Training in progress, epoch 1
Browse files- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- trainer_state.json +2186 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,10 +20,10 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"q_proj",
|
| 24 |
-
"o_proj",
|
| 25 |
"v_proj",
|
| 26 |
-
"k_proj"
|
|
|
|
|
|
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"v_proj",
|
| 24 |
+
"k_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
+
"q_proj"
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9034480
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d73fa1270f6e1d13138d6364238deeb7d555e149caedb8009b3dc6b4d539f71b
|
| 3 |
size 9034480
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2186 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3080,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.032467532467532464,
|
| 13 |
+
"grad_norm": 0.56201171875,
|
| 14 |
+
"learning_rate": 0.00019999479806942977,
|
| 15 |
+
"loss": 0.8091,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06493506493506493,
|
| 20 |
+
"grad_norm": 0.406005859375,
|
| 21 |
+
"learning_rate": 0.00019997919281892067,
|
| 22 |
+
"loss": 0.7613,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09740259740259741,
|
| 27 |
+
"grad_norm": 0.4560546875,
|
| 28 |
+
"learning_rate": 0.0001999531858720213,
|
| 29 |
+
"loss": 0.797,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12987012987012986,
|
| 34 |
+
"grad_norm": 0.424560546875,
|
| 35 |
+
"learning_rate": 0.0001999167799344583,
|
| 36 |
+
"loss": 0.7912,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16233766233766234,
|
| 41 |
+
"grad_norm": 0.468994140625,
|
| 42 |
+
"learning_rate": 0.00019986997879385487,
|
| 43 |
+
"loss": 0.7866,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19480519480519481,
|
| 48 |
+
"grad_norm": 0.38671875,
|
| 49 |
+
"learning_rate": 0.0001998127873193367,
|
| 50 |
+
"loss": 0.7504,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.22727272727272727,
|
| 55 |
+
"grad_norm": 0.389892578125,
|
| 56 |
+
"learning_rate": 0.00019974521146102537,
|
| 57 |
+
"loss": 0.8146,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2597402597402597,
|
| 62 |
+
"grad_norm": 0.39599609375,
|
| 63 |
+
"learning_rate": 0.00019966725824941932,
|
| 64 |
+
"loss": 0.7641,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2922077922077922,
|
| 69 |
+
"grad_norm": 0.421875,
|
| 70 |
+
"learning_rate": 0.00019957893579466252,
|
| 71 |
+
"loss": 0.7902,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3246753246753247,
|
| 76 |
+
"grad_norm": 0.3681640625,
|
| 77 |
+
"learning_rate": 0.00019948025328570042,
|
| 78 |
+
"loss": 0.7831,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.35714285714285715,
|
| 83 |
+
"grad_norm": 0.4013671875,
|
| 84 |
+
"learning_rate": 0.00019937122098932428,
|
| 85 |
+
"loss": 0.7268,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.38961038961038963,
|
| 90 |
+
"grad_norm": 0.41748046875,
|
| 91 |
+
"learning_rate": 0.00019925185024910277,
|
| 92 |
+
"loss": 0.7673,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.42207792207792205,
|
| 97 |
+
"grad_norm": 0.423583984375,
|
| 98 |
+
"learning_rate": 0.000199122153484202,
|
| 99 |
+
"loss": 0.7652,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.45454545454545453,
|
| 104 |
+
"grad_norm": 0.428955078125,
|
| 105 |
+
"learning_rate": 0.0001989821441880933,
|
| 106 |
+
"loss": 0.7806,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.487012987012987,
|
| 111 |
+
"grad_norm": 0.387939453125,
|
| 112 |
+
"learning_rate": 0.00019883183692714936,
|
| 113 |
+
"loss": 0.7474,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5194805194805194,
|
| 118 |
+
"grad_norm": 0.428955078125,
|
| 119 |
+
"learning_rate": 0.0001986712473391289,
|
| 120 |
+
"loss": 0.752,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.551948051948052,
|
| 125 |
+
"grad_norm": 0.38330078125,
|
| 126 |
+
"learning_rate": 0.00019850039213154973,
|
| 127 |
+
"loss": 0.785,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5844155844155844,
|
| 132 |
+
"grad_norm": 0.44677734375,
|
| 133 |
+
"learning_rate": 0.0001983192890799503,
|
| 134 |
+
"loss": 0.7612,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6168831168831169,
|
| 139 |
+
"grad_norm": 0.406982421875,
|
| 140 |
+
"learning_rate": 0.00019812795702604073,
|
| 141 |
+
"loss": 0.784,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6493506493506493,
|
| 146 |
+
"grad_norm": 0.410400390625,
|
| 147 |
+
"learning_rate": 0.00019792641587574212,
|
| 148 |
+
"loss": 0.7649,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6818181818181818,
|
| 153 |
+
"grad_norm": 0.4072265625,
|
| 154 |
+
"learning_rate": 0.00019771468659711595,
|
| 155 |
+
"loss": 0.7543,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7142857142857143,
|
| 160 |
+
"grad_norm": 0.37744140625,
|
| 161 |
+
"learning_rate": 0.00019749279121818235,
|
| 162 |
+
"loss": 0.7209,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7467532467532467,
|
| 167 |
+
"grad_norm": 0.431396484375,
|
| 168 |
+
"learning_rate": 0.00019726075282462845,
|
| 169 |
+
"loss": 0.7532,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7792207792207793,
|
| 174 |
+
"grad_norm": 0.3720703125,
|
| 175 |
+
"learning_rate": 0.00019701859555740648,
|
| 176 |
+
"loss": 0.7781,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8116883116883117,
|
| 181 |
+
"grad_norm": 0.420654296875,
|
| 182 |
+
"learning_rate": 0.00019676634461022224,
|
| 183 |
+
"loss": 0.7613,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8441558441558441,
|
| 188 |
+
"grad_norm": 0.350830078125,
|
| 189 |
+
"learning_rate": 0.000196504026226914,
|
| 190 |
+
"loss": 0.7439,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8766233766233766,
|
| 195 |
+
"grad_norm": 0.40087890625,
|
| 196 |
+
"learning_rate": 0.00019623166769872213,
|
| 197 |
+
"loss": 0.7324,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9090909090909091,
|
| 202 |
+
"grad_norm": 0.3779296875,
|
| 203 |
+
"learning_rate": 0.00019594929736144976,
|
| 204 |
+
"loss": 0.7481,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9415584415584416,
|
| 209 |
+
"grad_norm": 0.3818359375,
|
| 210 |
+
"learning_rate": 0.00019565694459251458,
|
| 211 |
+
"loss": 0.7584,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.974025974025974,
|
| 216 |
+
"grad_norm": 0.4423828125,
|
| 217 |
+
"learning_rate": 0.00019535463980789277,
|
| 218 |
+
"loss": 0.7364,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.0064935064935066,
|
| 223 |
+
"grad_norm": 0.42138671875,
|
| 224 |
+
"learning_rate": 0.00019504241445895436,
|
| 225 |
+
"loss": 0.7284,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0389610389610389,
|
| 230 |
+
"grad_norm": 0.4599609375,
|
| 231 |
+
"learning_rate": 0.000194720301029191,
|
| 232 |
+
"loss": 0.7216,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0714285714285714,
|
| 237 |
+
"grad_norm": 0.450927734375,
|
| 238 |
+
"learning_rate": 0.00019438833303083678,
|
| 239 |
+
"loss": 0.6916,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.103896103896104,
|
| 244 |
+
"grad_norm": 0.387451171875,
|
| 245 |
+
"learning_rate": 0.00019404654500138117,
|
| 246 |
+
"loss": 0.7093,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.1363636363636362,
|
| 251 |
+
"grad_norm": 0.46435546875,
|
| 252 |
+
"learning_rate": 0.0001936949724999762,
|
| 253 |
+
"loss": 0.7334,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1688311688311688,
|
| 258 |
+
"grad_norm": 0.423583984375,
|
| 259 |
+
"learning_rate": 0.0001933336521037367,
|
| 260 |
+
"loss": 0.7308,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.2012987012987013,
|
| 265 |
+
"grad_norm": 0.377197265625,
|
| 266 |
+
"learning_rate": 0.00019296262140393498,
|
| 267 |
+
"loss": 0.7008,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.2337662337662338,
|
| 272 |
+
"grad_norm": 0.427001953125,
|
| 273 |
+
"learning_rate": 0.0001925819190020898,
|
| 274 |
+
"loss": 0.7344,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2662337662337662,
|
| 279 |
+
"grad_norm": 0.397216796875,
|
| 280 |
+
"learning_rate": 0.0001921915845059505,
|
| 281 |
+
"loss": 0.6947,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2987012987012987,
|
| 286 |
+
"grad_norm": 0.42626953125,
|
| 287 |
+
"learning_rate": 0.00019179165852537596,
|
| 288 |
+
"loss": 0.7058,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.3311688311688312,
|
| 293 |
+
"grad_norm": 0.467529296875,
|
| 294 |
+
"learning_rate": 0.00019138218266810986,
|
| 295 |
+
"loss": 0.7244,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3636363636363638,
|
| 300 |
+
"grad_norm": 0.457275390625,
|
| 301 |
+
"learning_rate": 0.00019096319953545185,
|
| 302 |
+
"loss": 0.7003,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.396103896103896,
|
| 307 |
+
"grad_norm": 0.4013671875,
|
| 308 |
+
"learning_rate": 0.0001905347527178252,
|
| 309 |
+
"loss": 0.6961,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4285714285714286,
|
| 314 |
+
"grad_norm": 0.458740234375,
|
| 315 |
+
"learning_rate": 0.0001900968867902419,
|
| 316 |
+
"loss": 0.7052,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4610389610389611,
|
| 321 |
+
"grad_norm": 0.426025390625,
|
| 322 |
+
"learning_rate": 0.00018964964730766508,
|
| 323 |
+
"loss": 0.7059,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4935064935064934,
|
| 328 |
+
"grad_norm": 0.49609375,
|
| 329 |
+
"learning_rate": 0.0001891930808002694,
|
| 330 |
+
"loss": 0.6921,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.525974025974026,
|
| 335 |
+
"grad_norm": 0.46044921875,
|
| 336 |
+
"learning_rate": 0.00018872723476860034,
|
| 337 |
+
"loss": 0.7038,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.5584415584415585,
|
| 342 |
+
"grad_norm": 0.4140625,
|
| 343 |
+
"learning_rate": 0.00018825215767863214,
|
| 344 |
+
"loss": 0.7459,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5909090909090908,
|
| 349 |
+
"grad_norm": 0.46728515625,
|
| 350 |
+
"learning_rate": 0.00018776789895672558,
|
| 351 |
+
"loss": 0.726,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.6233766233766234,
|
| 356 |
+
"grad_norm": 0.486328125,
|
| 357 |
+
"learning_rate": 0.00018727450898448563,
|
| 358 |
+
"loss": 0.7192,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.655844155844156,
|
| 363 |
+
"grad_norm": 0.490966796875,
|
| 364 |
+
"learning_rate": 0.00018677203909351988,
|
| 365 |
+
"loss": 0.7186,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6883116883116882,
|
| 370 |
+
"grad_norm": 0.4443359375,
|
| 371 |
+
"learning_rate": 0.00018626054156009806,
|
| 372 |
+
"loss": 0.7177,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7207792207792207,
|
| 377 |
+
"grad_norm": 0.509765625,
|
| 378 |
+
"learning_rate": 0.00018574006959971333,
|
| 379 |
+
"loss": 0.6966,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.7532467532467533,
|
| 384 |
+
"grad_norm": 0.467529296875,
|
| 385 |
+
"learning_rate": 0.00018521067736154568,
|
| 386 |
+
"loss": 0.6668,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7857142857142856,
|
| 391 |
+
"grad_norm": 0.5068359375,
|
| 392 |
+
"learning_rate": 0.00018467241992282843,
|
| 393 |
+
"loss": 0.6856,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8181818181818183,
|
| 398 |
+
"grad_norm": 0.433837890625,
|
| 399 |
+
"learning_rate": 0.00018412535328311814,
|
| 400 |
+
"loss": 0.7126,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.8506493506493507,
|
| 405 |
+
"grad_norm": 0.46826171875,
|
| 406 |
+
"learning_rate": 0.0001835695343584683,
|
| 407 |
+
"loss": 0.7255,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.883116883116883,
|
| 412 |
+
"grad_norm": 0.452880859375,
|
| 413 |
+
"learning_rate": 0.00018300502097550806,
|
| 414 |
+
"loss": 0.6792,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.9155844155844157,
|
| 419 |
+
"grad_norm": 0.46142578125,
|
| 420 |
+
"learning_rate": 0.00018243187186542593,
|
| 421 |
+
"loss": 0.6891,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.948051948051948,
|
| 426 |
+
"grad_norm": 0.51904296875,
|
| 427 |
+
"learning_rate": 0.00018185014665785936,
|
| 428 |
+
"loss": 0.6959,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9805194805194806,
|
| 433 |
+
"grad_norm": 0.4833984375,
|
| 434 |
+
"learning_rate": 0.00018125990587469123,
|
| 435 |
+
"loss": 0.6975,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.012987012987013,
|
| 440 |
+
"grad_norm": 0.41796875,
|
| 441 |
+
"learning_rate": 0.000180661210923753,
|
| 442 |
+
"loss": 0.6652,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.0454545454545454,
|
| 447 |
+
"grad_norm": 0.458984375,
|
| 448 |
+
"learning_rate": 0.00018005412409243606,
|
| 449 |
+
"loss": 0.6559,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.0779220779220777,
|
| 454 |
+
"grad_norm": 0.580078125,
|
| 455 |
+
"learning_rate": 0.00017943870854121124,
|
| 456 |
+
"loss": 0.6489,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.1103896103896105,
|
| 461 |
+
"grad_norm": 0.469970703125,
|
| 462 |
+
"learning_rate": 0.00017881502829705808,
|
| 463 |
+
"loss": 0.6409,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.142857142857143,
|
| 468 |
+
"grad_norm": 0.51611328125,
|
| 469 |
+
"learning_rate": 0.000178183148246803,
|
| 470 |
+
"loss": 0.6835,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.175324675324675,
|
| 475 |
+
"grad_norm": 0.52099609375,
|
| 476 |
+
"learning_rate": 0.00017754313413036906,
|
| 477 |
+
"loss": 0.66,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.207792207792208,
|
| 482 |
+
"grad_norm": 0.490966796875,
|
| 483 |
+
"learning_rate": 0.0001768950525339362,
|
| 484 |
+
"loss": 0.6767,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.24025974025974,
|
| 489 |
+
"grad_norm": 0.5166015625,
|
| 490 |
+
"learning_rate": 0.00017623897088301385,
|
| 491 |
+
"loss": 0.6568,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.2727272727272725,
|
| 496 |
+
"grad_norm": 0.64111328125,
|
| 497 |
+
"learning_rate": 0.00017557495743542585,
|
| 498 |
+
"loss": 0.6585,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.3051948051948052,
|
| 503 |
+
"grad_norm": 0.52001953125,
|
| 504 |
+
"learning_rate": 0.00017490308127420928,
|
| 505 |
+
"loss": 0.6728,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.3376623376623376,
|
| 510 |
+
"grad_norm": 0.55908203125,
|
| 511 |
+
"learning_rate": 0.000174223412300427,
|
| 512 |
+
"loss": 0.6574,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.3701298701298703,
|
| 517 |
+
"grad_norm": 0.53076171875,
|
| 518 |
+
"learning_rate": 0.00017353602122589527,
|
| 519 |
+
"loss": 0.6486,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.4025974025974026,
|
| 524 |
+
"grad_norm": 0.51513671875,
|
| 525 |
+
"learning_rate": 0.00017284097956582692,
|
| 526 |
+
"loss": 0.6571,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.435064935064935,
|
| 531 |
+
"grad_norm": 0.482177734375,
|
| 532 |
+
"learning_rate": 0.0001721383596313912,
|
| 533 |
+
"loss": 0.666,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.4675324675324677,
|
| 538 |
+
"grad_norm": 0.492919921875,
|
| 539 |
+
"learning_rate": 0.00017142823452219038,
|
| 540 |
+
"loss": 0.665,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.5,
|
| 545 |
+
"grad_norm": 0.54541015625,
|
| 546 |
+
"learning_rate": 0.00017071067811865476,
|
| 547 |
+
"loss": 0.654,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.5324675324675323,
|
| 552 |
+
"grad_norm": 0.5810546875,
|
| 553 |
+
"learning_rate": 0.00016998576507435618,
|
| 554 |
+
"loss": 0.6573,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.564935064935065,
|
| 559 |
+
"grad_norm": 0.5234375,
|
| 560 |
+
"learning_rate": 0.00016925357080824118,
|
| 561 |
+
"loss": 0.6627,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.5974025974025974,
|
| 566 |
+
"grad_norm": 0.5185546875,
|
| 567 |
+
"learning_rate": 0.00016851417149678444,
|
| 568 |
+
"loss": 0.651,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.62987012987013,
|
| 573 |
+
"grad_norm": 0.603515625,
|
| 574 |
+
"learning_rate": 0.0001677676440660636,
|
| 575 |
+
"loss": 0.6616,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.6623376623376624,
|
| 580 |
+
"grad_norm": 1.826171875,
|
| 581 |
+
"learning_rate": 0.00016701406618375596,
|
| 582 |
+
"loss": 0.6668,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.6948051948051948,
|
| 587 |
+
"grad_norm": 0.541015625,
|
| 588 |
+
"learning_rate": 0.00016625351625105796,
|
| 589 |
+
"loss": 0.6661,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.7272727272727275,
|
| 594 |
+
"grad_norm": 0.61328125,
|
| 595 |
+
"learning_rate": 0.00016548607339452853,
|
| 596 |
+
"loss": 0.6378,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.75974025974026,
|
| 601 |
+
"grad_norm": 0.52001953125,
|
| 602 |
+
"learning_rate": 0.00016471181745785672,
|
| 603 |
+
"loss": 0.6672,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.792207792207792,
|
| 608 |
+
"grad_norm": 0.5888671875,
|
| 609 |
+
"learning_rate": 0.00016393082899355516,
|
| 610 |
+
"loss": 0.6573,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.824675324675325,
|
| 615 |
+
"grad_norm": 0.5673828125,
|
| 616 |
+
"learning_rate": 0.0001631431892545791,
|
| 617 |
+
"loss": 0.6471,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.857142857142857,
|
| 622 |
+
"grad_norm": 0.69482421875,
|
| 623 |
+
"learning_rate": 0.00016234898018587337,
|
| 624 |
+
"loss": 0.6604,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.8896103896103895,
|
| 629 |
+
"grad_norm": 0.5029296875,
|
| 630 |
+
"learning_rate": 0.00016154828441584655,
|
| 631 |
+
"loss": 0.6535,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.9220779220779223,
|
| 636 |
+
"grad_norm": 0.49462890625,
|
| 637 |
+
"learning_rate": 0.00016074118524777477,
|
| 638 |
+
"loss": 0.647,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.9545454545454546,
|
| 643 |
+
"grad_norm": 0.55322265625,
|
| 644 |
+
"learning_rate": 0.0001599277666511347,
|
| 645 |
+
"loss": 0.66,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.987012987012987,
|
| 650 |
+
"grad_norm": 0.5390625,
|
| 651 |
+
"learning_rate": 0.00015910811325286768,
|
| 652 |
+
"loss": 0.6381,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.0194805194805197,
|
| 657 |
+
"grad_norm": 0.509765625,
|
| 658 |
+
"learning_rate": 0.00015828231032857503,
|
| 659 |
+
"loss": 0.6559,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.051948051948052,
|
| 664 |
+
"grad_norm": 0.55517578125,
|
| 665 |
+
"learning_rate": 0.00015745044379364634,
|
| 666 |
+
"loss": 0.5904,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.0844155844155843,
|
| 671 |
+
"grad_norm": 0.58642578125,
|
| 672 |
+
"learning_rate": 0.00015661260019432077,
|
| 673 |
+
"loss": 0.6012,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.116883116883117,
|
| 678 |
+
"grad_norm": 0.56640625,
|
| 679 |
+
"learning_rate": 0.00015576886669868296,
|
| 680 |
+
"loss": 0.6161,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.1493506493506493,
|
| 685 |
+
"grad_norm": 0.521484375,
|
| 686 |
+
"learning_rate": 0.0001549193310875942,
|
| 687 |
+
"loss": 0.6019,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.1818181818181817,
|
| 692 |
+
"grad_norm": 0.55224609375,
|
| 693 |
+
"learning_rate": 0.00015406408174555976,
|
| 694 |
+
"loss": 0.5906,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.2142857142857144,
|
| 699 |
+
"grad_norm": 0.61865234375,
|
| 700 |
+
"learning_rate": 0.00015320320765153367,
|
| 701 |
+
"loss": 0.6128,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.2467532467532467,
|
| 706 |
+
"grad_norm": 0.642578125,
|
| 707 |
+
"learning_rate": 0.00015233679836966122,
|
| 708 |
+
"loss": 0.6012,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.279220779220779,
|
| 713 |
+
"grad_norm": 0.60107421875,
|
| 714 |
+
"learning_rate": 0.000151464944039961,
|
| 715 |
+
"loss": 0.6152,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.311688311688312,
|
| 720 |
+
"grad_norm": 0.5966796875,
|
| 721 |
+
"learning_rate": 0.00015058773536894685,
|
| 722 |
+
"loss": 0.5867,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.344155844155844,
|
| 727 |
+
"grad_norm": 0.5986328125,
|
| 728 |
+
"learning_rate": 0.00014970526362019079,
|
| 729 |
+
"loss": 0.6248,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.3766233766233764,
|
| 734 |
+
"grad_norm": 0.60009765625,
|
| 735 |
+
"learning_rate": 0.00014881762060482814,
|
| 736 |
+
"loss": 0.6134,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.409090909090909,
|
| 741 |
+
"grad_norm": 0.5986328125,
|
| 742 |
+
"learning_rate": 0.0001479248986720057,
|
| 743 |
+
"loss": 0.6078,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.4415584415584415,
|
| 748 |
+
"grad_norm": 0.5732421875,
|
| 749 |
+
"learning_rate": 0.0001470271906992737,
|
| 750 |
+
"loss": 0.6028,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.474025974025974,
|
| 755 |
+
"grad_norm": 0.63623046875,
|
| 756 |
+
"learning_rate": 0.00014612459008292306,
|
| 757 |
+
"loss": 0.6176,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.5064935064935066,
|
| 762 |
+
"grad_norm": 0.6259765625,
|
| 763 |
+
"learning_rate": 0.00014521719072826858,
|
| 764 |
+
"loss": 0.6287,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.538961038961039,
|
| 769 |
+
"grad_norm": 0.5595703125,
|
| 770 |
+
"learning_rate": 0.00014430508703987914,
|
| 771 |
+
"loss": 0.6007,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.571428571428571,
|
| 776 |
+
"grad_norm": 0.62109375,
|
| 777 |
+
"learning_rate": 0.00014338837391175582,
|
| 778 |
+
"loss": 0.6112,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.603896103896104,
|
| 783 |
+
"grad_norm": 0.59375,
|
| 784 |
+
"learning_rate": 0.00014246714671745965,
|
| 785 |
+
"loss": 0.6273,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.6363636363636362,
|
| 790 |
+
"grad_norm": 0.64501953125,
|
| 791 |
+
"learning_rate": 0.00014154150130018866,
|
| 792 |
+
"loss": 0.5894,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.6688311688311686,
|
| 797 |
+
"grad_norm": 0.5341796875,
|
| 798 |
+
"learning_rate": 0.00014061153396280674,
|
| 799 |
+
"loss": 0.6315,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.7012987012987013,
|
| 804 |
+
"grad_norm": 0.7060546875,
|
| 805 |
+
"learning_rate": 0.00013967734145782425,
|
| 806 |
+
"loss": 0.6377,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.7337662337662336,
|
| 811 |
+
"grad_norm": 0.626953125,
|
| 812 |
+
"learning_rate": 0.0001387390209773323,
|
| 813 |
+
"loss": 0.6212,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.7662337662337664,
|
| 818 |
+
"grad_norm": 0.59765625,
|
| 819 |
+
"learning_rate": 0.00013779667014289065,
|
| 820 |
+
"loss": 0.6126,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.7987012987012987,
|
| 825 |
+
"grad_norm": 0.67431640625,
|
| 826 |
+
"learning_rate": 0.00013685038699537164,
|
| 827 |
+
"loss": 0.6255,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.8311688311688314,
|
| 832 |
+
"grad_norm": 0.6494140625,
|
| 833 |
+
"learning_rate": 0.00013590026998475986,
|
| 834 |
+
"loss": 0.6199,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.8636363636363638,
|
| 839 |
+
"grad_norm": 0.6630859375,
|
| 840 |
+
"learning_rate": 0.00013494641795990986,
|
| 841 |
+
"loss": 0.6109,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.896103896103896,
|
| 846 |
+
"grad_norm": 0.63525390625,
|
| 847 |
+
"learning_rate": 0.00013398893015826167,
|
| 848 |
+
"loss": 0.6359,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.928571428571429,
|
| 853 |
+
"grad_norm": 0.5849609375,
|
| 854 |
+
"learning_rate": 0.00013302790619551674,
|
| 855 |
+
"loss": 0.6523,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.961038961038961,
|
| 860 |
+
"grad_norm": 0.59033203125,
|
| 861 |
+
"learning_rate": 0.00013206344605527355,
|
| 862 |
+
"loss": 0.5958,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.9935064935064934,
|
| 867 |
+
"grad_norm": 0.62060546875,
|
| 868 |
+
"learning_rate": 0.00013109565007862596,
|
| 869 |
+
"loss": 0.6284,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.025974025974026,
|
| 874 |
+
"grad_norm": 0.76904296875,
|
| 875 |
+
"learning_rate": 0.00013012461895372344,
|
| 876 |
+
"loss": 0.5536,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.058441558441558,
|
| 881 |
+
"grad_norm": 0.65966796875,
|
| 882 |
+
"learning_rate": 0.00012915045370529585,
|
| 883 |
+
"loss": 0.5586,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.090909090909091,
|
| 888 |
+
"grad_norm": 0.64599609375,
|
| 889 |
+
"learning_rate": 0.00012817325568414297,
|
| 890 |
+
"loss": 0.581,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.123376623376624,
|
| 895 |
+
"grad_norm": 0.681640625,
|
| 896 |
+
"learning_rate": 0.00012719312655658994,
|
| 897 |
+
"loss": 0.545,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.1558441558441555,
|
| 902 |
+
"grad_norm": 0.6435546875,
|
| 903 |
+
"learning_rate": 0.00012621016829391022,
|
| 904 |
+
"loss": 0.5603,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.188311688311688,
|
| 909 |
+
"grad_norm": 0.7587890625,
|
| 910 |
+
"learning_rate": 0.0001252244831617165,
|
| 911 |
+
"loss": 0.557,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.220779220779221,
|
| 916 |
+
"grad_norm": 0.642578125,
|
| 917 |
+
"learning_rate": 0.00012423617370932127,
|
| 918 |
+
"loss": 0.5634,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.253246753246753,
|
| 923 |
+
"grad_norm": 0.7373046875,
|
| 924 |
+
"learning_rate": 0.00012324534275906748,
|
| 925 |
+
"loss": 0.567,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.285714285714286,
|
| 930 |
+
"grad_norm": 0.6796875,
|
| 931 |
+
"learning_rate": 0.00012225209339563145,
|
| 932 |
+
"loss": 0.554,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.318181818181818,
|
| 937 |
+
"grad_norm": 0.68310546875,
|
| 938 |
+
"learning_rate": 0.00012125652895529766,
|
| 939 |
+
"loss": 0.5777,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.35064935064935,
|
| 944 |
+
"grad_norm": 0.85205078125,
|
| 945 |
+
"learning_rate": 0.0001202587530152081,
|
| 946 |
+
"loss": 0.5824,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.383116883116883,
|
| 951 |
+
"grad_norm": 0.70751953125,
|
| 952 |
+
"learning_rate": 0.00011925886938258605,
|
| 953 |
+
"loss": 0.5551,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.415584415584416,
|
| 958 |
+
"grad_norm": 0.69140625,
|
| 959 |
+
"learning_rate": 0.00011825698208393619,
|
| 960 |
+
"loss": 0.574,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.448051948051948,
|
| 965 |
+
"grad_norm": 0.708984375,
|
| 966 |
+
"learning_rate": 0.00011725319535422188,
|
| 967 |
+
"loss": 0.5919,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.48051948051948,
|
| 972 |
+
"grad_norm": 0.6796875,
|
| 973 |
+
"learning_rate": 0.00011624761362602061,
|
| 974 |
+
"loss": 0.5592,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.512987012987013,
|
| 979 |
+
"grad_norm": 0.66455078125,
|
| 980 |
+
"learning_rate": 0.00011524034151865903,
|
| 981 |
+
"loss": 0.5882,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.545454545454545,
|
| 986 |
+
"grad_norm": 0.7119140625,
|
| 987 |
+
"learning_rate": 0.00011423148382732853,
|
| 988 |
+
"loss": 0.5741,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.577922077922078,
|
| 993 |
+
"grad_norm": 0.66748046875,
|
| 994 |
+
"learning_rate": 0.00011322114551218239,
|
| 995 |
+
"loss": 0.5842,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.6103896103896105,
|
| 1000 |
+
"grad_norm": 0.69873046875,
|
| 1001 |
+
"learning_rate": 0.000112209431687416,
|
| 1002 |
+
"loss": 0.5626,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.642857142857143,
|
| 1007 |
+
"grad_norm": 0.6357421875,
|
| 1008 |
+
"learning_rate": 0.00011119644761033078,
|
| 1009 |
+
"loss": 0.5657,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.675324675324675,
|
| 1014 |
+
"grad_norm": 0.677734375,
|
| 1015 |
+
"learning_rate": 0.00011018229867038356,
|
| 1016 |
+
"loss": 0.5664,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.707792207792208,
|
| 1021 |
+
"grad_norm": 0.7998046875,
|
| 1022 |
+
"learning_rate": 0.00010916709037822173,
|
| 1023 |
+
"loss": 0.5758,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.740259740259741,
|
| 1028 |
+
"grad_norm": 0.68212890625,
|
| 1029 |
+
"learning_rate": 0.00010815092835470633,
|
| 1030 |
+
"loss": 0.5785,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.7727272727272725,
|
| 1035 |
+
"grad_norm": 0.75341796875,
|
| 1036 |
+
"learning_rate": 0.00010713391831992323,
|
| 1037 |
+
"loss": 0.5646,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.805194805194805,
|
| 1042 |
+
"grad_norm": 0.65478515625,
|
| 1043 |
+
"learning_rate": 0.00010611616608218429,
|
| 1044 |
+
"loss": 0.577,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.837662337662338,
|
| 1049 |
+
"grad_norm": 0.7626953125,
|
| 1050 |
+
"learning_rate": 0.00010509777752701899,
|
| 1051 |
+
"loss": 0.5878,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.87012987012987,
|
| 1056 |
+
"grad_norm": 0.88623046875,
|
| 1057 |
+
"learning_rate": 0.00010407885860615859,
|
| 1058 |
+
"loss": 0.5931,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.902597402597403,
|
| 1063 |
+
"grad_norm": 0.80419921875,
|
| 1064 |
+
"learning_rate": 0.00010305951532651261,
|
| 1065 |
+
"loss": 0.5946,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.935064935064935,
|
| 1070 |
+
"grad_norm": 0.71630859375,
|
| 1071 |
+
"learning_rate": 0.00010203985373914056,
|
| 1072 |
+
"loss": 0.5671,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.967532467532467,
|
| 1077 |
+
"grad_norm": 0.73779296875,
|
| 1078 |
+
"learning_rate": 0.00010101997992821797,
|
| 1079 |
+
"loss": 0.5942,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.0,
|
| 1084 |
+
"grad_norm": 0.67333984375,
|
| 1085 |
+
"learning_rate": 0.0001,
|
| 1086 |
+
"loss": 0.5654,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.032467532467533,
|
| 1091 |
+
"grad_norm": 0.7333984375,
|
| 1092 |
+
"learning_rate": 9.898002007178204e-05,
|
| 1093 |
+
"loss": 0.5399,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.064935064935065,
|
| 1098 |
+
"grad_norm": 0.77001953125,
|
| 1099 |
+
"learning_rate": 9.79601462608595e-05,
|
| 1100 |
+
"loss": 0.4992,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.097402597402597,
|
| 1105 |
+
"grad_norm": 0.74267578125,
|
| 1106 |
+
"learning_rate": 9.69404846734874e-05,
|
| 1107 |
+
"loss": 0.5386,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.12987012987013,
|
| 1112 |
+
"grad_norm": 0.87060546875,
|
| 1113 |
+
"learning_rate": 9.592114139384145e-05,
|
| 1114 |
+
"loss": 0.5219,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.162337662337662,
|
| 1119 |
+
"grad_norm": 0.8505859375,
|
| 1120 |
+
"learning_rate": 9.490222247298099e-05,
|
| 1121 |
+
"loss": 0.5218,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.194805194805195,
|
| 1126 |
+
"grad_norm": 0.7724609375,
|
| 1127 |
+
"learning_rate": 9.388383391781575e-05,
|
| 1128 |
+
"loss": 0.5193,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.2272727272727275,
|
| 1133 |
+
"grad_norm": 0.8701171875,
|
| 1134 |
+
"learning_rate": 9.286608168007678e-05,
|
| 1135 |
+
"loss": 0.5356,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.259740259740259,
|
| 1140 |
+
"grad_norm": 0.7861328125,
|
| 1141 |
+
"learning_rate": 9.184907164529368e-05,
|
| 1142 |
+
"loss": 0.5276,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.292207792207792,
|
| 1147 |
+
"grad_norm": 0.86083984375,
|
| 1148 |
+
"learning_rate": 9.083290962177828e-05,
|
| 1149 |
+
"loss": 0.5385,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.324675324675325,
|
| 1154 |
+
"grad_norm": 0.8212890625,
|
| 1155 |
+
"learning_rate": 8.981770132961649e-05,
|
| 1156 |
+
"loss": 0.5315,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.357142857142857,
|
| 1161 |
+
"grad_norm": 0.77001953125,
|
| 1162 |
+
"learning_rate": 8.880355238966923e-05,
|
| 1163 |
+
"loss": 0.5144,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.3896103896103895,
|
| 1168 |
+
"grad_norm": 0.8388671875,
|
| 1169 |
+
"learning_rate": 8.779056831258402e-05,
|
| 1170 |
+
"loss": 0.557,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.422077922077922,
|
| 1175 |
+
"grad_norm": 0.84326171875,
|
| 1176 |
+
"learning_rate": 8.67788544878176e-05,
|
| 1177 |
+
"loss": 0.5406,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.454545454545454,
|
| 1182 |
+
"grad_norm": 0.9423828125,
|
| 1183 |
+
"learning_rate": 8.57685161726715e-05,
|
| 1184 |
+
"loss": 0.5483,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.487012987012987,
|
| 1189 |
+
"grad_norm": 0.7421875,
|
| 1190 |
+
"learning_rate": 8.475965848134099e-05,
|
| 1191 |
+
"loss": 0.5399,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.51948051948052,
|
| 1196 |
+
"grad_norm": 0.8623046875,
|
| 1197 |
+
"learning_rate": 8.375238637397942e-05,
|
| 1198 |
+
"loss": 0.5427,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.551948051948052,
|
| 1203 |
+
"grad_norm": 0.796875,
|
| 1204 |
+
"learning_rate": 8.274680464577816e-05,
|
| 1205 |
+
"loss": 0.4993,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.584415584415584,
|
| 1210 |
+
"grad_norm": 0.6953125,
|
| 1211 |
+
"learning_rate": 8.174301791606385e-05,
|
| 1212 |
+
"loss": 0.5258,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.616883116883117,
|
| 1217 |
+
"grad_norm": 0.81591796875,
|
| 1218 |
+
"learning_rate": 8.074113061741397e-05,
|
| 1219 |
+
"loss": 0.5342,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.64935064935065,
|
| 1224 |
+
"grad_norm": 0.8369140625,
|
| 1225 |
+
"learning_rate": 7.974124698479192e-05,
|
| 1226 |
+
"loss": 0.5295,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.681818181818182,
|
| 1231 |
+
"grad_norm": 0.77392578125,
|
| 1232 |
+
"learning_rate": 7.874347104470234e-05,
|
| 1233 |
+
"loss": 0.5291,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.714285714285714,
|
| 1238 |
+
"grad_norm": 1.908203125,
|
| 1239 |
+
"learning_rate": 7.774790660436858e-05,
|
| 1240 |
+
"loss": 0.5367,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.746753246753247,
|
| 1245 |
+
"grad_norm": 0.82861328125,
|
| 1246 |
+
"learning_rate": 7.675465724093254e-05,
|
| 1247 |
+
"loss": 0.5437,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.779220779220779,
|
| 1252 |
+
"grad_norm": 0.84228515625,
|
| 1253 |
+
"learning_rate": 7.576382629067877e-05,
|
| 1254 |
+
"loss": 0.5168,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.811688311688312,
|
| 1259 |
+
"grad_norm": 0.7587890625,
|
| 1260 |
+
"learning_rate": 7.47755168382835e-05,
|
| 1261 |
+
"loss": 0.5411,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.8441558441558445,
|
| 1266 |
+
"grad_norm": 0.77001953125,
|
| 1267 |
+
"learning_rate": 7.378983170608982e-05,
|
| 1268 |
+
"loss": 0.5443,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.876623376623376,
|
| 1273 |
+
"grad_norm": 0.75048828125,
|
| 1274 |
+
"learning_rate": 7.280687344341007e-05,
|
| 1275 |
+
"loss": 0.5365,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.909090909090909,
|
| 1280 |
+
"grad_norm": 0.7861328125,
|
| 1281 |
+
"learning_rate": 7.182674431585704e-05,
|
| 1282 |
+
"loss": 0.5274,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.941558441558442,
|
| 1287 |
+
"grad_norm": 0.83984375,
|
| 1288 |
+
"learning_rate": 7.084954629470417e-05,
|
| 1289 |
+
"loss": 0.5325,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.974025974025974,
|
| 1294 |
+
"grad_norm": 0.986328125,
|
| 1295 |
+
"learning_rate": 6.98753810462766e-05,
|
| 1296 |
+
"loss": 0.5503,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 6.0064935064935066,
|
| 1301 |
+
"grad_norm": 0.72314453125,
|
| 1302 |
+
"learning_rate": 6.890434992137407e-05,
|
| 1303 |
+
"loss": 0.5213,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 6.038961038961039,
|
| 1308 |
+
"grad_norm": 0.74560546875,
|
| 1309 |
+
"learning_rate": 6.793655394472644e-05,
|
| 1310 |
+
"loss": 0.4754,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.071428571428571,
|
| 1315 |
+
"grad_norm": 0.826171875,
|
| 1316 |
+
"learning_rate": 6.697209380448333e-05,
|
| 1317 |
+
"loss": 0.4703,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.103896103896104,
|
| 1322 |
+
"grad_norm": 0.90283203125,
|
| 1323 |
+
"learning_rate": 6.601106984173835e-05,
|
| 1324 |
+
"loss": 0.5074,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.136363636363637,
|
| 1329 |
+
"grad_norm": 0.91650390625,
|
| 1330 |
+
"learning_rate": 6.505358204009017e-05,
|
| 1331 |
+
"loss": 0.4941,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.1688311688311686,
|
| 1336 |
+
"grad_norm": 0.8486328125,
|
| 1337 |
+
"learning_rate": 6.409973001524012e-05,
|
| 1338 |
+
"loss": 0.4937,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.201298701298701,
|
| 1343 |
+
"grad_norm": 0.92822265625,
|
| 1344 |
+
"learning_rate": 6.314961300462841e-05,
|
| 1345 |
+
"loss": 0.5005,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.233766233766234,
|
| 1350 |
+
"grad_norm": 0.85546875,
|
| 1351 |
+
"learning_rate": 6.220332985710936e-05,
|
| 1352 |
+
"loss": 0.4936,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.266233766233766,
|
| 1357 |
+
"grad_norm": 0.91162109375,
|
| 1358 |
+
"learning_rate": 6.126097902266772e-05,
|
| 1359 |
+
"loss": 0.4682,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.298701298701299,
|
| 1364 |
+
"grad_norm": 0.88232421875,
|
| 1365 |
+
"learning_rate": 6.0322658542175736e-05,
|
| 1366 |
+
"loss": 0.5096,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.3311688311688314,
|
| 1371 |
+
"grad_norm": 0.90576171875,
|
| 1372 |
+
"learning_rate": 5.938846603719332e-05,
|
| 1373 |
+
"loss": 0.5064,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.363636363636363,
|
| 1378 |
+
"grad_norm": 0.931640625,
|
| 1379 |
+
"learning_rate": 5.845849869981137e-05,
|
| 1380 |
+
"loss": 0.5013,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.396103896103896,
|
| 1385 |
+
"grad_norm": 0.8046875,
|
| 1386 |
+
"learning_rate": 5.7532853282540387e-05,
|
| 1387 |
+
"loss": 0.4929,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.428571428571429,
|
| 1392 |
+
"grad_norm": 0.8662109375,
|
| 1393 |
+
"learning_rate": 5.6611626088244194e-05,
|
| 1394 |
+
"loss": 0.5128,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.461038961038961,
|
| 1399 |
+
"grad_norm": 0.9287109375,
|
| 1400 |
+
"learning_rate": 5.569491296012095e-05,
|
| 1401 |
+
"loss": 0.4912,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.4935064935064934,
|
| 1406 |
+
"grad_norm": 0.76904296875,
|
| 1407 |
+
"learning_rate": 5.478280927173145e-05,
|
| 1408 |
+
"loss": 0.4999,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.525974025974026,
|
| 1413 |
+
"grad_norm": 0.93701171875,
|
| 1414 |
+
"learning_rate": 5.387540991707697e-05,
|
| 1415 |
+
"loss": 0.4988,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.558441558441558,
|
| 1420 |
+
"grad_norm": 0.8828125,
|
| 1421 |
+
"learning_rate": 5.297280930072632e-05,
|
| 1422 |
+
"loss": 0.4825,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.590909090909091,
|
| 1427 |
+
"grad_norm": 0.83984375,
|
| 1428 |
+
"learning_rate": 5.207510132799436e-05,
|
| 1429 |
+
"loss": 0.5054,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.623376623376624,
|
| 1434 |
+
"grad_norm": 0.88671875,
|
| 1435 |
+
"learning_rate": 5.11823793951719e-05,
|
| 1436 |
+
"loss": 0.4975,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.6558441558441555,
|
| 1441 |
+
"grad_norm": 0.9208984375,
|
| 1442 |
+
"learning_rate": 5.029473637980926e-05,
|
| 1443 |
+
"loss": 0.497,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.688311688311688,
|
| 1448 |
+
"grad_norm": 0.86865234375,
|
| 1449 |
+
"learning_rate": 4.9412264631053216e-05,
|
| 1450 |
+
"loss": 0.4983,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.720779220779221,
|
| 1455 |
+
"grad_norm": 0.9228515625,
|
| 1456 |
+
"learning_rate": 4.853505596003905e-05,
|
| 1457 |
+
"loss": 0.522,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.753246753246753,
|
| 1462 |
+
"grad_norm": 0.86865234375,
|
| 1463 |
+
"learning_rate": 4.7663201630338816e-05,
|
| 1464 |
+
"loss": 0.4981,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.785714285714286,
|
| 1469 |
+
"grad_norm": 0.87060546875,
|
| 1470 |
+
"learning_rate": 4.6796792348466356e-05,
|
| 1471 |
+
"loss": 0.4795,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.818181818181818,
|
| 1476 |
+
"grad_norm": 1.0458984375,
|
| 1477 |
+
"learning_rate": 4.593591825444028e-05,
|
| 1478 |
+
"loss": 0.5133,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.85064935064935,
|
| 1483 |
+
"grad_norm": 0.828125,
|
| 1484 |
+
"learning_rate": 4.5080668912405855e-05,
|
| 1485 |
+
"loss": 0.5019,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.883116883116883,
|
| 1490 |
+
"grad_norm": 0.87158203125,
|
| 1491 |
+
"learning_rate": 4.423113330131707e-05,
|
| 1492 |
+
"loss": 0.4884,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.915584415584416,
|
| 1497 |
+
"grad_norm": 0.82568359375,
|
| 1498 |
+
"learning_rate": 4.3387399805679255e-05,
|
| 1499 |
+
"loss": 0.492,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.948051948051948,
|
| 1504 |
+
"grad_norm": 0.87255859375,
|
| 1505 |
+
"learning_rate": 4.25495562063537e-05,
|
| 1506 |
+
"loss": 0.489,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.98051948051948,
|
| 1511 |
+
"grad_norm": 0.90869140625,
|
| 1512 |
+
"learning_rate": 4.1717689671425e-05,
|
| 1513 |
+
"loss": 0.4903,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.012987012987013,
|
| 1518 |
+
"grad_norm": 0.87060546875,
|
| 1519 |
+
"learning_rate": 4.089188674713236e-05,
|
| 1520 |
+
"loss": 0.4801,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.045454545454546,
|
| 1525 |
+
"grad_norm": 0.94970703125,
|
| 1526 |
+
"learning_rate": 4.007223334886531e-05,
|
| 1527 |
+
"loss": 0.4534,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.077922077922078,
|
| 1532 |
+
"grad_norm": 0.92138671875,
|
| 1533 |
+
"learning_rate": 3.9258814752225284e-05,
|
| 1534 |
+
"loss": 0.4601,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.1103896103896105,
|
| 1539 |
+
"grad_norm": 0.87841796875,
|
| 1540 |
+
"learning_rate": 3.845171558415348e-05,
|
| 1541 |
+
"loss": 0.4498,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.142857142857143,
|
| 1546 |
+
"grad_norm": 0.994140625,
|
| 1547 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 1548 |
+
"loss": 0.4679,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.175324675324675,
|
| 1553 |
+
"grad_norm": 0.98193359375,
|
| 1554 |
+
"learning_rate": 3.6856810745420886e-05,
|
| 1555 |
+
"loss": 0.468,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.207792207792208,
|
| 1560 |
+
"grad_norm": 0.89111328125,
|
| 1561 |
+
"learning_rate": 3.606917100644488e-05,
|
| 1562 |
+
"loss": 0.4599,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.240259740259741,
|
| 1567 |
+
"grad_norm": 1.123046875,
|
| 1568 |
+
"learning_rate": 3.528818254214329e-05,
|
| 1569 |
+
"loss": 0.4696,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.2727272727272725,
|
| 1574 |
+
"grad_norm": 0.85107421875,
|
| 1575 |
+
"learning_rate": 3.45139266054715e-05,
|
| 1576 |
+
"loss": 0.4594,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.305194805194805,
|
| 1581 |
+
"grad_norm": 0.86181640625,
|
| 1582 |
+
"learning_rate": 3.374648374894204e-05,
|
| 1583 |
+
"loss": 0.4488,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.337662337662338,
|
| 1588 |
+
"grad_norm": 0.97705078125,
|
| 1589 |
+
"learning_rate": 3.298593381624406e-05,
|
| 1590 |
+
"loss": 0.4619,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.37012987012987,
|
| 1595 |
+
"grad_norm": 0.90869140625,
|
| 1596 |
+
"learning_rate": 3.223235593393642e-05,
|
| 1597 |
+
"loss": 0.4717,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.402597402597403,
|
| 1602 |
+
"grad_norm": 0.974609375,
|
| 1603 |
+
"learning_rate": 3.1485828503215585e-05,
|
| 1604 |
+
"loss": 0.472,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.435064935064935,
|
| 1609 |
+
"grad_norm": 0.921875,
|
| 1610 |
+
"learning_rate": 3.074642919175883e-05,
|
| 1611 |
+
"loss": 0.478,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.467532467532467,
|
| 1616 |
+
"grad_norm": 0.91015625,
|
| 1617 |
+
"learning_rate": 3.0014234925643837e-05,
|
| 1618 |
+
"loss": 0.4605,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.5,
|
| 1623 |
+
"grad_norm": 1.04296875,
|
| 1624 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 1625 |
+
"loss": 0.4573,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.532467532467533,
|
| 1630 |
+
"grad_norm": 0.93994140625,
|
| 1631 |
+
"learning_rate": 2.8571765477809643e-05,
|
| 1632 |
+
"loss": 0.462,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.564935064935065,
|
| 1637 |
+
"grad_norm": 0.9765625,
|
| 1638 |
+
"learning_rate": 2.7861640368608844e-05,
|
| 1639 |
+
"loss": 0.4722,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.597402597402597,
|
| 1644 |
+
"grad_norm": 1.0263671875,
|
| 1645 |
+
"learning_rate": 2.71590204341731e-05,
|
| 1646 |
+
"loss": 0.4719,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.62987012987013,
|
| 1651 |
+
"grad_norm": 0.8818359375,
|
| 1652 |
+
"learning_rate": 2.6463978774104758e-05,
|
| 1653 |
+
"loss": 0.4476,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.662337662337662,
|
| 1658 |
+
"grad_norm": 1.0595703125,
|
| 1659 |
+
"learning_rate": 2.5776587699573006e-05,
|
| 1660 |
+
"loss": 0.4742,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.694805194805195,
|
| 1665 |
+
"grad_norm": 1.0439453125,
|
| 1666 |
+
"learning_rate": 2.509691872579075e-05,
|
| 1667 |
+
"loss": 0.4714,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.7272727272727275,
|
| 1672 |
+
"grad_norm": 0.89404296875,
|
| 1673 |
+
"learning_rate": 2.4425042564574184e-05,
|
| 1674 |
+
"loss": 0.4771,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.759740259740259,
|
| 1679 |
+
"grad_norm": 0.90673828125,
|
| 1680 |
+
"learning_rate": 2.3761029116986178e-05,
|
| 1681 |
+
"loss": 0.4537,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.792207792207792,
|
| 1686 |
+
"grad_norm": 0.958984375,
|
| 1687 |
+
"learning_rate": 2.3104947466063787e-05,
|
| 1688 |
+
"loss": 0.4678,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.824675324675325,
|
| 1693 |
+
"grad_norm": 1.1796875,
|
| 1694 |
+
"learning_rate": 2.2456865869630972e-05,
|
| 1695 |
+
"loss": 0.4819,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.857142857142857,
|
| 1700 |
+
"grad_norm": 0.95068359375,
|
| 1701 |
+
"learning_rate": 2.181685175319702e-05,
|
| 1702 |
+
"loss": 0.4759,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.8896103896103895,
|
| 1707 |
+
"grad_norm": 0.99755859375,
|
| 1708 |
+
"learning_rate": 2.118497170294195e-05,
|
| 1709 |
+
"loss": 0.4795,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.922077922077922,
|
| 1714 |
+
"grad_norm": 1.0068359375,
|
| 1715 |
+
"learning_rate": 2.0561291458788733e-05,
|
| 1716 |
+
"loss": 0.4593,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.954545454545455,
|
| 1721 |
+
"grad_norm": 1.1923828125,
|
| 1722 |
+
"learning_rate": 1.994587590756397e-05,
|
| 1723 |
+
"loss": 0.4653,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.987012987012987,
|
| 1728 |
+
"grad_norm": 0.9072265625,
|
| 1729 |
+
"learning_rate": 1.9338789076247e-05,
|
| 1730 |
+
"loss": 0.4776,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 8.019480519480519,
|
| 1735 |
+
"grad_norm": 0.8837890625,
|
| 1736 |
+
"learning_rate": 1.874009412530877e-05,
|
| 1737 |
+
"loss": 0.4631,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 8.051948051948052,
|
| 1742 |
+
"grad_norm": 0.9091796875,
|
| 1743 |
+
"learning_rate": 1.8149853342140645e-05,
|
| 1744 |
+
"loss": 0.4328,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 8.084415584415584,
|
| 1749 |
+
"grad_norm": 1.1220703125,
|
| 1750 |
+
"learning_rate": 1.7568128134574113e-05,
|
| 1751 |
+
"loss": 0.4424,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.116883116883116,
|
| 1756 |
+
"grad_norm": 0.95947265625,
|
| 1757 |
+
"learning_rate": 1.6994979024491942e-05,
|
| 1758 |
+
"loss": 0.4333,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.14935064935065,
|
| 1763 |
+
"grad_norm": 0.89794921875,
|
| 1764 |
+
"learning_rate": 1.64304656415317e-05,
|
| 1765 |
+
"loss": 0.4438,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 8.181818181818182,
|
| 1770 |
+
"grad_norm": 1.0498046875,
|
| 1771 |
+
"learning_rate": 1.587464671688187e-05,
|
| 1772 |
+
"loss": 0.4356,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 8.214285714285714,
|
| 1777 |
+
"grad_norm": 0.93212890625,
|
| 1778 |
+
"learning_rate": 1.5327580077171587e-05,
|
| 1779 |
+
"loss": 0.4312,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.246753246753247,
|
| 1784 |
+
"grad_norm": 0.90771484375,
|
| 1785 |
+
"learning_rate": 1.4789322638454351e-05,
|
| 1786 |
+
"loss": 0.4429,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 8.279220779220779,
|
| 1791 |
+
"grad_norm": 1.046875,
|
| 1792 |
+
"learning_rate": 1.4259930400286669e-05,
|
| 1793 |
+
"loss": 0.4469,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 8.311688311688311,
|
| 1798 |
+
"grad_norm": 1.0224609375,
|
| 1799 |
+
"learning_rate": 1.373945843990192e-05,
|
| 1800 |
+
"loss": 0.4497,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.344155844155845,
|
| 1805 |
+
"grad_norm": 1.712890625,
|
| 1806 |
+
"learning_rate": 1.322796090648013e-05,
|
| 1807 |
+
"loss": 0.4467,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.376623376623376,
|
| 1812 |
+
"grad_norm": 0.923828125,
|
| 1813 |
+
"learning_rate": 1.272549101551438e-05,
|
| 1814 |
+
"loss": 0.4437,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.409090909090908,
|
| 1819 |
+
"grad_norm": 1.134765625,
|
| 1820 |
+
"learning_rate": 1.2232101043274436e-05,
|
| 1821 |
+
"loss": 0.4374,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.441558441558442,
|
| 1826 |
+
"grad_norm": 1.0517578125,
|
| 1827 |
+
"learning_rate": 1.1747842321367886e-05,
|
| 1828 |
+
"loss": 0.4516,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.474025974025974,
|
| 1833 |
+
"grad_norm": 0.9755859375,
|
| 1834 |
+
"learning_rate": 1.1272765231399685e-05,
|
| 1835 |
+
"loss": 0.4382,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.506493506493506,
|
| 1840 |
+
"grad_norm": 1.1240234375,
|
| 1841 |
+
"learning_rate": 1.0806919199730615e-05,
|
| 1842 |
+
"loss": 0.4438,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.53896103896104,
|
| 1847 |
+
"grad_norm": 0.9833984375,
|
| 1848 |
+
"learning_rate": 1.035035269233493e-05,
|
| 1849 |
+
"loss": 0.446,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.571428571428571,
|
| 1854 |
+
"grad_norm": 0.97412109375,
|
| 1855 |
+
"learning_rate": 9.903113209758096e-06,
|
| 1856 |
+
"loss": 0.4419,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.603896103896105,
|
| 1861 |
+
"grad_norm": 1.0947265625,
|
| 1862 |
+
"learning_rate": 9.465247282174805e-06,
|
| 1863 |
+
"loss": 0.4607,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.636363636363637,
|
| 1868 |
+
"grad_norm": 0.912109375,
|
| 1869 |
+
"learning_rate": 9.036800464548157e-06,
|
| 1870 |
+
"loss": 0.4575,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.668831168831169,
|
| 1875 |
+
"grad_norm": 0.91845703125,
|
| 1876 |
+
"learning_rate": 8.617817331890154e-06,
|
| 1877 |
+
"loss": 0.4486,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.7012987012987,
|
| 1882 |
+
"grad_norm": 1.0595703125,
|
| 1883 |
+
"learning_rate": 8.208341474624071e-06,
|
| 1884 |
+
"loss": 0.455,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.733766233766234,
|
| 1889 |
+
"grad_norm": 0.98095703125,
|
| 1890 |
+
"learning_rate": 7.808415494049514e-06,
|
| 1891 |
+
"loss": 0.4565,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.766233766233766,
|
| 1896 |
+
"grad_norm": 0.9755859375,
|
| 1897 |
+
"learning_rate": 7.4180809979102036e-06,
|
| 1898 |
+
"loss": 0.447,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.7987012987013,
|
| 1903 |
+
"grad_norm": 1.0478515625,
|
| 1904 |
+
"learning_rate": 7.0373785960650475e-06,
|
| 1905 |
+
"loss": 0.4431,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.831168831168831,
|
| 1910 |
+
"grad_norm": 0.97802734375,
|
| 1911 |
+
"learning_rate": 6.666347896263325e-06,
|
| 1912 |
+
"loss": 0.4479,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.863636363636363,
|
| 1917 |
+
"grad_norm": 0.8994140625,
|
| 1918 |
+
"learning_rate": 6.3050275000238414e-06,
|
| 1919 |
+
"loss": 0.4558,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.896103896103895,
|
| 1924 |
+
"grad_norm": 1.087890625,
|
| 1925 |
+
"learning_rate": 5.953454998618857e-06,
|
| 1926 |
+
"loss": 0.4357,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.928571428571429,
|
| 1931 |
+
"grad_norm": 1.0263671875,
|
| 1932 |
+
"learning_rate": 5.611666969163243e-06,
|
| 1933 |
+
"loss": 0.4464,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.96103896103896,
|
| 1938 |
+
"grad_norm": 0.93505859375,
|
| 1939 |
+
"learning_rate": 5.27969897080901e-06,
|
| 1940 |
+
"loss": 0.4459,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.993506493506494,
|
| 1945 |
+
"grad_norm": 0.96435546875,
|
| 1946 |
+
"learning_rate": 4.957585541045684e-06,
|
| 1947 |
+
"loss": 0.4749,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 9.025974025974026,
|
| 1952 |
+
"grad_norm": 1.185546875,
|
| 1953 |
+
"learning_rate": 4.6453601921072395e-06,
|
| 1954 |
+
"loss": 0.4351,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 9.058441558441558,
|
| 1959 |
+
"grad_norm": 0.95849609375,
|
| 1960 |
+
"learning_rate": 4.34305540748543e-06,
|
| 1961 |
+
"loss": 0.429,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 9.090909090909092,
|
| 1966 |
+
"grad_norm": 0.93701171875,
|
| 1967 |
+
"learning_rate": 4.050702638550275e-06,
|
| 1968 |
+
"loss": 0.4335,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 9.123376623376624,
|
| 1973 |
+
"grad_norm": 0.982421875,
|
| 1974 |
+
"learning_rate": 3.768332301277866e-06,
|
| 1975 |
+
"loss": 0.445,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 9.155844155844155,
|
| 1980 |
+
"grad_norm": 0.935546875,
|
| 1981 |
+
"learning_rate": 3.495973773086014e-06,
|
| 1982 |
+
"loss": 0.4316,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 9.188311688311689,
|
| 1987 |
+
"grad_norm": 1.0439453125,
|
| 1988 |
+
"learning_rate": 3.233655389777801e-06,
|
| 1989 |
+
"loss": 0.4395,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 9.220779220779221,
|
| 1994 |
+
"grad_norm": 0.9423828125,
|
| 1995 |
+
"learning_rate": 2.9814044425935606e-06,
|
| 1996 |
+
"loss": 0.4207,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 9.253246753246753,
|
| 2001 |
+
"grad_norm": 1.09765625,
|
| 2002 |
+
"learning_rate": 2.739247175371562e-06,
|
| 2003 |
+
"loss": 0.4503,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 9.285714285714286,
|
| 2008 |
+
"grad_norm": 1.0166015625,
|
| 2009 |
+
"learning_rate": 2.5072087818176382e-06,
|
| 2010 |
+
"loss": 0.454,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 9.318181818181818,
|
| 2015 |
+
"grad_norm": 1.01171875,
|
| 2016 |
+
"learning_rate": 2.2853134028840594e-06,
|
| 2017 |
+
"loss": 0.426,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 9.35064935064935,
|
| 2022 |
+
"grad_norm": 1.130859375,
|
| 2023 |
+
"learning_rate": 2.073584124257899e-06,
|
| 2024 |
+
"loss": 0.4464,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.383116883116884,
|
| 2029 |
+
"grad_norm": 0.888671875,
|
| 2030 |
+
"learning_rate": 1.8720429739592982e-06,
|
| 2031 |
+
"loss": 0.4535,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 9.415584415584416,
|
| 2036 |
+
"grad_norm": 1.0634765625,
|
| 2037 |
+
"learning_rate": 1.6807109200496995e-06,
|
| 2038 |
+
"loss": 0.4556,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.448051948051948,
|
| 2043 |
+
"grad_norm": 0.9482421875,
|
| 2044 |
+
"learning_rate": 1.4996078684503144e-06,
|
| 2045 |
+
"loss": 0.4236,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.480519480519481,
|
| 2050 |
+
"grad_norm": 0.85888671875,
|
| 2051 |
+
"learning_rate": 1.3287526608711131e-06,
|
| 2052 |
+
"loss": 0.4366,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.512987012987013,
|
| 2057 |
+
"grad_norm": 1.01953125,
|
| 2058 |
+
"learning_rate": 1.1681630728506699e-06,
|
| 2059 |
+
"loss": 0.4236,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.545454545454545,
|
| 2064 |
+
"grad_norm": 1.0693359375,
|
| 2065 |
+
"learning_rate": 1.0178558119067315e-06,
|
| 2066 |
+
"loss": 0.4398,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.577922077922079,
|
| 2071 |
+
"grad_norm": 0.96337890625,
|
| 2072 |
+
"learning_rate": 8.778465157979976e-07,
|
| 2073 |
+
"loss": 0.4412,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 9.61038961038961,
|
| 2078 |
+
"grad_norm": 0.982421875,
|
| 2079 |
+
"learning_rate": 7.481497508972312e-07,
|
| 2080 |
+
"loss": 0.4307,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.642857142857142,
|
| 2085 |
+
"grad_norm": 0.9599609375,
|
| 2086 |
+
"learning_rate": 6.287790106757396e-07,
|
| 2087 |
+
"loss": 0.4415,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.675324675324676,
|
| 2092 |
+
"grad_norm": 1.0146484375,
|
| 2093 |
+
"learning_rate": 5.19746714299596e-07,
|
| 2094 |
+
"loss": 0.4412,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.707792207792208,
|
| 2099 |
+
"grad_norm": 0.95361328125,
|
| 2100 |
+
"learning_rate": 4.210642053375069e-07,
|
| 2101 |
+
"loss": 0.4416,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.74025974025974,
|
| 2106 |
+
"grad_norm": 1.00390625,
|
| 2107 |
+
"learning_rate": 3.3274175058067846e-07,
|
| 2108 |
+
"loss": 0.4383,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.772727272727273,
|
| 2113 |
+
"grad_norm": 0.947265625,
|
| 2114 |
+
"learning_rate": 2.547885389746485e-07,
|
| 2115 |
+
"loss": 0.4401,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.805194805194805,
|
| 2120 |
+
"grad_norm": 1.0166015625,
|
| 2121 |
+
"learning_rate": 1.8721268066330676e-07,
|
| 2122 |
+
"loss": 0.4332,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.837662337662337,
|
| 2127 |
+
"grad_norm": 1.0302734375,
|
| 2128 |
+
"learning_rate": 1.300212061451367e-07,
|
| 2129 |
+
"loss": 0.4358,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.87012987012987,
|
| 2134 |
+
"grad_norm": 1.0673828125,
|
| 2135 |
+
"learning_rate": 8.322006554171146e-08,
|
| 2136 |
+
"loss": 0.4412,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.902597402597403,
|
| 2141 |
+
"grad_norm": 0.96826171875,
|
| 2142 |
+
"learning_rate": 4.6814127978722644e-08,
|
| 2143 |
+
"loss": 0.4329,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.935064935064934,
|
| 2148 |
+
"grad_norm": 1.03515625,
|
| 2149 |
+
"learning_rate": 2.080718107935198e-08,
|
| 2150 |
+
"loss": 0.4401,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.967532467532468,
|
| 2155 |
+
"grad_norm": 1.044921875,
|
| 2156 |
+
"learning_rate": 5.201930570242208e-09,
|
| 2157 |
+
"loss": 0.4407,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 10.0,
|
| 2162 |
+
"grad_norm": 0.93701171875,
|
| 2163 |
+
"learning_rate": 0.0,
|
| 2164 |
+
"loss": 0.4513,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 10.0,
|
| 2169 |
+
"step": 3080,
|
| 2170 |
+
"total_flos": 7.864751040233472e+16,
|
| 2171 |
+
"train_loss": 0.5693606966501706,
|
| 2172 |
+
"train_runtime": 1451.854,
|
| 2173 |
+
"train_samples_per_second": 8.486,
|
| 2174 |
+
"train_steps_per_second": 2.121
|
| 2175 |
+
}
|
| 2176 |
+
],
|
| 2177 |
+
"logging_steps": 10,
|
| 2178 |
+
"max_steps": 3080,
|
| 2179 |
+
"num_input_tokens_seen": 0,
|
| 2180 |
+
"num_train_epochs": 10,
|
| 2181 |
+
"save_steps": 500,
|
| 2182 |
+
"total_flos": 7.864751040233472e+16,
|
| 2183 |
+
"train_batch_size": 1,
|
| 2184 |
+
"trial_name": null,
|
| 2185 |
+
"trial_params": null
|
| 2186 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5ebb2168c296a12d4cddde9b5fb5278e1e03683c75c57622a183944688eeec8
|
| 3 |
size 5048
|