checkpoint-26200-uploaded-manually
Browse files- bn_adapter_ultrachatconv25april2024/adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +2243 -3
bn_adapter_ultrachatconv25april2024/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1195438880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25609992204958fc1ee37024b6a504d718c9cadba289138d4c060ec421c86a54
|
| 3 |
size 1195438880
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1198237404
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09d92075491bf5f5606a3a5a371709841022fc2c5a91e033c7c081fd600d6d04
|
| 3 |
size 1198237404
|
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 sha256:0abb51f32da6588af521ea52e1d3bf2d6faf48968d797fd32303a7862fb8dd97
|
| 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 sha256:43cc6f7e24533ba24f16551452140098ac7955092206a834a0a31696d925450b
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -6937,6 +6937,2246 @@
|
|
| 6937 |
"learning_rate": 1.840239603620849e-05,
|
| 6938 |
"loss": 2.1362,
|
| 6939 |
"step": 19800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6940 |
}
|
| 6941 |
],
|
| 6942 |
"logging_steps": 20,
|
|
@@ -6944,7 +9184,7 @@
|
|
| 6944 |
"num_input_tokens_seen": 0,
|
| 6945 |
"num_train_epochs": 2,
|
| 6946 |
"save_steps": 200,
|
| 6947 |
-
"total_flos":
|
| 6948 |
"train_batch_size": 16,
|
| 6949 |
"trial_name": null,
|
| 6950 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.48275353773584906,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 26200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 6937 |
"learning_rate": 1.840239603620849e-05,
|
| 6938 |
"loss": 2.1362,
|
| 6939 |
"step": 19800
|
| 6940 |
+
},
|
| 6941 |
+
{
|
| 6942 |
+
"epoch": 0.36519752358490565,
|
| 6943 |
+
"grad_norm": 2.71875,
|
| 6944 |
+
"learning_rate": 1.839925595675872e-05,
|
| 6945 |
+
"loss": 2.1075,
|
| 6946 |
+
"step": 19820
|
| 6947 |
+
},
|
| 6948 |
+
{
|
| 6949 |
+
"epoch": 0.36556603773584906,
|
| 6950 |
+
"grad_norm": 2.21875,
|
| 6951 |
+
"learning_rate": 1.8396113062887467e-05,
|
| 6952 |
+
"loss": 2.1305,
|
| 6953 |
+
"step": 19840
|
| 6954 |
+
},
|
| 6955 |
+
{
|
| 6956 |
+
"epoch": 0.36593455188679247,
|
| 6957 |
+
"grad_norm": 2.53125,
|
| 6958 |
+
"learning_rate": 1.839296735564786e-05,
|
| 6959 |
+
"loss": 2.1248,
|
| 6960 |
+
"step": 19860
|
| 6961 |
+
},
|
| 6962 |
+
{
|
| 6963 |
+
"epoch": 0.3663030660377358,
|
| 6964 |
+
"grad_norm": 2.359375,
|
| 6965 |
+
"learning_rate": 1.8389818836093948e-05,
|
| 6966 |
+
"loss": 2.1572,
|
| 6967 |
+
"step": 19880
|
| 6968 |
+
},
|
| 6969 |
+
{
|
| 6970 |
+
"epoch": 0.36667158018867924,
|
| 6971 |
+
"grad_norm": 2.390625,
|
| 6972 |
+
"learning_rate": 1.8386667505280745e-05,
|
| 6973 |
+
"loss": 2.1612,
|
| 6974 |
+
"step": 19900
|
| 6975 |
+
},
|
| 6976 |
+
{
|
| 6977 |
+
"epoch": 0.36704009433962265,
|
| 6978 |
+
"grad_norm": 2.40625,
|
| 6979 |
+
"learning_rate": 1.83835133642642e-05,
|
| 6980 |
+
"loss": 2.158,
|
| 6981 |
+
"step": 19920
|
| 6982 |
+
},
|
| 6983 |
+
{
|
| 6984 |
+
"epoch": 0.36740860849056606,
|
| 6985 |
+
"grad_norm": 2.9375,
|
| 6986 |
+
"learning_rate": 1.8380356414101198e-05,
|
| 6987 |
+
"loss": 2.096,
|
| 6988 |
+
"step": 19940
|
| 6989 |
+
},
|
| 6990 |
+
{
|
| 6991 |
+
"epoch": 0.3677771226415094,
|
| 6992 |
+
"grad_norm": 2.609375,
|
| 6993 |
+
"learning_rate": 1.837719665584957e-05,
|
| 6994 |
+
"loss": 2.1365,
|
| 6995 |
+
"step": 19960
|
| 6996 |
+
},
|
| 6997 |
+
{
|
| 6998 |
+
"epoch": 0.3681456367924528,
|
| 6999 |
+
"grad_norm": 2.65625,
|
| 7000 |
+
"learning_rate": 1.837403409056809e-05,
|
| 7001 |
+
"loss": 2.1483,
|
| 7002 |
+
"step": 19980
|
| 7003 |
+
},
|
| 7004 |
+
{
|
| 7005 |
+
"epoch": 0.36851415094339623,
|
| 7006 |
+
"grad_norm": 3.03125,
|
| 7007 |
+
"learning_rate": 1.837086871931647e-05,
|
| 7008 |
+
"loss": 2.1442,
|
| 7009 |
+
"step": 20000
|
| 7010 |
+
},
|
| 7011 |
+
{
|
| 7012 |
+
"epoch": 0.36888266509433965,
|
| 7013 |
+
"grad_norm": 2.40625,
|
| 7014 |
+
"learning_rate": 1.8367700543155355e-05,
|
| 7015 |
+
"loss": 2.1285,
|
| 7016 |
+
"step": 20020
|
| 7017 |
+
},
|
| 7018 |
+
{
|
| 7019 |
+
"epoch": 0.369251179245283,
|
| 7020 |
+
"grad_norm": 2.25,
|
| 7021 |
+
"learning_rate": 1.8364529563146348e-05,
|
| 7022 |
+
"loss": 2.1639,
|
| 7023 |
+
"step": 20040
|
| 7024 |
+
},
|
| 7025 |
+
{
|
| 7026 |
+
"epoch": 0.3696196933962264,
|
| 7027 |
+
"grad_norm": 2.40625,
|
| 7028 |
+
"learning_rate": 1.836135578035197e-05,
|
| 7029 |
+
"loss": 2.1572,
|
| 7030 |
+
"step": 20060
|
| 7031 |
+
},
|
| 7032 |
+
{
|
| 7033 |
+
"epoch": 0.3699882075471698,
|
| 7034 |
+
"grad_norm": 2.328125,
|
| 7035 |
+
"learning_rate": 1.83581791958357e-05,
|
| 7036 |
+
"loss": 2.1444,
|
| 7037 |
+
"step": 20080
|
| 7038 |
+
},
|
| 7039 |
+
{
|
| 7040 |
+
"epoch": 0.37035672169811323,
|
| 7041 |
+
"grad_norm": 2.46875,
|
| 7042 |
+
"learning_rate": 1.8354999810661942e-05,
|
| 7043 |
+
"loss": 2.1119,
|
| 7044 |
+
"step": 20100
|
| 7045 |
+
},
|
| 7046 |
+
{
|
| 7047 |
+
"epoch": 0.3707252358490566,
|
| 7048 |
+
"grad_norm": 2.5625,
|
| 7049 |
+
"learning_rate": 1.8351817625896046e-05,
|
| 7050 |
+
"loss": 2.0858,
|
| 7051 |
+
"step": 20120
|
| 7052 |
+
},
|
| 7053 |
+
{
|
| 7054 |
+
"epoch": 0.37109375,
|
| 7055 |
+
"grad_norm": 2.625,
|
| 7056 |
+
"learning_rate": 1.8348632642604297e-05,
|
| 7057 |
+
"loss": 2.1187,
|
| 7058 |
+
"step": 20140
|
| 7059 |
+
},
|
| 7060 |
+
{
|
| 7061 |
+
"epoch": 0.3714622641509434,
|
| 7062 |
+
"grad_norm": 2.734375,
|
| 7063 |
+
"learning_rate": 1.8345444861853922e-05,
|
| 7064 |
+
"loss": 2.1463,
|
| 7065 |
+
"step": 20160
|
| 7066 |
+
},
|
| 7067 |
+
{
|
| 7068 |
+
"epoch": 0.37183077830188677,
|
| 7069 |
+
"grad_norm": 2.640625,
|
| 7070 |
+
"learning_rate": 1.834225428471308e-05,
|
| 7071 |
+
"loss": 2.1447,
|
| 7072 |
+
"step": 20180
|
| 7073 |
+
},
|
| 7074 |
+
{
|
| 7075 |
+
"epoch": 0.3721992924528302,
|
| 7076 |
+
"grad_norm": 2.5,
|
| 7077 |
+
"learning_rate": 1.833906091225087e-05,
|
| 7078 |
+
"loss": 2.1054,
|
| 7079 |
+
"step": 20200
|
| 7080 |
+
},
|
| 7081 |
+
{
|
| 7082 |
+
"epoch": 0.3725678066037736,
|
| 7083 |
+
"grad_norm": 3.0,
|
| 7084 |
+
"learning_rate": 1.8335864745537323e-05,
|
| 7085 |
+
"loss": 2.0707,
|
| 7086 |
+
"step": 20220
|
| 7087 |
+
},
|
| 7088 |
+
{
|
| 7089 |
+
"epoch": 0.372936320754717,
|
| 7090 |
+
"grad_norm": 2.296875,
|
| 7091 |
+
"learning_rate": 1.8332665785643414e-05,
|
| 7092 |
+
"loss": 2.1163,
|
| 7093 |
+
"step": 20240
|
| 7094 |
+
},
|
| 7095 |
+
{
|
| 7096 |
+
"epoch": 0.37330483490566035,
|
| 7097 |
+
"grad_norm": 2.5,
|
| 7098 |
+
"learning_rate": 1.8329464033641048e-05,
|
| 7099 |
+
"loss": 2.1219,
|
| 7100 |
+
"step": 20260
|
| 7101 |
+
},
|
| 7102 |
+
{
|
| 7103 |
+
"epoch": 0.37367334905660377,
|
| 7104 |
+
"grad_norm": 3.09375,
|
| 7105 |
+
"learning_rate": 1.832625949060307e-05,
|
| 7106 |
+
"loss": 2.1054,
|
| 7107 |
+
"step": 20280
|
| 7108 |
+
},
|
| 7109 |
+
{
|
| 7110 |
+
"epoch": 0.3740418632075472,
|
| 7111 |
+
"grad_norm": 2.359375,
|
| 7112 |
+
"learning_rate": 1.8323052157603256e-05,
|
| 7113 |
+
"loss": 2.1189,
|
| 7114 |
+
"step": 20300
|
| 7115 |
+
},
|
| 7116 |
+
{
|
| 7117 |
+
"epoch": 0.3744103773584906,
|
| 7118 |
+
"grad_norm": 2.171875,
|
| 7119 |
+
"learning_rate": 1.831984203571632e-05,
|
| 7120 |
+
"loss": 2.1239,
|
| 7121 |
+
"step": 20320
|
| 7122 |
+
},
|
| 7123 |
+
{
|
| 7124 |
+
"epoch": 0.37477889150943394,
|
| 7125 |
+
"grad_norm": 2.375,
|
| 7126 |
+
"learning_rate": 1.8316629126017906e-05,
|
| 7127 |
+
"loss": 2.1513,
|
| 7128 |
+
"step": 20340
|
| 7129 |
+
},
|
| 7130 |
+
{
|
| 7131 |
+
"epoch": 0.37514740566037735,
|
| 7132 |
+
"grad_norm": 2.40625,
|
| 7133 |
+
"learning_rate": 1.8313413429584596e-05,
|
| 7134 |
+
"loss": 2.1285,
|
| 7135 |
+
"step": 20360
|
| 7136 |
+
},
|
| 7137 |
+
{
|
| 7138 |
+
"epoch": 0.37551591981132076,
|
| 7139 |
+
"grad_norm": 2.40625,
|
| 7140 |
+
"learning_rate": 1.8310194947493907e-05,
|
| 7141 |
+
"loss": 2.1439,
|
| 7142 |
+
"step": 20380
|
| 7143 |
+
},
|
| 7144 |
+
{
|
| 7145 |
+
"epoch": 0.3758844339622642,
|
| 7146 |
+
"grad_norm": 2.640625,
|
| 7147 |
+
"learning_rate": 1.830697368082429e-05,
|
| 7148 |
+
"loss": 2.128,
|
| 7149 |
+
"step": 20400
|
| 7150 |
+
},
|
| 7151 |
+
{
|
| 7152 |
+
"epoch": 0.37625294811320753,
|
| 7153 |
+
"grad_norm": 2.203125,
|
| 7154 |
+
"learning_rate": 1.8303749630655125e-05,
|
| 7155 |
+
"loss": 2.1059,
|
| 7156 |
+
"step": 20420
|
| 7157 |
+
},
|
| 7158 |
+
{
|
| 7159 |
+
"epoch": 0.37662146226415094,
|
| 7160 |
+
"grad_norm": 2.296875,
|
| 7161 |
+
"learning_rate": 1.830052279806672e-05,
|
| 7162 |
+
"loss": 2.1378,
|
| 7163 |
+
"step": 20440
|
| 7164 |
+
},
|
| 7165 |
+
{
|
| 7166 |
+
"epoch": 0.37698997641509435,
|
| 7167 |
+
"grad_norm": 2.703125,
|
| 7168 |
+
"learning_rate": 1.829729318414033e-05,
|
| 7169 |
+
"loss": 2.1479,
|
| 7170 |
+
"step": 20460
|
| 7171 |
+
},
|
| 7172 |
+
{
|
| 7173 |
+
"epoch": 0.37735849056603776,
|
| 7174 |
+
"grad_norm": 2.3125,
|
| 7175 |
+
"learning_rate": 1.8294060789958128e-05,
|
| 7176 |
+
"loss": 2.1055,
|
| 7177 |
+
"step": 20480
|
| 7178 |
+
},
|
| 7179 |
+
{
|
| 7180 |
+
"epoch": 0.3777270047169811,
|
| 7181 |
+
"grad_norm": 2.453125,
|
| 7182 |
+
"learning_rate": 1.829082561660323e-05,
|
| 7183 |
+
"loss": 2.1284,
|
| 7184 |
+
"step": 20500
|
| 7185 |
+
},
|
| 7186 |
+
{
|
| 7187 |
+
"epoch": 0.37809551886792453,
|
| 7188 |
+
"grad_norm": 2.328125,
|
| 7189 |
+
"learning_rate": 1.8287587665159668e-05,
|
| 7190 |
+
"loss": 2.139,
|
| 7191 |
+
"step": 20520
|
| 7192 |
+
},
|
| 7193 |
+
{
|
| 7194 |
+
"epoch": 0.37846403301886794,
|
| 7195 |
+
"grad_norm": 2.453125,
|
| 7196 |
+
"learning_rate": 1.8284346936712425e-05,
|
| 7197 |
+
"loss": 2.1444,
|
| 7198 |
+
"step": 20540
|
| 7199 |
+
},
|
| 7200 |
+
{
|
| 7201 |
+
"epoch": 0.3788325471698113,
|
| 7202 |
+
"grad_norm": 2.5,
|
| 7203 |
+
"learning_rate": 1.8281103432347397e-05,
|
| 7204 |
+
"loss": 2.1615,
|
| 7205 |
+
"step": 20560
|
| 7206 |
+
},
|
| 7207 |
+
{
|
| 7208 |
+
"epoch": 0.3792010613207547,
|
| 7209 |
+
"grad_norm": 2.28125,
|
| 7210 |
+
"learning_rate": 1.827785715315142e-05,
|
| 7211 |
+
"loss": 2.0913,
|
| 7212 |
+
"step": 20580
|
| 7213 |
+
},
|
| 7214 |
+
{
|
| 7215 |
+
"epoch": 0.3795695754716981,
|
| 7216 |
+
"grad_norm": 2.3125,
|
| 7217 |
+
"learning_rate": 1.8274608100212253e-05,
|
| 7218 |
+
"loss": 2.1519,
|
| 7219 |
+
"step": 20600
|
| 7220 |
+
},
|
| 7221 |
+
{
|
| 7222 |
+
"epoch": 0.37993808962264153,
|
| 7223 |
+
"grad_norm": 2.28125,
|
| 7224 |
+
"learning_rate": 1.8271356274618594e-05,
|
| 7225 |
+
"loss": 2.108,
|
| 7226 |
+
"step": 20620
|
| 7227 |
+
},
|
| 7228 |
+
{
|
| 7229 |
+
"epoch": 0.3803066037735849,
|
| 7230 |
+
"grad_norm": 2.3125,
|
| 7231 |
+
"learning_rate": 1.8268101677460056e-05,
|
| 7232 |
+
"loss": 2.1208,
|
| 7233 |
+
"step": 20640
|
| 7234 |
+
},
|
| 7235 |
+
{
|
| 7236 |
+
"epoch": 0.3806751179245283,
|
| 7237 |
+
"grad_norm": 2.234375,
|
| 7238 |
+
"learning_rate": 1.8264844309827195e-05,
|
| 7239 |
+
"loss": 2.1263,
|
| 7240 |
+
"step": 20660
|
| 7241 |
+
},
|
| 7242 |
+
{
|
| 7243 |
+
"epoch": 0.3810436320754717,
|
| 7244 |
+
"grad_norm": 2.359375,
|
| 7245 |
+
"learning_rate": 1.8261584172811494e-05,
|
| 7246 |
+
"loss": 2.1344,
|
| 7247 |
+
"step": 20680
|
| 7248 |
+
},
|
| 7249 |
+
{
|
| 7250 |
+
"epoch": 0.3814121462264151,
|
| 7251 |
+
"grad_norm": 2.484375,
|
| 7252 |
+
"learning_rate": 1.825832126750535e-05,
|
| 7253 |
+
"loss": 2.1199,
|
| 7254 |
+
"step": 20700
|
| 7255 |
+
},
|
| 7256 |
+
{
|
| 7257 |
+
"epoch": 0.38178066037735847,
|
| 7258 |
+
"grad_norm": 2.6875,
|
| 7259 |
+
"learning_rate": 1.82550555950021e-05,
|
| 7260 |
+
"loss": 2.1269,
|
| 7261 |
+
"step": 20720
|
| 7262 |
+
},
|
| 7263 |
+
{
|
| 7264 |
+
"epoch": 0.3821491745283019,
|
| 7265 |
+
"grad_norm": 3.234375,
|
| 7266 |
+
"learning_rate": 1.8251787156396002e-05,
|
| 7267 |
+
"loss": 2.0986,
|
| 7268 |
+
"step": 20740
|
| 7269 |
+
},
|
| 7270 |
+
{
|
| 7271 |
+
"epoch": 0.3825176886792453,
|
| 7272 |
+
"grad_norm": 2.5625,
|
| 7273 |
+
"learning_rate": 1.8248515952782254e-05,
|
| 7274 |
+
"loss": 2.1599,
|
| 7275 |
+
"step": 20760
|
| 7276 |
+
},
|
| 7277 |
+
{
|
| 7278 |
+
"epoch": 0.3828862028301887,
|
| 7279 |
+
"grad_norm": 2.46875,
|
| 7280 |
+
"learning_rate": 1.824524198525696e-05,
|
| 7281 |
+
"loss": 2.137,
|
| 7282 |
+
"step": 20780
|
| 7283 |
+
},
|
| 7284 |
+
{
|
| 7285 |
+
"epoch": 0.38325471698113206,
|
| 7286 |
+
"grad_norm": 2.484375,
|
| 7287 |
+
"learning_rate": 1.8241965254917168e-05,
|
| 7288 |
+
"loss": 2.1332,
|
| 7289 |
+
"step": 20800
|
| 7290 |
+
},
|
| 7291 |
+
{
|
| 7292 |
+
"epoch": 0.38362323113207547,
|
| 7293 |
+
"grad_norm": 2.4375,
|
| 7294 |
+
"learning_rate": 1.8238685762860835e-05,
|
| 7295 |
+
"loss": 2.137,
|
| 7296 |
+
"step": 20820
|
| 7297 |
+
},
|
| 7298 |
+
{
|
| 7299 |
+
"epoch": 0.3839917452830189,
|
| 7300 |
+
"grad_norm": 2.28125,
|
| 7301 |
+
"learning_rate": 1.8235403510186863e-05,
|
| 7302 |
+
"loss": 2.1272,
|
| 7303 |
+
"step": 20840
|
| 7304 |
+
},
|
| 7305 |
+
{
|
| 7306 |
+
"epoch": 0.38436025943396224,
|
| 7307 |
+
"grad_norm": 2.515625,
|
| 7308 |
+
"learning_rate": 1.8232118497995058e-05,
|
| 7309 |
+
"loss": 2.0837,
|
| 7310 |
+
"step": 20860
|
| 7311 |
+
},
|
| 7312 |
+
{
|
| 7313 |
+
"epoch": 0.38472877358490565,
|
| 7314 |
+
"grad_norm": 2.890625,
|
| 7315 |
+
"learning_rate": 1.8228830727386175e-05,
|
| 7316 |
+
"loss": 2.1107,
|
| 7317 |
+
"step": 20880
|
| 7318 |
+
},
|
| 7319 |
+
{
|
| 7320 |
+
"epoch": 0.38509728773584906,
|
| 7321 |
+
"grad_norm": 2.546875,
|
| 7322 |
+
"learning_rate": 1.822554019946187e-05,
|
| 7323 |
+
"loss": 2.1133,
|
| 7324 |
+
"step": 20900
|
| 7325 |
+
},
|
| 7326 |
+
{
|
| 7327 |
+
"epoch": 0.38546580188679247,
|
| 7328 |
+
"grad_norm": 3.0625,
|
| 7329 |
+
"learning_rate": 1.8222246915324734e-05,
|
| 7330 |
+
"loss": 2.1449,
|
| 7331 |
+
"step": 20920
|
| 7332 |
+
},
|
| 7333 |
+
{
|
| 7334 |
+
"epoch": 0.3858343160377358,
|
| 7335 |
+
"grad_norm": 2.46875,
|
| 7336 |
+
"learning_rate": 1.821895087607828e-05,
|
| 7337 |
+
"loss": 2.1228,
|
| 7338 |
+
"step": 20940
|
| 7339 |
+
},
|
| 7340 |
+
{
|
| 7341 |
+
"epoch": 0.38620283018867924,
|
| 7342 |
+
"grad_norm": 2.375,
|
| 7343 |
+
"learning_rate": 1.8215652082826945e-05,
|
| 7344 |
+
"loss": 2.1257,
|
| 7345 |
+
"step": 20960
|
| 7346 |
+
},
|
| 7347 |
+
{
|
| 7348 |
+
"epoch": 0.38657134433962265,
|
| 7349 |
+
"grad_norm": 2.25,
|
| 7350 |
+
"learning_rate": 1.8212350536676093e-05,
|
| 7351 |
+
"loss": 2.114,
|
| 7352 |
+
"step": 20980
|
| 7353 |
+
},
|
| 7354 |
+
{
|
| 7355 |
+
"epoch": 0.38693985849056606,
|
| 7356 |
+
"grad_norm": 2.3125,
|
| 7357 |
+
"learning_rate": 1.8209046238731998e-05,
|
| 7358 |
+
"loss": 2.1355,
|
| 7359 |
+
"step": 21000
|
| 7360 |
+
},
|
| 7361 |
+
{
|
| 7362 |
+
"epoch": 0.3873083726415094,
|
| 7363 |
+
"grad_norm": 2.34375,
|
| 7364 |
+
"learning_rate": 1.820573919010187e-05,
|
| 7365 |
+
"loss": 2.0884,
|
| 7366 |
+
"step": 21020
|
| 7367 |
+
},
|
| 7368 |
+
{
|
| 7369 |
+
"epoch": 0.3876768867924528,
|
| 7370 |
+
"grad_norm": 2.46875,
|
| 7371 |
+
"learning_rate": 1.8202429391893826e-05,
|
| 7372 |
+
"loss": 2.1283,
|
| 7373 |
+
"step": 21040
|
| 7374 |
+
},
|
| 7375 |
+
{
|
| 7376 |
+
"epoch": 0.38804540094339623,
|
| 7377 |
+
"grad_norm": 2.953125,
|
| 7378 |
+
"learning_rate": 1.8199116845216923e-05,
|
| 7379 |
+
"loss": 2.1312,
|
| 7380 |
+
"step": 21060
|
| 7381 |
+
},
|
| 7382 |
+
{
|
| 7383 |
+
"epoch": 0.38841391509433965,
|
| 7384 |
+
"grad_norm": 2.90625,
|
| 7385 |
+
"learning_rate": 1.819580155118112e-05,
|
| 7386 |
+
"loss": 2.1085,
|
| 7387 |
+
"step": 21080
|
| 7388 |
+
},
|
| 7389 |
+
{
|
| 7390 |
+
"epoch": 0.388782429245283,
|
| 7391 |
+
"grad_norm": 2.5,
|
| 7392 |
+
"learning_rate": 1.819248351089731e-05,
|
| 7393 |
+
"loss": 2.1113,
|
| 7394 |
+
"step": 21100
|
| 7395 |
+
},
|
| 7396 |
+
{
|
| 7397 |
+
"epoch": 0.3891509433962264,
|
| 7398 |
+
"grad_norm": 2.171875,
|
| 7399 |
+
"learning_rate": 1.81891627254773e-05,
|
| 7400 |
+
"loss": 2.1113,
|
| 7401 |
+
"step": 21120
|
| 7402 |
+
},
|
| 7403 |
+
{
|
| 7404 |
+
"epoch": 0.3895194575471698,
|
| 7405 |
+
"grad_norm": 2.875,
|
| 7406 |
+
"learning_rate": 1.818583919603382e-05,
|
| 7407 |
+
"loss": 2.13,
|
| 7408 |
+
"step": 21140
|
| 7409 |
+
},
|
| 7410 |
+
{
|
| 7411 |
+
"epoch": 0.38988797169811323,
|
| 7412 |
+
"grad_norm": 2.1875,
|
| 7413 |
+
"learning_rate": 1.8182512923680515e-05,
|
| 7414 |
+
"loss": 2.1251,
|
| 7415 |
+
"step": 21160
|
| 7416 |
+
},
|
| 7417 |
+
{
|
| 7418 |
+
"epoch": 0.3902564858490566,
|
| 7419 |
+
"grad_norm": 2.375,
|
| 7420 |
+
"learning_rate": 1.8179183909531955e-05,
|
| 7421 |
+
"loss": 2.1498,
|
| 7422 |
+
"step": 21180
|
| 7423 |
+
},
|
| 7424 |
+
{
|
| 7425 |
+
"epoch": 0.390625,
|
| 7426 |
+
"grad_norm": 2.265625,
|
| 7427 |
+
"learning_rate": 1.8175852154703624e-05,
|
| 7428 |
+
"loss": 2.105,
|
| 7429 |
+
"step": 21200
|
| 7430 |
+
},
|
| 7431 |
+
{
|
| 7432 |
+
"epoch": 0.3909935141509434,
|
| 7433 |
+
"grad_norm": 2.71875,
|
| 7434 |
+
"learning_rate": 1.8172517660311926e-05,
|
| 7435 |
+
"loss": 2.0958,
|
| 7436 |
+
"step": 21220
|
| 7437 |
+
},
|
| 7438 |
+
{
|
| 7439 |
+
"epoch": 0.39136202830188677,
|
| 7440 |
+
"grad_norm": 2.671875,
|
| 7441 |
+
"learning_rate": 1.816918042747418e-05,
|
| 7442 |
+
"loss": 2.1024,
|
| 7443 |
+
"step": 21240
|
| 7444 |
+
},
|
| 7445 |
+
{
|
| 7446 |
+
"epoch": 0.3917305424528302,
|
| 7447 |
+
"grad_norm": 2.359375,
|
| 7448 |
+
"learning_rate": 1.816584045730863e-05,
|
| 7449 |
+
"loss": 2.1055,
|
| 7450 |
+
"step": 21260
|
| 7451 |
+
},
|
| 7452 |
+
{
|
| 7453 |
+
"epoch": 0.3920990566037736,
|
| 7454 |
+
"grad_norm": 2.5,
|
| 7455 |
+
"learning_rate": 1.816249775093443e-05,
|
| 7456 |
+
"loss": 2.1376,
|
| 7457 |
+
"step": 21280
|
| 7458 |
+
},
|
| 7459 |
+
{
|
| 7460 |
+
"epoch": 0.392467570754717,
|
| 7461 |
+
"grad_norm": 2.53125,
|
| 7462 |
+
"learning_rate": 1.8159152309471655e-05,
|
| 7463 |
+
"loss": 2.0971,
|
| 7464 |
+
"step": 21300
|
| 7465 |
+
},
|
| 7466 |
+
{
|
| 7467 |
+
"epoch": 0.39283608490566035,
|
| 7468 |
+
"grad_norm": 2.515625,
|
| 7469 |
+
"learning_rate": 1.8155804134041294e-05,
|
| 7470 |
+
"loss": 2.0963,
|
| 7471 |
+
"step": 21320
|
| 7472 |
+
},
|
| 7473 |
+
{
|
| 7474 |
+
"epoch": 0.39320459905660377,
|
| 7475 |
+
"grad_norm": 2.3125,
|
| 7476 |
+
"learning_rate": 1.8152453225765256e-05,
|
| 7477 |
+
"loss": 2.0779,
|
| 7478 |
+
"step": 21340
|
| 7479 |
+
},
|
| 7480 |
+
{
|
| 7481 |
+
"epoch": 0.3935731132075472,
|
| 7482 |
+
"grad_norm": 2.328125,
|
| 7483 |
+
"learning_rate": 1.8149099585766362e-05,
|
| 7484 |
+
"loss": 2.1209,
|
| 7485 |
+
"step": 21360
|
| 7486 |
+
},
|
| 7487 |
+
{
|
| 7488 |
+
"epoch": 0.3939416273584906,
|
| 7489 |
+
"grad_norm": 2.453125,
|
| 7490 |
+
"learning_rate": 1.8145743215168343e-05,
|
| 7491 |
+
"loss": 2.0993,
|
| 7492 |
+
"step": 21380
|
| 7493 |
+
},
|
| 7494 |
+
{
|
| 7495 |
+
"epoch": 0.39431014150943394,
|
| 7496 |
+
"grad_norm": 2.515625,
|
| 7497 |
+
"learning_rate": 1.8142384115095857e-05,
|
| 7498 |
+
"loss": 2.1092,
|
| 7499 |
+
"step": 21400
|
| 7500 |
+
},
|
| 7501 |
+
{
|
| 7502 |
+
"epoch": 0.39467865566037735,
|
| 7503 |
+
"grad_norm": 2.609375,
|
| 7504 |
+
"learning_rate": 1.8139022286674473e-05,
|
| 7505 |
+
"loss": 2.132,
|
| 7506 |
+
"step": 21420
|
| 7507 |
+
},
|
| 7508 |
+
{
|
| 7509 |
+
"epoch": 0.39504716981132076,
|
| 7510 |
+
"grad_norm": 2.5,
|
| 7511 |
+
"learning_rate": 1.8135657731030666e-05,
|
| 7512 |
+
"loss": 2.1542,
|
| 7513 |
+
"step": 21440
|
| 7514 |
+
},
|
| 7515 |
+
{
|
| 7516 |
+
"epoch": 0.3954156839622642,
|
| 7517 |
+
"grad_norm": 2.453125,
|
| 7518 |
+
"learning_rate": 1.8132290449291834e-05,
|
| 7519 |
+
"loss": 2.1324,
|
| 7520 |
+
"step": 21460
|
| 7521 |
+
},
|
| 7522 |
+
{
|
| 7523 |
+
"epoch": 0.39578419811320753,
|
| 7524 |
+
"grad_norm": 2.59375,
|
| 7525 |
+
"learning_rate": 1.8128920442586285e-05,
|
| 7526 |
+
"loss": 2.0921,
|
| 7527 |
+
"step": 21480
|
| 7528 |
+
},
|
| 7529 |
+
{
|
| 7530 |
+
"epoch": 0.39615271226415094,
|
| 7531 |
+
"grad_norm": 2.265625,
|
| 7532 |
+
"learning_rate": 1.812554771204324e-05,
|
| 7533 |
+
"loss": 2.1594,
|
| 7534 |
+
"step": 21500
|
| 7535 |
+
},
|
| 7536 |
+
{
|
| 7537 |
+
"epoch": 0.39652122641509435,
|
| 7538 |
+
"grad_norm": 2.484375,
|
| 7539 |
+
"learning_rate": 1.8122172258792835e-05,
|
| 7540 |
+
"loss": 2.1285,
|
| 7541 |
+
"step": 21520
|
| 7542 |
+
},
|
| 7543 |
+
{
|
| 7544 |
+
"epoch": 0.39688974056603776,
|
| 7545 |
+
"grad_norm": 3.265625,
|
| 7546 |
+
"learning_rate": 1.8118794083966112e-05,
|
| 7547 |
+
"loss": 2.1157,
|
| 7548 |
+
"step": 21540
|
| 7549 |
+
},
|
| 7550 |
+
{
|
| 7551 |
+
"epoch": 0.3972582547169811,
|
| 7552 |
+
"grad_norm": 2.578125,
|
| 7553 |
+
"learning_rate": 1.8115413188695032e-05,
|
| 7554 |
+
"loss": 2.121,
|
| 7555 |
+
"step": 21560
|
| 7556 |
+
},
|
| 7557 |
+
{
|
| 7558 |
+
"epoch": 0.39762676886792453,
|
| 7559 |
+
"grad_norm": 2.34375,
|
| 7560 |
+
"learning_rate": 1.8112029574112465e-05,
|
| 7561 |
+
"loss": 2.0751,
|
| 7562 |
+
"step": 21580
|
| 7563 |
+
},
|
| 7564 |
+
{
|
| 7565 |
+
"epoch": 0.39799528301886794,
|
| 7566 |
+
"grad_norm": 2.390625,
|
| 7567 |
+
"learning_rate": 1.810864324135219e-05,
|
| 7568 |
+
"loss": 2.1296,
|
| 7569 |
+
"step": 21600
|
| 7570 |
+
},
|
| 7571 |
+
{
|
| 7572 |
+
"epoch": 0.3983637971698113,
|
| 7573 |
+
"grad_norm": 2.390625,
|
| 7574 |
+
"learning_rate": 1.8105254191548907e-05,
|
| 7575 |
+
"loss": 2.1194,
|
| 7576 |
+
"step": 21620
|
| 7577 |
+
},
|
| 7578 |
+
{
|
| 7579 |
+
"epoch": 0.3987323113207547,
|
| 7580 |
+
"grad_norm": 2.46875,
|
| 7581 |
+
"learning_rate": 1.810186242583821e-05,
|
| 7582 |
+
"loss": 2.1254,
|
| 7583 |
+
"step": 21640
|
| 7584 |
+
},
|
| 7585 |
+
{
|
| 7586 |
+
"epoch": 0.3991008254716981,
|
| 7587 |
+
"grad_norm": 2.375,
|
| 7588 |
+
"learning_rate": 1.8098467945356612e-05,
|
| 7589 |
+
"loss": 2.1173,
|
| 7590 |
+
"step": 21660
|
| 7591 |
+
},
|
| 7592 |
+
{
|
| 7593 |
+
"epoch": 0.39946933962264153,
|
| 7594 |
+
"grad_norm": 2.34375,
|
| 7595 |
+
"learning_rate": 1.8095070751241537e-05,
|
| 7596 |
+
"loss": 2.1531,
|
| 7597 |
+
"step": 21680
|
| 7598 |
+
},
|
| 7599 |
+
{
|
| 7600 |
+
"epoch": 0.3998378537735849,
|
| 7601 |
+
"grad_norm": 2.34375,
|
| 7602 |
+
"learning_rate": 1.8091670844631317e-05,
|
| 7603 |
+
"loss": 2.1302,
|
| 7604 |
+
"step": 21700
|
| 7605 |
+
},
|
| 7606 |
+
{
|
| 7607 |
+
"epoch": 0.4002063679245283,
|
| 7608 |
+
"grad_norm": 2.265625,
|
| 7609 |
+
"learning_rate": 1.808826822666519e-05,
|
| 7610 |
+
"loss": 2.0457,
|
| 7611 |
+
"step": 21720
|
| 7612 |
+
},
|
| 7613 |
+
{
|
| 7614 |
+
"epoch": 0.4005748820754717,
|
| 7615 |
+
"grad_norm": 2.296875,
|
| 7616 |
+
"learning_rate": 1.8084862898483304e-05,
|
| 7617 |
+
"loss": 2.1384,
|
| 7618 |
+
"step": 21740
|
| 7619 |
+
},
|
| 7620 |
+
{
|
| 7621 |
+
"epoch": 0.4009433962264151,
|
| 7622 |
+
"grad_norm": 2.375,
|
| 7623 |
+
"learning_rate": 1.8081454861226723e-05,
|
| 7624 |
+
"loss": 2.1201,
|
| 7625 |
+
"step": 21760
|
| 7626 |
+
},
|
| 7627 |
+
{
|
| 7628 |
+
"epoch": 0.40131191037735847,
|
| 7629 |
+
"grad_norm": 2.984375,
|
| 7630 |
+
"learning_rate": 1.80780441160374e-05,
|
| 7631 |
+
"loss": 2.0842,
|
| 7632 |
+
"step": 21780
|
| 7633 |
+
},
|
| 7634 |
+
{
|
| 7635 |
+
"epoch": 0.4016804245283019,
|
| 7636 |
+
"grad_norm": 2.3125,
|
| 7637 |
+
"learning_rate": 1.8074630664058218e-05,
|
| 7638 |
+
"loss": 2.1259,
|
| 7639 |
+
"step": 21800
|
| 7640 |
+
},
|
| 7641 |
+
{
|
| 7642 |
+
"epoch": 0.4020489386792453,
|
| 7643 |
+
"grad_norm": 2.671875,
|
| 7644 |
+
"learning_rate": 1.807121450643295e-05,
|
| 7645 |
+
"loss": 2.1188,
|
| 7646 |
+
"step": 21820
|
| 7647 |
+
},
|
| 7648 |
+
{
|
| 7649 |
+
"epoch": 0.4024174528301887,
|
| 7650 |
+
"grad_norm": 2.3125,
|
| 7651 |
+
"learning_rate": 1.8067795644306284e-05,
|
| 7652 |
+
"loss": 2.1493,
|
| 7653 |
+
"step": 21840
|
| 7654 |
+
},
|
| 7655 |
+
{
|
| 7656 |
+
"epoch": 0.40278596698113206,
|
| 7657 |
+
"grad_norm": 2.390625,
|
| 7658 |
+
"learning_rate": 1.8064374078823807e-05,
|
| 7659 |
+
"loss": 2.1433,
|
| 7660 |
+
"step": 21860
|
| 7661 |
+
},
|
| 7662 |
+
{
|
| 7663 |
+
"epoch": 0.40315448113207547,
|
| 7664 |
+
"grad_norm": 2.390625,
|
| 7665 |
+
"learning_rate": 1.8060949811132024e-05,
|
| 7666 |
+
"loss": 2.1383,
|
| 7667 |
+
"step": 21880
|
| 7668 |
+
},
|
| 7669 |
+
{
|
| 7670 |
+
"epoch": 0.4035229952830189,
|
| 7671 |
+
"grad_norm": 2.21875,
|
| 7672 |
+
"learning_rate": 1.8057522842378333e-05,
|
| 7673 |
+
"loss": 2.1132,
|
| 7674 |
+
"step": 21900
|
| 7675 |
+
},
|
| 7676 |
+
{
|
| 7677 |
+
"epoch": 0.40389150943396224,
|
| 7678 |
+
"grad_norm": 2.328125,
|
| 7679 |
+
"learning_rate": 1.8054093173711046e-05,
|
| 7680 |
+
"loss": 2.0957,
|
| 7681 |
+
"step": 21920
|
| 7682 |
+
},
|
| 7683 |
+
{
|
| 7684 |
+
"epoch": 0.40426002358490565,
|
| 7685 |
+
"grad_norm": 2.65625,
|
| 7686 |
+
"learning_rate": 1.8050660806279374e-05,
|
| 7687 |
+
"loss": 2.135,
|
| 7688 |
+
"step": 21940
|
| 7689 |
+
},
|
| 7690 |
+
{
|
| 7691 |
+
"epoch": 0.40462853773584906,
|
| 7692 |
+
"grad_norm": 2.6875,
|
| 7693 |
+
"learning_rate": 1.804722574123343e-05,
|
| 7694 |
+
"loss": 2.1502,
|
| 7695 |
+
"step": 21960
|
| 7696 |
+
},
|
| 7697 |
+
{
|
| 7698 |
+
"epoch": 0.40499705188679247,
|
| 7699 |
+
"grad_norm": 2.484375,
|
| 7700 |
+
"learning_rate": 1.8043787979724242e-05,
|
| 7701 |
+
"loss": 2.0767,
|
| 7702 |
+
"step": 21980
|
| 7703 |
+
},
|
| 7704 |
+
{
|
| 7705 |
+
"epoch": 0.4053655660377358,
|
| 7706 |
+
"grad_norm": 2.78125,
|
| 7707 |
+
"learning_rate": 1.804034752290373e-05,
|
| 7708 |
+
"loss": 2.115,
|
| 7709 |
+
"step": 22000
|
| 7710 |
+
},
|
| 7711 |
+
{
|
| 7712 |
+
"epoch": 0.40573408018867924,
|
| 7713 |
+
"grad_norm": 2.421875,
|
| 7714 |
+
"learning_rate": 1.8036904371924724e-05,
|
| 7715 |
+
"loss": 2.1322,
|
| 7716 |
+
"step": 22020
|
| 7717 |
+
},
|
| 7718 |
+
{
|
| 7719 |
+
"epoch": 0.40610259433962265,
|
| 7720 |
+
"grad_norm": 2.71875,
|
| 7721 |
+
"learning_rate": 1.8033458527940956e-05,
|
| 7722 |
+
"loss": 2.1141,
|
| 7723 |
+
"step": 22040
|
| 7724 |
+
},
|
| 7725 |
+
{
|
| 7726 |
+
"epoch": 0.40647110849056606,
|
| 7727 |
+
"grad_norm": 2.515625,
|
| 7728 |
+
"learning_rate": 1.8030009992107055e-05,
|
| 7729 |
+
"loss": 2.0886,
|
| 7730 |
+
"step": 22060
|
| 7731 |
+
},
|
| 7732 |
+
{
|
| 7733 |
+
"epoch": 0.4068396226415094,
|
| 7734 |
+
"grad_norm": 2.609375,
|
| 7735 |
+
"learning_rate": 1.802655876557856e-05,
|
| 7736 |
+
"loss": 2.1382,
|
| 7737 |
+
"step": 22080
|
| 7738 |
+
},
|
| 7739 |
+
{
|
| 7740 |
+
"epoch": 0.4072081367924528,
|
| 7741 |
+
"grad_norm": 2.484375,
|
| 7742 |
+
"learning_rate": 1.8023104849511902e-05,
|
| 7743 |
+
"loss": 2.1051,
|
| 7744 |
+
"step": 22100
|
| 7745 |
+
},
|
| 7746 |
+
{
|
| 7747 |
+
"epoch": 0.40757665094339623,
|
| 7748 |
+
"grad_norm": 2.5,
|
| 7749 |
+
"learning_rate": 1.8019648245064423e-05,
|
| 7750 |
+
"loss": 2.0943,
|
| 7751 |
+
"step": 22120
|
| 7752 |
+
},
|
| 7753 |
+
{
|
| 7754 |
+
"epoch": 0.40794516509433965,
|
| 7755 |
+
"grad_norm": 2.484375,
|
| 7756 |
+
"learning_rate": 1.801618895339436e-05,
|
| 7757 |
+
"loss": 2.1503,
|
| 7758 |
+
"step": 22140
|
| 7759 |
+
},
|
| 7760 |
+
{
|
| 7761 |
+
"epoch": 0.408313679245283,
|
| 7762 |
+
"grad_norm": 2.703125,
|
| 7763 |
+
"learning_rate": 1.8012726975660856e-05,
|
| 7764 |
+
"loss": 2.1316,
|
| 7765 |
+
"step": 22160
|
| 7766 |
+
},
|
| 7767 |
+
{
|
| 7768 |
+
"epoch": 0.4086821933962264,
|
| 7769 |
+
"grad_norm": 2.484375,
|
| 7770 |
+
"learning_rate": 1.800926231302394e-05,
|
| 7771 |
+
"loss": 2.1229,
|
| 7772 |
+
"step": 22180
|
| 7773 |
+
},
|
| 7774 |
+
{
|
| 7775 |
+
"epoch": 0.4090507075471698,
|
| 7776 |
+
"grad_norm": 2.5,
|
| 7777 |
+
"learning_rate": 1.800579496664456e-05,
|
| 7778 |
+
"loss": 2.1029,
|
| 7779 |
+
"step": 22200
|
| 7780 |
+
},
|
| 7781 |
+
{
|
| 7782 |
+
"epoch": 0.40941922169811323,
|
| 7783 |
+
"grad_norm": 2.625,
|
| 7784 |
+
"learning_rate": 1.8002324937684552e-05,
|
| 7785 |
+
"loss": 2.1008,
|
| 7786 |
+
"step": 22220
|
| 7787 |
+
},
|
| 7788 |
+
{
|
| 7789 |
+
"epoch": 0.4097877358490566,
|
| 7790 |
+
"grad_norm": 2.703125,
|
| 7791 |
+
"learning_rate": 1.7998852227306655e-05,
|
| 7792 |
+
"loss": 2.1325,
|
| 7793 |
+
"step": 22240
|
| 7794 |
+
},
|
| 7795 |
+
{
|
| 7796 |
+
"epoch": 0.41015625,
|
| 7797 |
+
"grad_norm": 2.4375,
|
| 7798 |
+
"learning_rate": 1.7995376836674495e-05,
|
| 7799 |
+
"loss": 2.1312,
|
| 7800 |
+
"step": 22260
|
| 7801 |
+
},
|
| 7802 |
+
{
|
| 7803 |
+
"epoch": 0.4105247641509434,
|
| 7804 |
+
"grad_norm": 2.15625,
|
| 7805 |
+
"learning_rate": 1.7991898766952614e-05,
|
| 7806 |
+
"loss": 2.1224,
|
| 7807 |
+
"step": 22280
|
| 7808 |
+
},
|
| 7809 |
+
{
|
| 7810 |
+
"epoch": 0.41089327830188677,
|
| 7811 |
+
"grad_norm": 2.34375,
|
| 7812 |
+
"learning_rate": 1.798841801930644e-05,
|
| 7813 |
+
"loss": 2.1377,
|
| 7814 |
+
"step": 22300
|
| 7815 |
+
},
|
| 7816 |
+
{
|
| 7817 |
+
"epoch": 0.4112617924528302,
|
| 7818 |
+
"grad_norm": 2.390625,
|
| 7819 |
+
"learning_rate": 1.7984934594902302e-05,
|
| 7820 |
+
"loss": 2.1548,
|
| 7821 |
+
"step": 22320
|
| 7822 |
+
},
|
| 7823 |
+
{
|
| 7824 |
+
"epoch": 0.4116303066037736,
|
| 7825 |
+
"grad_norm": 2.78125,
|
| 7826 |
+
"learning_rate": 1.7981448494907424e-05,
|
| 7827 |
+
"loss": 2.1323,
|
| 7828 |
+
"step": 22340
|
| 7829 |
+
},
|
| 7830 |
+
{
|
| 7831 |
+
"epoch": 0.411998820754717,
|
| 7832 |
+
"grad_norm": 2.890625,
|
| 7833 |
+
"learning_rate": 1.7977959720489932e-05,
|
| 7834 |
+
"loss": 2.1364,
|
| 7835 |
+
"step": 22360
|
| 7836 |
+
},
|
| 7837 |
+
{
|
| 7838 |
+
"epoch": 0.41236733490566035,
|
| 7839 |
+
"grad_norm": 2.703125,
|
| 7840 |
+
"learning_rate": 1.7974468272818844e-05,
|
| 7841 |
+
"loss": 2.1422,
|
| 7842 |
+
"step": 22380
|
| 7843 |
+
},
|
| 7844 |
+
{
|
| 7845 |
+
"epoch": 0.41273584905660377,
|
| 7846 |
+
"grad_norm": 2.46875,
|
| 7847 |
+
"learning_rate": 1.7970974153064068e-05,
|
| 7848 |
+
"loss": 2.1491,
|
| 7849 |
+
"step": 22400
|
| 7850 |
+
},
|
| 7851 |
+
{
|
| 7852 |
+
"epoch": 0.4131043632075472,
|
| 7853 |
+
"grad_norm": 2.328125,
|
| 7854 |
+
"learning_rate": 1.7967477362396413e-05,
|
| 7855 |
+
"loss": 2.1264,
|
| 7856 |
+
"step": 22420
|
| 7857 |
+
},
|
| 7858 |
+
{
|
| 7859 |
+
"epoch": 0.4134728773584906,
|
| 7860 |
+
"grad_norm": 2.765625,
|
| 7861 |
+
"learning_rate": 1.796397790198759e-05,
|
| 7862 |
+
"loss": 2.1195,
|
| 7863 |
+
"step": 22440
|
| 7864 |
+
},
|
| 7865 |
+
{
|
| 7866 |
+
"epoch": 0.41384139150943394,
|
| 7867 |
+
"grad_norm": 2.359375,
|
| 7868 |
+
"learning_rate": 1.7960475773010193e-05,
|
| 7869 |
+
"loss": 2.112,
|
| 7870 |
+
"step": 22460
|
| 7871 |
+
},
|
| 7872 |
+
{
|
| 7873 |
+
"epoch": 0.41420990566037735,
|
| 7874 |
+
"grad_norm": 2.40625,
|
| 7875 |
+
"learning_rate": 1.7956970976637715e-05,
|
| 7876 |
+
"loss": 2.1132,
|
| 7877 |
+
"step": 22480
|
| 7878 |
+
},
|
| 7879 |
+
{
|
| 7880 |
+
"epoch": 0.41457841981132076,
|
| 7881 |
+
"grad_norm": 2.4375,
|
| 7882 |
+
"learning_rate": 1.7953463514044545e-05,
|
| 7883 |
+
"loss": 2.12,
|
| 7884 |
+
"step": 22500
|
| 7885 |
+
},
|
| 7886 |
+
{
|
| 7887 |
+
"epoch": 0.4149469339622642,
|
| 7888 |
+
"grad_norm": 2.4375,
|
| 7889 |
+
"learning_rate": 1.7949953386405962e-05,
|
| 7890 |
+
"loss": 2.116,
|
| 7891 |
+
"step": 22520
|
| 7892 |
+
},
|
| 7893 |
+
{
|
| 7894 |
+
"epoch": 0.41531544811320753,
|
| 7895 |
+
"grad_norm": 2.546875,
|
| 7896 |
+
"learning_rate": 1.794644059489814e-05,
|
| 7897 |
+
"loss": 2.0904,
|
| 7898 |
+
"step": 22540
|
| 7899 |
+
},
|
| 7900 |
+
{
|
| 7901 |
+
"epoch": 0.41568396226415094,
|
| 7902 |
+
"grad_norm": 2.421875,
|
| 7903 |
+
"learning_rate": 1.794292514069814e-05,
|
| 7904 |
+
"loss": 2.1462,
|
| 7905 |
+
"step": 22560
|
| 7906 |
+
},
|
| 7907 |
+
{
|
| 7908 |
+
"epoch": 0.41605247641509435,
|
| 7909 |
+
"grad_norm": 2.609375,
|
| 7910 |
+
"learning_rate": 1.7939407024983927e-05,
|
| 7911 |
+
"loss": 2.0715,
|
| 7912 |
+
"step": 22580
|
| 7913 |
+
},
|
| 7914 |
+
{
|
| 7915 |
+
"epoch": 0.41642099056603776,
|
| 7916 |
+
"grad_norm": 2.421875,
|
| 7917 |
+
"learning_rate": 1.793588624893434e-05,
|
| 7918 |
+
"loss": 2.1402,
|
| 7919 |
+
"step": 22600
|
| 7920 |
+
},
|
| 7921 |
+
{
|
| 7922 |
+
"epoch": 0.4167895047169811,
|
| 7923 |
+
"grad_norm": 2.671875,
|
| 7924 |
+
"learning_rate": 1.7932362813729134e-05,
|
| 7925 |
+
"loss": 2.098,
|
| 7926 |
+
"step": 22620
|
| 7927 |
+
},
|
| 7928 |
+
{
|
| 7929 |
+
"epoch": 0.41715801886792453,
|
| 7930 |
+
"grad_norm": 2.546875,
|
| 7931 |
+
"learning_rate": 1.792883672054893e-05,
|
| 7932 |
+
"loss": 2.1199,
|
| 7933 |
+
"step": 22640
|
| 7934 |
+
},
|
| 7935 |
+
{
|
| 7936 |
+
"epoch": 0.41752653301886794,
|
| 7937 |
+
"grad_norm": 2.1875,
|
| 7938 |
+
"learning_rate": 1.7925307970575262e-05,
|
| 7939 |
+
"loss": 2.0988,
|
| 7940 |
+
"step": 22660
|
| 7941 |
+
},
|
| 7942 |
+
{
|
| 7943 |
+
"epoch": 0.4178950471698113,
|
| 7944 |
+
"grad_norm": 2.484375,
|
| 7945 |
+
"learning_rate": 1.7921776564990528e-05,
|
| 7946 |
+
"loss": 2.1301,
|
| 7947 |
+
"step": 22680
|
| 7948 |
+
},
|
| 7949 |
+
{
|
| 7950 |
+
"epoch": 0.4182635613207547,
|
| 7951 |
+
"grad_norm": 2.640625,
|
| 7952 |
+
"learning_rate": 1.7918242504978047e-05,
|
| 7953 |
+
"loss": 2.1171,
|
| 7954 |
+
"step": 22700
|
| 7955 |
+
},
|
| 7956 |
+
{
|
| 7957 |
+
"epoch": 0.4186320754716981,
|
| 7958 |
+
"grad_norm": 2.484375,
|
| 7959 |
+
"learning_rate": 1.7914705791721994e-05,
|
| 7960 |
+
"loss": 2.117,
|
| 7961 |
+
"step": 22720
|
| 7962 |
+
},
|
| 7963 |
+
{
|
| 7964 |
+
"epoch": 0.41900058962264153,
|
| 7965 |
+
"grad_norm": 3.0625,
|
| 7966 |
+
"learning_rate": 1.7911166426407467e-05,
|
| 7967 |
+
"loss": 2.1198,
|
| 7968 |
+
"step": 22740
|
| 7969 |
+
},
|
| 7970 |
+
{
|
| 7971 |
+
"epoch": 0.4193691037735849,
|
| 7972 |
+
"grad_norm": 2.46875,
|
| 7973 |
+
"learning_rate": 1.7907624410220425e-05,
|
| 7974 |
+
"loss": 2.1664,
|
| 7975 |
+
"step": 22760
|
| 7976 |
+
},
|
| 7977 |
+
{
|
| 7978 |
+
"epoch": 0.4197376179245283,
|
| 7979 |
+
"grad_norm": 2.453125,
|
| 7980 |
+
"learning_rate": 1.7904079744347732e-05,
|
| 7981 |
+
"loss": 2.1384,
|
| 7982 |
+
"step": 22780
|
| 7983 |
+
},
|
| 7984 |
+
{
|
| 7985 |
+
"epoch": 0.4201061320754717,
|
| 7986 |
+
"grad_norm": 2.546875,
|
| 7987 |
+
"learning_rate": 1.790053242997713e-05,
|
| 7988 |
+
"loss": 2.0966,
|
| 7989 |
+
"step": 22800
|
| 7990 |
+
},
|
| 7991 |
+
{
|
| 7992 |
+
"epoch": 0.4204746462264151,
|
| 7993 |
+
"grad_norm": 2.46875,
|
| 7994 |
+
"learning_rate": 1.7896982468297255e-05,
|
| 7995 |
+
"loss": 2.1267,
|
| 7996 |
+
"step": 22820
|
| 7997 |
+
},
|
| 7998 |
+
{
|
| 7999 |
+
"epoch": 0.42084316037735847,
|
| 8000 |
+
"grad_norm": 2.375,
|
| 8001 |
+
"learning_rate": 1.789342986049763e-05,
|
| 8002 |
+
"loss": 2.1452,
|
| 8003 |
+
"step": 22840
|
| 8004 |
+
},
|
| 8005 |
+
{
|
| 8006 |
+
"epoch": 0.4212116745283019,
|
| 8007 |
+
"grad_norm": 2.34375,
|
| 8008 |
+
"learning_rate": 1.7889874607768656e-05,
|
| 8009 |
+
"loss": 2.0968,
|
| 8010 |
+
"step": 22860
|
| 8011 |
+
},
|
| 8012 |
+
{
|
| 8013 |
+
"epoch": 0.4215801886792453,
|
| 8014 |
+
"grad_norm": 2.515625,
|
| 8015 |
+
"learning_rate": 1.7886316711301632e-05,
|
| 8016 |
+
"loss": 2.1147,
|
| 8017 |
+
"step": 22880
|
| 8018 |
+
},
|
| 8019 |
+
{
|
| 8020 |
+
"epoch": 0.4219487028301887,
|
| 8021 |
+
"grad_norm": 2.71875,
|
| 8022 |
+
"learning_rate": 1.7882756172288735e-05,
|
| 8023 |
+
"loss": 2.1206,
|
| 8024 |
+
"step": 22900
|
| 8025 |
+
},
|
| 8026 |
+
{
|
| 8027 |
+
"epoch": 0.42231721698113206,
|
| 8028 |
+
"grad_norm": 2.8125,
|
| 8029 |
+
"learning_rate": 1.787919299192303e-05,
|
| 8030 |
+
"loss": 2.1124,
|
| 8031 |
+
"step": 22920
|
| 8032 |
+
},
|
| 8033 |
+
{
|
| 8034 |
+
"epoch": 0.42268573113207547,
|
| 8035 |
+
"grad_norm": 2.609375,
|
| 8036 |
+
"learning_rate": 1.787562717139847e-05,
|
| 8037 |
+
"loss": 2.1247,
|
| 8038 |
+
"step": 22940
|
| 8039 |
+
},
|
| 8040 |
+
{
|
| 8041 |
+
"epoch": 0.4230542452830189,
|
| 8042 |
+
"grad_norm": 2.5625,
|
| 8043 |
+
"learning_rate": 1.7872058711909884e-05,
|
| 8044 |
+
"loss": 2.1297,
|
| 8045 |
+
"step": 22960
|
| 8046 |
+
},
|
| 8047 |
+
{
|
| 8048 |
+
"epoch": 0.42342275943396224,
|
| 8049 |
+
"grad_norm": 2.5,
|
| 8050 |
+
"learning_rate": 1.7868487614653e-05,
|
| 8051 |
+
"loss": 2.1258,
|
| 8052 |
+
"step": 22980
|
| 8053 |
+
},
|
| 8054 |
+
{
|
| 8055 |
+
"epoch": 0.42379127358490565,
|
| 8056 |
+
"grad_norm": 2.8125,
|
| 8057 |
+
"learning_rate": 1.786491388082441e-05,
|
| 8058 |
+
"loss": 2.12,
|
| 8059 |
+
"step": 23000
|
| 8060 |
+
},
|
| 8061 |
+
{
|
| 8062 |
+
"epoch": 0.42415978773584906,
|
| 8063 |
+
"grad_norm": 2.25,
|
| 8064 |
+
"learning_rate": 1.7861337511621604e-05,
|
| 8065 |
+
"loss": 2.1035,
|
| 8066 |
+
"step": 23020
|
| 8067 |
+
},
|
| 8068 |
+
{
|
| 8069 |
+
"epoch": 0.42452830188679247,
|
| 8070 |
+
"grad_norm": 2.40625,
|
| 8071 |
+
"learning_rate": 1.7857758508242956e-05,
|
| 8072 |
+
"loss": 2.0851,
|
| 8073 |
+
"step": 23040
|
| 8074 |
+
},
|
| 8075 |
+
{
|
| 8076 |
+
"epoch": 0.4248968160377358,
|
| 8077 |
+
"grad_norm": 3.515625,
|
| 8078 |
+
"learning_rate": 1.7854176871887713e-05,
|
| 8079 |
+
"loss": 2.1357,
|
| 8080 |
+
"step": 23060
|
| 8081 |
+
},
|
| 8082 |
+
{
|
| 8083 |
+
"epoch": 0.42526533018867924,
|
| 8084 |
+
"grad_norm": 2.921875,
|
| 8085 |
+
"learning_rate": 1.7850592603756008e-05,
|
| 8086 |
+
"loss": 2.107,
|
| 8087 |
+
"step": 23080
|
| 8088 |
+
},
|
| 8089 |
+
{
|
| 8090 |
+
"epoch": 0.42563384433962265,
|
| 8091 |
+
"grad_norm": 2.71875,
|
| 8092 |
+
"learning_rate": 1.784700570504886e-05,
|
| 8093 |
+
"loss": 2.1241,
|
| 8094 |
+
"step": 23100
|
| 8095 |
+
},
|
| 8096 |
+
{
|
| 8097 |
+
"epoch": 0.42600235849056606,
|
| 8098 |
+
"grad_norm": 2.640625,
|
| 8099 |
+
"learning_rate": 1.7843416176968167e-05,
|
| 8100 |
+
"loss": 2.1463,
|
| 8101 |
+
"step": 23120
|
| 8102 |
+
},
|
| 8103 |
+
{
|
| 8104 |
+
"epoch": 0.4263708726415094,
|
| 8105 |
+
"grad_norm": 2.328125,
|
| 8106 |
+
"learning_rate": 1.783982402071671e-05,
|
| 8107 |
+
"loss": 2.1319,
|
| 8108 |
+
"step": 23140
|
| 8109 |
+
},
|
| 8110 |
+
{
|
| 8111 |
+
"epoch": 0.4267393867924528,
|
| 8112 |
+
"grad_norm": 2.65625,
|
| 8113 |
+
"learning_rate": 1.7836229237498138e-05,
|
| 8114 |
+
"loss": 2.1138,
|
| 8115 |
+
"step": 23160
|
| 8116 |
+
},
|
| 8117 |
+
{
|
| 8118 |
+
"epoch": 0.42710790094339623,
|
| 8119 |
+
"grad_norm": 2.515625,
|
| 8120 |
+
"learning_rate": 1.7832631828517004e-05,
|
| 8121 |
+
"loss": 2.1099,
|
| 8122 |
+
"step": 23180
|
| 8123 |
+
},
|
| 8124 |
+
{
|
| 8125 |
+
"epoch": 0.42747641509433965,
|
| 8126 |
+
"grad_norm": 2.71875,
|
| 8127 |
+
"learning_rate": 1.7829031794978717e-05,
|
| 8128 |
+
"loss": 2.1536,
|
| 8129 |
+
"step": 23200
|
| 8130 |
+
},
|
| 8131 |
+
{
|
| 8132 |
+
"epoch": 0.427844929245283,
|
| 8133 |
+
"grad_norm": 2.6875,
|
| 8134 |
+
"learning_rate": 1.782542913808958e-05,
|
| 8135 |
+
"loss": 2.0897,
|
| 8136 |
+
"step": 23220
|
| 8137 |
+
},
|
| 8138 |
+
{
|
| 8139 |
+
"epoch": 0.4282134433962264,
|
| 8140 |
+
"grad_norm": 2.203125,
|
| 8141 |
+
"learning_rate": 1.7821823859056772e-05,
|
| 8142 |
+
"loss": 2.1334,
|
| 8143 |
+
"step": 23240
|
| 8144 |
+
},
|
| 8145 |
+
{
|
| 8146 |
+
"epoch": 0.4285819575471698,
|
| 8147 |
+
"grad_norm": 2.703125,
|
| 8148 |
+
"learning_rate": 1.7818215959088345e-05,
|
| 8149 |
+
"loss": 2.1019,
|
| 8150 |
+
"step": 23260
|
| 8151 |
+
},
|
| 8152 |
+
{
|
| 8153 |
+
"epoch": 0.42895047169811323,
|
| 8154 |
+
"grad_norm": 2.234375,
|
| 8155 |
+
"learning_rate": 1.7814605439393233e-05,
|
| 8156 |
+
"loss": 2.1333,
|
| 8157 |
+
"step": 23280
|
| 8158 |
+
},
|
| 8159 |
+
{
|
| 8160 |
+
"epoch": 0.4293189858490566,
|
| 8161 |
+
"grad_norm": 2.828125,
|
| 8162 |
+
"learning_rate": 1.7810992301181254e-05,
|
| 8163 |
+
"loss": 2.1413,
|
| 8164 |
+
"step": 23300
|
| 8165 |
+
},
|
| 8166 |
+
{
|
| 8167 |
+
"epoch": 0.4296875,
|
| 8168 |
+
"grad_norm": 2.828125,
|
| 8169 |
+
"learning_rate": 1.7807376545663096e-05,
|
| 8170 |
+
"loss": 2.103,
|
| 8171 |
+
"step": 23320
|
| 8172 |
+
},
|
| 8173 |
+
{
|
| 8174 |
+
"epoch": 0.4300560141509434,
|
| 8175 |
+
"grad_norm": 2.21875,
|
| 8176 |
+
"learning_rate": 1.7803758174050325e-05,
|
| 8177 |
+
"loss": 2.1395,
|
| 8178 |
+
"step": 23340
|
| 8179 |
+
},
|
| 8180 |
+
{
|
| 8181 |
+
"epoch": 0.43042452830188677,
|
| 8182 |
+
"grad_norm": 2.359375,
|
| 8183 |
+
"learning_rate": 1.7800137187555385e-05,
|
| 8184 |
+
"loss": 2.0909,
|
| 8185 |
+
"step": 23360
|
| 8186 |
+
},
|
| 8187 |
+
{
|
| 8188 |
+
"epoch": 0.4307930424528302,
|
| 8189 |
+
"grad_norm": 2.28125,
|
| 8190 |
+
"learning_rate": 1.7796513587391593e-05,
|
| 8191 |
+
"loss": 2.1422,
|
| 8192 |
+
"step": 23380
|
| 8193 |
+
},
|
| 8194 |
+
{
|
| 8195 |
+
"epoch": 0.4311615566037736,
|
| 8196 |
+
"grad_norm": 2.203125,
|
| 8197 |
+
"learning_rate": 1.7792887374773148e-05,
|
| 8198 |
+
"loss": 2.1051,
|
| 8199 |
+
"step": 23400
|
| 8200 |
+
},
|
| 8201 |
+
{
|
| 8202 |
+
"epoch": 0.431530070754717,
|
| 8203 |
+
"grad_norm": 2.265625,
|
| 8204 |
+
"learning_rate": 1.778925855091512e-05,
|
| 8205 |
+
"loss": 2.1094,
|
| 8206 |
+
"step": 23420
|
| 8207 |
+
},
|
| 8208 |
+
{
|
| 8209 |
+
"epoch": 0.43189858490566035,
|
| 8210 |
+
"grad_norm": 2.53125,
|
| 8211 |
+
"learning_rate": 1.7785627117033453e-05,
|
| 8212 |
+
"loss": 2.1083,
|
| 8213 |
+
"step": 23440
|
| 8214 |
+
},
|
| 8215 |
+
{
|
| 8216 |
+
"epoch": 0.43226709905660377,
|
| 8217 |
+
"grad_norm": 2.453125,
|
| 8218 |
+
"learning_rate": 1.7781993074344967e-05,
|
| 8219 |
+
"loss": 2.1161,
|
| 8220 |
+
"step": 23460
|
| 8221 |
+
},
|
| 8222 |
+
{
|
| 8223 |
+
"epoch": 0.4326356132075472,
|
| 8224 |
+
"grad_norm": 2.359375,
|
| 8225 |
+
"learning_rate": 1.777835642406736e-05,
|
| 8226 |
+
"loss": 2.1066,
|
| 8227 |
+
"step": 23480
|
| 8228 |
+
},
|
| 8229 |
+
{
|
| 8230 |
+
"epoch": 0.4330041273584906,
|
| 8231 |
+
"grad_norm": 2.34375,
|
| 8232 |
+
"learning_rate": 1.7774717167419197e-05,
|
| 8233 |
+
"loss": 2.1006,
|
| 8234 |
+
"step": 23500
|
| 8235 |
+
},
|
| 8236 |
+
{
|
| 8237 |
+
"epoch": 0.43337264150943394,
|
| 8238 |
+
"grad_norm": 2.4375,
|
| 8239 |
+
"learning_rate": 1.777107530561992e-05,
|
| 8240 |
+
"loss": 2.1097,
|
| 8241 |
+
"step": 23520
|
| 8242 |
+
},
|
| 8243 |
+
{
|
| 8244 |
+
"epoch": 0.43374115566037735,
|
| 8245 |
+
"grad_norm": 2.546875,
|
| 8246 |
+
"learning_rate": 1.7767430839889848e-05,
|
| 8247 |
+
"loss": 2.1047,
|
| 8248 |
+
"step": 23540
|
| 8249 |
+
},
|
| 8250 |
+
{
|
| 8251 |
+
"epoch": 0.43410966981132076,
|
| 8252 |
+
"grad_norm": 2.71875,
|
| 8253 |
+
"learning_rate": 1.7763783771450162e-05,
|
| 8254 |
+
"loss": 2.0918,
|
| 8255 |
+
"step": 23560
|
| 8256 |
+
},
|
| 8257 |
+
{
|
| 8258 |
+
"epoch": 0.4344781839622642,
|
| 8259 |
+
"grad_norm": 2.265625,
|
| 8260 |
+
"learning_rate": 1.7760134101522925e-05,
|
| 8261 |
+
"loss": 2.1057,
|
| 8262 |
+
"step": 23580
|
| 8263 |
+
},
|
| 8264 |
+
{
|
| 8265 |
+
"epoch": 0.43484669811320753,
|
| 8266 |
+
"grad_norm": 2.625,
|
| 8267 |
+
"learning_rate": 1.7756481831331064e-05,
|
| 8268 |
+
"loss": 2.109,
|
| 8269 |
+
"step": 23600
|
| 8270 |
+
},
|
| 8271 |
+
{
|
| 8272 |
+
"epoch": 0.43521521226415094,
|
| 8273 |
+
"grad_norm": 2.421875,
|
| 8274 |
+
"learning_rate": 1.7752826962098387e-05,
|
| 8275 |
+
"loss": 2.0851,
|
| 8276 |
+
"step": 23620
|
| 8277 |
+
},
|
| 8278 |
+
{
|
| 8279 |
+
"epoch": 0.43558372641509435,
|
| 8280 |
+
"grad_norm": 2.453125,
|
| 8281 |
+
"learning_rate": 1.7749169495049563e-05,
|
| 8282 |
+
"loss": 2.0989,
|
| 8283 |
+
"step": 23640
|
| 8284 |
+
},
|
| 8285 |
+
{
|
| 8286 |
+
"epoch": 0.43595224056603776,
|
| 8287 |
+
"grad_norm": 2.578125,
|
| 8288 |
+
"learning_rate": 1.7745509431410136e-05,
|
| 8289 |
+
"loss": 2.1,
|
| 8290 |
+
"step": 23660
|
| 8291 |
+
},
|
| 8292 |
+
{
|
| 8293 |
+
"epoch": 0.4363207547169811,
|
| 8294 |
+
"grad_norm": 2.359375,
|
| 8295 |
+
"learning_rate": 1.7741846772406518e-05,
|
| 8296 |
+
"loss": 2.151,
|
| 8297 |
+
"step": 23680
|
| 8298 |
+
},
|
| 8299 |
+
{
|
| 8300 |
+
"epoch": 0.43668926886792453,
|
| 8301 |
+
"grad_norm": 2.359375,
|
| 8302 |
+
"learning_rate": 1.7738181519265993e-05,
|
| 8303 |
+
"loss": 2.1182,
|
| 8304 |
+
"step": 23700
|
| 8305 |
+
},
|
| 8306 |
+
{
|
| 8307 |
+
"epoch": 0.43705778301886794,
|
| 8308 |
+
"grad_norm": 2.5,
|
| 8309 |
+
"learning_rate": 1.7734513673216714e-05,
|
| 8310 |
+
"loss": 2.1046,
|
| 8311 |
+
"step": 23720
|
| 8312 |
+
},
|
| 8313 |
+
{
|
| 8314 |
+
"epoch": 0.4374262971698113,
|
| 8315 |
+
"grad_norm": 2.859375,
|
| 8316 |
+
"learning_rate": 1.7730843235487706e-05,
|
| 8317 |
+
"loss": 2.0966,
|
| 8318 |
+
"step": 23740
|
| 8319 |
+
},
|
| 8320 |
+
{
|
| 8321 |
+
"epoch": 0.4377948113207547,
|
| 8322 |
+
"grad_norm": 2.390625,
|
| 8323 |
+
"learning_rate": 1.7727170207308857e-05,
|
| 8324 |
+
"loss": 2.1083,
|
| 8325 |
+
"step": 23760
|
| 8326 |
+
},
|
| 8327 |
+
{
|
| 8328 |
+
"epoch": 0.4381633254716981,
|
| 8329 |
+
"grad_norm": 2.625,
|
| 8330 |
+
"learning_rate": 1.772349458991092e-05,
|
| 8331 |
+
"loss": 2.143,
|
| 8332 |
+
"step": 23780
|
| 8333 |
+
},
|
| 8334 |
+
{
|
| 8335 |
+
"epoch": 0.43853183962264153,
|
| 8336 |
+
"grad_norm": 2.4375,
|
| 8337 |
+
"learning_rate": 1.7719816384525524e-05,
|
| 8338 |
+
"loss": 2.1452,
|
| 8339 |
+
"step": 23800
|
| 8340 |
+
},
|
| 8341 |
+
{
|
| 8342 |
+
"epoch": 0.4389003537735849,
|
| 8343 |
+
"grad_norm": 2.390625,
|
| 8344 |
+
"learning_rate": 1.7716135592385164e-05,
|
| 8345 |
+
"loss": 2.1432,
|
| 8346 |
+
"step": 23820
|
| 8347 |
+
},
|
| 8348 |
+
{
|
| 8349 |
+
"epoch": 0.4392688679245283,
|
| 8350 |
+
"grad_norm": 2.578125,
|
| 8351 |
+
"learning_rate": 1.7712452214723197e-05,
|
| 8352 |
+
"loss": 2.1193,
|
| 8353 |
+
"step": 23840
|
| 8354 |
+
},
|
| 8355 |
+
{
|
| 8356 |
+
"epoch": 0.4396373820754717,
|
| 8357 |
+
"grad_norm": 2.6875,
|
| 8358 |
+
"learning_rate": 1.7708766252773845e-05,
|
| 8359 |
+
"loss": 2.1585,
|
| 8360 |
+
"step": 23860
|
| 8361 |
+
},
|
| 8362 |
+
{
|
| 8363 |
+
"epoch": 0.4400058962264151,
|
| 8364 |
+
"grad_norm": 2.4375,
|
| 8365 |
+
"learning_rate": 1.7705077707772204e-05,
|
| 8366 |
+
"loss": 2.1349,
|
| 8367 |
+
"step": 23880
|
| 8368 |
+
},
|
| 8369 |
+
{
|
| 8370 |
+
"epoch": 0.44037441037735847,
|
| 8371 |
+
"grad_norm": 2.328125,
|
| 8372 |
+
"learning_rate": 1.7701386580954228e-05,
|
| 8373 |
+
"loss": 2.1203,
|
| 8374 |
+
"step": 23900
|
| 8375 |
+
},
|
| 8376 |
+
{
|
| 8377 |
+
"epoch": 0.4407429245283019,
|
| 8378 |
+
"grad_norm": 2.609375,
|
| 8379 |
+
"learning_rate": 1.7697692873556747e-05,
|
| 8380 |
+
"loss": 2.099,
|
| 8381 |
+
"step": 23920
|
| 8382 |
+
},
|
| 8383 |
+
{
|
| 8384 |
+
"epoch": 0.4411114386792453,
|
| 8385 |
+
"grad_norm": 2.46875,
|
| 8386 |
+
"learning_rate": 1.7693996586817436e-05,
|
| 8387 |
+
"loss": 2.0907,
|
| 8388 |
+
"step": 23940
|
| 8389 |
+
},
|
| 8390 |
+
{
|
| 8391 |
+
"epoch": 0.4414799528301887,
|
| 8392 |
+
"grad_norm": 2.515625,
|
| 8393 |
+
"learning_rate": 1.7690297721974852e-05,
|
| 8394 |
+
"loss": 2.1221,
|
| 8395 |
+
"step": 23960
|
| 8396 |
+
},
|
| 8397 |
+
{
|
| 8398 |
+
"epoch": 0.44184846698113206,
|
| 8399 |
+
"grad_norm": 2.609375,
|
| 8400 |
+
"learning_rate": 1.7686596280268417e-05,
|
| 8401 |
+
"loss": 2.1181,
|
| 8402 |
+
"step": 23980
|
| 8403 |
+
},
|
| 8404 |
+
{
|
| 8405 |
+
"epoch": 0.44221698113207547,
|
| 8406 |
+
"grad_norm": 2.40625,
|
| 8407 |
+
"learning_rate": 1.76828922629384e-05,
|
| 8408 |
+
"loss": 2.0741,
|
| 8409 |
+
"step": 24000
|
| 8410 |
+
},
|
| 8411 |
+
{
|
| 8412 |
+
"epoch": 0.4425854952830189,
|
| 8413 |
+
"grad_norm": 2.9375,
|
| 8414 |
+
"learning_rate": 1.7679185671225944e-05,
|
| 8415 |
+
"loss": 2.1043,
|
| 8416 |
+
"step": 24020
|
| 8417 |
+
},
|
| 8418 |
+
{
|
| 8419 |
+
"epoch": 0.44295400943396224,
|
| 8420 |
+
"grad_norm": 2.703125,
|
| 8421 |
+
"learning_rate": 1.7675476506373055e-05,
|
| 8422 |
+
"loss": 2.1082,
|
| 8423 |
+
"step": 24040
|
| 8424 |
+
},
|
| 8425 |
+
{
|
| 8426 |
+
"epoch": 0.44332252358490565,
|
| 8427 |
+
"grad_norm": 2.734375,
|
| 8428 |
+
"learning_rate": 1.76717647696226e-05,
|
| 8429 |
+
"loss": 2.0884,
|
| 8430 |
+
"step": 24060
|
| 8431 |
+
},
|
| 8432 |
+
{
|
| 8433 |
+
"epoch": 0.44369103773584906,
|
| 8434 |
+
"grad_norm": 2.78125,
|
| 8435 |
+
"learning_rate": 1.7668050462218308e-05,
|
| 8436 |
+
"loss": 2.0943,
|
| 8437 |
+
"step": 24080
|
| 8438 |
+
},
|
| 8439 |
+
{
|
| 8440 |
+
"epoch": 0.44405955188679247,
|
| 8441 |
+
"grad_norm": 2.578125,
|
| 8442 |
+
"learning_rate": 1.7664333585404768e-05,
|
| 8443 |
+
"loss": 2.1311,
|
| 8444 |
+
"step": 24100
|
| 8445 |
+
},
|
| 8446 |
+
{
|
| 8447 |
+
"epoch": 0.4444280660377358,
|
| 8448 |
+
"grad_norm": 3.125,
|
| 8449 |
+
"learning_rate": 1.7660614140427427e-05,
|
| 8450 |
+
"loss": 2.1166,
|
| 8451 |
+
"step": 24120
|
| 8452 |
+
},
|
| 8453 |
+
{
|
| 8454 |
+
"epoch": 0.44479658018867924,
|
| 8455 |
+
"grad_norm": 2.859375,
|
| 8456 |
+
"learning_rate": 1.7656892128532596e-05,
|
| 8457 |
+
"loss": 2.0731,
|
| 8458 |
+
"step": 24140
|
| 8459 |
+
},
|
| 8460 |
+
{
|
| 8461 |
+
"epoch": 0.44516509433962265,
|
| 8462 |
+
"grad_norm": 2.65625,
|
| 8463 |
+
"learning_rate": 1.7653167550967453e-05,
|
| 8464 |
+
"loss": 2.1201,
|
| 8465 |
+
"step": 24160
|
| 8466 |
+
},
|
| 8467 |
+
{
|
| 8468 |
+
"epoch": 0.44553360849056606,
|
| 8469 |
+
"grad_norm": 2.59375,
|
| 8470 |
+
"learning_rate": 1.7649440408980026e-05,
|
| 8471 |
+
"loss": 2.1099,
|
| 8472 |
+
"step": 24180
|
| 8473 |
+
},
|
| 8474 |
+
{
|
| 8475 |
+
"epoch": 0.4459021226415094,
|
| 8476 |
+
"grad_norm": 2.34375,
|
| 8477 |
+
"learning_rate": 1.7645710703819202e-05,
|
| 8478 |
+
"loss": 2.1279,
|
| 8479 |
+
"step": 24200
|
| 8480 |
+
},
|
| 8481 |
+
{
|
| 8482 |
+
"epoch": 0.4462706367924528,
|
| 8483 |
+
"grad_norm": 2.40625,
|
| 8484 |
+
"learning_rate": 1.7641978436734732e-05,
|
| 8485 |
+
"loss": 2.0748,
|
| 8486 |
+
"step": 24220
|
| 8487 |
+
},
|
| 8488 |
+
{
|
| 8489 |
+
"epoch": 0.44663915094339623,
|
| 8490 |
+
"grad_norm": 2.1875,
|
| 8491 |
+
"learning_rate": 1.7638243608977225e-05,
|
| 8492 |
+
"loss": 2.0422,
|
| 8493 |
+
"step": 24240
|
| 8494 |
+
},
|
| 8495 |
+
{
|
| 8496 |
+
"epoch": 0.44700766509433965,
|
| 8497 |
+
"grad_norm": 2.390625,
|
| 8498 |
+
"learning_rate": 1.7634506221798144e-05,
|
| 8499 |
+
"loss": 2.1285,
|
| 8500 |
+
"step": 24260
|
| 8501 |
+
},
|
| 8502 |
+
{
|
| 8503 |
+
"epoch": 0.447376179245283,
|
| 8504 |
+
"grad_norm": 2.515625,
|
| 8505 |
+
"learning_rate": 1.7630766276449813e-05,
|
| 8506 |
+
"loss": 2.111,
|
| 8507 |
+
"step": 24280
|
| 8508 |
+
},
|
| 8509 |
+
{
|
| 8510 |
+
"epoch": 0.4477446933962264,
|
| 8511 |
+
"grad_norm": 2.375,
|
| 8512 |
+
"learning_rate": 1.7627023774185412e-05,
|
| 8513 |
+
"loss": 2.1067,
|
| 8514 |
+
"step": 24300
|
| 8515 |
+
},
|
| 8516 |
+
{
|
| 8517 |
+
"epoch": 0.4481132075471698,
|
| 8518 |
+
"grad_norm": 2.46875,
|
| 8519 |
+
"learning_rate": 1.762327871625898e-05,
|
| 8520 |
+
"loss": 2.128,
|
| 8521 |
+
"step": 24320
|
| 8522 |
+
},
|
| 8523 |
+
{
|
| 8524 |
+
"epoch": 0.44848172169811323,
|
| 8525 |
+
"grad_norm": 2.390625,
|
| 8526 |
+
"learning_rate": 1.761953110392541e-05,
|
| 8527 |
+
"loss": 2.1402,
|
| 8528 |
+
"step": 24340
|
| 8529 |
+
},
|
| 8530 |
+
{
|
| 8531 |
+
"epoch": 0.4488502358490566,
|
| 8532 |
+
"grad_norm": 2.46875,
|
| 8533 |
+
"learning_rate": 1.761578093844045e-05,
|
| 8534 |
+
"loss": 2.1319,
|
| 8535 |
+
"step": 24360
|
| 8536 |
+
},
|
| 8537 |
+
{
|
| 8538 |
+
"epoch": 0.44921875,
|
| 8539 |
+
"grad_norm": 2.546875,
|
| 8540 |
+
"learning_rate": 1.7612028221060705e-05,
|
| 8541 |
+
"loss": 2.1004,
|
| 8542 |
+
"step": 24380
|
| 8543 |
+
},
|
| 8544 |
+
{
|
| 8545 |
+
"epoch": 0.4495872641509434,
|
| 8546 |
+
"grad_norm": 2.34375,
|
| 8547 |
+
"learning_rate": 1.7608272953043636e-05,
|
| 8548 |
+
"loss": 2.1283,
|
| 8549 |
+
"step": 24400
|
| 8550 |
+
},
|
| 8551 |
+
{
|
| 8552 |
+
"epoch": 0.44995577830188677,
|
| 8553 |
+
"grad_norm": 2.578125,
|
| 8554 |
+
"learning_rate": 1.760451513564756e-05,
|
| 8555 |
+
"loss": 2.1034,
|
| 8556 |
+
"step": 24420
|
| 8557 |
+
},
|
| 8558 |
+
{
|
| 8559 |
+
"epoch": 0.4503242924528302,
|
| 8560 |
+
"grad_norm": 2.265625,
|
| 8561 |
+
"learning_rate": 1.760075477013164e-05,
|
| 8562 |
+
"loss": 2.0715,
|
| 8563 |
+
"step": 24440
|
| 8564 |
+
},
|
| 8565 |
+
{
|
| 8566 |
+
"epoch": 0.4506928066037736,
|
| 8567 |
+
"grad_norm": 2.71875,
|
| 8568 |
+
"learning_rate": 1.7596991857755903e-05,
|
| 8569 |
+
"loss": 2.1336,
|
| 8570 |
+
"step": 24460
|
| 8571 |
+
},
|
| 8572 |
+
{
|
| 8573 |
+
"epoch": 0.451061320754717,
|
| 8574 |
+
"grad_norm": 2.484375,
|
| 8575 |
+
"learning_rate": 1.7593226399781222e-05,
|
| 8576 |
+
"loss": 2.1419,
|
| 8577 |
+
"step": 24480
|
| 8578 |
+
},
|
| 8579 |
+
{
|
| 8580 |
+
"epoch": 0.45142983490566035,
|
| 8581 |
+
"grad_norm": 2.734375,
|
| 8582 |
+
"learning_rate": 1.7589458397469333e-05,
|
| 8583 |
+
"loss": 2.1267,
|
| 8584 |
+
"step": 24500
|
| 8585 |
+
},
|
| 8586 |
+
{
|
| 8587 |
+
"epoch": 0.45179834905660377,
|
| 8588 |
+
"grad_norm": 2.375,
|
| 8589 |
+
"learning_rate": 1.758568785208281e-05,
|
| 8590 |
+
"loss": 2.1158,
|
| 8591 |
+
"step": 24520
|
| 8592 |
+
},
|
| 8593 |
+
{
|
| 8594 |
+
"epoch": 0.4521668632075472,
|
| 8595 |
+
"grad_norm": 2.484375,
|
| 8596 |
+
"learning_rate": 1.758191476488509e-05,
|
| 8597 |
+
"loss": 2.1312,
|
| 8598 |
+
"step": 24540
|
| 8599 |
+
},
|
| 8600 |
+
{
|
| 8601 |
+
"epoch": 0.4525353773584906,
|
| 8602 |
+
"grad_norm": 3.21875,
|
| 8603 |
+
"learning_rate": 1.7578139137140455e-05,
|
| 8604 |
+
"loss": 2.1509,
|
| 8605 |
+
"step": 24560
|
| 8606 |
+
},
|
| 8607 |
+
{
|
| 8608 |
+
"epoch": 0.45290389150943394,
|
| 8609 |
+
"grad_norm": 2.234375,
|
| 8610 |
+
"learning_rate": 1.757436097011405e-05,
|
| 8611 |
+
"loss": 2.1073,
|
| 8612 |
+
"step": 24580
|
| 8613 |
+
},
|
| 8614 |
+
{
|
| 8615 |
+
"epoch": 0.45327240566037735,
|
| 8616 |
+
"grad_norm": 2.6875,
|
| 8617 |
+
"learning_rate": 1.7570580265071855e-05,
|
| 8618 |
+
"loss": 2.127,
|
| 8619 |
+
"step": 24600
|
| 8620 |
+
},
|
| 8621 |
+
{
|
| 8622 |
+
"epoch": 0.45364091981132076,
|
| 8623 |
+
"grad_norm": 2.828125,
|
| 8624 |
+
"learning_rate": 1.7566797023280712e-05,
|
| 8625 |
+
"loss": 2.1007,
|
| 8626 |
+
"step": 24620
|
| 8627 |
+
},
|
| 8628 |
+
{
|
| 8629 |
+
"epoch": 0.4540094339622642,
|
| 8630 |
+
"grad_norm": 2.40625,
|
| 8631 |
+
"learning_rate": 1.7563011246008307e-05,
|
| 8632 |
+
"loss": 2.1209,
|
| 8633 |
+
"step": 24640
|
| 8634 |
+
},
|
| 8635 |
+
{
|
| 8636 |
+
"epoch": 0.45437794811320753,
|
| 8637 |
+
"grad_norm": 2.46875,
|
| 8638 |
+
"learning_rate": 1.7559222934523177e-05,
|
| 8639 |
+
"loss": 2.0866,
|
| 8640 |
+
"step": 24660
|
| 8641 |
+
},
|
| 8642 |
+
{
|
| 8643 |
+
"epoch": 0.45474646226415094,
|
| 8644 |
+
"grad_norm": 2.546875,
|
| 8645 |
+
"learning_rate": 1.7555432090094716e-05,
|
| 8646 |
+
"loss": 2.1239,
|
| 8647 |
+
"step": 24680
|
| 8648 |
+
},
|
| 8649 |
+
{
|
| 8650 |
+
"epoch": 0.45511497641509435,
|
| 8651 |
+
"grad_norm": 2.46875,
|
| 8652 |
+
"learning_rate": 1.7551638713993145e-05,
|
| 8653 |
+
"loss": 2.0982,
|
| 8654 |
+
"step": 24700
|
| 8655 |
+
},
|
| 8656 |
+
{
|
| 8657 |
+
"epoch": 0.45548349056603776,
|
| 8658 |
+
"grad_norm": 2.53125,
|
| 8659 |
+
"learning_rate": 1.7547842807489566e-05,
|
| 8660 |
+
"loss": 2.086,
|
| 8661 |
+
"step": 24720
|
| 8662 |
+
},
|
| 8663 |
+
{
|
| 8664 |
+
"epoch": 0.4558520047169811,
|
| 8665 |
+
"grad_norm": 2.328125,
|
| 8666 |
+
"learning_rate": 1.7544044371855897e-05,
|
| 8667 |
+
"loss": 2.1182,
|
| 8668 |
+
"step": 24740
|
| 8669 |
+
},
|
| 8670 |
+
{
|
| 8671 |
+
"epoch": 0.45622051886792453,
|
| 8672 |
+
"grad_norm": 2.328125,
|
| 8673 |
+
"learning_rate": 1.7540243408364924e-05,
|
| 8674 |
+
"loss": 2.1121,
|
| 8675 |
+
"step": 24760
|
| 8676 |
+
},
|
| 8677 |
+
{
|
| 8678 |
+
"epoch": 0.45658903301886794,
|
| 8679 |
+
"grad_norm": 2.46875,
|
| 8680 |
+
"learning_rate": 1.7536439918290272e-05,
|
| 8681 |
+
"loss": 2.1324,
|
| 8682 |
+
"step": 24780
|
| 8683 |
+
},
|
| 8684 |
+
{
|
| 8685 |
+
"epoch": 0.4569575471698113,
|
| 8686 |
+
"grad_norm": 2.5,
|
| 8687 |
+
"learning_rate": 1.7532633902906414e-05,
|
| 8688 |
+
"loss": 2.1164,
|
| 8689 |
+
"step": 24800
|
| 8690 |
+
},
|
| 8691 |
+
{
|
| 8692 |
+
"epoch": 0.4573260613207547,
|
| 8693 |
+
"grad_norm": 2.703125,
|
| 8694 |
+
"learning_rate": 1.7528825363488664e-05,
|
| 8695 |
+
"loss": 2.0858,
|
| 8696 |
+
"step": 24820
|
| 8697 |
+
},
|
| 8698 |
+
{
|
| 8699 |
+
"epoch": 0.4576945754716981,
|
| 8700 |
+
"grad_norm": 2.46875,
|
| 8701 |
+
"learning_rate": 1.752501430131319e-05,
|
| 8702 |
+
"loss": 2.131,
|
| 8703 |
+
"step": 24840
|
| 8704 |
+
},
|
| 8705 |
+
{
|
| 8706 |
+
"epoch": 0.45806308962264153,
|
| 8707 |
+
"grad_norm": 2.40625,
|
| 8708 |
+
"learning_rate": 1.752120071765701e-05,
|
| 8709 |
+
"loss": 2.153,
|
| 8710 |
+
"step": 24860
|
| 8711 |
+
},
|
| 8712 |
+
{
|
| 8713 |
+
"epoch": 0.4584316037735849,
|
| 8714 |
+
"grad_norm": 2.546875,
|
| 8715 |
+
"learning_rate": 1.751738461379797e-05,
|
| 8716 |
+
"loss": 2.1512,
|
| 8717 |
+
"step": 24880
|
| 8718 |
+
},
|
| 8719 |
+
{
|
| 8720 |
+
"epoch": 0.4588001179245283,
|
| 8721 |
+
"grad_norm": 2.625,
|
| 8722 |
+
"learning_rate": 1.751356599101477e-05,
|
| 8723 |
+
"loss": 2.1456,
|
| 8724 |
+
"step": 24900
|
| 8725 |
+
},
|
| 8726 |
+
{
|
| 8727 |
+
"epoch": 0.4591686320754717,
|
| 8728 |
+
"grad_norm": 2.421875,
|
| 8729 |
+
"learning_rate": 1.750974485058696e-05,
|
| 8730 |
+
"loss": 2.0839,
|
| 8731 |
+
"step": 24920
|
| 8732 |
+
},
|
| 8733 |
+
{
|
| 8734 |
+
"epoch": 0.4595371462264151,
|
| 8735 |
+
"grad_norm": 2.5,
|
| 8736 |
+
"learning_rate": 1.7505921193794916e-05,
|
| 8737 |
+
"loss": 2.1421,
|
| 8738 |
+
"step": 24940
|
| 8739 |
+
},
|
| 8740 |
+
{
|
| 8741 |
+
"epoch": 0.45990566037735847,
|
| 8742 |
+
"grad_norm": 2.484375,
|
| 8743 |
+
"learning_rate": 1.750209502191988e-05,
|
| 8744 |
+
"loss": 2.0954,
|
| 8745 |
+
"step": 24960
|
| 8746 |
+
},
|
| 8747 |
+
{
|
| 8748 |
+
"epoch": 0.4602741745283019,
|
| 8749 |
+
"grad_norm": 2.3125,
|
| 8750 |
+
"learning_rate": 1.749826633624392e-05,
|
| 8751 |
+
"loss": 2.1094,
|
| 8752 |
+
"step": 24980
|
| 8753 |
+
},
|
| 8754 |
+
{
|
| 8755 |
+
"epoch": 0.4606426886792453,
|
| 8756 |
+
"grad_norm": 2.71875,
|
| 8757 |
+
"learning_rate": 1.749443513804995e-05,
|
| 8758 |
+
"loss": 2.0905,
|
| 8759 |
+
"step": 25000
|
| 8760 |
+
},
|
| 8761 |
+
{
|
| 8762 |
+
"epoch": 0.4610112028301887,
|
| 8763 |
+
"grad_norm": 2.359375,
|
| 8764 |
+
"learning_rate": 1.749060142862173e-05,
|
| 8765 |
+
"loss": 2.1255,
|
| 8766 |
+
"step": 25020
|
| 8767 |
+
},
|
| 8768 |
+
{
|
| 8769 |
+
"epoch": 0.46137971698113206,
|
| 8770 |
+
"grad_norm": 2.5,
|
| 8771 |
+
"learning_rate": 1.7486765209243858e-05,
|
| 8772 |
+
"loss": 2.1198,
|
| 8773 |
+
"step": 25040
|
| 8774 |
+
},
|
| 8775 |
+
{
|
| 8776 |
+
"epoch": 0.46174823113207547,
|
| 8777 |
+
"grad_norm": 2.40625,
|
| 8778 |
+
"learning_rate": 1.7482926481201775e-05,
|
| 8779 |
+
"loss": 2.1223,
|
| 8780 |
+
"step": 25060
|
| 8781 |
+
},
|
| 8782 |
+
{
|
| 8783 |
+
"epoch": 0.4621167452830189,
|
| 8784 |
+
"grad_norm": 2.28125,
|
| 8785 |
+
"learning_rate": 1.7479085245781754e-05,
|
| 8786 |
+
"loss": 2.0708,
|
| 8787 |
+
"step": 25080
|
| 8788 |
+
},
|
| 8789 |
+
{
|
| 8790 |
+
"epoch": 0.46248525943396224,
|
| 8791 |
+
"grad_norm": 2.390625,
|
| 8792 |
+
"learning_rate": 1.747524150427093e-05,
|
| 8793 |
+
"loss": 2.1043,
|
| 8794 |
+
"step": 25100
|
| 8795 |
+
},
|
| 8796 |
+
{
|
| 8797 |
+
"epoch": 0.46285377358490565,
|
| 8798 |
+
"grad_norm": 2.640625,
|
| 8799 |
+
"learning_rate": 1.747139525795725e-05,
|
| 8800 |
+
"loss": 2.1392,
|
| 8801 |
+
"step": 25120
|
| 8802 |
+
},
|
| 8803 |
+
{
|
| 8804 |
+
"epoch": 0.46322228773584906,
|
| 8805 |
+
"grad_norm": 2.75,
|
| 8806 |
+
"learning_rate": 1.746754650812952e-05,
|
| 8807 |
+
"loss": 2.1248,
|
| 8808 |
+
"step": 25140
|
| 8809 |
+
},
|
| 8810 |
+
{
|
| 8811 |
+
"epoch": 0.46359080188679247,
|
| 8812 |
+
"grad_norm": 2.53125,
|
| 8813 |
+
"learning_rate": 1.7463695256077376e-05,
|
| 8814 |
+
"loss": 2.1419,
|
| 8815 |
+
"step": 25160
|
| 8816 |
+
},
|
| 8817 |
+
{
|
| 8818 |
+
"epoch": 0.4639593160377358,
|
| 8819 |
+
"grad_norm": 2.3125,
|
| 8820 |
+
"learning_rate": 1.7459841503091296e-05,
|
| 8821 |
+
"loss": 2.1347,
|
| 8822 |
+
"step": 25180
|
| 8823 |
+
},
|
| 8824 |
+
{
|
| 8825 |
+
"epoch": 0.46432783018867924,
|
| 8826 |
+
"grad_norm": 2.4375,
|
| 8827 |
+
"learning_rate": 1.7455985250462597e-05,
|
| 8828 |
+
"loss": 2.0906,
|
| 8829 |
+
"step": 25200
|
| 8830 |
+
},
|
| 8831 |
+
{
|
| 8832 |
+
"epoch": 0.46469634433962265,
|
| 8833 |
+
"grad_norm": 2.796875,
|
| 8834 |
+
"learning_rate": 1.745212649948343e-05,
|
| 8835 |
+
"loss": 2.1114,
|
| 8836 |
+
"step": 25220
|
| 8837 |
+
},
|
| 8838 |
+
{
|
| 8839 |
+
"epoch": 0.46506485849056606,
|
| 8840 |
+
"grad_norm": 2.46875,
|
| 8841 |
+
"learning_rate": 1.7448265251446783e-05,
|
| 8842 |
+
"loss": 2.0949,
|
| 8843 |
+
"step": 25240
|
| 8844 |
+
},
|
| 8845 |
+
{
|
| 8846 |
+
"epoch": 0.4654333726415094,
|
| 8847 |
+
"grad_norm": 2.59375,
|
| 8848 |
+
"learning_rate": 1.7444401507646488e-05,
|
| 8849 |
+
"loss": 2.0722,
|
| 8850 |
+
"step": 25260
|
| 8851 |
+
},
|
| 8852 |
+
{
|
| 8853 |
+
"epoch": 0.4658018867924528,
|
| 8854 |
+
"grad_norm": 2.484375,
|
| 8855 |
+
"learning_rate": 1.7440535269377198e-05,
|
| 8856 |
+
"loss": 2.1266,
|
| 8857 |
+
"step": 25280
|
| 8858 |
+
},
|
| 8859 |
+
{
|
| 8860 |
+
"epoch": 0.46617040094339623,
|
| 8861 |
+
"grad_norm": 2.5,
|
| 8862 |
+
"learning_rate": 1.743666653793442e-05,
|
| 8863 |
+
"loss": 2.1079,
|
| 8864 |
+
"step": 25300
|
| 8865 |
+
},
|
| 8866 |
+
{
|
| 8867 |
+
"epoch": 0.46653891509433965,
|
| 8868 |
+
"grad_norm": 2.34375,
|
| 8869 |
+
"learning_rate": 1.7432795314614484e-05,
|
| 8870 |
+
"loss": 2.086,
|
| 8871 |
+
"step": 25320
|
| 8872 |
+
},
|
| 8873 |
+
{
|
| 8874 |
+
"epoch": 0.466907429245283,
|
| 8875 |
+
"grad_norm": 2.65625,
|
| 8876 |
+
"learning_rate": 1.7428921600714566e-05,
|
| 8877 |
+
"loss": 2.1129,
|
| 8878 |
+
"step": 25340
|
| 8879 |
+
},
|
| 8880 |
+
{
|
| 8881 |
+
"epoch": 0.4672759433962264,
|
| 8882 |
+
"grad_norm": 2.421875,
|
| 8883 |
+
"learning_rate": 1.7425045397532657e-05,
|
| 8884 |
+
"loss": 2.0769,
|
| 8885 |
+
"step": 25360
|
| 8886 |
+
},
|
| 8887 |
+
{
|
| 8888 |
+
"epoch": 0.4676444575471698,
|
| 8889 |
+
"grad_norm": 2.53125,
|
| 8890 |
+
"learning_rate": 1.7421166706367604e-05,
|
| 8891 |
+
"loss": 2.0878,
|
| 8892 |
+
"step": 25380
|
| 8893 |
+
},
|
| 8894 |
+
{
|
| 8895 |
+
"epoch": 0.46801297169811323,
|
| 8896 |
+
"grad_norm": 2.75,
|
| 8897 |
+
"learning_rate": 1.7417285528519078e-05,
|
| 8898 |
+
"loss": 2.142,
|
| 8899 |
+
"step": 25400
|
| 8900 |
+
},
|
| 8901 |
+
{
|
| 8902 |
+
"epoch": 0.4683814858490566,
|
| 8903 |
+
"grad_norm": 2.71875,
|
| 8904 |
+
"learning_rate": 1.7413401865287576e-05,
|
| 8905 |
+
"loss": 2.135,
|
| 8906 |
+
"step": 25420
|
| 8907 |
+
},
|
| 8908 |
+
{
|
| 8909 |
+
"epoch": 0.46875,
|
| 8910 |
+
"grad_norm": 2.84375,
|
| 8911 |
+
"learning_rate": 1.740951571797444e-05,
|
| 8912 |
+
"loss": 2.1272,
|
| 8913 |
+
"step": 25440
|
| 8914 |
+
},
|
| 8915 |
+
{
|
| 8916 |
+
"epoch": 0.4691185141509434,
|
| 8917 |
+
"grad_norm": 2.375,
|
| 8918 |
+
"learning_rate": 1.7405627087881846e-05,
|
| 8919 |
+
"loss": 2.0918,
|
| 8920 |
+
"step": 25460
|
| 8921 |
+
},
|
| 8922 |
+
{
|
| 8923 |
+
"epoch": 0.46948702830188677,
|
| 8924 |
+
"grad_norm": 2.453125,
|
| 8925 |
+
"learning_rate": 1.7401735976312786e-05,
|
| 8926 |
+
"loss": 2.0877,
|
| 8927 |
+
"step": 25480
|
| 8928 |
+
},
|
| 8929 |
+
{
|
| 8930 |
+
"epoch": 0.4698555424528302,
|
| 8931 |
+
"grad_norm": 2.453125,
|
| 8932 |
+
"learning_rate": 1.7397842384571098e-05,
|
| 8933 |
+
"loss": 2.1101,
|
| 8934 |
+
"step": 25500
|
| 8935 |
+
},
|
| 8936 |
+
{
|
| 8937 |
+
"epoch": 0.4702240566037736,
|
| 8938 |
+
"grad_norm": 2.578125,
|
| 8939 |
+
"learning_rate": 1.7393946313961444e-05,
|
| 8940 |
+
"loss": 2.1004,
|
| 8941 |
+
"step": 25520
|
| 8942 |
+
},
|
| 8943 |
+
{
|
| 8944 |
+
"epoch": 0.470592570754717,
|
| 8945 |
+
"grad_norm": 2.515625,
|
| 8946 |
+
"learning_rate": 1.739004776578932e-05,
|
| 8947 |
+
"loss": 2.076,
|
| 8948 |
+
"step": 25540
|
| 8949 |
+
},
|
| 8950 |
+
{
|
| 8951 |
+
"epoch": 0.47096108490566035,
|
| 8952 |
+
"grad_norm": 2.296875,
|
| 8953 |
+
"learning_rate": 1.738614674136105e-05,
|
| 8954 |
+
"loss": 2.1314,
|
| 8955 |
+
"step": 25560
|
| 8956 |
+
},
|
| 8957 |
+
{
|
| 8958 |
+
"epoch": 0.47132959905660377,
|
| 8959 |
+
"grad_norm": 2.484375,
|
| 8960 |
+
"learning_rate": 1.7382243241983793e-05,
|
| 8961 |
+
"loss": 2.082,
|
| 8962 |
+
"step": 25580
|
| 8963 |
+
},
|
| 8964 |
+
{
|
| 8965 |
+
"epoch": 0.4716981132075472,
|
| 8966 |
+
"grad_norm": 2.59375,
|
| 8967 |
+
"learning_rate": 1.7378337268965525e-05,
|
| 8968 |
+
"loss": 2.0995,
|
| 8969 |
+
"step": 25600
|
| 8970 |
+
},
|
| 8971 |
+
{
|
| 8972 |
+
"epoch": 0.4720666273584906,
|
| 8973 |
+
"grad_norm": 2.6875,
|
| 8974 |
+
"learning_rate": 1.7374428823615065e-05,
|
| 8975 |
+
"loss": 2.1468,
|
| 8976 |
+
"step": 25620
|
| 8977 |
+
},
|
| 8978 |
+
{
|
| 8979 |
+
"epoch": 0.47243514150943394,
|
| 8980 |
+
"grad_norm": 2.359375,
|
| 8981 |
+
"learning_rate": 1.7370517907242054e-05,
|
| 8982 |
+
"loss": 2.0789,
|
| 8983 |
+
"step": 25640
|
| 8984 |
+
},
|
| 8985 |
+
{
|
| 8986 |
+
"epoch": 0.47280365566037735,
|
| 8987 |
+
"grad_norm": 2.15625,
|
| 8988 |
+
"learning_rate": 1.7366604521156958e-05,
|
| 8989 |
+
"loss": 2.1137,
|
| 8990 |
+
"step": 25660
|
| 8991 |
+
},
|
| 8992 |
+
{
|
| 8993 |
+
"epoch": 0.47317216981132076,
|
| 8994 |
+
"grad_norm": 2.734375,
|
| 8995 |
+
"learning_rate": 1.7362688666671077e-05,
|
| 8996 |
+
"loss": 2.0753,
|
| 8997 |
+
"step": 25680
|
| 8998 |
+
},
|
| 8999 |
+
{
|
| 9000 |
+
"epoch": 0.4735406839622642,
|
| 9001 |
+
"grad_norm": 2.671875,
|
| 9002 |
+
"learning_rate": 1.7358770345096536e-05,
|
| 9003 |
+
"loss": 2.0988,
|
| 9004 |
+
"step": 25700
|
| 9005 |
+
},
|
| 9006 |
+
{
|
| 9007 |
+
"epoch": 0.47390919811320753,
|
| 9008 |
+
"grad_norm": 2.46875,
|
| 9009 |
+
"learning_rate": 1.7354849557746275e-05,
|
| 9010 |
+
"loss": 2.0755,
|
| 9011 |
+
"step": 25720
|
| 9012 |
+
},
|
| 9013 |
+
{
|
| 9014 |
+
"epoch": 0.47427771226415094,
|
| 9015 |
+
"grad_norm": 2.75,
|
| 9016 |
+
"learning_rate": 1.7350926305934085e-05,
|
| 9017 |
+
"loss": 2.1032,
|
| 9018 |
+
"step": 25740
|
| 9019 |
+
},
|
| 9020 |
+
{
|
| 9021 |
+
"epoch": 0.47464622641509435,
|
| 9022 |
+
"grad_norm": 2.40625,
|
| 9023 |
+
"learning_rate": 1.7347000590974564e-05,
|
| 9024 |
+
"loss": 2.1105,
|
| 9025 |
+
"step": 25760
|
| 9026 |
+
},
|
| 9027 |
+
{
|
| 9028 |
+
"epoch": 0.47501474056603776,
|
| 9029 |
+
"grad_norm": 2.8125,
|
| 9030 |
+
"learning_rate": 1.7343072414183136e-05,
|
| 9031 |
+
"loss": 2.1139,
|
| 9032 |
+
"step": 25780
|
| 9033 |
+
},
|
| 9034 |
+
{
|
| 9035 |
+
"epoch": 0.4753832547169811,
|
| 9036 |
+
"grad_norm": 2.765625,
|
| 9037 |
+
"learning_rate": 1.733914177687606e-05,
|
| 9038 |
+
"loss": 2.1169,
|
| 9039 |
+
"step": 25800
|
| 9040 |
+
},
|
| 9041 |
+
{
|
| 9042 |
+
"epoch": 0.47575176886792453,
|
| 9043 |
+
"grad_norm": 2.765625,
|
| 9044 |
+
"learning_rate": 1.733520868037041e-05,
|
| 9045 |
+
"loss": 2.1285,
|
| 9046 |
+
"step": 25820
|
| 9047 |
+
},
|
| 9048 |
+
{
|
| 9049 |
+
"epoch": 0.47612028301886794,
|
| 9050 |
+
"grad_norm": 2.328125,
|
| 9051 |
+
"learning_rate": 1.7331273125984086e-05,
|
| 9052 |
+
"loss": 2.1631,
|
| 9053 |
+
"step": 25840
|
| 9054 |
+
},
|
| 9055 |
+
{
|
| 9056 |
+
"epoch": 0.4764887971698113,
|
| 9057 |
+
"grad_norm": 2.3125,
|
| 9058 |
+
"learning_rate": 1.7327335115035814e-05,
|
| 9059 |
+
"loss": 2.0671,
|
| 9060 |
+
"step": 25860
|
| 9061 |
+
},
|
| 9062 |
+
{
|
| 9063 |
+
"epoch": 0.4768573113207547,
|
| 9064 |
+
"grad_norm": 2.609375,
|
| 9065 |
+
"learning_rate": 1.7323394648845145e-05,
|
| 9066 |
+
"loss": 2.0903,
|
| 9067 |
+
"step": 25880
|
| 9068 |
+
},
|
| 9069 |
+
{
|
| 9070 |
+
"epoch": 0.4772258254716981,
|
| 9071 |
+
"grad_norm": 2.515625,
|
| 9072 |
+
"learning_rate": 1.7319451728732448e-05,
|
| 9073 |
+
"loss": 2.1003,
|
| 9074 |
+
"step": 25900
|
| 9075 |
+
},
|
| 9076 |
+
{
|
| 9077 |
+
"epoch": 0.47759433962264153,
|
| 9078 |
+
"grad_norm": 3.125,
|
| 9079 |
+
"learning_rate": 1.7315506356018912e-05,
|
| 9080 |
+
"loss": 2.1062,
|
| 9081 |
+
"step": 25920
|
| 9082 |
+
},
|
| 9083 |
+
{
|
| 9084 |
+
"epoch": 0.4779628537735849,
|
| 9085 |
+
"grad_norm": 2.578125,
|
| 9086 |
+
"learning_rate": 1.731155853202656e-05,
|
| 9087 |
+
"loss": 2.1029,
|
| 9088 |
+
"step": 25940
|
| 9089 |
+
},
|
| 9090 |
+
{
|
| 9091 |
+
"epoch": 0.4783313679245283,
|
| 9092 |
+
"grad_norm": 2.609375,
|
| 9093 |
+
"learning_rate": 1.730760825807822e-05,
|
| 9094 |
+
"loss": 2.1113,
|
| 9095 |
+
"step": 25960
|
| 9096 |
+
},
|
| 9097 |
+
{
|
| 9098 |
+
"epoch": 0.4786998820754717,
|
| 9099 |
+
"grad_norm": 2.375,
|
| 9100 |
+
"learning_rate": 1.7303655535497555e-05,
|
| 9101 |
+
"loss": 2.1399,
|
| 9102 |
+
"step": 25980
|
| 9103 |
+
},
|
| 9104 |
+
{
|
| 9105 |
+
"epoch": 0.4790683962264151,
|
| 9106 |
+
"grad_norm": 2.5625,
|
| 9107 |
+
"learning_rate": 1.729970036560904e-05,
|
| 9108 |
+
"loss": 2.1394,
|
| 9109 |
+
"step": 26000
|
| 9110 |
+
},
|
| 9111 |
+
{
|
| 9112 |
+
"epoch": 0.47943691037735847,
|
| 9113 |
+
"grad_norm": 2.53125,
|
| 9114 |
+
"learning_rate": 1.7295742749737975e-05,
|
| 9115 |
+
"loss": 2.12,
|
| 9116 |
+
"step": 26020
|
| 9117 |
+
},
|
| 9118 |
+
{
|
| 9119 |
+
"epoch": 0.4798054245283019,
|
| 9120 |
+
"grad_norm": 2.890625,
|
| 9121 |
+
"learning_rate": 1.7291782689210475e-05,
|
| 9122 |
+
"loss": 2.1166,
|
| 9123 |
+
"step": 26040
|
| 9124 |
+
},
|
| 9125 |
+
{
|
| 9126 |
+
"epoch": 0.4801739386792453,
|
| 9127 |
+
"grad_norm": 2.453125,
|
| 9128 |
+
"learning_rate": 1.7287820185353478e-05,
|
| 9129 |
+
"loss": 2.107,
|
| 9130 |
+
"step": 26060
|
| 9131 |
+
},
|
| 9132 |
+
{
|
| 9133 |
+
"epoch": 0.4805424528301887,
|
| 9134 |
+
"grad_norm": 2.609375,
|
| 9135 |
+
"learning_rate": 1.728385523949474e-05,
|
| 9136 |
+
"loss": 2.1413,
|
| 9137 |
+
"step": 26080
|
| 9138 |
+
},
|
| 9139 |
+
{
|
| 9140 |
+
"epoch": 0.48091096698113206,
|
| 9141 |
+
"grad_norm": 2.5625,
|
| 9142 |
+
"learning_rate": 1.727988785296283e-05,
|
| 9143 |
+
"loss": 2.0822,
|
| 9144 |
+
"step": 26100
|
| 9145 |
+
},
|
| 9146 |
+
{
|
| 9147 |
+
"epoch": 0.48127948113207547,
|
| 9148 |
+
"grad_norm": 2.4375,
|
| 9149 |
+
"learning_rate": 1.7275918027087146e-05,
|
| 9150 |
+
"loss": 2.1056,
|
| 9151 |
+
"step": 26120
|
| 9152 |
+
},
|
| 9153 |
+
{
|
| 9154 |
+
"epoch": 0.4816479952830189,
|
| 9155 |
+
"grad_norm": 2.4375,
|
| 9156 |
+
"learning_rate": 1.727194576319789e-05,
|
| 9157 |
+
"loss": 2.1218,
|
| 9158 |
+
"step": 26140
|
| 9159 |
+
},
|
| 9160 |
+
{
|
| 9161 |
+
"epoch": 0.48201650943396224,
|
| 9162 |
+
"grad_norm": 2.5,
|
| 9163 |
+
"learning_rate": 1.7267971062626095e-05,
|
| 9164 |
+
"loss": 2.1634,
|
| 9165 |
+
"step": 26160
|
| 9166 |
+
},
|
| 9167 |
+
{
|
| 9168 |
+
"epoch": 0.48238502358490565,
|
| 9169 |
+
"grad_norm": 2.84375,
|
| 9170 |
+
"learning_rate": 1.7263993926703593e-05,
|
| 9171 |
+
"loss": 2.1486,
|
| 9172 |
+
"step": 26180
|
| 9173 |
+
},
|
| 9174 |
+
{
|
| 9175 |
+
"epoch": 0.48275353773584906,
|
| 9176 |
+
"grad_norm": 2.234375,
|
| 9177 |
+
"learning_rate": 1.726001435676305e-05,
|
| 9178 |
+
"loss": 2.1075,
|
| 9179 |
+
"step": 26200
|
| 9180 |
}
|
| 9181 |
],
|
| 9182 |
"logging_steps": 20,
|
|
|
|
| 9184 |
"num_input_tokens_seen": 0,
|
| 9185 |
"num_train_epochs": 2,
|
| 9186 |
"save_steps": 200,
|
| 9187 |
+
"total_flos": 2.6285411908110385e+19,
|
| 9188 |
"train_batch_size": 16,
|
| 9189 |
"trial_name": null,
|
| 9190 |
"trial_params": null
|