|
| 4546 |
+
"learning_rate": 3.957219251336899e-06,
|
| 4547 |
+
"loss": 2.2022,
|
| 4548 |
+
"step": 310500
|
| 4549 |
+
},
|
| 4550 |
+
{
|
| 4551 |
+
"epoch": 24.10292180113152,
|
| 4552 |
+
"grad_norm": 6.386953353881836,
|
| 4553 |
+
"learning_rate": 3.931385465912321e-06,
|
| 4554 |
+
"loss": 2.1941,
|
| 4555 |
+
"step": 311000
|
| 4556 |
+
},
|
| 4557 |
+
{
|
| 4558 |
+
"epoch": 24.141672479268387,
|
| 4559 |
+
"grad_norm": 6.9508376121521,
|
| 4560 |
+
"learning_rate": 3.905551680487742e-06,
|
| 4561 |
+
"loss": 2.1991,
|
| 4562 |
+
"step": 311500
|
| 4563 |
+
},
|
| 4564 |
+
{
|
| 4565 |
+
"epoch": 24.180423157405254,
|
| 4566 |
+
"grad_norm": 7.1515727043151855,
|
| 4567 |
+
"learning_rate": 3.879717895063164e-06,
|
| 4568 |
+
"loss": 2.2118,
|
| 4569 |
+
"step": 312000
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 24.219173835542122,
|
| 4573 |
+
"grad_norm": 6.807953357696533,
|
| 4574 |
+
"learning_rate": 3.853884109638585e-06,
|
| 4575 |
+
"loss": 2.2158,
|
| 4576 |
+
"step": 312500
|
| 4577 |
+
},
|
| 4578 |
+
{
|
| 4579 |
+
"epoch": 24.25792451367899,
|
| 4580 |
+
"grad_norm": 7.41762638092041,
|
| 4581 |
+
"learning_rate": 3.828050324214007e-06,
|
| 4582 |
+
"loss": 2.1948,
|
| 4583 |
+
"step": 313000
|
| 4584 |
+
},
|
| 4585 |
+
{
|
| 4586 |
+
"epoch": 24.296675191815858,
|
| 4587 |
+
"grad_norm": 7.462344646453857,
|
| 4588 |
+
"learning_rate": 3.802216538789429e-06,
|
| 4589 |
+
"loss": 2.2061,
|
| 4590 |
+
"step": 313500
|
| 4591 |
+
},
|
| 4592 |
+
{
|
| 4593 |
+
"epoch": 24.335425869952726,
|
| 4594 |
+
"grad_norm": 6.6912384033203125,
|
| 4595 |
+
"learning_rate": 3.7763827533648505e-06,
|
| 4596 |
+
"loss": 2.1932,
|
| 4597 |
+
"step": 314000
|
| 4598 |
+
},
|
| 4599 |
+
{
|
| 4600 |
+
"epoch": 24.37417654808959,
|
| 4601 |
+
"grad_norm": 6.79492712020874,
|
| 4602 |
+
"learning_rate": 3.7505489679402724e-06,
|
| 4603 |
+
"loss": 2.193,
|
| 4604 |
+
"step": 314500
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 24.412927226226458,
|
| 4608 |
+
"grad_norm": 6.873208522796631,
|
| 4609 |
+
"learning_rate": 3.724715182515694e-06,
|
| 4610 |
+
"loss": 2.1756,
|
| 4611 |
+
"step": 315000
|
| 4612 |
+
},
|
| 4613 |
+
{
|
| 4614 |
+
"epoch": 24.451677904363326,
|
| 4615 |
+
"grad_norm": 6.520395278930664,
|
| 4616 |
+
"learning_rate": 3.6988813970911158e-06,
|
| 4617 |
+
"loss": 2.2019,
|
| 4618 |
+
"step": 315500
|
| 4619 |
+
},
|
| 4620 |
+
{
|
| 4621 |
+
"epoch": 24.490428582500193,
|
| 4622 |
+
"grad_norm": 7.425100326538086,
|
| 4623 |
+
"learning_rate": 3.6730476116665377e-06,
|
| 4624 |
+
"loss": 2.1933,
|
| 4625 |
+
"step": 316000
|
| 4626 |
+
},
|
| 4627 |
+
{
|
| 4628 |
+
"epoch": 24.52917926063706,
|
| 4629 |
+
"grad_norm": 6.990531921386719,
|
| 4630 |
+
"learning_rate": 3.647213826241959e-06,
|
| 4631 |
+
"loss": 2.1953,
|
| 4632 |
+
"step": 316500
|
| 4633 |
+
},
|
| 4634 |
+
{
|
| 4635 |
+
"epoch": 24.56792993877393,
|
| 4636 |
+
"grad_norm": 6.99529504776001,
|
| 4637 |
+
"learning_rate": 3.621380040817381e-06,
|
| 4638 |
+
"loss": 2.1668,
|
| 4639 |
+
"step": 317000
|
| 4640 |
+
},
|
| 4641 |
+
{
|
| 4642 |
+
"epoch": 24.606680616910797,
|
| 4643 |
+
"grad_norm": 7.046565532684326,
|
| 4644 |
+
"learning_rate": 3.595546255392803e-06,
|
| 4645 |
+
"loss": 2.2185,
|
| 4646 |
+
"step": 317500
|
| 4647 |
+
},
|
| 4648 |
+
{
|
| 4649 |
+
"epoch": 24.645431295047665,
|
| 4650 |
+
"grad_norm": 7.261152744293213,
|
| 4651 |
+
"learning_rate": 3.5697124699682244e-06,
|
| 4652 |
+
"loss": 2.1776,
|
| 4653 |
+
"step": 318000
|
| 4654 |
+
},
|
| 4655 |
+
{
|
| 4656 |
+
"epoch": 24.68418197318453,
|
| 4657 |
+
"grad_norm": 7.088150978088379,
|
| 4658 |
+
"learning_rate": 3.5438786845436463e-06,
|
| 4659 |
+
"loss": 2.1939,
|
| 4660 |
+
"step": 318500
|
| 4661 |
+
},
|
| 4662 |
+
{
|
| 4663 |
+
"epoch": 24.722932651321397,
|
| 4664 |
+
"grad_norm": 7.677366733551025,
|
| 4665 |
+
"learning_rate": 3.518044899119068e-06,
|
| 4666 |
+
"loss": 2.1916,
|
| 4667 |
+
"step": 319000
|
| 4668 |
+
},
|
| 4669 |
+
{
|
| 4670 |
+
"epoch": 24.761683329458265,
|
| 4671 |
+
"grad_norm": 7.108632564544678,
|
| 4672 |
+
"learning_rate": 3.4922111136944897e-06,
|
| 4673 |
+
"loss": 2.1851,
|
| 4674 |
+
"step": 319500
|
| 4675 |
+
},
|
| 4676 |
+
{
|
| 4677 |
+
"epoch": 24.800434007595133,
|
| 4678 |
+
"grad_norm": 7.283915996551514,
|
| 4679 |
+
"learning_rate": 3.4663773282699116e-06,
|
| 4680 |
+
"loss": 2.2015,
|
| 4681 |
+
"step": 320000
|
| 4682 |
+
},
|
| 4683 |
+
{
|
| 4684 |
+
"epoch": 24.839184685732,
|
| 4685 |
+
"grad_norm": 7.392533779144287,
|
| 4686 |
+
"learning_rate": 3.440543542845333e-06,
|
| 4687 |
+
"loss": 2.1915,
|
| 4688 |
+
"step": 320500
|
| 4689 |
+
},
|
| 4690 |
+
{
|
| 4691 |
+
"epoch": 24.87793536386887,
|
| 4692 |
+
"grad_norm": 6.849175453186035,
|
| 4693 |
+
"learning_rate": 3.414709757420755e-06,
|
| 4694 |
+
"loss": 2.1931,
|
| 4695 |
+
"step": 321000
|
| 4696 |
+
},
|
| 4697 |
+
{
|
| 4698 |
+
"epoch": 24.916686042005736,
|
| 4699 |
+
"grad_norm": 6.42083740234375,
|
| 4700 |
+
"learning_rate": 3.388875971996177e-06,
|
| 4701 |
+
"loss": 2.198,
|
| 4702 |
+
"step": 321500
|
| 4703 |
+
},
|
| 4704 |
+
{
|
| 4705 |
+
"epoch": 24.955436720142604,
|
| 4706 |
+
"grad_norm": 6.040030002593994,
|
| 4707 |
+
"learning_rate": 3.3630421865715983e-06,
|
| 4708 |
+
"loss": 2.1802,
|
| 4709 |
+
"step": 322000
|
| 4710 |
+
},
|
| 4711 |
+
{
|
| 4712 |
+
"epoch": 24.99418739827947,
|
| 4713 |
+
"grad_norm": 7.585995674133301,
|
| 4714 |
+
"learning_rate": 3.3372084011470202e-06,
|
| 4715 |
+
"loss": 2.1946,
|
| 4716 |
+
"step": 322500
|
| 4717 |
+
},
|
| 4718 |
+
{
|
| 4719 |
+
"epoch": 25.0,
|
| 4720 |
+
"eval_loss": 2.159193992614746,
|
| 4721 |
+
"eval_runtime": 270.695,
|
| 4722 |
+
"eval_samples_per_second": 762.722,
|
| 4723 |
+
"eval_steps_per_second": 11.921,
|
| 4724 |
+
"step": 322575
|
| 4725 |
+
},
|
| 4726 |
+
{
|
| 4727 |
+
"epoch": 25.032938076416336,
|
| 4728 |
+
"grad_norm": 7.309504985809326,
|
| 4729 |
+
"learning_rate": 3.3113746157224417e-06,
|
| 4730 |
+
"loss": 2.1854,
|
| 4731 |
+
"step": 323000
|
| 4732 |
+
},
|
| 4733 |
+
{
|
| 4734 |
+
"epoch": 25.071688754553204,
|
| 4735 |
+
"grad_norm": 6.450008869171143,
|
| 4736 |
+
"learning_rate": 3.2855408302978636e-06,
|
| 4737 |
+
"loss": 2.1827,
|
| 4738 |
+
"step": 323500
|
| 4739 |
+
},
|
| 4740 |
+
{
|
| 4741 |
+
"epoch": 25.11043943269007,
|
| 4742 |
+
"grad_norm": 6.82379674911499,
|
| 4743 |
+
"learning_rate": 3.2597070448732855e-06,
|
| 4744 |
+
"loss": 2.1838,
|
| 4745 |
+
"step": 324000
|
| 4746 |
+
},
|
| 4747 |
+
{
|
| 4748 |
+
"epoch": 25.14919011082694,
|
| 4749 |
+
"grad_norm": 7.034087657928467,
|
| 4750 |
+
"learning_rate": 3.233873259448707e-06,
|
| 4751 |
+
"loss": 2.1618,
|
| 4752 |
+
"step": 324500
|
| 4753 |
+
},
|
| 4754 |
+
{
|
| 4755 |
+
"epoch": 25.187940788963807,
|
| 4756 |
+
"grad_norm": 7.005911827087402,
|
| 4757 |
+
"learning_rate": 3.208039474024129e-06,
|
| 4758 |
+
"loss": 2.1912,
|
| 4759 |
+
"step": 325000
|
| 4760 |
+
},
|
| 4761 |
+
{
|
| 4762 |
+
"epoch": 25.226691467100675,
|
| 4763 |
+
"grad_norm": 6.7085394859313965,
|
| 4764 |
+
"learning_rate": 3.1822056885995508e-06,
|
| 4765 |
+
"loss": 2.1795,
|
| 4766 |
+
"step": 325500
|
| 4767 |
+
},
|
| 4768 |
+
{
|
| 4769 |
+
"epoch": 25.265442145237543,
|
| 4770 |
+
"grad_norm": 6.773245334625244,
|
| 4771 |
+
"learning_rate": 3.1563719031749722e-06,
|
| 4772 |
+
"loss": 2.1965,
|
| 4773 |
+
"step": 326000
|
| 4774 |
+
},
|
| 4775 |
+
{
|
| 4776 |
+
"epoch": 25.304192823374407,
|
| 4777 |
+
"grad_norm": 6.718632698059082,
|
| 4778 |
+
"learning_rate": 3.130538117750394e-06,
|
| 4779 |
+
"loss": 2.1976,
|
| 4780 |
+
"step": 326500
|
| 4781 |
+
},
|
| 4782 |
+
{
|
| 4783 |
+
"epoch": 25.342943501511275,
|
| 4784 |
+
"grad_norm": 8.191710472106934,
|
| 4785 |
+
"learning_rate": 3.1047043323258156e-06,
|
| 4786 |
+
"loss": 2.1762,
|
| 4787 |
+
"step": 327000
|
| 4788 |
+
},
|
| 4789 |
+
{
|
| 4790 |
+
"epoch": 25.381694179648143,
|
| 4791 |
+
"grad_norm": 7.172983169555664,
|
| 4792 |
+
"learning_rate": 3.0788705469012375e-06,
|
| 4793 |
+
"loss": 2.1703,
|
| 4794 |
+
"step": 327500
|
| 4795 |
+
},
|
| 4796 |
+
{
|
| 4797 |
+
"epoch": 25.42044485778501,
|
| 4798 |
+
"grad_norm": 6.283721446990967,
|
| 4799 |
+
"learning_rate": 3.0530367614766594e-06,
|
| 4800 |
+
"loss": 2.1692,
|
| 4801 |
+
"step": 328000
|
| 4802 |
+
},
|
| 4803 |
+
{
|
| 4804 |
+
"epoch": 25.45919553592188,
|
| 4805 |
+
"grad_norm": 6.850103855133057,
|
| 4806 |
+
"learning_rate": 3.027202976052081e-06,
|
| 4807 |
+
"loss": 2.1914,
|
| 4808 |
+
"step": 328500
|
| 4809 |
+
},
|
| 4810 |
+
{
|
| 4811 |
+
"epoch": 25.497946214058747,
|
| 4812 |
+
"grad_norm": 6.31437873840332,
|
| 4813 |
+
"learning_rate": 3.0013691906275028e-06,
|
| 4814 |
+
"loss": 2.1692,
|
| 4815 |
+
"step": 329000
|
| 4816 |
+
},
|
| 4817 |
+
{
|
| 4818 |
+
"epoch": 25.536696892195614,
|
| 4819 |
+
"grad_norm": 6.947432994842529,
|
| 4820 |
+
"learning_rate": 2.9755354052029247e-06,
|
| 4821 |
+
"loss": 2.1848,
|
| 4822 |
+
"step": 329500
|
| 4823 |
+
},
|
| 4824 |
+
{
|
| 4825 |
+
"epoch": 25.575447570332482,
|
| 4826 |
+
"grad_norm": 6.133412837982178,
|
| 4827 |
+
"learning_rate": 2.949701619778346e-06,
|
| 4828 |
+
"loss": 2.1827,
|
| 4829 |
+
"step": 330000
|
| 4830 |
+
},
|
| 4831 |
+
{
|
| 4832 |
+
"epoch": 25.61419824846935,
|
| 4833 |
+
"grad_norm": 7.019827365875244,
|
| 4834 |
+
"learning_rate": 2.923867834353768e-06,
|
| 4835 |
+
"loss": 2.1654,
|
| 4836 |
+
"step": 330500
|
| 4837 |
+
},
|
| 4838 |
+
{
|
| 4839 |
+
"epoch": 25.652948926606214,
|
| 4840 |
+
"grad_norm": 7.326742172241211,
|
| 4841 |
+
"learning_rate": 2.8980340489291895e-06,
|
| 4842 |
+
"loss": 2.1929,
|
| 4843 |
+
"step": 331000
|
| 4844 |
+
},
|
| 4845 |
+
{
|
| 4846 |
+
"epoch": 25.691699604743082,
|
| 4847 |
+
"grad_norm": 7.231571674346924,
|
| 4848 |
+
"learning_rate": 2.8722002635046114e-06,
|
| 4849 |
+
"loss": 2.1913,
|
| 4850 |
+
"step": 331500
|
| 4851 |
+
},
|
| 4852 |
+
{
|
| 4853 |
+
"epoch": 25.73045028287995,
|
| 4854 |
+
"grad_norm": 7.050189971923828,
|
| 4855 |
+
"learning_rate": 2.8463664780800333e-06,
|
| 4856 |
+
"loss": 2.191,
|
| 4857 |
+
"step": 332000
|
| 4858 |
+
},
|
| 4859 |
+
{
|
| 4860 |
+
"epoch": 25.769200961016818,
|
| 4861 |
+
"grad_norm": 6.654092311859131,
|
| 4862 |
+
"learning_rate": 2.8205326926554548e-06,
|
| 4863 |
+
"loss": 2.1871,
|
| 4864 |
+
"step": 332500
|
| 4865 |
+
},
|
| 4866 |
+
{
|
| 4867 |
+
"epoch": 25.807951639153686,
|
| 4868 |
+
"grad_norm": 7.114500522613525,
|
| 4869 |
+
"learning_rate": 2.7946989072308767e-06,
|
| 4870 |
+
"loss": 2.1842,
|
| 4871 |
+
"step": 333000
|
| 4872 |
+
},
|
| 4873 |
+
{
|
| 4874 |
+
"epoch": 25.846702317290553,
|
| 4875 |
+
"grad_norm": 6.987917900085449,
|
| 4876 |
+
"learning_rate": 2.7688651218062986e-06,
|
| 4877 |
+
"loss": 2.1782,
|
| 4878 |
+
"step": 333500
|
| 4879 |
+
},
|
| 4880 |
+
{
|
| 4881 |
+
"epoch": 25.88545299542742,
|
| 4882 |
+
"grad_norm": 6.479386806488037,
|
| 4883 |
+
"learning_rate": 2.74303133638172e-06,
|
| 4884 |
+
"loss": 2.1904,
|
| 4885 |
+
"step": 334000
|
| 4886 |
+
},
|
| 4887 |
+
{
|
| 4888 |
+
"epoch": 25.924203673564286,
|
| 4889 |
+
"grad_norm": 6.597611904144287,
|
| 4890 |
+
"learning_rate": 2.717197550957142e-06,
|
| 4891 |
+
"loss": 2.1782,
|
| 4892 |
+
"step": 334500
|
| 4893 |
+
},
|
| 4894 |
+
{
|
| 4895 |
+
"epoch": 25.962954351701153,
|
| 4896 |
+
"grad_norm": 7.492031097412109,
|
| 4897 |
+
"learning_rate": 2.6913637655325634e-06,
|
| 4898 |
+
"loss": 2.1976,
|
| 4899 |
+
"step": 335000
|
| 4900 |
+
},
|
| 4901 |
+
{
|
| 4902 |
+
"epoch": 26.0,
|
| 4903 |
+
"eval_loss": 2.144183874130249,
|
| 4904 |
+
"eval_runtime": 268.6578,
|
| 4905 |
+
"eval_samples_per_second": 768.505,
|
| 4906 |
+
"eval_steps_per_second": 12.012,
|
| 4907 |
+
"step": 335478
|
| 4908 |
+
},
|
| 4909 |
+
{
|
| 4910 |
+
"epoch": 26.00170502983802,
|
| 4911 |
+
"grad_norm": 7.5874552726745605,
|
| 4912 |
+
"learning_rate": 2.6655299801079853e-06,
|
| 4913 |
+
"loss": 2.1755,
|
| 4914 |
+
"step": 335500
|
| 4915 |
+
},
|
| 4916 |
+
{
|
| 4917 |
+
"epoch": 26.04045570797489,
|
| 4918 |
+
"grad_norm": 7.499856948852539,
|
| 4919 |
+
"learning_rate": 2.6396961946834072e-06,
|
| 4920 |
+
"loss": 2.1885,
|
| 4921 |
+
"step": 336000
|
| 4922 |
+
},
|
| 4923 |
+
{
|
| 4924 |
+
"epoch": 26.079206386111757,
|
| 4925 |
+
"grad_norm": 7.2821946144104,
|
| 4926 |
+
"learning_rate": 2.6138624092588287e-06,
|
| 4927 |
+
"loss": 2.1782,
|
| 4928 |
+
"step": 336500
|
| 4929 |
+
},
|
| 4930 |
+
{
|
| 4931 |
+
"epoch": 26.117957064248625,
|
| 4932 |
+
"grad_norm": 7.0137834548950195,
|
| 4933 |
+
"learning_rate": 2.5880286238342506e-06,
|
| 4934 |
+
"loss": 2.1688,
|
| 4935 |
+
"step": 337000
|
| 4936 |
+
},
|
| 4937 |
+
{
|
| 4938 |
+
"epoch": 26.156707742385493,
|
| 4939 |
+
"grad_norm": 6.468008518218994,
|
| 4940 |
+
"learning_rate": 2.5621948384096725e-06,
|
| 4941 |
+
"loss": 2.1735,
|
| 4942 |
+
"step": 337500
|
| 4943 |
+
},
|
| 4944 |
+
{
|
| 4945 |
+
"epoch": 26.19545842052236,
|
| 4946 |
+
"grad_norm": 6.922983169555664,
|
| 4947 |
+
"learning_rate": 2.536361052985094e-06,
|
| 4948 |
+
"loss": 2.1643,
|
| 4949 |
+
"step": 338000
|
| 4950 |
+
},
|
| 4951 |
+
{
|
| 4952 |
+
"epoch": 26.23420909865923,
|
| 4953 |
+
"grad_norm": 6.963326454162598,
|
| 4954 |
+
"learning_rate": 2.510527267560516e-06,
|
| 4955 |
+
"loss": 2.1569,
|
| 4956 |
+
"step": 338500
|
| 4957 |
+
},
|
| 4958 |
+
{
|
| 4959 |
+
"epoch": 26.272959776796093,
|
| 4960 |
+
"grad_norm": 6.4791579246521,
|
| 4961 |
+
"learning_rate": 2.4846934821359373e-06,
|
| 4962 |
+
"loss": 2.1816,
|
| 4963 |
+
"step": 339000
|
| 4964 |
+
},
|
| 4965 |
+
{
|
| 4966 |
+
"epoch": 26.31171045493296,
|
| 4967 |
+
"grad_norm": 7.289137840270996,
|
| 4968 |
+
"learning_rate": 2.4588596967113592e-06,
|
| 4969 |
+
"loss": 2.1628,
|
| 4970 |
+
"step": 339500
|
| 4971 |
+
},
|
| 4972 |
+
{
|
| 4973 |
+
"epoch": 26.350461133069828,
|
| 4974 |
+
"grad_norm": 7.020922660827637,
|
| 4975 |
+
"learning_rate": 2.433025911286781e-06,
|
| 4976 |
+
"loss": 2.1608,
|
| 4977 |
+
"step": 340000
|
| 4978 |
+
},
|
| 4979 |
+
{
|
| 4980 |
+
"epoch": 26.389211811206696,
|
| 4981 |
+
"grad_norm": 6.522220134735107,
|
| 4982 |
+
"learning_rate": 2.4071921258622026e-06,
|
| 4983 |
+
"loss": 2.1736,
|
| 4984 |
+
"step": 340500
|
| 4985 |
+
},
|
| 4986 |
+
{
|
| 4987 |
+
"epoch": 26.427962489343564,
|
| 4988 |
+
"grad_norm": 7.149320602416992,
|
| 4989 |
+
"learning_rate": 2.3813583404376245e-06,
|
| 4990 |
+
"loss": 2.1761,
|
| 4991 |
+
"step": 341000
|
| 4992 |
+
},
|
| 4993 |
+
{
|
| 4994 |
+
"epoch": 26.46671316748043,
|
| 4995 |
+
"grad_norm": 7.04742431640625,
|
| 4996 |
+
"learning_rate": 2.3555245550130464e-06,
|
| 4997 |
+
"loss": 2.168,
|
| 4998 |
+
"step": 341500
|
| 4999 |
+
},
|
| 5000 |
+
{
|
| 5001 |
+
"epoch": 26.5054638456173,
|
| 5002 |
+
"grad_norm": 7.135145664215088,
|
| 5003 |
+
"learning_rate": 2.329690769588468e-06,
|
| 5004 |
+
"loss": 2.1928,
|
| 5005 |
+
"step": 342000
|
| 5006 |
+
},
|
| 5007 |
+
{
|
| 5008 |
+
"epoch": 26.544214523754167,
|
| 5009 |
+
"grad_norm": 7.492802619934082,
|
| 5010 |
+
"learning_rate": 2.3038569841638898e-06,
|
| 5011 |
+
"loss": 2.1764,
|
| 5012 |
+
"step": 342500
|
| 5013 |
+
},
|
| 5014 |
+
{
|
| 5015 |
+
"epoch": 26.58296520189103,
|
| 5016 |
+
"grad_norm": 6.618491172790527,
|
| 5017 |
+
"learning_rate": 2.2780231987393112e-06,
|
| 5018 |
+
"loss": 2.1768,
|
| 5019 |
+
"step": 343000
|
| 5020 |
+
},
|
| 5021 |
+
{
|
| 5022 |
+
"epoch": 26.6217158800279,
|
| 5023 |
+
"grad_norm": 6.808167457580566,
|
| 5024 |
+
"learning_rate": 2.252189413314733e-06,
|
| 5025 |
+
"loss": 2.1623,
|
| 5026 |
+
"step": 343500
|
| 5027 |
+
},
|
| 5028 |
+
{
|
| 5029 |
+
"epoch": 26.660466558164767,
|
| 5030 |
+
"grad_norm": 6.65431022644043,
|
| 5031 |
+
"learning_rate": 2.226355627890155e-06,
|
| 5032 |
+
"loss": 2.1658,
|
| 5033 |
+
"step": 344000
|
| 5034 |
+
},
|
| 5035 |
+
{
|
| 5036 |
+
"epoch": 26.699217236301635,
|
| 5037 |
+
"grad_norm": 7.762594699859619,
|
| 5038 |
+
"learning_rate": 2.2005218424655765e-06,
|
| 5039 |
+
"loss": 2.1794,
|
| 5040 |
+
"step": 344500
|
| 5041 |
+
},
|
| 5042 |
+
{
|
| 5043 |
+
"epoch": 26.737967914438503,
|
| 5044 |
+
"grad_norm": 6.6927056312561035,
|
| 5045 |
+
"learning_rate": 2.1746880570409984e-06,
|
| 5046 |
+
"loss": 2.1624,
|
| 5047 |
+
"step": 345000
|
| 5048 |
+
},
|
| 5049 |
+
{
|
| 5050 |
+
"epoch": 26.77671859257537,
|
| 5051 |
+
"grad_norm": 6.606927394866943,
|
| 5052 |
+
"learning_rate": 2.1488542716164203e-06,
|
| 5053 |
+
"loss": 2.1741,
|
| 5054 |
+
"step": 345500
|
| 5055 |
+
},
|
| 5056 |
+
{
|
| 5057 |
+
"epoch": 26.81546927071224,
|
| 5058 |
+
"grad_norm": 6.104671955108643,
|
| 5059 |
+
"learning_rate": 2.1230204861918418e-06,
|
| 5060 |
+
"loss": 2.1716,
|
| 5061 |
+
"step": 346000
|
| 5062 |
+
},
|
| 5063 |
+
{
|
| 5064 |
+
"epoch": 26.854219948849106,
|
| 5065 |
+
"grad_norm": 5.965663433074951,
|
| 5066 |
+
"learning_rate": 2.0971867007672637e-06,
|
| 5067 |
+
"loss": 2.1674,
|
| 5068 |
+
"step": 346500
|
| 5069 |
+
},
|
| 5070 |
+
{
|
| 5071 |
+
"epoch": 26.89297062698597,
|
| 5072 |
+
"grad_norm": 6.041355133056641,
|
| 5073 |
+
"learning_rate": 2.071352915342685e-06,
|
| 5074 |
+
"loss": 2.181,
|
| 5075 |
+
"step": 347000
|
| 5076 |
+
},
|
| 5077 |
+
{
|
| 5078 |
+
"epoch": 26.93172130512284,
|
| 5079 |
+
"grad_norm": 7.279519557952881,
|
| 5080 |
+
"learning_rate": 2.045519129918107e-06,
|
| 5081 |
+
"loss": 2.1661,
|
| 5082 |
+
"step": 347500
|
| 5083 |
+
},
|
| 5084 |
+
{
|
| 5085 |
+
"epoch": 26.970471983259706,
|
| 5086 |
+
"grad_norm": 6.790727615356445,
|
| 5087 |
+
"learning_rate": 2.019685344493529e-06,
|
| 5088 |
+
"loss": 2.1658,
|
| 5089 |
+
"step": 348000
|
| 5090 |
+
},
|
| 5091 |
+
{
|
| 5092 |
+
"epoch": 27.0,
|
| 5093 |
+
"eval_loss": 2.137254238128662,
|
| 5094 |
+
"eval_runtime": 268.8992,
|
| 5095 |
+
"eval_samples_per_second": 767.815,
|
| 5096 |
+
"eval_steps_per_second": 12.001,
|
| 5097 |
+
"step": 348381
|
| 5098 |
+
},
|
| 5099 |
+
{
|
| 5100 |
+
"epoch": 27.009222661396574,
|
| 5101 |
+
"grad_norm": 6.905515193939209,
|
| 5102 |
+
"learning_rate": 1.9938515590689504e-06,
|
| 5103 |
+
"loss": 2.1569,
|
| 5104 |
+
"step": 348500
|
| 5105 |
+
},
|
| 5106 |
+
{
|
| 5107 |
+
"epoch": 27.047973339533442,
|
| 5108 |
+
"grad_norm": 6.515853404998779,
|
| 5109 |
+
"learning_rate": 1.9680177736443723e-06,
|
| 5110 |
+
"loss": 2.1713,
|
| 5111 |
+
"step": 349000
|
| 5112 |
+
},
|
| 5113 |
+
{
|
| 5114 |
+
"epoch": 27.08672401767031,
|
| 5115 |
+
"grad_norm": 6.981870651245117,
|
| 5116 |
+
"learning_rate": 1.942183988219794e-06,
|
| 5117 |
+
"loss": 2.1745,
|
| 5118 |
+
"step": 349500
|
| 5119 |
+
},
|
| 5120 |
+
{
|
| 5121 |
+
"epoch": 27.125474695807178,
|
| 5122 |
+
"grad_norm": 6.35358190536499,
|
| 5123 |
+
"learning_rate": 1.9163502027952157e-06,
|
| 5124 |
+
"loss": 2.1644,
|
| 5125 |
+
"step": 350000
|
| 5126 |
+
},
|
| 5127 |
+
{
|
| 5128 |
+
"epoch": 27.164225373944046,
|
| 5129 |
+
"grad_norm": 7.149428844451904,
|
| 5130 |
+
"learning_rate": 1.8905164173706376e-06,
|
| 5131 |
+
"loss": 2.1816,
|
| 5132 |
+
"step": 350500
|
| 5133 |
+
},
|
| 5134 |
+
{
|
| 5135 |
+
"epoch": 27.20297605208091,
|
| 5136 |
+
"grad_norm": 7.136536121368408,
|
| 5137 |
+
"learning_rate": 1.8646826319460593e-06,
|
| 5138 |
+
"loss": 2.1562,
|
| 5139 |
+
"step": 351000
|
| 5140 |
+
},
|
| 5141 |
+
{
|
| 5142 |
+
"epoch": 27.241726730217778,
|
| 5143 |
+
"grad_norm": 6.473196506500244,
|
| 5144 |
+
"learning_rate": 1.838848846521481e-06,
|
| 5145 |
+
"loss": 2.167,
|
| 5146 |
+
"step": 351500
|
| 5147 |
+
},
|
| 5148 |
+
{
|
| 5149 |
+
"epoch": 27.280477408354646,
|
| 5150 |
+
"grad_norm": 6.8429694175720215,
|
| 5151 |
+
"learning_rate": 1.8130150610969026e-06,
|
| 5152 |
+
"loss": 2.1587,
|
| 5153 |
+
"step": 352000
|
| 5154 |
+
},
|
| 5155 |
+
{
|
| 5156 |
+
"epoch": 27.319228086491513,
|
| 5157 |
+
"grad_norm": 6.667392253875732,
|
| 5158 |
+
"learning_rate": 1.7871812756723245e-06,
|
| 5159 |
+
"loss": 2.1575,
|
| 5160 |
+
"step": 352500
|
| 5161 |
+
},
|
| 5162 |
+
{
|
| 5163 |
+
"epoch": 27.35797876462838,
|
| 5164 |
+
"grad_norm": 7.551825046539307,
|
| 5165 |
+
"learning_rate": 1.7613474902477462e-06,
|
| 5166 |
+
"loss": 2.1567,
|
| 5167 |
+
"step": 353000
|
| 5168 |
+
},
|
| 5169 |
+
{
|
| 5170 |
+
"epoch": 27.39672944276525,
|
| 5171 |
+
"grad_norm": 7.393056392669678,
|
| 5172 |
+
"learning_rate": 1.735513704823168e-06,
|
| 5173 |
+
"loss": 2.163,
|
| 5174 |
+
"step": 353500
|
| 5175 |
+
},
|
| 5176 |
+
{
|
| 5177 |
+
"epoch": 27.435480120902117,
|
| 5178 |
+
"grad_norm": 6.7227678298950195,
|
| 5179 |
+
"learning_rate": 1.7096799193985896e-06,
|
| 5180 |
+
"loss": 2.1678,
|
| 5181 |
+
"step": 354000
|
| 5182 |
+
},
|
| 5183 |
+
{
|
| 5184 |
+
"epoch": 27.474230799038985,
|
| 5185 |
+
"grad_norm": 6.587380409240723,
|
| 5186 |
+
"learning_rate": 1.6838461339740115e-06,
|
| 5187 |
+
"loss": 2.1611,
|
| 5188 |
+
"step": 354500
|
| 5189 |
+
},
|
| 5190 |
+
{
|
| 5191 |
+
"epoch": 27.51298147717585,
|
| 5192 |
+
"grad_norm": 7.290678977966309,
|
| 5193 |
+
"learning_rate": 1.6580123485494332e-06,
|
| 5194 |
+
"loss": 2.1555,
|
| 5195 |
+
"step": 355000
|
| 5196 |
+
},
|
| 5197 |
+
{
|
| 5198 |
+
"epoch": 27.551732155312717,
|
| 5199 |
+
"grad_norm": 6.52154016494751,
|
| 5200 |
+
"learning_rate": 1.6321785631248548e-06,
|
| 5201 |
+
"loss": 2.1487,
|
| 5202 |
+
"step": 355500
|
| 5203 |
+
},
|
| 5204 |
+
{
|
| 5205 |
+
"epoch": 27.590482833449585,
|
| 5206 |
+
"grad_norm": 6.613160610198975,
|
| 5207 |
+
"learning_rate": 1.6063447777002765e-06,
|
| 5208 |
+
"loss": 2.1599,
|
| 5209 |
+
"step": 356000
|
| 5210 |
+
},
|
| 5211 |
+
{
|
| 5212 |
+
"epoch": 27.629233511586452,
|
| 5213 |
+
"grad_norm": 7.148532390594482,
|
| 5214 |
+
"learning_rate": 1.5805109922756984e-06,
|
| 5215 |
+
"loss": 2.1731,
|
| 5216 |
+
"step": 356500
|
| 5217 |
+
},
|
| 5218 |
+
{
|
| 5219 |
+
"epoch": 27.66798418972332,
|
| 5220 |
+
"grad_norm": 6.29647159576416,
|
| 5221 |
+
"learning_rate": 1.5546772068511201e-06,
|
| 5222 |
+
"loss": 2.1641,
|
| 5223 |
+
"step": 357000
|
| 5224 |
+
},
|
| 5225 |
+
{
|
| 5226 |
+
"epoch": 27.706734867860188,
|
| 5227 |
+
"grad_norm": 6.647765636444092,
|
| 5228 |
+
"learning_rate": 1.5288434214265418e-06,
|
| 5229 |
+
"loss": 2.1756,
|
| 5230 |
+
"step": 357500
|
| 5231 |
+
},
|
| 5232 |
+
{
|
| 5233 |
+
"epoch": 27.745485545997056,
|
| 5234 |
+
"grad_norm": 6.541094779968262,
|
| 5235 |
+
"learning_rate": 1.5030096360019635e-06,
|
| 5236 |
+
"loss": 2.1584,
|
| 5237 |
+
"step": 358000
|
| 5238 |
+
},
|
| 5239 |
+
{
|
| 5240 |
+
"epoch": 27.784236224133924,
|
| 5241 |
+
"grad_norm": 7.08396053314209,
|
| 5242 |
+
"learning_rate": 1.4771758505773854e-06,
|
| 5243 |
+
"loss": 2.1551,
|
| 5244 |
+
"step": 358500
|
| 5245 |
+
},
|
| 5246 |
+
{
|
| 5247 |
+
"epoch": 27.822986902270788,
|
| 5248 |
+
"grad_norm": 6.8339643478393555,
|
| 5249 |
+
"learning_rate": 1.451342065152807e-06,
|
| 5250 |
+
"loss": 2.1575,
|
| 5251 |
+
"step": 359000
|
| 5252 |
+
},
|
| 5253 |
+
{
|
| 5254 |
+
"epoch": 27.861737580407656,
|
| 5255 |
+
"grad_norm": 6.175314903259277,
|
| 5256 |
+
"learning_rate": 1.4255082797282288e-06,
|
| 5257 |
+
"loss": 2.1312,
|
| 5258 |
+
"step": 359500
|
| 5259 |
+
},
|
| 5260 |
+
{
|
| 5261 |
+
"epoch": 27.900488258544524,
|
| 5262 |
+
"grad_norm": 6.25184965133667,
|
| 5263 |
+
"learning_rate": 1.3996744943036504e-06,
|
| 5264 |
+
"loss": 2.1509,
|
| 5265 |
+
"step": 360000
|
| 5266 |
+
},
|
| 5267 |
+
{
|
| 5268 |
+
"epoch": 27.93923893668139,
|
| 5269 |
+
"grad_norm": 7.08027982711792,
|
| 5270 |
+
"learning_rate": 1.3738407088790723e-06,
|
| 5271 |
+
"loss": 2.159,
|
| 5272 |
+
"step": 360500
|
| 5273 |
+
},
|
| 5274 |
+
{
|
| 5275 |
+
"epoch": 27.97798961481826,
|
| 5276 |
+
"grad_norm": 6.8008880615234375,
|
| 5277 |
+
"learning_rate": 1.348006923454494e-06,
|
| 5278 |
+
"loss": 2.1634,
|
| 5279 |
+
"step": 361000
|
| 5280 |
+
},
|
| 5281 |
+
{
|
| 5282 |
+
"epoch": 28.0,
|
| 5283 |
+
"eval_loss": 2.130622386932373,
|
| 5284 |
+
"eval_runtime": 269.6844,
|
| 5285 |
+
"eval_samples_per_second": 765.58,
|
| 5286 |
+
"eval_steps_per_second": 11.966,
|
| 5287 |
+
"step": 361284
|
| 5288 |
+
},
|
| 5289 |
+
{
|
| 5290 |
+
"epoch": 28.016740292955127,
|
| 5291 |
+
"grad_norm": 7.46795654296875,
|
| 5292 |
+
"learning_rate": 1.3221731380299157e-06,
|
| 5293 |
+
"loss": 2.1363,
|
| 5294 |
+
"step": 361500
|
| 5295 |
+
},
|
| 5296 |
+
{
|
| 5297 |
+
"epoch": 28.055490971091995,
|
| 5298 |
+
"grad_norm": 7.271740436553955,
|
| 5299 |
+
"learning_rate": 1.2963393526053374e-06,
|
| 5300 |
+
"loss": 2.1604,
|
| 5301 |
+
"step": 362000
|
| 5302 |
+
},
|
| 5303 |
+
{
|
| 5304 |
+
"epoch": 28.094241649228863,
|
| 5305 |
+
"grad_norm": 6.692265510559082,
|
| 5306 |
+
"learning_rate": 1.2705055671807593e-06,
|
| 5307 |
+
"loss": 2.1596,
|
| 5308 |
+
"step": 362500
|
| 5309 |
+
},
|
| 5310 |
+
{
|
| 5311 |
+
"epoch": 28.132992327365727,
|
| 5312 |
+
"grad_norm": 6.122591018676758,
|
| 5313 |
+
"learning_rate": 1.2446717817561808e-06,
|
| 5314 |
+
"loss": 2.1524,
|
| 5315 |
+
"step": 363000
|
| 5316 |
+
},
|
| 5317 |
+
{
|
| 5318 |
+
"epoch": 28.171743005502595,
|
| 5319 |
+
"grad_norm": 6.683858394622803,
|
| 5320 |
+
"learning_rate": 1.2188379963316027e-06,
|
| 5321 |
+
"loss": 2.1598,
|
| 5322 |
+
"step": 363500
|
| 5323 |
+
},
|
| 5324 |
+
{
|
| 5325 |
+
"epoch": 28.210493683639463,
|
| 5326 |
+
"grad_norm": 6.768929958343506,
|
| 5327 |
+
"learning_rate": 1.1930042109070243e-06,
|
| 5328 |
+
"loss": 2.1515,
|
| 5329 |
+
"step": 364000
|
| 5330 |
+
},
|
| 5331 |
+
{
|
| 5332 |
+
"epoch": 28.24924436177633,
|
| 5333 |
+
"grad_norm": 6.956704139709473,
|
| 5334 |
+
"learning_rate": 1.167170425482446e-06,
|
| 5335 |
+
"loss": 2.1552,
|
| 5336 |
+
"step": 364500
|
| 5337 |
+
},
|
| 5338 |
+
{
|
| 5339 |
+
"epoch": 28.2879950399132,
|
| 5340 |
+
"grad_norm": 6.655780792236328,
|
| 5341 |
+
"learning_rate": 1.1413366400578677e-06,
|
| 5342 |
+
"loss": 2.1551,
|
| 5343 |
+
"step": 365000
|
| 5344 |
+
},
|
| 5345 |
+
{
|
| 5346 |
+
"epoch": 28.326745718050066,
|
| 5347 |
+
"grad_norm": 7.394413471221924,
|
| 5348 |
+
"learning_rate": 1.1155028546332896e-06,
|
| 5349 |
+
"loss": 2.1465,
|
| 5350 |
+
"step": 365500
|
| 5351 |
+
},
|
| 5352 |
+
{
|
| 5353 |
+
"epoch": 28.365496396186934,
|
| 5354 |
+
"grad_norm": 7.250267505645752,
|
| 5355 |
+
"learning_rate": 1.0896690692087113e-06,
|
| 5356 |
+
"loss": 2.1729,
|
| 5357 |
+
"step": 366000
|
| 5358 |
+
},
|
| 5359 |
+
{
|
| 5360 |
+
"epoch": 28.404247074323802,
|
| 5361 |
+
"grad_norm": 6.102252960205078,
|
| 5362 |
+
"learning_rate": 1.063835283784133e-06,
|
| 5363 |
+
"loss": 2.1556,
|
| 5364 |
+
"step": 366500
|
| 5365 |
+
},
|
| 5366 |
+
{
|
| 5367 |
+
"epoch": 28.44299775246067,
|
| 5368 |
+
"grad_norm": 6.5598297119140625,
|
| 5369 |
+
"learning_rate": 1.0380014983595547e-06,
|
| 5370 |
+
"loss": 2.1473,
|
| 5371 |
+
"step": 367000
|
| 5372 |
+
},
|
| 5373 |
+
{
|
| 5374 |
+
"epoch": 28.481748430597534,
|
| 5375 |
+
"grad_norm": 7.368846416473389,
|
| 5376 |
+
"learning_rate": 1.0121677129349766e-06,
|
| 5377 |
+
"loss": 2.1552,
|
| 5378 |
+
"step": 367500
|
| 5379 |
+
},
|
| 5380 |
+
{
|
| 5381 |
+
"epoch": 28.520499108734402,
|
| 5382 |
+
"grad_norm": 6.635545253753662,
|
| 5383 |
+
"learning_rate": 9.863339275103983e-07,
|
| 5384 |
+
"loss": 2.1584,
|
| 5385 |
+
"step": 368000
|
| 5386 |
+
},
|
| 5387 |
+
{
|
| 5388 |
+
"epoch": 28.55924978687127,
|
| 5389 |
+
"grad_norm": 6.502518177032471,
|
| 5390 |
+
"learning_rate": 9.6050014208582e-07,
|
| 5391 |
+
"loss": 2.1669,
|
| 5392 |
+
"step": 368500
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 28.598000465008138,
|
| 5396 |
+
"grad_norm": 7.150147914886475,
|
| 5397 |
+
"learning_rate": 9.346663566612417e-07,
|
| 5398 |
+
"loss": 2.158,
|
| 5399 |
+
"step": 369000
|
| 5400 |
+
},
|
| 5401 |
+
{
|
| 5402 |
+
"epoch": 28.636751143145005,
|
| 5403 |
+
"grad_norm": 6.391610622406006,
|
| 5404 |
+
"learning_rate": 9.088325712366634e-07,
|
| 5405 |
+
"loss": 2.1464,
|
| 5406 |
+
"step": 369500
|
| 5407 |
+
},
|
| 5408 |
+
{
|
| 5409 |
+
"epoch": 28.675501821281873,
|
| 5410 |
+
"grad_norm": 6.436591625213623,
|
| 5411 |
+
"learning_rate": 8.829987858120852e-07,
|
| 5412 |
+
"loss": 2.1438,
|
| 5413 |
+
"step": 370000
|
| 5414 |
+
},
|
| 5415 |
+
{
|
| 5416 |
+
"epoch": 28.71425249941874,
|
| 5417 |
+
"grad_norm": 6.646981716156006,
|
| 5418 |
+
"learning_rate": 8.571650003875069e-07,
|
| 5419 |
+
"loss": 2.1507,
|
| 5420 |
+
"step": 370500
|
| 5421 |
+
},
|
| 5422 |
+
{
|
| 5423 |
+
"epoch": 28.753003177555605,
|
| 5424 |
+
"grad_norm": 6.943175792694092,
|
| 5425 |
+
"learning_rate": 8.313312149629287e-07,
|
| 5426 |
+
"loss": 2.1483,
|
| 5427 |
+
"step": 371000
|
| 5428 |
+
},
|
| 5429 |
+
{
|
| 5430 |
+
"epoch": 28.791753855692473,
|
| 5431 |
+
"grad_norm": 6.345837116241455,
|
| 5432 |
+
"learning_rate": 8.054974295383504e-07,
|
| 5433 |
+
"loss": 2.1662,
|
| 5434 |
+
"step": 371500
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 28.83050453382934,
|
| 5438 |
+
"grad_norm": 6.562370300292969,
|
| 5439 |
+
"learning_rate": 7.79663644113772e-07,
|
| 5440 |
+
"loss": 2.1554,
|
| 5441 |
+
"step": 372000
|
| 5442 |
+
},
|
| 5443 |
+
{
|
| 5444 |
+
"epoch": 28.86925521196621,
|
| 5445 |
+
"grad_norm": 6.556326866149902,
|
| 5446 |
+
"learning_rate": 7.538298586891937e-07,
|
| 5447 |
+
"loss": 2.1448,
|
| 5448 |
+
"step": 372500
|
| 5449 |
+
},
|
| 5450 |
+
{
|
| 5451 |
+
"epoch": 28.908005890103077,
|
| 5452 |
+
"grad_norm": 6.487407684326172,
|
| 5453 |
+
"learning_rate": 7.279960732646154e-07,
|
| 5454 |
+
"loss": 2.1425,
|
| 5455 |
+
"step": 373000
|
| 5456 |
+
},
|
| 5457 |
+
{
|
| 5458 |
+
"epoch": 28.946756568239945,
|
| 5459 |
+
"grad_norm": 7.614674091339111,
|
| 5460 |
+
"learning_rate": 7.021622878400372e-07,
|
| 5461 |
+
"loss": 2.1565,
|
| 5462 |
+
"step": 373500
|
| 5463 |
+
},
|
| 5464 |
+
{
|
| 5465 |
+
"epoch": 28.985507246376812,
|
| 5466 |
+
"grad_norm": 6.897189140319824,
|
| 5467 |
+
"learning_rate": 6.763285024154589e-07,
|
| 5468 |
+
"loss": 2.1438,
|
| 5469 |
+
"step": 374000
|
| 5470 |
+
},
|
| 5471 |
+
{
|
| 5472 |
+
"epoch": 29.0,
|
| 5473 |
+
"eval_loss": 2.1226651668548584,
|
| 5474 |
+
"eval_runtime": 266.9511,
|
| 5475 |
+
"eval_samples_per_second": 773.419,
|
| 5476 |
+
"eval_steps_per_second": 12.088,
|
| 5477 |
+
"step": 374187
|
| 5478 |
+
},
|
| 5479 |
+
{
|
| 5480 |
+
"epoch": 29.02425792451368,
|
| 5481 |
+
"grad_norm": 6.869750499725342,
|
| 5482 |
+
"learning_rate": 6.504947169908807e-07,
|
| 5483 |
+
"loss": 2.138,
|
| 5484 |
+
"step": 374500
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 29.063008602650548,
|
| 5488 |
+
"grad_norm": 7.1249589920043945,
|
| 5489 |
+
"learning_rate": 6.246609315663025e-07,
|
| 5490 |
+
"loss": 2.1527,
|
| 5491 |
+
"step": 375000
|
| 5492 |
+
},
|
| 5493 |
+
{
|
| 5494 |
+
"epoch": 29.101759280787412,
|
| 5495 |
+
"grad_norm": 7.201192378997803,
|
| 5496 |
+
"learning_rate": 5.988271461417243e-07,
|
| 5497 |
+
"loss": 2.1517,
|
| 5498 |
+
"step": 375500
|
| 5499 |
+
},
|
| 5500 |
+
{
|
| 5501 |
+
"epoch": 29.14050995892428,
|
| 5502 |
+
"grad_norm": 6.720222473144531,
|
| 5503 |
+
"learning_rate": 5.729933607171459e-07,
|
| 5504 |
+
"loss": 2.1526,
|
| 5505 |
+
"step": 376000
|
| 5506 |
+
},
|
| 5507 |
+
{
|
| 5508 |
+
"epoch": 29.179260637061148,
|
| 5509 |
+
"grad_norm": 6.9030866622924805,
|
| 5510 |
+
"learning_rate": 5.471595752925676e-07,
|
| 5511 |
+
"loss": 2.1532,
|
| 5512 |
+
"step": 376500
|
| 5513 |
+
},
|
| 5514 |
+
{
|
| 5515 |
+
"epoch": 29.218011315198016,
|
| 5516 |
+
"grad_norm": 5.900801181793213,
|
| 5517 |
+
"learning_rate": 5.213257898679893e-07,
|
| 5518 |
+
"loss": 2.1534,
|
| 5519 |
+
"step": 377000
|
| 5520 |
+
},
|
| 5521 |
+
{
|
| 5522 |
+
"epoch": 29.256761993334884,
|
| 5523 |
+
"grad_norm": 6.259501934051514,
|
| 5524 |
+
"learning_rate": 4.954920044434111e-07,
|
| 5525 |
+
"loss": 2.1621,
|
| 5526 |
+
"step": 377500
|
| 5527 |
+
},
|
| 5528 |
+
{
|
| 5529 |
+
"epoch": 29.29551267147175,
|
| 5530 |
+
"grad_norm": 6.566405296325684,
|
| 5531 |
+
"learning_rate": 4.6965821901883286e-07,
|
| 5532 |
+
"loss": 2.1621,
|
| 5533 |
+
"step": 378000
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 29.33426334960862,
|
| 5537 |
+
"grad_norm": 6.553793430328369,
|
| 5538 |
+
"learning_rate": 4.438244335942546e-07,
|
| 5539 |
+
"loss": 2.1631,
|
| 5540 |
+
"step": 378500
|
| 5541 |
+
},
|
| 5542 |
+
{
|
| 5543 |
+
"epoch": 29.373014027745487,
|
| 5544 |
+
"grad_norm": 6.773620128631592,
|
| 5545 |
+
"learning_rate": 4.1799064816967634e-07,
|
| 5546 |
+
"loss": 2.1556,
|
| 5547 |
+
"step": 379000
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 29.41176470588235,
|
| 5551 |
+
"grad_norm": 6.494615077972412,
|
| 5552 |
+
"learning_rate": 3.921568627450981e-07,
|
| 5553 |
+
"loss": 2.1554,
|
| 5554 |
+
"step": 379500
|
| 5555 |
+
},
|
| 5556 |
+
{
|
| 5557 |
+
"epoch": 29.45051538401922,
|
| 5558 |
+
"grad_norm": 7.172949314117432,
|
| 5559 |
+
"learning_rate": 3.663230773205198e-07,
|
| 5560 |
+
"loss": 2.1494,
|
| 5561 |
+
"step": 380000
|
| 5562 |
+
},
|
| 5563 |
+
{
|
| 5564 |
+
"epoch": 29.489266062156087,
|
| 5565 |
+
"grad_norm": 6.8991241455078125,
|
| 5566 |
+
"learning_rate": 3.4048929189594155e-07,
|
| 5567 |
+
"loss": 2.1406,
|
| 5568 |
+
"step": 380500
|
| 5569 |
+
},
|
| 5570 |
+
{
|
| 5571 |
+
"epoch": 29.528016740292955,
|
| 5572 |
+
"grad_norm": 7.046799182891846,
|
| 5573 |
+
"learning_rate": 3.146555064713633e-07,
|
| 5574 |
+
"loss": 2.1493,
|
| 5575 |
+
"step": 381000
|
| 5576 |
+
},
|
| 5577 |
+
{
|
| 5578 |
+
"epoch": 29.566767418429823,
|
| 5579 |
+
"grad_norm": 6.826701641082764,
|
| 5580 |
+
"learning_rate": 2.8882172104678503e-07,
|
| 5581 |
+
"loss": 2.1459,
|
| 5582 |
+
"step": 381500
|
| 5583 |
+
},
|
| 5584 |
+
{
|
| 5585 |
+
"epoch": 29.60551809656669,
|
| 5586 |
+
"grad_norm": 6.649389743804932,
|
| 5587 |
+
"learning_rate": 2.6298793562220677e-07,
|
| 5588 |
+
"loss": 2.1593,
|
| 5589 |
+
"step": 382000
|
| 5590 |
+
},
|
| 5591 |
+
{
|
| 5592 |
+
"epoch": 29.64426877470356,
|
| 5593 |
+
"grad_norm": 6.10260009765625,
|
| 5594 |
+
"learning_rate": 2.3715415019762845e-07,
|
| 5595 |
+
"loss": 2.151,
|
| 5596 |
+
"step": 382500
|
| 5597 |
+
},
|
| 5598 |
+
{
|
| 5599 |
+
"epoch": 29.683019452840426,
|
| 5600 |
+
"grad_norm": 6.9101128578186035,
|
| 5601 |
+
"learning_rate": 2.113203647730502e-07,
|
| 5602 |
+
"loss": 2.1625,
|
| 5603 |
+
"step": 383000
|
| 5604 |
+
},
|
| 5605 |
+
{
|
| 5606 |
+
"epoch": 29.72177013097729,
|
| 5607 |
+
"grad_norm": 6.671387672424316,
|
| 5608 |
+
"learning_rate": 1.8548657934847193e-07,
|
| 5609 |
+
"loss": 2.1496,
|
| 5610 |
+
"step": 383500
|
| 5611 |
+
},
|
| 5612 |
+
{
|
| 5613 |
+
"epoch": 29.76052080911416,
|
| 5614 |
+
"grad_norm": 7.705864429473877,
|
| 5615 |
+
"learning_rate": 1.5965279392389367e-07,
|
| 5616 |
+
"loss": 2.1496,
|
| 5617 |
+
"step": 384000
|
| 5618 |
+
},
|
| 5619 |
+
{
|
| 5620 |
+
"epoch": 29.799271487251026,
|
| 5621 |
+
"grad_norm": 6.319827079772949,
|
| 5622 |
+
"learning_rate": 1.338190084993154e-07,
|
| 5623 |
+
"loss": 2.149,
|
| 5624 |
+
"step": 384500
|
| 5625 |
+
},
|
| 5626 |
+
{
|
| 5627 |
+
"epoch": 29.838022165387894,
|
| 5628 |
+
"grad_norm": 6.850592613220215,
|
| 5629 |
+
"learning_rate": 1.0798522307473716e-07,
|
| 5630 |
+
"loss": 2.1511,
|
| 5631 |
+
"step": 385000
|
| 5632 |
+
},
|
| 5633 |
+
{
|
| 5634 |
+
"epoch": 29.876772843524762,
|
| 5635 |
+
"grad_norm": 7.166690826416016,
|
| 5636 |
+
"learning_rate": 8.21514376501589e-08,
|
| 5637 |
+
"loss": 2.1492,
|
| 5638 |
+
"step": 385500
|
| 5639 |
+
},
|
| 5640 |
+
{
|
| 5641 |
+
"epoch": 29.91552352166163,
|
| 5642 |
+
"grad_norm": 6.324465274810791,
|
| 5643 |
+
"learning_rate": 5.631765222558062e-08,
|
| 5644 |
+
"loss": 2.1523,
|
| 5645 |
+
"step": 386000
|
| 5646 |
+
},
|
| 5647 |
+
{
|
| 5648 |
+
"epoch": 29.954274199798498,
|
| 5649 |
+
"grad_norm": 7.214087009429932,
|
| 5650 |
+
"learning_rate": 3.048386680100235e-08,
|
| 5651 |
+
"loss": 2.1628,
|
| 5652 |
+
"step": 386500
|
| 5653 |
+
},
|
| 5654 |
+
{
|
| 5655 |
+
"epoch": 29.993024877935365,
|
| 5656 |
+
"grad_norm": 6.88369607925415,
|
| 5657 |
+
"learning_rate": 4.6500813764240875e-09,
|
| 5658 |
+
"loss": 2.144,
|
| 5659 |
+
"step": 387000
|
| 5660 |
+
},
|
| 5661 |
+
{
|
| 5662 |
+
"epoch": 30.0,
|
| 5663 |
+
"eval_loss": 2.125218629837036,
|
| 5664 |
+
"eval_runtime": 267.0237,
|
| 5665 |
+
"eval_samples_per_second": 773.209,
|
| 5666 |
+
"eval_steps_per_second": 12.085,
|
| 5667 |
+
"step": 387090
|
| 5668 |
}
|
| 5669 |
],
|
| 5670 |
"logging_steps": 500,
|
| 5671 |
+
"max_steps": 387090,
|
| 5672 |
"num_input_tokens_seen": 0,
|
| 5673 |
+
"num_train_epochs": 30,
|
| 5674 |
"save_steps": 500,
|
| 5675 |
"stateful_callbacks": {
|
| 5676 |
"TrainerControl": {
|
|
|
|
| 5684 |
"attributes": {}
|
| 5685 |
}
|
| 5686 |
},
|
| 5687 |
+
"total_flos": 9.706577784666885e+17,
|
| 5688 |
"train_batch_size": 64,
|
| 5689 |
"trial_name": null,
|
| 5690 |
"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 sha256:26bd31272e273002d078f94ae220a0225f0a03244f57837a66ef9002c2b6fb24
|
| 3 |
size 5048
|