Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1403 -3
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 83946192
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2df44592754fe7188dd677ae0bc32281f8bcfad5f4a6c037e3ae72d516e0f92
|
| 3 |
size 83946192
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 168150290
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d4416518671bd3a5927dad1918e6feb19b8689187dec1647b573a84566f45c4
|
| 3 |
size 168150290
|
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 sha256:89b8115cb129c536e89826f236d6c4894075bda82a59ceefb56b1d1ccf0de0fb
|
| 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 sha256:167f19cbea1c5998444d9bc539623c9f6f524cce3680072969e6130b6d4da06a
|
| 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,
|
|
@@ -3234,6 +3234,1406 @@
|
|
| 3234 |
"learning_rate": 0.00016652690605760775,
|
| 3235 |
"loss": 0.9739,
|
| 3236 |
"step": 2300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3237 |
}
|
| 3238 |
],
|
| 3239 |
"logging_steps": 5,
|
|
@@ -3241,7 +4641,7 @@
|
|
| 3241 |
"num_input_tokens_seen": 0,
|
| 3242 |
"num_train_epochs": 1,
|
| 3243 |
"save_steps": 100,
|
| 3244 |
-
"total_flos":
|
| 3245 |
"train_batch_size": 2,
|
| 3246 |
"trial_name": null,
|
| 3247 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.4900141064667013,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3300,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3234 |
"learning_rate": 0.00016652690605760775,
|
| 3235 |
"loss": 0.9739,
|
| 3236 |
"step": 2300
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.34,
|
| 3240 |
+
"grad_norm": 0.333984375,
|
| 3241 |
+
"learning_rate": 0.00016633315764368818,
|
| 3242 |
+
"loss": 0.9428,
|
| 3243 |
+
"step": 2305
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.34,
|
| 3247 |
+
"grad_norm": 0.341796875,
|
| 3248 |
+
"learning_rate": 0.0001661389635479332,
|
| 3249 |
+
"loss": 0.938,
|
| 3250 |
+
"step": 2310
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.34,
|
| 3254 |
+
"grad_norm": 0.35546875,
|
| 3255 |
+
"learning_rate": 0.00016594432507510175,
|
| 3256 |
+
"loss": 0.9647,
|
| 3257 |
+
"step": 2315
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.34,
|
| 3261 |
+
"grad_norm": 0.333984375,
|
| 3262 |
+
"learning_rate": 0.00016574924353293845,
|
| 3263 |
+
"loss": 0.9413,
|
| 3264 |
+
"step": 2320
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.35,
|
| 3268 |
+
"grad_norm": 0.3515625,
|
| 3269 |
+
"learning_rate": 0.0001655537202321649,
|
| 3270 |
+
"loss": 0.9443,
|
| 3271 |
+
"step": 2325
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.35,
|
| 3275 |
+
"grad_norm": 0.341796875,
|
| 3276 |
+
"learning_rate": 0.00016535775648647075,
|
| 3277 |
+
"loss": 0.9407,
|
| 3278 |
+
"step": 2330
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.35,
|
| 3282 |
+
"grad_norm": 0.337890625,
|
| 3283 |
+
"learning_rate": 0.0001651613536125049,
|
| 3284 |
+
"loss": 0.9803,
|
| 3285 |
+
"step": 2335
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.35,
|
| 3289 |
+
"grad_norm": 0.34375,
|
| 3290 |
+
"learning_rate": 0.0001649645129298668,
|
| 3291 |
+
"loss": 0.9349,
|
| 3292 |
+
"step": 2340
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.35,
|
| 3296 |
+
"grad_norm": 0.328125,
|
| 3297 |
+
"learning_rate": 0.0001647672357610973,
|
| 3298 |
+
"loss": 0.9581,
|
| 3299 |
+
"step": 2345
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.35,
|
| 3303 |
+
"grad_norm": 0.33984375,
|
| 3304 |
+
"learning_rate": 0.00016456952343167007,
|
| 3305 |
+
"loss": 0.9243,
|
| 3306 |
+
"step": 2350
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.35,
|
| 3310 |
+
"grad_norm": 0.34765625,
|
| 3311 |
+
"learning_rate": 0.00016437137726998255,
|
| 3312 |
+
"loss": 0.9591,
|
| 3313 |
+
"step": 2355
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.35,
|
| 3317 |
+
"grad_norm": 0.34375,
|
| 3318 |
+
"learning_rate": 0.00016417279860734692,
|
| 3319 |
+
"loss": 0.9286,
|
| 3320 |
+
"step": 2360
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.35,
|
| 3324 |
+
"grad_norm": 0.345703125,
|
| 3325 |
+
"learning_rate": 0.00016397378877798134,
|
| 3326 |
+
"loss": 0.9695,
|
| 3327 |
+
"step": 2365
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.35,
|
| 3331 |
+
"grad_norm": 0.3515625,
|
| 3332 |
+
"learning_rate": 0.0001637743491190009,
|
| 3333 |
+
"loss": 0.9696,
|
| 3334 |
+
"step": 2370
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.35,
|
| 3338 |
+
"grad_norm": 0.3359375,
|
| 3339 |
+
"learning_rate": 0.00016357448097040867,
|
| 3340 |
+
"loss": 0.9465,
|
| 3341 |
+
"step": 2375
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.35,
|
| 3345 |
+
"grad_norm": 0.333984375,
|
| 3346 |
+
"learning_rate": 0.00016337418567508665,
|
| 3347 |
+
"loss": 0.9405,
|
| 3348 |
+
"step": 2380
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.35,
|
| 3352 |
+
"grad_norm": 0.337890625,
|
| 3353 |
+
"learning_rate": 0.00016317346457878675,
|
| 3354 |
+
"loss": 0.9151,
|
| 3355 |
+
"step": 2385
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.35,
|
| 3359 |
+
"grad_norm": 0.361328125,
|
| 3360 |
+
"learning_rate": 0.0001629723190301218,
|
| 3361 |
+
"loss": 0.9319,
|
| 3362 |
+
"step": 2390
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.36,
|
| 3366 |
+
"grad_norm": 0.345703125,
|
| 3367 |
+
"learning_rate": 0.00016277075038055634,
|
| 3368 |
+
"loss": 0.956,
|
| 3369 |
+
"step": 2395
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.36,
|
| 3373 |
+
"grad_norm": 0.3359375,
|
| 3374 |
+
"learning_rate": 0.00016256875998439796,
|
| 3375 |
+
"loss": 0.9475,
|
| 3376 |
+
"step": 2400
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.36,
|
| 3380 |
+
"grad_norm": 0.3359375,
|
| 3381 |
+
"learning_rate": 0.00016236634919878756,
|
| 3382 |
+
"loss": 0.9747,
|
| 3383 |
+
"step": 2405
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.36,
|
| 3387 |
+
"grad_norm": 0.3359375,
|
| 3388 |
+
"learning_rate": 0.00016216351938369073,
|
| 3389 |
+
"loss": 0.9405,
|
| 3390 |
+
"step": 2410
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.36,
|
| 3394 |
+
"grad_norm": 0.341796875,
|
| 3395 |
+
"learning_rate": 0.00016196027190188848,
|
| 3396 |
+
"loss": 0.9888,
|
| 3397 |
+
"step": 2415
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.36,
|
| 3401 |
+
"grad_norm": 0.34765625,
|
| 3402 |
+
"learning_rate": 0.00016175660811896796,
|
| 3403 |
+
"loss": 0.9063,
|
| 3404 |
+
"step": 2420
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.36,
|
| 3408 |
+
"grad_norm": 0.345703125,
|
| 3409 |
+
"learning_rate": 0.00016155252940331342,
|
| 3410 |
+
"loss": 0.9454,
|
| 3411 |
+
"step": 2425
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.36,
|
| 3415 |
+
"grad_norm": 0.341796875,
|
| 3416 |
+
"learning_rate": 0.000161348037126097,
|
| 3417 |
+
"loss": 0.9783,
|
| 3418 |
+
"step": 2430
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.36,
|
| 3422 |
+
"grad_norm": 0.34375,
|
| 3423 |
+
"learning_rate": 0.0001611431326612695,
|
| 3424 |
+
"loss": 0.9951,
|
| 3425 |
+
"step": 2435
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.36,
|
| 3429 |
+
"grad_norm": 0.33984375,
|
| 3430 |
+
"learning_rate": 0.00016093781738555106,
|
| 3431 |
+
"loss": 0.9194,
|
| 3432 |
+
"step": 2440
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.36,
|
| 3436 |
+
"grad_norm": 0.345703125,
|
| 3437 |
+
"learning_rate": 0.0001607320926784221,
|
| 3438 |
+
"loss": 0.9692,
|
| 3439 |
+
"step": 2445
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.36,
|
| 3443 |
+
"grad_norm": 0.349609375,
|
| 3444 |
+
"learning_rate": 0.00016052595992211387,
|
| 3445 |
+
"loss": 0.8929,
|
| 3446 |
+
"step": 2450
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.36,
|
| 3450 |
+
"grad_norm": 0.345703125,
|
| 3451 |
+
"learning_rate": 0.0001603194205015993,
|
| 3452 |
+
"loss": 0.9456,
|
| 3453 |
+
"step": 2455
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.37,
|
| 3457 |
+
"grad_norm": 0.337890625,
|
| 3458 |
+
"learning_rate": 0.00016011247580458355,
|
| 3459 |
+
"loss": 0.956,
|
| 3460 |
+
"step": 2460
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.37,
|
| 3464 |
+
"grad_norm": 0.341796875,
|
| 3465 |
+
"learning_rate": 0.00015990512722149482,
|
| 3466 |
+
"loss": 0.9113,
|
| 3467 |
+
"step": 2465
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.37,
|
| 3471 |
+
"grad_norm": 0.333984375,
|
| 3472 |
+
"learning_rate": 0.00015969737614547494,
|
| 3473 |
+
"loss": 0.9445,
|
| 3474 |
+
"step": 2470
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.37,
|
| 3478 |
+
"grad_norm": 0.357421875,
|
| 3479 |
+
"learning_rate": 0.00015948922397237007,
|
| 3480 |
+
"loss": 0.957,
|
| 3481 |
+
"step": 2475
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.37,
|
| 3485 |
+
"grad_norm": 0.349609375,
|
| 3486 |
+
"learning_rate": 0.00015928067210072122,
|
| 3487 |
+
"loss": 0.924,
|
| 3488 |
+
"step": 2480
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.37,
|
| 3492 |
+
"grad_norm": 0.35546875,
|
| 3493 |
+
"learning_rate": 0.0001590717219317549,
|
| 3494 |
+
"loss": 0.9558,
|
| 3495 |
+
"step": 2485
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.37,
|
| 3499 |
+
"grad_norm": 0.33984375,
|
| 3500 |
+
"learning_rate": 0.00015886237486937378,
|
| 3501 |
+
"loss": 0.9632,
|
| 3502 |
+
"step": 2490
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.37,
|
| 3506 |
+
"grad_norm": 0.375,
|
| 3507 |
+
"learning_rate": 0.00015865263232014715,
|
| 3508 |
+
"loss": 0.9477,
|
| 3509 |
+
"step": 2495
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.37,
|
| 3513 |
+
"grad_norm": 0.341796875,
|
| 3514 |
+
"learning_rate": 0.0001584424956933015,
|
| 3515 |
+
"loss": 0.9443,
|
| 3516 |
+
"step": 2500
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.37,
|
| 3520 |
+
"grad_norm": 0.333984375,
|
| 3521 |
+
"learning_rate": 0.0001582319664007111,
|
| 3522 |
+
"loss": 0.9316,
|
| 3523 |
+
"step": 2505
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.37,
|
| 3527 |
+
"grad_norm": 0.3359375,
|
| 3528 |
+
"learning_rate": 0.00015802104585688851,
|
| 3529 |
+
"loss": 0.9205,
|
| 3530 |
+
"step": 2510
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.37,
|
| 3534 |
+
"grad_norm": 0.333984375,
|
| 3535 |
+
"learning_rate": 0.00015780973547897494,
|
| 3536 |
+
"loss": 0.924,
|
| 3537 |
+
"step": 2515
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.37,
|
| 3541 |
+
"grad_norm": 0.333984375,
|
| 3542 |
+
"learning_rate": 0.0001575980366867309,
|
| 3543 |
+
"loss": 0.9498,
|
| 3544 |
+
"step": 2520
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.37,
|
| 3548 |
+
"grad_norm": 0.34765625,
|
| 3549 |
+
"learning_rate": 0.00015738595090252657,
|
| 3550 |
+
"loss": 0.9677,
|
| 3551 |
+
"step": 2525
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.38,
|
| 3555 |
+
"grad_norm": 0.353515625,
|
| 3556 |
+
"learning_rate": 0.00015717347955133233,
|
| 3557 |
+
"loss": 0.9786,
|
| 3558 |
+
"step": 2530
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.38,
|
| 3562 |
+
"grad_norm": 0.345703125,
|
| 3563 |
+
"learning_rate": 0.00015696062406070902,
|
| 3564 |
+
"loss": 0.9523,
|
| 3565 |
+
"step": 2535
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.38,
|
| 3569 |
+
"grad_norm": 0.3359375,
|
| 3570 |
+
"learning_rate": 0.00015674738586079852,
|
| 3571 |
+
"loss": 0.9496,
|
| 3572 |
+
"step": 2540
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.38,
|
| 3576 |
+
"grad_norm": 0.349609375,
|
| 3577 |
+
"learning_rate": 0.0001565337663843141,
|
| 3578 |
+
"loss": 0.9439,
|
| 3579 |
+
"step": 2545
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.38,
|
| 3583 |
+
"grad_norm": 0.33203125,
|
| 3584 |
+
"learning_rate": 0.00015631976706653063,
|
| 3585 |
+
"loss": 0.9496,
|
| 3586 |
+
"step": 2550
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.38,
|
| 3590 |
+
"grad_norm": 0.330078125,
|
| 3591 |
+
"learning_rate": 0.00015610538934527526,
|
| 3592 |
+
"loss": 0.9315,
|
| 3593 |
+
"step": 2555
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.38,
|
| 3597 |
+
"grad_norm": 0.345703125,
|
| 3598 |
+
"learning_rate": 0.00015589063466091743,
|
| 3599 |
+
"loss": 0.9595,
|
| 3600 |
+
"step": 2560
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.38,
|
| 3604 |
+
"grad_norm": 0.333984375,
|
| 3605 |
+
"learning_rate": 0.0001556755044563594,
|
| 3606 |
+
"loss": 0.9524,
|
| 3607 |
+
"step": 2565
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.38,
|
| 3611 |
+
"grad_norm": 0.341796875,
|
| 3612 |
+
"learning_rate": 0.00015546000017702648,
|
| 3613 |
+
"loss": 0.9442,
|
| 3614 |
+
"step": 2570
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.38,
|
| 3618 |
+
"grad_norm": 0.357421875,
|
| 3619 |
+
"learning_rate": 0.00015524412327085725,
|
| 3620 |
+
"loss": 0.9807,
|
| 3621 |
+
"step": 2575
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.38,
|
| 3625 |
+
"grad_norm": 0.318359375,
|
| 3626 |
+
"learning_rate": 0.00015502787518829406,
|
| 3627 |
+
"loss": 0.9385,
|
| 3628 |
+
"step": 2580
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.38,
|
| 3632 |
+
"grad_norm": 0.328125,
|
| 3633 |
+
"learning_rate": 0.00015481125738227305,
|
| 3634 |
+
"loss": 0.9618,
|
| 3635 |
+
"step": 2585
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.38,
|
| 3639 |
+
"grad_norm": 0.3515625,
|
| 3640 |
+
"learning_rate": 0.00015459427130821442,
|
| 3641 |
+
"loss": 0.9443,
|
| 3642 |
+
"step": 2590
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.39,
|
| 3646 |
+
"grad_norm": 0.35546875,
|
| 3647 |
+
"learning_rate": 0.0001543769184240128,
|
| 3648 |
+
"loss": 0.9428,
|
| 3649 |
+
"step": 2595
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.39,
|
| 3653 |
+
"grad_norm": 0.361328125,
|
| 3654 |
+
"learning_rate": 0.00015415920019002736,
|
| 3655 |
+
"loss": 0.9349,
|
| 3656 |
+
"step": 2600
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.39,
|
| 3660 |
+
"grad_norm": 0.3359375,
|
| 3661 |
+
"learning_rate": 0.00015394111806907188,
|
| 3662 |
+
"loss": 0.9325,
|
| 3663 |
+
"step": 2605
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.39,
|
| 3667 |
+
"grad_norm": 0.345703125,
|
| 3668 |
+
"learning_rate": 0.00015372267352640513,
|
| 3669 |
+
"loss": 0.9622,
|
| 3670 |
+
"step": 2610
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.39,
|
| 3674 |
+
"grad_norm": 0.34765625,
|
| 3675 |
+
"learning_rate": 0.00015350386802972097,
|
| 3676 |
+
"loss": 0.934,
|
| 3677 |
+
"step": 2615
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.39,
|
| 3681 |
+
"grad_norm": 0.349609375,
|
| 3682 |
+
"learning_rate": 0.00015328470304913833,
|
| 3683 |
+
"loss": 0.9729,
|
| 3684 |
+
"step": 2620
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.39,
|
| 3688 |
+
"grad_norm": 0.330078125,
|
| 3689 |
+
"learning_rate": 0.00015306518005719157,
|
| 3690 |
+
"loss": 0.9473,
|
| 3691 |
+
"step": 2625
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.39,
|
| 3695 |
+
"grad_norm": 0.3359375,
|
| 3696 |
+
"learning_rate": 0.00015284530052882045,
|
| 3697 |
+
"loss": 0.9268,
|
| 3698 |
+
"step": 2630
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.39,
|
| 3702 |
+
"grad_norm": 0.34765625,
|
| 3703 |
+
"learning_rate": 0.00015262506594136016,
|
| 3704 |
+
"loss": 0.9737,
|
| 3705 |
+
"step": 2635
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.39,
|
| 3709 |
+
"grad_norm": 0.34375,
|
| 3710 |
+
"learning_rate": 0.00015240447777453153,
|
| 3711 |
+
"loss": 0.9231,
|
| 3712 |
+
"step": 2640
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.39,
|
| 3716 |
+
"grad_norm": 0.345703125,
|
| 3717 |
+
"learning_rate": 0.00015218353751043107,
|
| 3718 |
+
"loss": 0.9359,
|
| 3719 |
+
"step": 2645
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.39,
|
| 3723 |
+
"grad_norm": 0.353515625,
|
| 3724 |
+
"learning_rate": 0.00015196224663352093,
|
| 3725 |
+
"loss": 0.9394,
|
| 3726 |
+
"step": 2650
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.39,
|
| 3730 |
+
"grad_norm": 0.3671875,
|
| 3731 |
+
"learning_rate": 0.00015174060663061898,
|
| 3732 |
+
"loss": 0.9556,
|
| 3733 |
+
"step": 2655
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.39,
|
| 3737 |
+
"grad_norm": 0.3359375,
|
| 3738 |
+
"learning_rate": 0.00015151861899088877,
|
| 3739 |
+
"loss": 0.9345,
|
| 3740 |
+
"step": 2660
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.4,
|
| 3744 |
+
"grad_norm": 0.357421875,
|
| 3745 |
+
"learning_rate": 0.0001512962852058297,
|
| 3746 |
+
"loss": 0.9363,
|
| 3747 |
+
"step": 2665
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.4,
|
| 3751 |
+
"grad_norm": 0.357421875,
|
| 3752 |
+
"learning_rate": 0.00015107360676926666,
|
| 3753 |
+
"loss": 0.9575,
|
| 3754 |
+
"step": 2670
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.4,
|
| 3758 |
+
"grad_norm": 0.33984375,
|
| 3759 |
+
"learning_rate": 0.00015085058517734043,
|
| 3760 |
+
"loss": 0.9824,
|
| 3761 |
+
"step": 2675
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.4,
|
| 3765 |
+
"grad_norm": 0.341796875,
|
| 3766 |
+
"learning_rate": 0.0001506272219284972,
|
| 3767 |
+
"loss": 0.95,
|
| 3768 |
+
"step": 2680
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.4,
|
| 3772 |
+
"grad_norm": 0.33984375,
|
| 3773 |
+
"learning_rate": 0.00015040351852347878,
|
| 3774 |
+
"loss": 0.9348,
|
| 3775 |
+
"step": 2685
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.4,
|
| 3779 |
+
"grad_norm": 0.34765625,
|
| 3780 |
+
"learning_rate": 0.0001501794764653124,
|
| 3781 |
+
"loss": 0.9585,
|
| 3782 |
+
"step": 2690
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.4,
|
| 3786 |
+
"grad_norm": 0.345703125,
|
| 3787 |
+
"learning_rate": 0.00014995509725930078,
|
| 3788 |
+
"loss": 0.9342,
|
| 3789 |
+
"step": 2695
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.4,
|
| 3793 |
+
"grad_norm": 0.34375,
|
| 3794 |
+
"learning_rate": 0.0001497303824130117,
|
| 3795 |
+
"loss": 0.9266,
|
| 3796 |
+
"step": 2700
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.4,
|
| 3800 |
+
"grad_norm": 0.33984375,
|
| 3801 |
+
"learning_rate": 0.00014950533343626812,
|
| 3802 |
+
"loss": 0.9552,
|
| 3803 |
+
"step": 2705
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.4,
|
| 3807 |
+
"grad_norm": 0.33984375,
|
| 3808 |
+
"learning_rate": 0.000149279951841138,
|
| 3809 |
+
"loss": 0.9514,
|
| 3810 |
+
"step": 2710
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.4,
|
| 3814 |
+
"grad_norm": 0.341796875,
|
| 3815 |
+
"learning_rate": 0.00014905423914192412,
|
| 3816 |
+
"loss": 0.9697,
|
| 3817 |
+
"step": 2715
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.4,
|
| 3821 |
+
"grad_norm": 0.34375,
|
| 3822 |
+
"learning_rate": 0.0001488281968551538,
|
| 3823 |
+
"loss": 0.9494,
|
| 3824 |
+
"step": 2720
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.4,
|
| 3828 |
+
"grad_norm": 0.357421875,
|
| 3829 |
+
"learning_rate": 0.00014860182649956892,
|
| 3830 |
+
"loss": 0.9667,
|
| 3831 |
+
"step": 2725
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.41,
|
| 3835 |
+
"grad_norm": 0.35546875,
|
| 3836 |
+
"learning_rate": 0.0001483751295961156,
|
| 3837 |
+
"loss": 0.9504,
|
| 3838 |
+
"step": 2730
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.41,
|
| 3842 |
+
"grad_norm": 0.3515625,
|
| 3843 |
+
"learning_rate": 0.0001481481076679338,
|
| 3844 |
+
"loss": 0.953,
|
| 3845 |
+
"step": 2735
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.41,
|
| 3849 |
+
"grad_norm": 0.34375,
|
| 3850 |
+
"learning_rate": 0.00014792076224034753,
|
| 3851 |
+
"loss": 0.9548,
|
| 3852 |
+
"step": 2740
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.41,
|
| 3856 |
+
"grad_norm": 0.34375,
|
| 3857 |
+
"learning_rate": 0.00014769309484085412,
|
| 3858 |
+
"loss": 0.9632,
|
| 3859 |
+
"step": 2745
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.41,
|
| 3863 |
+
"grad_norm": 0.33984375,
|
| 3864 |
+
"learning_rate": 0.00014746510699911432,
|
| 3865 |
+
"loss": 0.9619,
|
| 3866 |
+
"step": 2750
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.41,
|
| 3870 |
+
"grad_norm": 0.34375,
|
| 3871 |
+
"learning_rate": 0.00014723680024694184,
|
| 3872 |
+
"loss": 0.9155,
|
| 3873 |
+
"step": 2755
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.41,
|
| 3877 |
+
"grad_norm": 0.353515625,
|
| 3878 |
+
"learning_rate": 0.00014700817611829308,
|
| 3879 |
+
"loss": 0.9234,
|
| 3880 |
+
"step": 2760
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.41,
|
| 3884 |
+
"grad_norm": 0.337890625,
|
| 3885 |
+
"learning_rate": 0.00014677923614925685,
|
| 3886 |
+
"loss": 0.934,
|
| 3887 |
+
"step": 2765
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 0.41,
|
| 3891 |
+
"grad_norm": 0.3359375,
|
| 3892 |
+
"learning_rate": 0.000146549981878044,
|
| 3893 |
+
"loss": 0.9354,
|
| 3894 |
+
"step": 2770
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 0.41,
|
| 3898 |
+
"grad_norm": 0.333984375,
|
| 3899 |
+
"learning_rate": 0.00014632041484497727,
|
| 3900 |
+
"loss": 0.9554,
|
| 3901 |
+
"step": 2775
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 0.41,
|
| 3905 |
+
"grad_norm": 0.353515625,
|
| 3906 |
+
"learning_rate": 0.00014609053659248058,
|
| 3907 |
+
"loss": 0.9443,
|
| 3908 |
+
"step": 2780
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 0.41,
|
| 3912 |
+
"grad_norm": 0.341796875,
|
| 3913 |
+
"learning_rate": 0.00014586034866506906,
|
| 3914 |
+
"loss": 0.9584,
|
| 3915 |
+
"step": 2785
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 0.41,
|
| 3919 |
+
"grad_norm": 0.34375,
|
| 3920 |
+
"learning_rate": 0.00014562985260933845,
|
| 3921 |
+
"loss": 0.9035,
|
| 3922 |
+
"step": 2790
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 0.42,
|
| 3926 |
+
"grad_norm": 0.34765625,
|
| 3927 |
+
"learning_rate": 0.00014539904997395468,
|
| 3928 |
+
"loss": 0.9215,
|
| 3929 |
+
"step": 2795
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.42,
|
| 3933 |
+
"grad_norm": 0.341796875,
|
| 3934 |
+
"learning_rate": 0.00014516794230964365,
|
| 3935 |
+
"loss": 0.9279,
|
| 3936 |
+
"step": 2800
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 0.42,
|
| 3940 |
+
"grad_norm": 0.349609375,
|
| 3941 |
+
"learning_rate": 0.00014493653116918066,
|
| 3942 |
+
"loss": 0.9395,
|
| 3943 |
+
"step": 2805
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.42,
|
| 3947 |
+
"grad_norm": 0.341796875,
|
| 3948 |
+
"learning_rate": 0.0001447048181073799,
|
| 3949 |
+
"loss": 0.9314,
|
| 3950 |
+
"step": 2810
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.42,
|
| 3954 |
+
"grad_norm": 0.3359375,
|
| 3955 |
+
"learning_rate": 0.00014447280468108436,
|
| 3956 |
+
"loss": 0.9497,
|
| 3957 |
+
"step": 2815
|
| 3958 |
+
},
|
| 3959 |
+
{
|
| 3960 |
+
"epoch": 0.42,
|
| 3961 |
+
"grad_norm": 0.345703125,
|
| 3962 |
+
"learning_rate": 0.00014424049244915493,
|
| 3963 |
+
"loss": 0.9324,
|
| 3964 |
+
"step": 2820
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 0.42,
|
| 3968 |
+
"grad_norm": 0.3515625,
|
| 3969 |
+
"learning_rate": 0.00014400788297246024,
|
| 3970 |
+
"loss": 0.9351,
|
| 3971 |
+
"step": 2825
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 0.42,
|
| 3975 |
+
"grad_norm": 0.345703125,
|
| 3976 |
+
"learning_rate": 0.0001437749778138659,
|
| 3977 |
+
"loss": 0.9482,
|
| 3978 |
+
"step": 2830
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 0.42,
|
| 3982 |
+
"grad_norm": 0.34375,
|
| 3983 |
+
"learning_rate": 0.00014354177853822443,
|
| 3984 |
+
"loss": 0.938,
|
| 3985 |
+
"step": 2835
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.42,
|
| 3989 |
+
"grad_norm": 0.326171875,
|
| 3990 |
+
"learning_rate": 0.00014330828671236425,
|
| 3991 |
+
"loss": 0.9478,
|
| 3992 |
+
"step": 2840
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 0.42,
|
| 3996 |
+
"grad_norm": 0.33984375,
|
| 3997 |
+
"learning_rate": 0.0001430745039050794,
|
| 3998 |
+
"loss": 0.9596,
|
| 3999 |
+
"step": 2845
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 0.42,
|
| 4003 |
+
"grad_norm": 0.34765625,
|
| 4004 |
+
"learning_rate": 0.00014284043168711906,
|
| 4005 |
+
"loss": 0.9186,
|
| 4006 |
+
"step": 2850
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 0.42,
|
| 4010 |
+
"grad_norm": 0.337890625,
|
| 4011 |
+
"learning_rate": 0.00014260607163117694,
|
| 4012 |
+
"loss": 0.9631,
|
| 4013 |
+
"step": 2855
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.42,
|
| 4017 |
+
"grad_norm": 0.341796875,
|
| 4018 |
+
"learning_rate": 0.00014237142531188055,
|
| 4019 |
+
"loss": 0.9411,
|
| 4020 |
+
"step": 2860
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 0.43,
|
| 4024 |
+
"grad_norm": 0.357421875,
|
| 4025 |
+
"learning_rate": 0.00014213649430578083,
|
| 4026 |
+
"loss": 0.9558,
|
| 4027 |
+
"step": 2865
|
| 4028 |
+
},
|
| 4029 |
+
{
|
| 4030 |
+
"epoch": 0.43,
|
| 4031 |
+
"grad_norm": 0.341796875,
|
| 4032 |
+
"learning_rate": 0.00014190128019134153,
|
| 4033 |
+
"loss": 0.9684,
|
| 4034 |
+
"step": 2870
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 0.43,
|
| 4038 |
+
"grad_norm": 0.33984375,
|
| 4039 |
+
"learning_rate": 0.00014166578454892853,
|
| 4040 |
+
"loss": 0.9226,
|
| 4041 |
+
"step": 2875
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 0.43,
|
| 4045 |
+
"grad_norm": 0.341796875,
|
| 4046 |
+
"learning_rate": 0.00014143000896079918,
|
| 4047 |
+
"loss": 0.9103,
|
| 4048 |
+
"step": 2880
|
| 4049 |
+
},
|
| 4050 |
+
{
|
| 4051 |
+
"epoch": 0.43,
|
| 4052 |
+
"grad_norm": 0.337890625,
|
| 4053 |
+
"learning_rate": 0.00014119395501109182,
|
| 4054 |
+
"loss": 0.9262,
|
| 4055 |
+
"step": 2885
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 0.43,
|
| 4059 |
+
"grad_norm": 0.349609375,
|
| 4060 |
+
"learning_rate": 0.00014095762428581506,
|
| 4061 |
+
"loss": 0.9444,
|
| 4062 |
+
"step": 2890
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 0.43,
|
| 4066 |
+
"grad_norm": 0.349609375,
|
| 4067 |
+
"learning_rate": 0.0001407210183728371,
|
| 4068 |
+
"loss": 0.9438,
|
| 4069 |
+
"step": 2895
|
| 4070 |
+
},
|
| 4071 |
+
{
|
| 4072 |
+
"epoch": 0.43,
|
| 4073 |
+
"grad_norm": 0.357421875,
|
| 4074 |
+
"learning_rate": 0.00014048413886187503,
|
| 4075 |
+
"loss": 0.9699,
|
| 4076 |
+
"step": 2900
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 0.43,
|
| 4080 |
+
"grad_norm": 0.33203125,
|
| 4081 |
+
"learning_rate": 0.00014024698734448431,
|
| 4082 |
+
"loss": 0.9255,
|
| 4083 |
+
"step": 2905
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 0.43,
|
| 4087 |
+
"grad_norm": 0.361328125,
|
| 4088 |
+
"learning_rate": 0.00014000956541404785,
|
| 4089 |
+
"loss": 0.9733,
|
| 4090 |
+
"step": 2910
|
| 4091 |
+
},
|
| 4092 |
+
{
|
| 4093 |
+
"epoch": 0.43,
|
| 4094 |
+
"grad_norm": 0.353515625,
|
| 4095 |
+
"learning_rate": 0.0001397718746657655,
|
| 4096 |
+
"loss": 0.9554,
|
| 4097 |
+
"step": 2915
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 0.43,
|
| 4101 |
+
"grad_norm": 0.34765625,
|
| 4102 |
+
"learning_rate": 0.0001395339166966433,
|
| 4103 |
+
"loss": 0.9273,
|
| 4104 |
+
"step": 2920
|
| 4105 |
+
},
|
| 4106 |
+
{
|
| 4107 |
+
"epoch": 0.43,
|
| 4108 |
+
"grad_norm": 0.34375,
|
| 4109 |
+
"learning_rate": 0.0001392956931054825,
|
| 4110 |
+
"loss": 0.9364,
|
| 4111 |
+
"step": 2925
|
| 4112 |
+
},
|
| 4113 |
+
{
|
| 4114 |
+
"epoch": 0.44,
|
| 4115 |
+
"grad_norm": 0.3359375,
|
| 4116 |
+
"learning_rate": 0.00013905720549286932,
|
| 4117 |
+
"loss": 0.9433,
|
| 4118 |
+
"step": 2930
|
| 4119 |
+
},
|
| 4120 |
+
{
|
| 4121 |
+
"epoch": 0.44,
|
| 4122 |
+
"grad_norm": 0.341796875,
|
| 4123 |
+
"learning_rate": 0.0001388184554611636,
|
| 4124 |
+
"loss": 0.9288,
|
| 4125 |
+
"step": 2935
|
| 4126 |
+
},
|
| 4127 |
+
{
|
| 4128 |
+
"epoch": 0.44,
|
| 4129 |
+
"grad_norm": 0.34375,
|
| 4130 |
+
"learning_rate": 0.0001385794446144885,
|
| 4131 |
+
"loss": 0.9338,
|
| 4132 |
+
"step": 2940
|
| 4133 |
+
},
|
| 4134 |
+
{
|
| 4135 |
+
"epoch": 0.44,
|
| 4136 |
+
"grad_norm": 0.34765625,
|
| 4137 |
+
"learning_rate": 0.0001383401745587196,
|
| 4138 |
+
"loss": 1.0093,
|
| 4139 |
+
"step": 2945
|
| 4140 |
+
},
|
| 4141 |
+
{
|
| 4142 |
+
"epoch": 0.44,
|
| 4143 |
+
"grad_norm": 0.337890625,
|
| 4144 |
+
"learning_rate": 0.00013810064690147387,
|
| 4145 |
+
"loss": 0.924,
|
| 4146 |
+
"step": 2950
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 0.44,
|
| 4150 |
+
"grad_norm": 0.345703125,
|
| 4151 |
+
"learning_rate": 0.0001378608632520993,
|
| 4152 |
+
"loss": 0.9731,
|
| 4153 |
+
"step": 2955
|
| 4154 |
+
},
|
| 4155 |
+
{
|
| 4156 |
+
"epoch": 0.44,
|
| 4157 |
+
"grad_norm": 0.341796875,
|
| 4158 |
+
"learning_rate": 0.00013762082522166363,
|
| 4159 |
+
"loss": 0.9236,
|
| 4160 |
+
"step": 2960
|
| 4161 |
+
},
|
| 4162 |
+
{
|
| 4163 |
+
"epoch": 0.44,
|
| 4164 |
+
"grad_norm": 0.33984375,
|
| 4165 |
+
"learning_rate": 0.0001373805344229439,
|
| 4166 |
+
"loss": 0.9378,
|
| 4167 |
+
"step": 2965
|
| 4168 |
+
},
|
| 4169 |
+
{
|
| 4170 |
+
"epoch": 0.44,
|
| 4171 |
+
"grad_norm": 0.349609375,
|
| 4172 |
+
"learning_rate": 0.00013713999247041533,
|
| 4173 |
+
"loss": 0.929,
|
| 4174 |
+
"step": 2970
|
| 4175 |
+
},
|
| 4176 |
+
{
|
| 4177 |
+
"epoch": 0.44,
|
| 4178 |
+
"grad_norm": 0.353515625,
|
| 4179 |
+
"learning_rate": 0.00013689920098024078,
|
| 4180 |
+
"loss": 0.9338,
|
| 4181 |
+
"step": 2975
|
| 4182 |
+
},
|
| 4183 |
+
{
|
| 4184 |
+
"epoch": 0.44,
|
| 4185 |
+
"grad_norm": 0.34765625,
|
| 4186 |
+
"learning_rate": 0.0001366581615702596,
|
| 4187 |
+
"loss": 0.9437,
|
| 4188 |
+
"step": 2980
|
| 4189 |
+
},
|
| 4190 |
+
{
|
| 4191 |
+
"epoch": 0.44,
|
| 4192 |
+
"grad_norm": 0.345703125,
|
| 4193 |
+
"learning_rate": 0.00013641687585997677,
|
| 4194 |
+
"loss": 0.9331,
|
| 4195 |
+
"step": 2985
|
| 4196 |
+
},
|
| 4197 |
+
{
|
| 4198 |
+
"epoch": 0.44,
|
| 4199 |
+
"grad_norm": 0.34765625,
|
| 4200 |
+
"learning_rate": 0.00013617534547055236,
|
| 4201 |
+
"loss": 0.9655,
|
| 4202 |
+
"step": 2990
|
| 4203 |
+
},
|
| 4204 |
+
{
|
| 4205 |
+
"epoch": 0.44,
|
| 4206 |
+
"grad_norm": 0.341796875,
|
| 4207 |
+
"learning_rate": 0.0001359335720247902,
|
| 4208 |
+
"loss": 0.9384,
|
| 4209 |
+
"step": 2995
|
| 4210 |
+
},
|
| 4211 |
+
{
|
| 4212 |
+
"epoch": 0.45,
|
| 4213 |
+
"grad_norm": 0.3515625,
|
| 4214 |
+
"learning_rate": 0.0001356915571471273,
|
| 4215 |
+
"loss": 0.9497,
|
| 4216 |
+
"step": 3000
|
| 4217 |
+
},
|
| 4218 |
+
{
|
| 4219 |
+
"epoch": 0.45,
|
| 4220 |
+
"grad_norm": 0.353515625,
|
| 4221 |
+
"learning_rate": 0.0001354493024636227,
|
| 4222 |
+
"loss": 0.934,
|
| 4223 |
+
"step": 3005
|
| 4224 |
+
},
|
| 4225 |
+
{
|
| 4226 |
+
"epoch": 0.45,
|
| 4227 |
+
"grad_norm": 0.345703125,
|
| 4228 |
+
"learning_rate": 0.0001352068096019468,
|
| 4229 |
+
"loss": 0.9272,
|
| 4230 |
+
"step": 3010
|
| 4231 |
+
},
|
| 4232 |
+
{
|
| 4233 |
+
"epoch": 0.45,
|
| 4234 |
+
"grad_norm": 0.341796875,
|
| 4235 |
+
"learning_rate": 0.00013496408019137018,
|
| 4236 |
+
"loss": 0.9428,
|
| 4237 |
+
"step": 3015
|
| 4238 |
+
},
|
| 4239 |
+
{
|
| 4240 |
+
"epoch": 0.45,
|
| 4241 |
+
"grad_norm": 0.33203125,
|
| 4242 |
+
"learning_rate": 0.00013472111586275274,
|
| 4243 |
+
"loss": 0.9461,
|
| 4244 |
+
"step": 3020
|
| 4245 |
+
},
|
| 4246 |
+
{
|
| 4247 |
+
"epoch": 0.45,
|
| 4248 |
+
"grad_norm": 0.341796875,
|
| 4249 |
+
"learning_rate": 0.0001344779182485328,
|
| 4250 |
+
"loss": 0.9403,
|
| 4251 |
+
"step": 3025
|
| 4252 |
+
},
|
| 4253 |
+
{
|
| 4254 |
+
"epoch": 0.45,
|
| 4255 |
+
"grad_norm": 0.34375,
|
| 4256 |
+
"learning_rate": 0.0001342344889827161,
|
| 4257 |
+
"loss": 0.9324,
|
| 4258 |
+
"step": 3030
|
| 4259 |
+
},
|
| 4260 |
+
{
|
| 4261 |
+
"epoch": 0.45,
|
| 4262 |
+
"grad_norm": 0.345703125,
|
| 4263 |
+
"learning_rate": 0.0001339908297008648,
|
| 4264 |
+
"loss": 0.9427,
|
| 4265 |
+
"step": 3035
|
| 4266 |
+
},
|
| 4267 |
+
{
|
| 4268 |
+
"epoch": 0.45,
|
| 4269 |
+
"grad_norm": 0.345703125,
|
| 4270 |
+
"learning_rate": 0.00013374694204008647,
|
| 4271 |
+
"loss": 0.9369,
|
| 4272 |
+
"step": 3040
|
| 4273 |
+
},
|
| 4274 |
+
{
|
| 4275 |
+
"epoch": 0.45,
|
| 4276 |
+
"grad_norm": 0.33984375,
|
| 4277 |
+
"learning_rate": 0.00013350282763902315,
|
| 4278 |
+
"loss": 0.9251,
|
| 4279 |
+
"step": 3045
|
| 4280 |
+
},
|
| 4281 |
+
{
|
| 4282 |
+
"epoch": 0.45,
|
| 4283 |
+
"grad_norm": 0.33984375,
|
| 4284 |
+
"learning_rate": 0.0001332584881378403,
|
| 4285 |
+
"loss": 0.9365,
|
| 4286 |
+
"step": 3050
|
| 4287 |
+
},
|
| 4288 |
+
{
|
| 4289 |
+
"epoch": 0.45,
|
| 4290 |
+
"grad_norm": 0.345703125,
|
| 4291 |
+
"learning_rate": 0.00013301392517821577,
|
| 4292 |
+
"loss": 0.9184,
|
| 4293 |
+
"step": 3055
|
| 4294 |
+
},
|
| 4295 |
+
{
|
| 4296 |
+
"epoch": 0.45,
|
| 4297 |
+
"grad_norm": 0.3359375,
|
| 4298 |
+
"learning_rate": 0.00013276914040332889,
|
| 4299 |
+
"loss": 0.9231,
|
| 4300 |
+
"step": 3060
|
| 4301 |
+
},
|
| 4302 |
+
{
|
| 4303 |
+
"epoch": 0.46,
|
| 4304 |
+
"grad_norm": 0.349609375,
|
| 4305 |
+
"learning_rate": 0.0001325241354578492,
|
| 4306 |
+
"loss": 0.9217,
|
| 4307 |
+
"step": 3065
|
| 4308 |
+
},
|
| 4309 |
+
{
|
| 4310 |
+
"epoch": 0.46,
|
| 4311 |
+
"grad_norm": 0.341796875,
|
| 4312 |
+
"learning_rate": 0.0001322789119879256,
|
| 4313 |
+
"loss": 0.9712,
|
| 4314 |
+
"step": 3070
|
| 4315 |
+
},
|
| 4316 |
+
{
|
| 4317 |
+
"epoch": 0.46,
|
| 4318 |
+
"grad_norm": 0.3515625,
|
| 4319 |
+
"learning_rate": 0.00013203347164117524,
|
| 4320 |
+
"loss": 0.9239,
|
| 4321 |
+
"step": 3075
|
| 4322 |
+
},
|
| 4323 |
+
{
|
| 4324 |
+
"epoch": 0.46,
|
| 4325 |
+
"grad_norm": 0.341796875,
|
| 4326 |
+
"learning_rate": 0.00013178781606667234,
|
| 4327 |
+
"loss": 0.9567,
|
| 4328 |
+
"step": 3080
|
| 4329 |
+
},
|
| 4330 |
+
{
|
| 4331 |
+
"epoch": 0.46,
|
| 4332 |
+
"grad_norm": 0.34375,
|
| 4333 |
+
"learning_rate": 0.00013154194691493732,
|
| 4334 |
+
"loss": 0.9389,
|
| 4335 |
+
"step": 3085
|
| 4336 |
+
},
|
| 4337 |
+
{
|
| 4338 |
+
"epoch": 0.46,
|
| 4339 |
+
"grad_norm": 0.337890625,
|
| 4340 |
+
"learning_rate": 0.0001312958658379255,
|
| 4341 |
+
"loss": 0.9429,
|
| 4342 |
+
"step": 3090
|
| 4343 |
+
},
|
| 4344 |
+
{
|
| 4345 |
+
"epoch": 0.46,
|
| 4346 |
+
"grad_norm": 0.341796875,
|
| 4347 |
+
"learning_rate": 0.00013104957448901614,
|
| 4348 |
+
"loss": 0.9226,
|
| 4349 |
+
"step": 3095
|
| 4350 |
+
},
|
| 4351 |
+
{
|
| 4352 |
+
"epoch": 0.46,
|
| 4353 |
+
"grad_norm": 0.34375,
|
| 4354 |
+
"learning_rate": 0.00013080307452300127,
|
| 4355 |
+
"loss": 0.9155,
|
| 4356 |
+
"step": 3100
|
| 4357 |
+
},
|
| 4358 |
+
{
|
| 4359 |
+
"epoch": 0.46,
|
| 4360 |
+
"grad_norm": 0.34765625,
|
| 4361 |
+
"learning_rate": 0.00013055636759607458,
|
| 4362 |
+
"loss": 0.9516,
|
| 4363 |
+
"step": 3105
|
| 4364 |
+
},
|
| 4365 |
+
{
|
| 4366 |
+
"epoch": 0.46,
|
| 4367 |
+
"grad_norm": 0.337890625,
|
| 4368 |
+
"learning_rate": 0.00013030945536582025,
|
| 4369 |
+
"loss": 0.9284,
|
| 4370 |
+
"step": 3110
|
| 4371 |
+
},
|
| 4372 |
+
{
|
| 4373 |
+
"epoch": 0.46,
|
| 4374 |
+
"grad_norm": 0.33984375,
|
| 4375 |
+
"learning_rate": 0.00013006233949120199,
|
| 4376 |
+
"loss": 0.9113,
|
| 4377 |
+
"step": 3115
|
| 4378 |
+
},
|
| 4379 |
+
{
|
| 4380 |
+
"epoch": 0.46,
|
| 4381 |
+
"grad_norm": 0.341796875,
|
| 4382 |
+
"learning_rate": 0.00012981502163255166,
|
| 4383 |
+
"loss": 0.9471,
|
| 4384 |
+
"step": 3120
|
| 4385 |
+
},
|
| 4386 |
+
{
|
| 4387 |
+
"epoch": 0.46,
|
| 4388 |
+
"grad_norm": 0.34765625,
|
| 4389 |
+
"learning_rate": 0.0001295675034515582,
|
| 4390 |
+
"loss": 0.9416,
|
| 4391 |
+
"step": 3125
|
| 4392 |
+
},
|
| 4393 |
+
{
|
| 4394 |
+
"epoch": 0.46,
|
| 4395 |
+
"grad_norm": 0.359375,
|
| 4396 |
+
"learning_rate": 0.00012931978661125655,
|
| 4397 |
+
"loss": 0.9395,
|
| 4398 |
+
"step": 3130
|
| 4399 |
+
},
|
| 4400 |
+
{
|
| 4401 |
+
"epoch": 0.47,
|
| 4402 |
+
"grad_norm": 0.361328125,
|
| 4403 |
+
"learning_rate": 0.00012907187277601641,
|
| 4404 |
+
"loss": 0.9367,
|
| 4405 |
+
"step": 3135
|
| 4406 |
+
},
|
| 4407 |
+
{
|
| 4408 |
+
"epoch": 0.47,
|
| 4409 |
+
"grad_norm": 0.35546875,
|
| 4410 |
+
"learning_rate": 0.00012882376361153102,
|
| 4411 |
+
"loss": 0.923,
|
| 4412 |
+
"step": 3140
|
| 4413 |
+
},
|
| 4414 |
+
{
|
| 4415 |
+
"epoch": 0.47,
|
| 4416 |
+
"grad_norm": 0.3515625,
|
| 4417 |
+
"learning_rate": 0.00012857546078480598,
|
| 4418 |
+
"loss": 0.9115,
|
| 4419 |
+
"step": 3145
|
| 4420 |
+
},
|
| 4421 |
+
{
|
| 4422 |
+
"epoch": 0.47,
|
| 4423 |
+
"grad_norm": 0.345703125,
|
| 4424 |
+
"learning_rate": 0.00012832696596414817,
|
| 4425 |
+
"loss": 0.9182,
|
| 4426 |
+
"step": 3150
|
| 4427 |
+
},
|
| 4428 |
+
{
|
| 4429 |
+
"epoch": 0.47,
|
| 4430 |
+
"grad_norm": 0.3359375,
|
| 4431 |
+
"learning_rate": 0.00012807828081915436,
|
| 4432 |
+
"loss": 0.913,
|
| 4433 |
+
"step": 3155
|
| 4434 |
+
},
|
| 4435 |
+
{
|
| 4436 |
+
"epoch": 0.47,
|
| 4437 |
+
"grad_norm": 0.34765625,
|
| 4438 |
+
"learning_rate": 0.0001278294070207001,
|
| 4439 |
+
"loss": 0.9438,
|
| 4440 |
+
"step": 3160
|
| 4441 |
+
},
|
| 4442 |
+
{
|
| 4443 |
+
"epoch": 0.47,
|
| 4444 |
+
"grad_norm": 0.35546875,
|
| 4445 |
+
"learning_rate": 0.0001275803462409285,
|
| 4446 |
+
"loss": 0.9487,
|
| 4447 |
+
"step": 3165
|
| 4448 |
+
},
|
| 4449 |
+
{
|
| 4450 |
+
"epoch": 0.47,
|
| 4451 |
+
"grad_norm": 0.349609375,
|
| 4452 |
+
"learning_rate": 0.00012733110015323898,
|
| 4453 |
+
"loss": 0.9521,
|
| 4454 |
+
"step": 3170
|
| 4455 |
+
},
|
| 4456 |
+
{
|
| 4457 |
+
"epoch": 0.47,
|
| 4458 |
+
"grad_norm": 0.333984375,
|
| 4459 |
+
"learning_rate": 0.0001270816704322759,
|
| 4460 |
+
"loss": 0.9146,
|
| 4461 |
+
"step": 3175
|
| 4462 |
+
},
|
| 4463 |
+
{
|
| 4464 |
+
"epoch": 0.47,
|
| 4465 |
+
"grad_norm": 0.34765625,
|
| 4466 |
+
"learning_rate": 0.00012683205875391754,
|
| 4467 |
+
"loss": 0.9211,
|
| 4468 |
+
"step": 3180
|
| 4469 |
+
},
|
| 4470 |
+
{
|
| 4471 |
+
"epoch": 0.47,
|
| 4472 |
+
"grad_norm": 0.3515625,
|
| 4473 |
+
"learning_rate": 0.00012658226679526476,
|
| 4474 |
+
"loss": 0.9297,
|
| 4475 |
+
"step": 3185
|
| 4476 |
+
},
|
| 4477 |
+
{
|
| 4478 |
+
"epoch": 0.47,
|
| 4479 |
+
"grad_norm": 0.34375,
|
| 4480 |
+
"learning_rate": 0.00012633229623462951,
|
| 4481 |
+
"loss": 0.9186,
|
| 4482 |
+
"step": 3190
|
| 4483 |
+
},
|
| 4484 |
+
{
|
| 4485 |
+
"epoch": 0.47,
|
| 4486 |
+
"grad_norm": 0.35546875,
|
| 4487 |
+
"learning_rate": 0.00012608214875152392,
|
| 4488 |
+
"loss": 0.9443,
|
| 4489 |
+
"step": 3195
|
| 4490 |
+
},
|
| 4491 |
+
{
|
| 4492 |
+
"epoch": 0.48,
|
| 4493 |
+
"grad_norm": 0.341796875,
|
| 4494 |
+
"learning_rate": 0.00012583182602664877,
|
| 4495 |
+
"loss": 0.9399,
|
| 4496 |
+
"step": 3200
|
| 4497 |
+
},
|
| 4498 |
+
{
|
| 4499 |
+
"epoch": 0.48,
|
| 4500 |
+
"grad_norm": 0.337890625,
|
| 4501 |
+
"learning_rate": 0.00012558132974188223,
|
| 4502 |
+
"loss": 0.9342,
|
| 4503 |
+
"step": 3205
|
| 4504 |
+
},
|
| 4505 |
+
{
|
| 4506 |
+
"epoch": 0.48,
|
| 4507 |
+
"grad_norm": 0.330078125,
|
| 4508 |
+
"learning_rate": 0.00012533066158026862,
|
| 4509 |
+
"loss": 0.9088,
|
| 4510 |
+
"step": 3210
|
| 4511 |
+
},
|
| 4512 |
+
{
|
| 4513 |
+
"epoch": 0.48,
|
| 4514 |
+
"grad_norm": 0.3515625,
|
| 4515 |
+
"learning_rate": 0.00012507982322600703,
|
| 4516 |
+
"loss": 0.9777,
|
| 4517 |
+
"step": 3215
|
| 4518 |
+
},
|
| 4519 |
+
{
|
| 4520 |
+
"epoch": 0.48,
|
| 4521 |
+
"grad_norm": 0.33984375,
|
| 4522 |
+
"learning_rate": 0.00012482881636444014,
|
| 4523 |
+
"loss": 0.944,
|
| 4524 |
+
"step": 3220
|
| 4525 |
+
},
|
| 4526 |
+
{
|
| 4527 |
+
"epoch": 0.48,
|
| 4528 |
+
"grad_norm": 0.33203125,
|
| 4529 |
+
"learning_rate": 0.00012457764268204277,
|
| 4530 |
+
"loss": 0.9307,
|
| 4531 |
+
"step": 3225
|
| 4532 |
+
},
|
| 4533 |
+
{
|
| 4534 |
+
"epoch": 0.48,
|
| 4535 |
+
"grad_norm": 0.34375,
|
| 4536 |
+
"learning_rate": 0.0001243263038664105,
|
| 4537 |
+
"loss": 0.9396,
|
| 4538 |
+
"step": 3230
|
| 4539 |
+
},
|
| 4540 |
+
{
|
| 4541 |
+
"epoch": 0.48,
|
| 4542 |
+
"grad_norm": 0.34375,
|
| 4543 |
+
"learning_rate": 0.00012407480160624848,
|
| 4544 |
+
"loss": 0.9138,
|
| 4545 |
+
"step": 3235
|
| 4546 |
+
},
|
| 4547 |
+
{
|
| 4548 |
+
"epoch": 0.48,
|
| 4549 |
+
"grad_norm": 0.345703125,
|
| 4550 |
+
"learning_rate": 0.0001238231375913601,
|
| 4551 |
+
"loss": 0.9583,
|
| 4552 |
+
"step": 3240
|
| 4553 |
+
},
|
| 4554 |
+
{
|
| 4555 |
+
"epoch": 0.48,
|
| 4556 |
+
"grad_norm": 0.3515625,
|
| 4557 |
+
"learning_rate": 0.00012357131351263537,
|
| 4558 |
+
"loss": 0.9799,
|
| 4559 |
+
"step": 3245
|
| 4560 |
+
},
|
| 4561 |
+
{
|
| 4562 |
+
"epoch": 0.48,
|
| 4563 |
+
"grad_norm": 0.35546875,
|
| 4564 |
+
"learning_rate": 0.00012331933106203986,
|
| 4565 |
+
"loss": 0.9619,
|
| 4566 |
+
"step": 3250
|
| 4567 |
+
},
|
| 4568 |
+
{
|
| 4569 |
+
"epoch": 0.48,
|
| 4570 |
+
"grad_norm": 0.349609375,
|
| 4571 |
+
"learning_rate": 0.00012306719193260323,
|
| 4572 |
+
"loss": 0.9188,
|
| 4573 |
+
"step": 3255
|
| 4574 |
+
},
|
| 4575 |
+
{
|
| 4576 |
+
"epoch": 0.48,
|
| 4577 |
+
"grad_norm": 0.35546875,
|
| 4578 |
+
"learning_rate": 0.00012281489781840781,
|
| 4579 |
+
"loss": 0.9593,
|
| 4580 |
+
"step": 3260
|
| 4581 |
+
},
|
| 4582 |
+
{
|
| 4583 |
+
"epoch": 0.48,
|
| 4584 |
+
"grad_norm": 0.349609375,
|
| 4585 |
+
"learning_rate": 0.0001225624504145772,
|
| 4586 |
+
"loss": 0.938,
|
| 4587 |
+
"step": 3265
|
| 4588 |
+
},
|
| 4589 |
+
{
|
| 4590 |
+
"epoch": 0.49,
|
| 4591 |
+
"grad_norm": 0.349609375,
|
| 4592 |
+
"learning_rate": 0.00012230985141726498,
|
| 4593 |
+
"loss": 0.9237,
|
| 4594 |
+
"step": 3270
|
| 4595 |
+
},
|
| 4596 |
+
{
|
| 4597 |
+
"epoch": 0.49,
|
| 4598 |
+
"grad_norm": 0.357421875,
|
| 4599 |
+
"learning_rate": 0.00012205710252364329,
|
| 4600 |
+
"loss": 0.9314,
|
| 4601 |
+
"step": 3275
|
| 4602 |
+
},
|
| 4603 |
+
{
|
| 4604 |
+
"epoch": 0.49,
|
| 4605 |
+
"grad_norm": 0.34765625,
|
| 4606 |
+
"learning_rate": 0.00012180420543189131,
|
| 4607 |
+
"loss": 0.9093,
|
| 4608 |
+
"step": 3280
|
| 4609 |
+
},
|
| 4610 |
+
{
|
| 4611 |
+
"epoch": 0.49,
|
| 4612 |
+
"grad_norm": 0.349609375,
|
| 4613 |
+
"learning_rate": 0.00012155116184118402,
|
| 4614 |
+
"loss": 0.9335,
|
| 4615 |
+
"step": 3285
|
| 4616 |
+
},
|
| 4617 |
+
{
|
| 4618 |
+
"epoch": 0.49,
|
| 4619 |
+
"grad_norm": 0.341796875,
|
| 4620 |
+
"learning_rate": 0.00012129797345168073,
|
| 4621 |
+
"loss": 0.9643,
|
| 4622 |
+
"step": 3290
|
| 4623 |
+
},
|
| 4624 |
+
{
|
| 4625 |
+
"epoch": 0.49,
|
| 4626 |
+
"grad_norm": 0.341796875,
|
| 4627 |
+
"learning_rate": 0.00012104464196451353,
|
| 4628 |
+
"loss": 0.9507,
|
| 4629 |
+
"step": 3295
|
| 4630 |
+
},
|
| 4631 |
+
{
|
| 4632 |
+
"epoch": 0.49,
|
| 4633 |
+
"grad_norm": 0.34375,
|
| 4634 |
+
"learning_rate": 0.00012079116908177593,
|
| 4635 |
+
"loss": 0.9201,
|
| 4636 |
+
"step": 3300
|
| 4637 |
}
|
| 4638 |
],
|
| 4639 |
"logging_steps": 5,
|
|
|
|
| 4641 |
"num_input_tokens_seen": 0,
|
| 4642 |
"num_train_epochs": 1,
|
| 4643 |
"save_steps": 100,
|
| 4644 |
+
"total_flos": 4.640654868195836e+18,
|
| 4645 |
"train_batch_size": 2,
|
| 4646 |
"trial_name": null,
|
| 4647 |
"trial_params": null
|