Training in progress, epoch 1
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- trainer_state.json +2494 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,12 +20,12 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"v_proj",
|
| 24 |
-
"q_proj",
|
| 25 |
-
"dense",
|
| 26 |
"k_proj",
|
| 27 |
"fc2",
|
| 28 |
-
"fc1"
|
|
|
|
|
|
|
|
|
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
| 23 |
"k_proj",
|
| 24 |
"fc2",
|
| 25 |
+
"fc1",
|
| 26 |
+
"dense",
|
| 27 |
+
"v_proj",
|
| 28 |
+
"q_proj"
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"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 47236352
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b315d32978bf65a12cbf3069168071bfb3b0c02de07c11d231d177a23884237
|
| 3 |
size 47236352
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2494 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3520,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.028409090909090908,
|
| 13 |
+
"grad_norm": 1.4697265625,
|
| 14 |
+
"learning_rate": 0.00019999601726381413,
|
| 15 |
+
"loss": 2.9105,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.056818181818181816,
|
| 20 |
+
"grad_norm": 1.3994140625,
|
| 21 |
+
"learning_rate": 0.00019998406937250034,
|
| 22 |
+
"loss": 2.5482,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08522727272727272,
|
| 27 |
+
"grad_norm": 1.5947265625,
|
| 28 |
+
"learning_rate": 0.00019996415727776455,
|
| 29 |
+
"loss": 2.5436,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11363636363636363,
|
| 34 |
+
"grad_norm": 1.22265625,
|
| 35 |
+
"learning_rate": 0.0001999362825656992,
|
| 36 |
+
"loss": 2.5159,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.14204545454545456,
|
| 41 |
+
"grad_norm": 1.2119140625,
|
| 42 |
+
"learning_rate": 0.00019990044745665672,
|
| 43 |
+
"loss": 2.4464,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.17045454545454544,
|
| 48 |
+
"grad_norm": 1.23828125,
|
| 49 |
+
"learning_rate": 0.0001998566548050729,
|
| 50 |
+
"loss": 2.3599,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19886363636363635,
|
| 55 |
+
"grad_norm": 1.2744140625,
|
| 56 |
+
"learning_rate": 0.00019980490809923926,
|
| 57 |
+
"loss": 2.5164,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.22727272727272727,
|
| 62 |
+
"grad_norm": 1.1611328125,
|
| 63 |
+
"learning_rate": 0.00019974521146102537,
|
| 64 |
+
"loss": 2.4555,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2556818181818182,
|
| 69 |
+
"grad_norm": 1.2080078125,
|
| 70 |
+
"learning_rate": 0.00019967756964555045,
|
| 71 |
+
"loss": 2.3519,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2840909090909091,
|
| 76 |
+
"grad_norm": 1.1796875,
|
| 77 |
+
"learning_rate": 0.0001996019880408046,
|
| 78 |
+
"loss": 2.3518,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3125,
|
| 83 |
+
"grad_norm": 1.4345703125,
|
| 84 |
+
"learning_rate": 0.0001995184726672197,
|
| 85 |
+
"loss": 2.3879,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3409090909090909,
|
| 90 |
+
"grad_norm": 1.16015625,
|
| 91 |
+
"learning_rate": 0.00019942703017718975,
|
| 92 |
+
"loss": 2.3626,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3693181818181818,
|
| 97 |
+
"grad_norm": 1.103515625,
|
| 98 |
+
"learning_rate": 0.000199327667854541,
|
| 99 |
+
"loss": 2.3314,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3977272727272727,
|
| 104 |
+
"grad_norm": 1.140625,
|
| 105 |
+
"learning_rate": 0.00019922039361395185,
|
| 106 |
+
"loss": 2.4043,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.42613636363636365,
|
| 111 |
+
"grad_norm": 1.0341796875,
|
| 112 |
+
"learning_rate": 0.00019910521600032227,
|
| 113 |
+
"loss": 2.3263,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.45454545454545453,
|
| 118 |
+
"grad_norm": 0.986328125,
|
| 119 |
+
"learning_rate": 0.0001989821441880933,
|
| 120 |
+
"loss": 2.3119,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.48295454545454547,
|
| 125 |
+
"grad_norm": 1.0888671875,
|
| 126 |
+
"learning_rate": 0.00019885118798051605,
|
| 127 |
+
"loss": 2.2697,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5113636363636364,
|
| 132 |
+
"grad_norm": 0.9833984375,
|
| 133 |
+
"learning_rate": 0.00019871235780887113,
|
| 134 |
+
"loss": 2.2761,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5397727272727273,
|
| 139 |
+
"grad_norm": 1.068359375,
|
| 140 |
+
"learning_rate": 0.00019856566473163746,
|
| 141 |
+
"loss": 2.3054,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5681818181818182,
|
| 146 |
+
"grad_norm": 1.10546875,
|
| 147 |
+
"learning_rate": 0.0001984111204336116,
|
| 148 |
+
"loss": 2.2658,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5965909090909091,
|
| 153 |
+
"grad_norm": 1.021484375,
|
| 154 |
+
"learning_rate": 0.00019824873722497694,
|
| 155 |
+
"loss": 2.2965,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.625,
|
| 160 |
+
"grad_norm": 0.986328125,
|
| 161 |
+
"learning_rate": 0.00019807852804032305,
|
| 162 |
+
"loss": 2.303,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6534090909090909,
|
| 167 |
+
"grad_norm": 0.9755859375,
|
| 168 |
+
"learning_rate": 0.00019790050643761552,
|
| 169 |
+
"loss": 2.2398,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6818181818181818,
|
| 174 |
+
"grad_norm": 0.94677734375,
|
| 175 |
+
"learning_rate": 0.00019771468659711595,
|
| 176 |
+
"loss": 2.257,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7102272727272727,
|
| 181 |
+
"grad_norm": 0.953125,
|
| 182 |
+
"learning_rate": 0.0001975210833202524,
|
| 183 |
+
"loss": 2.288,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7386363636363636,
|
| 188 |
+
"grad_norm": 1.1494140625,
|
| 189 |
+
"learning_rate": 0.00019731971202844036,
|
| 190 |
+
"loss": 2.264,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7670454545454546,
|
| 195 |
+
"grad_norm": 0.98779296875,
|
| 196 |
+
"learning_rate": 0.00019711058876185447,
|
| 197 |
+
"loss": 2.2359,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7954545454545454,
|
| 202 |
+
"grad_norm": 1.0556640625,
|
| 203 |
+
"learning_rate": 0.00019689373017815073,
|
| 204 |
+
"loss": 2.287,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8238636363636364,
|
| 209 |
+
"grad_norm": 0.85009765625,
|
| 210 |
+
"learning_rate": 0.00019666915355113975,
|
| 211 |
+
"loss": 2.2565,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.8522727272727273,
|
| 216 |
+
"grad_norm": 1.0673828125,
|
| 217 |
+
"learning_rate": 0.00019643687676941068,
|
| 218 |
+
"loss": 2.2853,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8806818181818182,
|
| 223 |
+
"grad_norm": 0.9814453125,
|
| 224 |
+
"learning_rate": 0.00019619691833490643,
|
| 225 |
+
"loss": 2.2209,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9090909090909091,
|
| 230 |
+
"grad_norm": 0.87744140625,
|
| 231 |
+
"learning_rate": 0.00019594929736144976,
|
| 232 |
+
"loss": 2.2515,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9375,
|
| 237 |
+
"grad_norm": 0.9345703125,
|
| 238 |
+
"learning_rate": 0.0001956940335732209,
|
| 239 |
+
"loss": 2.193,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.9659090909090909,
|
| 244 |
+
"grad_norm": 0.87451171875,
|
| 245 |
+
"learning_rate": 0.0001954311473031864,
|
| 246 |
+
"loss": 2.3074,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9943181818181818,
|
| 251 |
+
"grad_norm": 0.91650390625,
|
| 252 |
+
"learning_rate": 0.00019516065949147943,
|
| 253 |
+
"loss": 2.2416,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0227272727272727,
|
| 258 |
+
"grad_norm": 0.9345703125,
|
| 259 |
+
"learning_rate": 0.00019488259168373197,
|
| 260 |
+
"loss": 2.2029,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.0511363636363635,
|
| 265 |
+
"grad_norm": 1.904296875,
|
| 266 |
+
"learning_rate": 0.00019459696602935837,
|
| 267 |
+
"loss": 2.0812,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.0795454545454546,
|
| 272 |
+
"grad_norm": 0.93603515625,
|
| 273 |
+
"learning_rate": 0.00019430380527979123,
|
| 274 |
+
"loss": 2.1138,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1079545454545454,
|
| 279 |
+
"grad_norm": 1.015625,
|
| 280 |
+
"learning_rate": 0.00019400313278666902,
|
| 281 |
+
"loss": 2.1646,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.1363636363636362,
|
| 286 |
+
"grad_norm": 1.1142578125,
|
| 287 |
+
"learning_rate": 0.0001936949724999762,
|
| 288 |
+
"loss": 2.0984,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.1647727272727273,
|
| 293 |
+
"grad_norm": 1.2470703125,
|
| 294 |
+
"learning_rate": 0.00019337934896613516,
|
| 295 |
+
"loss": 2.0974,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.1931818181818181,
|
| 300 |
+
"grad_norm": 0.96044921875,
|
| 301 |
+
"learning_rate": 0.00019305628732605137,
|
| 302 |
+
"loss": 2.095,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.2215909090909092,
|
| 307 |
+
"grad_norm": 0.9833984375,
|
| 308 |
+
"learning_rate": 0.00019272581331311047,
|
| 309 |
+
"loss": 2.1586,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.25,
|
| 314 |
+
"grad_norm": 1.068359375,
|
| 315 |
+
"learning_rate": 0.0001923879532511287,
|
| 316 |
+
"loss": 2.1677,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.2784090909090908,
|
| 321 |
+
"grad_norm": 1.0478515625,
|
| 322 |
+
"learning_rate": 0.00019204273405225587,
|
| 323 |
+
"loss": 2.1594,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.3068181818181819,
|
| 328 |
+
"grad_norm": 1.2109375,
|
| 329 |
+
"learning_rate": 0.00019169018321483198,
|
| 330 |
+
"loss": 2.1347,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.3352272727272727,
|
| 335 |
+
"grad_norm": 1.091796875,
|
| 336 |
+
"learning_rate": 0.00019133032882119653,
|
| 337 |
+
"loss": 2.0907,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.3636363636363638,
|
| 342 |
+
"grad_norm": 1.09375,
|
| 343 |
+
"learning_rate": 0.00019096319953545185,
|
| 344 |
+
"loss": 2.1914,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.3920454545454546,
|
| 349 |
+
"grad_norm": 1.025390625,
|
| 350 |
+
"learning_rate": 0.00019058882460117974,
|
| 351 |
+
"loss": 2.1875,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.4204545454545454,
|
| 356 |
+
"grad_norm": 1.0380859375,
|
| 357 |
+
"learning_rate": 0.00019020723383911215,
|
| 358 |
+
"loss": 2.1611,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.4488636363636362,
|
| 363 |
+
"grad_norm": 0.9951171875,
|
| 364 |
+
"learning_rate": 0.00018981845764475582,
|
| 365 |
+
"loss": 2.113,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.4772727272727273,
|
| 370 |
+
"grad_norm": 1.208984375,
|
| 371 |
+
"learning_rate": 0.00018942252698597113,
|
| 372 |
+
"loss": 2.1395,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.5056818181818183,
|
| 377 |
+
"grad_norm": 1.0732421875,
|
| 378 |
+
"learning_rate": 0.00018901947340050528,
|
| 379 |
+
"loss": 2.1425,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.5340909090909092,
|
| 384 |
+
"grad_norm": 1.0830078125,
|
| 385 |
+
"learning_rate": 0.00018860932899348028,
|
| 386 |
+
"loss": 2.0784,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.5625,
|
| 391 |
+
"grad_norm": 1.23046875,
|
| 392 |
+
"learning_rate": 0.0001881921264348355,
|
| 393 |
+
"loss": 2.0838,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.5909090909090908,
|
| 398 |
+
"grad_norm": 1.30859375,
|
| 399 |
+
"learning_rate": 0.00018776789895672558,
|
| 400 |
+
"loss": 2.1142,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.6193181818181817,
|
| 405 |
+
"grad_norm": 1.4462890625,
|
| 406 |
+
"learning_rate": 0.00018733668035087302,
|
| 407 |
+
"loss": 2.1271,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.6477272727272727,
|
| 412 |
+
"grad_norm": 1.123046875,
|
| 413 |
+
"learning_rate": 0.00018689850496587674,
|
| 414 |
+
"loss": 2.0622,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.6761363636363638,
|
| 419 |
+
"grad_norm": 1.119140625,
|
| 420 |
+
"learning_rate": 0.00018645340770447595,
|
| 421 |
+
"loss": 2.1017,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.7045454545454546,
|
| 426 |
+
"grad_norm": 0.984375,
|
| 427 |
+
"learning_rate": 0.00018600142402077006,
|
| 428 |
+
"loss": 2.0523,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.7329545454545454,
|
| 433 |
+
"grad_norm": 1.0556640625,
|
| 434 |
+
"learning_rate": 0.00018554258991739452,
|
| 435 |
+
"loss": 2.1738,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.7613636363636362,
|
| 440 |
+
"grad_norm": 1.1376953125,
|
| 441 |
+
"learning_rate": 0.0001850769419426531,
|
| 442 |
+
"loss": 2.107,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.7897727272727273,
|
| 447 |
+
"grad_norm": 1.1875,
|
| 448 |
+
"learning_rate": 0.0001846045171876065,
|
| 449 |
+
"loss": 2.1752,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.8181818181818183,
|
| 454 |
+
"grad_norm": 1.0634765625,
|
| 455 |
+
"learning_rate": 0.00018412535328311814,
|
| 456 |
+
"loss": 2.0615,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.8465909090909092,
|
| 461 |
+
"grad_norm": 1.205078125,
|
| 462 |
+
"learning_rate": 0.00018363948839685636,
|
| 463 |
+
"loss": 2.101,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.875,
|
| 468 |
+
"grad_norm": 1.05859375,
|
| 469 |
+
"learning_rate": 0.00018314696123025454,
|
| 470 |
+
"loss": 2.1308,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.9034090909090908,
|
| 475 |
+
"grad_norm": 1.078125,
|
| 476 |
+
"learning_rate": 0.000182647811015428,
|
| 477 |
+
"loss": 2.1138,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.9318181818181817,
|
| 482 |
+
"grad_norm": 1.015625,
|
| 483 |
+
"learning_rate": 0.00018214207751204918,
|
| 484 |
+
"loss": 2.0551,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.9602272727272727,
|
| 489 |
+
"grad_norm": 1.2529296875,
|
| 490 |
+
"learning_rate": 0.0001816298010041806,
|
| 491 |
+
"loss": 2.1243,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.9886363636363638,
|
| 496 |
+
"grad_norm": 1.0625,
|
| 497 |
+
"learning_rate": 0.0001811110222970659,
|
| 498 |
+
"loss": 2.0663,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.0170454545454546,
|
| 503 |
+
"grad_norm": 1.345703125,
|
| 504 |
+
"learning_rate": 0.0001805857827138798,
|
| 505 |
+
"loss": 2.1003,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.0454545454545454,
|
| 510 |
+
"grad_norm": 1.0517578125,
|
| 511 |
+
"learning_rate": 0.00018005412409243606,
|
| 512 |
+
"loss": 2.0274,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.0738636363636362,
|
| 517 |
+
"grad_norm": 1.021484375,
|
| 518 |
+
"learning_rate": 0.0001795160887818553,
|
| 519 |
+
"loss": 1.9693,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.102272727272727,
|
| 524 |
+
"grad_norm": 1.1298828125,
|
| 525 |
+
"learning_rate": 0.0001789717196391916,
|
| 526 |
+
"loss": 2.0038,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.1306818181818183,
|
| 531 |
+
"grad_norm": 1.080078125,
|
| 532 |
+
"learning_rate": 0.00017842106002601856,
|
| 533 |
+
"loss": 2.0249,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.159090909090909,
|
| 538 |
+
"grad_norm": 1.146484375,
|
| 539 |
+
"learning_rate": 0.00017786415380497553,
|
| 540 |
+
"loss": 1.965,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.1875,
|
| 545 |
+
"grad_norm": 1.2890625,
|
| 546 |
+
"learning_rate": 0.0001773010453362737,
|
| 547 |
+
"loss": 1.9959,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.215909090909091,
|
| 552 |
+
"grad_norm": 1.1259765625,
|
| 553 |
+
"learning_rate": 0.00017673177947416258,
|
| 554 |
+
"loss": 1.9325,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.2443181818181817,
|
| 559 |
+
"grad_norm": 1.15625,
|
| 560 |
+
"learning_rate": 0.00017615640156335712,
|
| 561 |
+
"loss": 2.031,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.2727272727272725,
|
| 566 |
+
"grad_norm": 1.24609375,
|
| 567 |
+
"learning_rate": 0.00017557495743542585,
|
| 568 |
+
"loss": 1.9761,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.3011363636363638,
|
| 573 |
+
"grad_norm": 1.3046875,
|
| 574 |
+
"learning_rate": 0.0001749874934051401,
|
| 575 |
+
"loss": 2.0198,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.3295454545454546,
|
| 580 |
+
"grad_norm": 1.25390625,
|
| 581 |
+
"learning_rate": 0.00017439405626678496,
|
| 582 |
+
"loss": 1.9408,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.3579545454545454,
|
| 587 |
+
"grad_norm": 1.1611328125,
|
| 588 |
+
"learning_rate": 0.00017379469329043167,
|
| 589 |
+
"loss": 2.0076,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.3863636363636362,
|
| 594 |
+
"grad_norm": 1.1669921875,
|
| 595 |
+
"learning_rate": 0.00017318945221817255,
|
| 596 |
+
"loss": 2.0166,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.4147727272727275,
|
| 601 |
+
"grad_norm": 1.212890625,
|
| 602 |
+
"learning_rate": 0.00017257838126031797,
|
| 603 |
+
"loss": 2.0593,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.4431818181818183,
|
| 608 |
+
"grad_norm": 1.2314453125,
|
| 609 |
+
"learning_rate": 0.00017196152909155628,
|
| 610 |
+
"loss": 1.9916,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.471590909090909,
|
| 615 |
+
"grad_norm": 1.1845703125,
|
| 616 |
+
"learning_rate": 0.00017133894484707655,
|
| 617 |
+
"loss": 1.9788,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.5,
|
| 622 |
+
"grad_norm": 1.087890625,
|
| 623 |
+
"learning_rate": 0.00017071067811865476,
|
| 624 |
+
"loss": 1.9791,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.528409090909091,
|
| 629 |
+
"grad_norm": 1.2236328125,
|
| 630 |
+
"learning_rate": 0.00017007677895070357,
|
| 631 |
+
"loss": 2.0564,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.5568181818181817,
|
| 636 |
+
"grad_norm": 1.2998046875,
|
| 637 |
+
"learning_rate": 0.00016943729783628608,
|
| 638 |
+
"loss": 1.9301,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.5852272727272725,
|
| 643 |
+
"grad_norm": 1.1318359375,
|
| 644 |
+
"learning_rate": 0.00016879228571309377,
|
| 645 |
+
"loss": 1.9204,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.6136363636363638,
|
| 650 |
+
"grad_norm": 1.1044921875,
|
| 651 |
+
"learning_rate": 0.00016814179395938913,
|
| 652 |
+
"loss": 2.0031,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.6420454545454546,
|
| 657 |
+
"grad_norm": 1.2138671875,
|
| 658 |
+
"learning_rate": 0.00016748587438991303,
|
| 659 |
+
"loss": 1.9994,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.6704545454545454,
|
| 664 |
+
"grad_norm": 1.1650390625,
|
| 665 |
+
"learning_rate": 0.00016682457925175763,
|
| 666 |
+
"loss": 2.0172,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.6988636363636362,
|
| 671 |
+
"grad_norm": 1.26953125,
|
| 672 |
+
"learning_rate": 0.00016615796122020443,
|
| 673 |
+
"loss": 1.9542,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.7272727272727275,
|
| 678 |
+
"grad_norm": 1.423828125,
|
| 679 |
+
"learning_rate": 0.00016548607339452853,
|
| 680 |
+
"loss": 1.9322,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.7556818181818183,
|
| 685 |
+
"grad_norm": 1.3388671875,
|
| 686 |
+
"learning_rate": 0.00016480896929376907,
|
| 687 |
+
"loss": 1.9912,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.784090909090909,
|
| 692 |
+
"grad_norm": 1.2529296875,
|
| 693 |
+
"learning_rate": 0.0001641267028524661,
|
| 694 |
+
"loss": 1.9818,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.8125,
|
| 699 |
+
"grad_norm": 1.2421875,
|
| 700 |
+
"learning_rate": 0.00016343932841636456,
|
| 701 |
+
"loss": 2.0132,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.840909090909091,
|
| 706 |
+
"grad_norm": 1.4873046875,
|
| 707 |
+
"learning_rate": 0.0001627469007380852,
|
| 708 |
+
"loss": 1.9735,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.8693181818181817,
|
| 713 |
+
"grad_norm": 1.1220703125,
|
| 714 |
+
"learning_rate": 0.00016204947497276345,
|
| 715 |
+
"loss": 1.9642,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.8977272727272725,
|
| 720 |
+
"grad_norm": 1.1171875,
|
| 721 |
+
"learning_rate": 0.00016134710667365596,
|
| 722 |
+
"loss": 1.9564,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.9261363636363638,
|
| 727 |
+
"grad_norm": 1.142578125,
|
| 728 |
+
"learning_rate": 0.00016063985178771555,
|
| 729 |
+
"loss": 1.9701,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.9545454545454546,
|
| 734 |
+
"grad_norm": 1.17578125,
|
| 735 |
+
"learning_rate": 0.0001599277666511347,
|
| 736 |
+
"loss": 1.913,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.9829545454545454,
|
| 741 |
+
"grad_norm": 1.1396484375,
|
| 742 |
+
"learning_rate": 0.00015921090798485832,
|
| 743 |
+
"loss": 2.0151,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.0113636363636362,
|
| 748 |
+
"grad_norm": 1.2216796875,
|
| 749 |
+
"learning_rate": 0.0001584893328900653,
|
| 750 |
+
"loss": 1.9299,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.039772727272727,
|
| 755 |
+
"grad_norm": 1.1787109375,
|
| 756 |
+
"learning_rate": 0.0001577630988436206,
|
| 757 |
+
"loss": 1.899,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.0681818181818183,
|
| 762 |
+
"grad_norm": 1.31640625,
|
| 763 |
+
"learning_rate": 0.0001570322636934964,
|
| 764 |
+
"loss": 1.8602,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.096590909090909,
|
| 769 |
+
"grad_norm": 1.1279296875,
|
| 770 |
+
"learning_rate": 0.00015629688565416478,
|
| 771 |
+
"loss": 1.8292,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.125,
|
| 776 |
+
"grad_norm": 1.2705078125,
|
| 777 |
+
"learning_rate": 0.00015555702330196023,
|
| 778 |
+
"loss": 1.891,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.153409090909091,
|
| 783 |
+
"grad_norm": 1.208984375,
|
| 784 |
+
"learning_rate": 0.000154812735570414,
|
| 785 |
+
"loss": 1.8526,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.1818181818181817,
|
| 790 |
+
"grad_norm": 1.4580078125,
|
| 791 |
+
"learning_rate": 0.00015406408174555976,
|
| 792 |
+
"loss": 1.8514,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.210227272727273,
|
| 797 |
+
"grad_norm": 1.330078125,
|
| 798 |
+
"learning_rate": 0.00015331112146121104,
|
| 799 |
+
"loss": 1.8956,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.2386363636363638,
|
| 804 |
+
"grad_norm": 1.771484375,
|
| 805 |
+
"learning_rate": 0.00015255391469421128,
|
| 806 |
+
"loss": 1.8339,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.2670454545454546,
|
| 811 |
+
"grad_norm": 1.7021484375,
|
| 812 |
+
"learning_rate": 0.00015179252175965633,
|
| 813 |
+
"loss": 1.8952,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.2954545454545454,
|
| 818 |
+
"grad_norm": 1.5029296875,
|
| 819 |
+
"learning_rate": 0.00015102700330609,
|
| 820 |
+
"loss": 1.8727,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.3238636363636362,
|
| 825 |
+
"grad_norm": 1.296875,
|
| 826 |
+
"learning_rate": 0.00015025742031067317,
|
| 827 |
+
"loss": 1.8974,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.3522727272727275,
|
| 832 |
+
"grad_norm": 1.6708984375,
|
| 833 |
+
"learning_rate": 0.00014948383407432678,
|
| 834 |
+
"loss": 1.8314,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.3806818181818183,
|
| 839 |
+
"grad_norm": 1.6494140625,
|
| 840 |
+
"learning_rate": 0.00014870630621684872,
|
| 841 |
+
"loss": 1.8768,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.409090909090909,
|
| 846 |
+
"grad_norm": 1.1337890625,
|
| 847 |
+
"learning_rate": 0.0001479248986720057,
|
| 848 |
+
"loss": 1.8845,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.4375,
|
| 853 |
+
"grad_norm": 1.2841796875,
|
| 854 |
+
"learning_rate": 0.0001471396736825998,
|
| 855 |
+
"loss": 1.8475,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.465909090909091,
|
| 860 |
+
"grad_norm": 1.537109375,
|
| 861 |
+
"learning_rate": 0.00014635069379551055,
|
| 862 |
+
"loss": 1.8741,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.4943181818181817,
|
| 867 |
+
"grad_norm": 1.58203125,
|
| 868 |
+
"learning_rate": 0.00014555802185671298,
|
| 869 |
+
"loss": 1.8274,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.5227272727272725,
|
| 874 |
+
"grad_norm": 1.306640625,
|
| 875 |
+
"learning_rate": 0.00014476172100627127,
|
| 876 |
+
"loss": 1.8671,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.5511363636363638,
|
| 881 |
+
"grad_norm": 1.4541015625,
|
| 882 |
+
"learning_rate": 0.00014396185467330973,
|
| 883 |
+
"loss": 1.8998,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.5795454545454546,
|
| 888 |
+
"grad_norm": 1.3642578125,
|
| 889 |
+
"learning_rate": 0.00014315848657096004,
|
| 890 |
+
"loss": 1.8484,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.6079545454545454,
|
| 895 |
+
"grad_norm": 1.31640625,
|
| 896 |
+
"learning_rate": 0.00014235168069128657,
|
| 897 |
+
"loss": 1.9098,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.6363636363636362,
|
| 902 |
+
"grad_norm": 1.51171875,
|
| 903 |
+
"learning_rate": 0.00014154150130018866,
|
| 904 |
+
"loss": 1.8746,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.6647727272727275,
|
| 909 |
+
"grad_norm": 1.1796875,
|
| 910 |
+
"learning_rate": 0.00014072801293228188,
|
| 911 |
+
"loss": 1.9036,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.6931818181818183,
|
| 916 |
+
"grad_norm": 1.205078125,
|
| 917 |
+
"learning_rate": 0.00013991128038575741,
|
| 918 |
+
"loss": 1.9155,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.721590909090909,
|
| 923 |
+
"grad_norm": 1.373046875,
|
| 924 |
+
"learning_rate": 0.00013909136871722067,
|
| 925 |
+
"loss": 1.8956,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.75,
|
| 930 |
+
"grad_norm": 1.2607421875,
|
| 931 |
+
"learning_rate": 0.000138268343236509,
|
| 932 |
+
"loss": 1.8862,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.778409090909091,
|
| 937 |
+
"grad_norm": 1.5869140625,
|
| 938 |
+
"learning_rate": 0.00013744226950148972,
|
| 939 |
+
"loss": 1.8749,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.8068181818181817,
|
| 944 |
+
"grad_norm": 1.3154296875,
|
| 945 |
+
"learning_rate": 0.00013661321331283796,
|
| 946 |
+
"loss": 1.8831,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.8352272727272725,
|
| 951 |
+
"grad_norm": 1.25,
|
| 952 |
+
"learning_rate": 0.00013578124070879534,
|
| 953 |
+
"loss": 1.8459,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.8636363636363638,
|
| 958 |
+
"grad_norm": 1.1337890625,
|
| 959 |
+
"learning_rate": 0.00013494641795990986,
|
| 960 |
+
"loss": 1.8644,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.8920454545454546,
|
| 965 |
+
"grad_norm": 1.2890625,
|
| 966 |
+
"learning_rate": 0.00013410881156375684,
|
| 967 |
+
"loss": 1.8569,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.9204545454545454,
|
| 972 |
+
"grad_norm": 1.76953125,
|
| 973 |
+
"learning_rate": 0.00013326848823964243,
|
| 974 |
+
"loss": 1.884,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.9488636363636362,
|
| 979 |
+
"grad_norm": 1.3408203125,
|
| 980 |
+
"learning_rate": 0.00013242551492328875,
|
| 981 |
+
"loss": 1.8759,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.9772727272727275,
|
| 986 |
+
"grad_norm": 1.546875,
|
| 987 |
+
"learning_rate": 0.0001315799587615025,
|
| 988 |
+
"loss": 1.8104,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.005681818181818,
|
| 993 |
+
"grad_norm": 1.169921875,
|
| 994 |
+
"learning_rate": 0.0001307318871068261,
|
| 995 |
+
"loss": 1.8617,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.034090909090909,
|
| 1000 |
+
"grad_norm": 1.2158203125,
|
| 1001 |
+
"learning_rate": 0.00012988136751217291,
|
| 1002 |
+
"loss": 1.7548,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.0625,
|
| 1007 |
+
"grad_norm": 1.30078125,
|
| 1008 |
+
"learning_rate": 0.00012902846772544624,
|
| 1009 |
+
"loss": 1.7547,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.090909090909091,
|
| 1014 |
+
"grad_norm": 1.2236328125,
|
| 1015 |
+
"learning_rate": 0.00012817325568414297,
|
| 1016 |
+
"loss": 1.7734,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.119318181818182,
|
| 1021 |
+
"grad_norm": 1.259765625,
|
| 1022 |
+
"learning_rate": 0.000127315799509942,
|
| 1023 |
+
"loss": 1.7146,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.1477272727272725,
|
| 1028 |
+
"grad_norm": 1.4580078125,
|
| 1029 |
+
"learning_rate": 0.0001264561675032779,
|
| 1030 |
+
"loss": 1.788,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.176136363636363,
|
| 1035 |
+
"grad_norm": 1.5205078125,
|
| 1036 |
+
"learning_rate": 0.00012559442813790076,
|
| 1037 |
+
"loss": 1.8247,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.204545454545454,
|
| 1042 |
+
"grad_norm": 1.3359375,
|
| 1043 |
+
"learning_rate": 0.00012473065005542155,
|
| 1044 |
+
"loss": 1.7217,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.232954545454546,
|
| 1049 |
+
"grad_norm": 1.28125,
|
| 1050 |
+
"learning_rate": 0.00012386490205984488,
|
| 1051 |
+
"loss": 1.772,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.261363636363637,
|
| 1056 |
+
"grad_norm": 1.2919921875,
|
| 1057 |
+
"learning_rate": 0.00012299725311208808,
|
| 1058 |
+
"loss": 1.7837,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.2897727272727275,
|
| 1063 |
+
"grad_norm": 1.265625,
|
| 1064 |
+
"learning_rate": 0.00012212777232448838,
|
| 1065 |
+
"loss": 1.7954,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.318181818181818,
|
| 1070 |
+
"grad_norm": 1.5068359375,
|
| 1071 |
+
"learning_rate": 0.00012125652895529766,
|
| 1072 |
+
"loss": 1.7202,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.346590909090909,
|
| 1077 |
+
"grad_norm": 1.56640625,
|
| 1078 |
+
"learning_rate": 0.00012038359240316589,
|
| 1079 |
+
"loss": 1.7618,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.375,
|
| 1084 |
+
"grad_norm": 1.388671875,
|
| 1085 |
+
"learning_rate": 0.00011950903220161285,
|
| 1086 |
+
"loss": 1.758,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.403409090909091,
|
| 1091 |
+
"grad_norm": 1.6416015625,
|
| 1092 |
+
"learning_rate": 0.0001186329180134898,
|
| 1093 |
+
"loss": 1.8011,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.431818181818182,
|
| 1098 |
+
"grad_norm": 1.4912109375,
|
| 1099 |
+
"learning_rate": 0.00011775531962543036,
|
| 1100 |
+
"loss": 1.7798,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.4602272727272725,
|
| 1105 |
+
"grad_norm": 1.65234375,
|
| 1106 |
+
"learning_rate": 0.0001168763069422916,
|
| 1107 |
+
"loss": 1.7664,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.488636363636363,
|
| 1112 |
+
"grad_norm": 1.37109375,
|
| 1113 |
+
"learning_rate": 0.00011599594998158602,
|
| 1114 |
+
"loss": 1.7301,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.517045454545455,
|
| 1119 |
+
"grad_norm": 1.984375,
|
| 1120 |
+
"learning_rate": 0.00011511431886790407,
|
| 1121 |
+
"loss": 1.7749,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.545454545454545,
|
| 1126 |
+
"grad_norm": 1.3369140625,
|
| 1127 |
+
"learning_rate": 0.00011423148382732853,
|
| 1128 |
+
"loss": 1.7646,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.573863636363637,
|
| 1133 |
+
"grad_norm": 1.2080078125,
|
| 1134 |
+
"learning_rate": 0.00011334751518184061,
|
| 1135 |
+
"loss": 1.7962,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.6022727272727275,
|
| 1140 |
+
"grad_norm": 1.3974609375,
|
| 1141 |
+
"learning_rate": 0.0001124624833437186,
|
| 1142 |
+
"loss": 1.805,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.630681818181818,
|
| 1147 |
+
"grad_norm": 1.392578125,
|
| 1148 |
+
"learning_rate": 0.00011157645880992902,
|
| 1149 |
+
"loss": 1.7486,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.659090909090909,
|
| 1154 |
+
"grad_norm": 1.404296875,
|
| 1155 |
+
"learning_rate": 0.00011068951215651132,
|
| 1156 |
+
"loss": 1.7962,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.6875,
|
| 1161 |
+
"grad_norm": 1.380859375,
|
| 1162 |
+
"learning_rate": 0.0001098017140329561,
|
| 1163 |
+
"loss": 1.7722,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.715909090909091,
|
| 1168 |
+
"grad_norm": 1.580078125,
|
| 1169 |
+
"learning_rate": 0.0001089131351565776,
|
| 1170 |
+
"loss": 1.7477,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.744318181818182,
|
| 1175 |
+
"grad_norm": 1.490234375,
|
| 1176 |
+
"learning_rate": 0.00010802384630688078,
|
| 1177 |
+
"loss": 1.7825,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.7727272727272725,
|
| 1182 |
+
"grad_norm": 1.330078125,
|
| 1183 |
+
"learning_rate": 0.00010713391831992323,
|
| 1184 |
+
"loss": 1.7713,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.801136363636363,
|
| 1189 |
+
"grad_norm": 1.4033203125,
|
| 1190 |
+
"learning_rate": 0.00010624342208267292,
|
| 1191 |
+
"loss": 1.74,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.829545454545455,
|
| 1196 |
+
"grad_norm": 1.6806640625,
|
| 1197 |
+
"learning_rate": 0.00010535242852736151,
|
| 1198 |
+
"loss": 1.8173,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.857954545454545,
|
| 1203 |
+
"grad_norm": 1.603515625,
|
| 1204 |
+
"learning_rate": 0.00010446100862583459,
|
| 1205 |
+
"loss": 1.7703,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.886363636363637,
|
| 1210 |
+
"grad_norm": 1.4755859375,
|
| 1211 |
+
"learning_rate": 0.00010356923338389806,
|
| 1212 |
+
"loss": 1.8032,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.9147727272727275,
|
| 1217 |
+
"grad_norm": 1.4833984375,
|
| 1218 |
+
"learning_rate": 0.00010267717383566246,
|
| 1219 |
+
"loss": 1.8321,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.943181818181818,
|
| 1224 |
+
"grad_norm": 1.4462890625,
|
| 1225 |
+
"learning_rate": 0.0001017849010378846,
|
| 1226 |
+
"loss": 1.792,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.971590909090909,
|
| 1231 |
+
"grad_norm": 1.333984375,
|
| 1232 |
+
"learning_rate": 0.00010089248606430775,
|
| 1233 |
+
"loss": 1.7765,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.0,
|
| 1238 |
+
"grad_norm": 1.4609375,
|
| 1239 |
+
"learning_rate": 0.0001,
|
| 1240 |
+
"loss": 1.8083,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.028409090909091,
|
| 1245 |
+
"grad_norm": 1.482421875,
|
| 1246 |
+
"learning_rate": 9.910751393569229e-05,
|
| 1247 |
+
"loss": 1.7054,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.056818181818182,
|
| 1252 |
+
"grad_norm": 1.30859375,
|
| 1253 |
+
"learning_rate": 9.821509896211539e-05,
|
| 1254 |
+
"loss": 1.6561,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.0852272727272725,
|
| 1259 |
+
"grad_norm": 1.3037109375,
|
| 1260 |
+
"learning_rate": 9.732282616433756e-05,
|
| 1261 |
+
"loss": 1.6723,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.113636363636363,
|
| 1266 |
+
"grad_norm": 1.283203125,
|
| 1267 |
+
"learning_rate": 9.643076661610196e-05,
|
| 1268 |
+
"loss": 1.6675,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.142045454545454,
|
| 1273 |
+
"grad_norm": 1.3359375,
|
| 1274 |
+
"learning_rate": 9.553899137416545e-05,
|
| 1275 |
+
"loss": 1.6865,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.170454545454546,
|
| 1280 |
+
"grad_norm": 1.6669921875,
|
| 1281 |
+
"learning_rate": 9.464757147263849e-05,
|
| 1282 |
+
"loss": 1.7227,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.198863636363637,
|
| 1287 |
+
"grad_norm": 1.6875,
|
| 1288 |
+
"learning_rate": 9.37565779173271e-05,
|
| 1289 |
+
"loss": 1.7106,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.2272727272727275,
|
| 1294 |
+
"grad_norm": 1.4619140625,
|
| 1295 |
+
"learning_rate": 9.286608168007678e-05,
|
| 1296 |
+
"loss": 1.7399,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.255681818181818,
|
| 1301 |
+
"grad_norm": 1.4873046875,
|
| 1302 |
+
"learning_rate": 9.197615369311925e-05,
|
| 1303 |
+
"loss": 1.6774,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.284090909090909,
|
| 1308 |
+
"grad_norm": 1.3671875,
|
| 1309 |
+
"learning_rate": 9.108686484342241e-05,
|
| 1310 |
+
"loss": 1.675,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.3125,
|
| 1315 |
+
"grad_norm": 1.3916015625,
|
| 1316 |
+
"learning_rate": 9.019828596704394e-05,
|
| 1317 |
+
"loss": 1.6782,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.340909090909091,
|
| 1322 |
+
"grad_norm": 1.376953125,
|
| 1323 |
+
"learning_rate": 8.931048784348875e-05,
|
| 1324 |
+
"loss": 1.6725,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.369318181818182,
|
| 1329 |
+
"grad_norm": 1.50390625,
|
| 1330 |
+
"learning_rate": 8.842354119007099e-05,
|
| 1331 |
+
"loss": 1.6885,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.3977272727272725,
|
| 1336 |
+
"grad_norm": 1.3623046875,
|
| 1337 |
+
"learning_rate": 8.753751665628141e-05,
|
| 1338 |
+
"loss": 1.6701,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.426136363636363,
|
| 1343 |
+
"grad_norm": 2.06640625,
|
| 1344 |
+
"learning_rate": 8.665248481815941e-05,
|
| 1345 |
+
"loss": 1.7138,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.454545454545454,
|
| 1350 |
+
"grad_norm": 1.341796875,
|
| 1351 |
+
"learning_rate": 8.57685161726715e-05,
|
| 1352 |
+
"loss": 1.6776,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.482954545454546,
|
| 1357 |
+
"grad_norm": 1.62109375,
|
| 1358 |
+
"learning_rate": 8.488568113209593e-05,
|
| 1359 |
+
"loss": 1.6245,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.511363636363637,
|
| 1364 |
+
"grad_norm": 1.4208984375,
|
| 1365 |
+
"learning_rate": 8.400405001841399e-05,
|
| 1366 |
+
"loss": 1.6695,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.5397727272727275,
|
| 1371 |
+
"grad_norm": 1.439453125,
|
| 1372 |
+
"learning_rate": 8.312369305770843e-05,
|
| 1373 |
+
"loss": 1.7594,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.568181818181818,
|
| 1378 |
+
"grad_norm": 1.873046875,
|
| 1379 |
+
"learning_rate": 8.224468037456969e-05,
|
| 1380 |
+
"loss": 1.6707,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.596590909090909,
|
| 1385 |
+
"grad_norm": 1.6162109375,
|
| 1386 |
+
"learning_rate": 8.136708198651021e-05,
|
| 1387 |
+
"loss": 1.7541,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.625,
|
| 1392 |
+
"grad_norm": 1.51171875,
|
| 1393 |
+
"learning_rate": 8.049096779838719e-05,
|
| 1394 |
+
"loss": 1.6912,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.653409090909091,
|
| 1399 |
+
"grad_norm": 1.6669921875,
|
| 1400 |
+
"learning_rate": 7.961640759683416e-05,
|
| 1401 |
+
"loss": 1.6736,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.681818181818182,
|
| 1406 |
+
"grad_norm": 1.654296875,
|
| 1407 |
+
"learning_rate": 7.874347104470234e-05,
|
| 1408 |
+
"loss": 1.6927,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.7102272727272725,
|
| 1413 |
+
"grad_norm": 1.8779296875,
|
| 1414 |
+
"learning_rate": 7.787222767551164e-05,
|
| 1415 |
+
"loss": 1.6773,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.738636363636363,
|
| 1420 |
+
"grad_norm": 1.55078125,
|
| 1421 |
+
"learning_rate": 7.700274688791196e-05,
|
| 1422 |
+
"loss": 1.6698,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.767045454545455,
|
| 1427 |
+
"grad_norm": 1.400390625,
|
| 1428 |
+
"learning_rate": 7.613509794015517e-05,
|
| 1429 |
+
"loss": 1.6778,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.795454545454545,
|
| 1434 |
+
"grad_norm": 1.3837890625,
|
| 1435 |
+
"learning_rate": 7.526934994457844e-05,
|
| 1436 |
+
"loss": 1.7331,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.823863636363637,
|
| 1441 |
+
"grad_norm": 1.6669921875,
|
| 1442 |
+
"learning_rate": 7.440557186209926e-05,
|
| 1443 |
+
"loss": 1.6646,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.8522727272727275,
|
| 1448 |
+
"grad_norm": 1.4169921875,
|
| 1449 |
+
"learning_rate": 7.354383249672212e-05,
|
| 1450 |
+
"loss": 1.6603,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.880681818181818,
|
| 1455 |
+
"grad_norm": 2.63671875,
|
| 1456 |
+
"learning_rate": 7.268420049005807e-05,
|
| 1457 |
+
"loss": 1.6668,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.909090909090909,
|
| 1462 |
+
"grad_norm": 1.6923828125,
|
| 1463 |
+
"learning_rate": 7.182674431585704e-05,
|
| 1464 |
+
"loss": 1.6975,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.9375,
|
| 1469 |
+
"grad_norm": 1.7412109375,
|
| 1470 |
+
"learning_rate": 7.097153227455379e-05,
|
| 1471 |
+
"loss": 1.6874,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.965909090909091,
|
| 1476 |
+
"grad_norm": 1.4111328125,
|
| 1477 |
+
"learning_rate": 7.011863248782711e-05,
|
| 1478 |
+
"loss": 1.7143,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.994318181818182,
|
| 1483 |
+
"grad_norm": 1.3232421875,
|
| 1484 |
+
"learning_rate": 6.92681128931739e-05,
|
| 1485 |
+
"loss": 1.6843,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.0227272727272725,
|
| 1490 |
+
"grad_norm": 1.3046875,
|
| 1491 |
+
"learning_rate": 6.842004123849752e-05,
|
| 1492 |
+
"loss": 1.5913,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.051136363636363,
|
| 1497 |
+
"grad_norm": 1.6064453125,
|
| 1498 |
+
"learning_rate": 6.757448507671128e-05,
|
| 1499 |
+
"loss": 1.6041,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.079545454545454,
|
| 1504 |
+
"grad_norm": 1.4404296875,
|
| 1505 |
+
"learning_rate": 6.673151176035762e-05,
|
| 1506 |
+
"loss": 1.6045,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.107954545454546,
|
| 1511 |
+
"grad_norm": 1.39453125,
|
| 1512 |
+
"learning_rate": 6.589118843624315e-05,
|
| 1513 |
+
"loss": 1.5673,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.136363636363637,
|
| 1518 |
+
"grad_norm": 1.68359375,
|
| 1519 |
+
"learning_rate": 6.505358204009017e-05,
|
| 1520 |
+
"loss": 1.603,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.1647727272727275,
|
| 1525 |
+
"grad_norm": 1.4921875,
|
| 1526 |
+
"learning_rate": 6.421875929120469e-05,
|
| 1527 |
+
"loss": 1.6078,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.193181818181818,
|
| 1532 |
+
"grad_norm": 2.29296875,
|
| 1533 |
+
"learning_rate": 6.338678668716209e-05,
|
| 1534 |
+
"loss": 1.5896,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.221590909090909,
|
| 1539 |
+
"grad_norm": 1.5146484375,
|
| 1540 |
+
"learning_rate": 6.25577304985103e-05,
|
| 1541 |
+
"loss": 1.6124,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.25,
|
| 1546 |
+
"grad_norm": 1.359375,
|
| 1547 |
+
"learning_rate": 6.173165676349103e-05,
|
| 1548 |
+
"loss": 1.6001,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.278409090909091,
|
| 1553 |
+
"grad_norm": 1.6513671875,
|
| 1554 |
+
"learning_rate": 6.090863128277938e-05,
|
| 1555 |
+
"loss": 1.6603,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.306818181818182,
|
| 1560 |
+
"grad_norm": 1.4775390625,
|
| 1561 |
+
"learning_rate": 6.008871961424258e-05,
|
| 1562 |
+
"loss": 1.6338,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.3352272727272725,
|
| 1567 |
+
"grad_norm": 1.5576171875,
|
| 1568 |
+
"learning_rate": 5.9271987067718125e-05,
|
| 1569 |
+
"loss": 1.6557,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.363636363636363,
|
| 1574 |
+
"grad_norm": 1.44921875,
|
| 1575 |
+
"learning_rate": 5.845849869981137e-05,
|
| 1576 |
+
"loss": 1.5855,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.392045454545454,
|
| 1581 |
+
"grad_norm": 1.35546875,
|
| 1582 |
+
"learning_rate": 5.7648319308713464e-05,
|
| 1583 |
+
"loss": 1.6695,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.420454545454546,
|
| 1588 |
+
"grad_norm": 1.3837890625,
|
| 1589 |
+
"learning_rate": 5.684151342903992e-05,
|
| 1590 |
+
"loss": 1.6349,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.448863636363637,
|
| 1595 |
+
"grad_norm": 1.669921875,
|
| 1596 |
+
"learning_rate": 5.603814532669032e-05,
|
| 1597 |
+
"loss": 1.6279,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.4772727272727275,
|
| 1602 |
+
"grad_norm": 1.5185546875,
|
| 1603 |
+
"learning_rate": 5.5238278993728756e-05,
|
| 1604 |
+
"loss": 1.5647,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.505681818181818,
|
| 1609 |
+
"grad_norm": 1.9384765625,
|
| 1610 |
+
"learning_rate": 5.4441978143287066e-05,
|
| 1611 |
+
"loss": 1.6135,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.534090909090909,
|
| 1616 |
+
"grad_norm": 1.4892578125,
|
| 1617 |
+
"learning_rate": 5.364930620448946e-05,
|
| 1618 |
+
"loss": 1.5748,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 6.5625,
|
| 1623 |
+
"grad_norm": 1.4716796875,
|
| 1624 |
+
"learning_rate": 5.286032631740023e-05,
|
| 1625 |
+
"loss": 1.644,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 6.590909090909091,
|
| 1630 |
+
"grad_norm": 1.513671875,
|
| 1631 |
+
"learning_rate": 5.207510132799436e-05,
|
| 1632 |
+
"loss": 1.6049,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 6.619318181818182,
|
| 1637 |
+
"grad_norm": 1.392578125,
|
| 1638 |
+
"learning_rate": 5.1293693783151275e-05,
|
| 1639 |
+
"loss": 1.6798,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 6.6477272727272725,
|
| 1644 |
+
"grad_norm": 1.748046875,
|
| 1645 |
+
"learning_rate": 5.051616592567323e-05,
|
| 1646 |
+
"loss": 1.6548,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 6.676136363636363,
|
| 1651 |
+
"grad_norm": 1.4150390625,
|
| 1652 |
+
"learning_rate": 4.9742579689326874e-05,
|
| 1653 |
+
"loss": 1.599,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 6.704545454545455,
|
| 1658 |
+
"grad_norm": 2.0625,
|
| 1659 |
+
"learning_rate": 4.8972996693910054e-05,
|
| 1660 |
+
"loss": 1.6521,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 6.732954545454545,
|
| 1665 |
+
"grad_norm": 1.7119140625,
|
| 1666 |
+
"learning_rate": 4.82074782403437e-05,
|
| 1667 |
+
"loss": 1.6407,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 6.761363636363637,
|
| 1672 |
+
"grad_norm": 1.619140625,
|
| 1673 |
+
"learning_rate": 4.7446085305788725e-05,
|
| 1674 |
+
"loss": 1.6925,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 6.7897727272727275,
|
| 1679 |
+
"grad_norm": 1.900390625,
|
| 1680 |
+
"learning_rate": 4.668887853878896e-05,
|
| 1681 |
+
"loss": 1.6109,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 6.818181818181818,
|
| 1686 |
+
"grad_norm": 1.6572265625,
|
| 1687 |
+
"learning_rate": 4.593591825444028e-05,
|
| 1688 |
+
"loss": 1.6055,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 6.846590909090909,
|
| 1693 |
+
"grad_norm": 1.3701171875,
|
| 1694 |
+
"learning_rate": 4.518726442958599e-05,
|
| 1695 |
+
"loss": 1.5787,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 6.875,
|
| 1700 |
+
"grad_norm": 1.6171875,
|
| 1701 |
+
"learning_rate": 4.444297669803981e-05,
|
| 1702 |
+
"loss": 1.6368,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 6.903409090909091,
|
| 1707 |
+
"grad_norm": 1.49609375,
|
| 1708 |
+
"learning_rate": 4.370311434583525e-05,
|
| 1709 |
+
"loss": 1.6239,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 6.931818181818182,
|
| 1714 |
+
"grad_norm": 1.7333984375,
|
| 1715 |
+
"learning_rate": 4.296773630650358e-05,
|
| 1716 |
+
"loss": 1.5953,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 6.9602272727272725,
|
| 1721 |
+
"grad_norm": 1.302734375,
|
| 1722 |
+
"learning_rate": 4.223690115637944e-05,
|
| 1723 |
+
"loss": 1.5964,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.988636363636363,
|
| 1728 |
+
"grad_norm": 1.4521484375,
|
| 1729 |
+
"learning_rate": 4.15106671099347e-05,
|
| 1730 |
+
"loss": 1.6531,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.017045454545454,
|
| 1735 |
+
"grad_norm": 1.59375,
|
| 1736 |
+
"learning_rate": 4.0789092015141714e-05,
|
| 1737 |
+
"loss": 1.5872,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.045454545454546,
|
| 1742 |
+
"grad_norm": 1.4765625,
|
| 1743 |
+
"learning_rate": 4.007223334886531e-05,
|
| 1744 |
+
"loss": 1.5554,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.073863636363637,
|
| 1749 |
+
"grad_norm": 1.517578125,
|
| 1750 |
+
"learning_rate": 3.9360148212284475e-05,
|
| 1751 |
+
"loss": 1.5189,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.1022727272727275,
|
| 1756 |
+
"grad_norm": 1.4599609375,
|
| 1757 |
+
"learning_rate": 3.865289332634407e-05,
|
| 1758 |
+
"loss": 1.5523,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.130681818181818,
|
| 1763 |
+
"grad_norm": 1.6494140625,
|
| 1764 |
+
"learning_rate": 3.7950525027236585e-05,
|
| 1765 |
+
"loss": 1.5453,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.159090909090909,
|
| 1770 |
+
"grad_norm": 1.4560546875,
|
| 1771 |
+
"learning_rate": 3.725309926191479e-05,
|
| 1772 |
+
"loss": 1.5612,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.1875,
|
| 1777 |
+
"grad_norm": 1.630859375,
|
| 1778 |
+
"learning_rate": 3.6560671583635467e-05,
|
| 1779 |
+
"loss": 1.6003,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.215909090909091,
|
| 1784 |
+
"grad_norm": 1.3662109375,
|
| 1785 |
+
"learning_rate": 3.5873297147533915e-05,
|
| 1786 |
+
"loss": 1.5927,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.244318181818182,
|
| 1791 |
+
"grad_norm": 1.43359375,
|
| 1792 |
+
"learning_rate": 3.519103070623096e-05,
|
| 1793 |
+
"loss": 1.545,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.2727272727272725,
|
| 1798 |
+
"grad_norm": 1.5654296875,
|
| 1799 |
+
"learning_rate": 3.45139266054715e-05,
|
| 1800 |
+
"loss": 1.5439,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.301136363636363,
|
| 1805 |
+
"grad_norm": 1.845703125,
|
| 1806 |
+
"learning_rate": 3.384203877979559e-05,
|
| 1807 |
+
"loss": 1.5953,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.329545454545454,
|
| 1812 |
+
"grad_norm": 1.80859375,
|
| 1813 |
+
"learning_rate": 3.3175420748242406e-05,
|
| 1814 |
+
"loss": 1.5458,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.357954545454546,
|
| 1819 |
+
"grad_norm": 1.4873046875,
|
| 1820 |
+
"learning_rate": 3.2514125610086955e-05,
|
| 1821 |
+
"loss": 1.5612,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.386363636363637,
|
| 1826 |
+
"grad_norm": 1.5927734375,
|
| 1827 |
+
"learning_rate": 3.185820604061088e-05,
|
| 1828 |
+
"loss": 1.5528,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 7.4147727272727275,
|
| 1833 |
+
"grad_norm": 1.53125,
|
| 1834 |
+
"learning_rate": 3.1207714286906256e-05,
|
| 1835 |
+
"loss": 1.5178,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 7.443181818181818,
|
| 1840 |
+
"grad_norm": 1.455078125,
|
| 1841 |
+
"learning_rate": 3.056270216371395e-05,
|
| 1842 |
+
"loss": 1.5612,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 7.471590909090909,
|
| 1847 |
+
"grad_norm": 1.580078125,
|
| 1848 |
+
"learning_rate": 2.9923221049296446e-05,
|
| 1849 |
+
"loss": 1.5789,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 7.5,
|
| 1854 |
+
"grad_norm": 1.5400390625,
|
| 1855 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 1856 |
+
"loss": 1.5545,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 7.528409090909091,
|
| 1861 |
+
"grad_norm": 1.4794921875,
|
| 1862 |
+
"learning_rate": 2.8661055152923456e-05,
|
| 1863 |
+
"loss": 1.5833,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 7.556818181818182,
|
| 1868 |
+
"grad_norm": 1.5126953125,
|
| 1869 |
+
"learning_rate": 2.8038470908443714e-05,
|
| 1870 |
+
"loss": 1.5997,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 7.5852272727272725,
|
| 1875 |
+
"grad_norm": 1.5869140625,
|
| 1876 |
+
"learning_rate": 2.742161873968202e-05,
|
| 1877 |
+
"loss": 1.5628,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 7.613636363636363,
|
| 1882 |
+
"grad_norm": 1.4697265625,
|
| 1883 |
+
"learning_rate": 2.681054778182748e-05,
|
| 1884 |
+
"loss": 1.5693,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 7.642045454545455,
|
| 1889 |
+
"grad_norm": 1.572265625,
|
| 1890 |
+
"learning_rate": 2.6205306709568354e-05,
|
| 1891 |
+
"loss": 1.6006,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 7.670454545454545,
|
| 1896 |
+
"grad_norm": 1.6005859375,
|
| 1897 |
+
"learning_rate": 2.5605943733215042e-05,
|
| 1898 |
+
"loss": 1.5808,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 7.698863636363637,
|
| 1903 |
+
"grad_norm": 1.3994140625,
|
| 1904 |
+
"learning_rate": 2.501250659485992e-05,
|
| 1905 |
+
"loss": 1.5769,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 7.7272727272727275,
|
| 1910 |
+
"grad_norm": 1.6845703125,
|
| 1911 |
+
"learning_rate": 2.4425042564574184e-05,
|
| 1912 |
+
"loss": 1.6105,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 7.755681818181818,
|
| 1917 |
+
"grad_norm": 1.5673828125,
|
| 1918 |
+
"learning_rate": 2.38435984366429e-05,
|
| 1919 |
+
"loss": 1.5196,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 7.784090909090909,
|
| 1924 |
+
"grad_norm": 2.056640625,
|
| 1925 |
+
"learning_rate": 2.3268220525837437e-05,
|
| 1926 |
+
"loss": 1.5696,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 7.8125,
|
| 1931 |
+
"grad_norm": 1.521484375,
|
| 1932 |
+
"learning_rate": 2.26989546637263e-05,
|
| 1933 |
+
"loss": 1.5564,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 7.840909090909091,
|
| 1938 |
+
"grad_norm": 1.744140625,
|
| 1939 |
+
"learning_rate": 2.2135846195024513e-05,
|
| 1940 |
+
"loss": 1.5608,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 7.869318181818182,
|
| 1945 |
+
"grad_norm": 1.6572265625,
|
| 1946 |
+
"learning_rate": 2.157893997398146e-05,
|
| 1947 |
+
"loss": 1.5823,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 7.8977272727272725,
|
| 1952 |
+
"grad_norm": 1.552734375,
|
| 1953 |
+
"learning_rate": 2.1028280360808407e-05,
|
| 1954 |
+
"loss": 1.5475,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 7.926136363636363,
|
| 1959 |
+
"grad_norm": 1.919921875,
|
| 1960 |
+
"learning_rate": 2.0483911218144715e-05,
|
| 1961 |
+
"loss": 1.5616,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 7.954545454545455,
|
| 1966 |
+
"grad_norm": 1.5625,
|
| 1967 |
+
"learning_rate": 1.994587590756397e-05,
|
| 1968 |
+
"loss": 1.5512,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 7.982954545454545,
|
| 1973 |
+
"grad_norm": 1.666015625,
|
| 1974 |
+
"learning_rate": 1.941421728612023e-05,
|
| 1975 |
+
"loss": 1.6264,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.011363636363637,
|
| 1980 |
+
"grad_norm": 1.4365234375,
|
| 1981 |
+
"learning_rate": 1.8888977702934085e-05,
|
| 1982 |
+
"loss": 1.5251,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.039772727272727,
|
| 1987 |
+
"grad_norm": 1.52734375,
|
| 1988 |
+
"learning_rate": 1.837019899581943e-05,
|
| 1989 |
+
"loss": 1.5552,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.068181818181818,
|
| 1994 |
+
"grad_norm": 1.443359375,
|
| 1995 |
+
"learning_rate": 1.7857922487950874e-05,
|
| 1996 |
+
"loss": 1.5443,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.096590909090908,
|
| 2001 |
+
"grad_norm": 1.4052734375,
|
| 2002 |
+
"learning_rate": 1.7352188984572026e-05,
|
| 2003 |
+
"loss": 1.5339,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.125,
|
| 2008 |
+
"grad_norm": 1.7041015625,
|
| 2009 |
+
"learning_rate": 1.6853038769745467e-05,
|
| 2010 |
+
"loss": 1.5153,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.153409090909092,
|
| 2015 |
+
"grad_norm": 1.615234375,
|
| 2016 |
+
"learning_rate": 1.6360511603143648e-05,
|
| 2017 |
+
"loss": 1.5135,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.181818181818182,
|
| 2022 |
+
"grad_norm": 1.568359375,
|
| 2023 |
+
"learning_rate": 1.587464671688187e-05,
|
| 2024 |
+
"loss": 1.5008,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 8.210227272727273,
|
| 2029 |
+
"grad_norm": 1.654296875,
|
| 2030 |
+
"learning_rate": 1.5395482812393514e-05,
|
| 2031 |
+
"loss": 1.5273,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 8.238636363636363,
|
| 2036 |
+
"grad_norm": 2.16015625,
|
| 2037 |
+
"learning_rate": 1.4923058057346929e-05,
|
| 2038 |
+
"loss": 1.548,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 8.267045454545455,
|
| 2043 |
+
"grad_norm": 1.5595703125,
|
| 2044 |
+
"learning_rate": 1.4457410082605483e-05,
|
| 2045 |
+
"loss": 1.516,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.295454545454545,
|
| 2050 |
+
"grad_norm": 1.5810546875,
|
| 2051 |
+
"learning_rate": 1.3998575979229944e-05,
|
| 2052 |
+
"loss": 1.5658,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 8.323863636363637,
|
| 2057 |
+
"grad_norm": 1.5517578125,
|
| 2058 |
+
"learning_rate": 1.3546592295524074e-05,
|
| 2059 |
+
"loss": 1.5368,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 8.352272727272727,
|
| 2064 |
+
"grad_norm": 1.4638671875,
|
| 2065 |
+
"learning_rate": 1.3101495034123313e-05,
|
| 2066 |
+
"loss": 1.5138,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 8.380681818181818,
|
| 2071 |
+
"grad_norm": 1.9521484375,
|
| 2072 |
+
"learning_rate": 1.2663319649127026e-05,
|
| 2073 |
+
"loss": 1.5193,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 8.409090909090908,
|
| 2078 |
+
"grad_norm": 1.3642578125,
|
| 2079 |
+
"learning_rate": 1.2232101043274436e-05,
|
| 2080 |
+
"loss": 1.5136,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 8.4375,
|
| 2085 |
+
"grad_norm": 1.376953125,
|
| 2086 |
+
"learning_rate": 1.1807873565164506e-05,
|
| 2087 |
+
"loss": 1.487,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 8.465909090909092,
|
| 2092 |
+
"grad_norm": 1.869140625,
|
| 2093 |
+
"learning_rate": 1.139067100651976e-05,
|
| 2094 |
+
"loss": 1.6046,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 8.494318181818182,
|
| 2099 |
+
"grad_norm": 1.9716796875,
|
| 2100 |
+
"learning_rate": 1.0980526599494733e-05,
|
| 2101 |
+
"loss": 1.5021,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 8.522727272727273,
|
| 2106 |
+
"grad_norm": 1.427734375,
|
| 2107 |
+
"learning_rate": 1.057747301402887e-05,
|
| 2108 |
+
"loss": 1.5226,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 8.551136363636363,
|
| 2113 |
+
"grad_norm": 2.681640625,
|
| 2114 |
+
"learning_rate": 1.0181542355244167e-05,
|
| 2115 |
+
"loss": 1.5702,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 8.579545454545455,
|
| 2120 |
+
"grad_norm": 1.568359375,
|
| 2121 |
+
"learning_rate": 9.792766160887868e-06,
|
| 2122 |
+
"loss": 1.5281,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 8.607954545454545,
|
| 2127 |
+
"grad_norm": 1.7783203125,
|
| 2128 |
+
"learning_rate": 9.411175398820271e-06,
|
| 2129 |
+
"loss": 1.5334,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 8.636363636363637,
|
| 2134 |
+
"grad_norm": 1.3994140625,
|
| 2135 |
+
"learning_rate": 9.036800464548157e-06,
|
| 2136 |
+
"loss": 1.5025,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 8.664772727272727,
|
| 2141 |
+
"grad_norm": 1.4384765625,
|
| 2142 |
+
"learning_rate": 8.669671178803485e-06,
|
| 2143 |
+
"loss": 1.6133,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 8.693181818181818,
|
| 2148 |
+
"grad_norm": 1.6689453125,
|
| 2149 |
+
"learning_rate": 8.309816785168034e-06,
|
| 2150 |
+
"loss": 1.4961,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 8.721590909090908,
|
| 2155 |
+
"grad_norm": 1.4951171875,
|
| 2156 |
+
"learning_rate": 7.95726594774413e-06,
|
| 2157 |
+
"loss": 1.5019,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 8.75,
|
| 2162 |
+
"grad_norm": 1.66796875,
|
| 2163 |
+
"learning_rate": 7.612046748871327e-06,
|
| 2164 |
+
"loss": 1.4871,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 8.778409090909092,
|
| 2169 |
+
"grad_norm": 1.859375,
|
| 2170 |
+
"learning_rate": 7.2741866868895395e-06,
|
| 2171 |
+
"loss": 1.5037,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 8.806818181818182,
|
| 2176 |
+
"grad_norm": 1.6630859375,
|
| 2177 |
+
"learning_rate": 6.943712673948644e-06,
|
| 2178 |
+
"loss": 1.5642,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 8.835227272727273,
|
| 2183 |
+
"grad_norm": 1.6640625,
|
| 2184 |
+
"learning_rate": 6.620651033864844e-06,
|
| 2185 |
+
"loss": 1.5448,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 8.863636363636363,
|
| 2190 |
+
"grad_norm": 1.607421875,
|
| 2191 |
+
"learning_rate": 6.3050275000238414e-06,
|
| 2192 |
+
"loss": 1.5506,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 8.892045454545455,
|
| 2197 |
+
"grad_norm": 1.5,
|
| 2198 |
+
"learning_rate": 5.996867213330992e-06,
|
| 2199 |
+
"loss": 1.5726,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 8.920454545454545,
|
| 2204 |
+
"grad_norm": 1.669921875,
|
| 2205 |
+
"learning_rate": 5.696194720208792e-06,
|
| 2206 |
+
"loss": 1.5486,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 8.948863636363637,
|
| 2211 |
+
"grad_norm": 1.9248046875,
|
| 2212 |
+
"learning_rate": 5.403033970641647e-06,
|
| 2213 |
+
"loss": 1.5235,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 8.977272727272727,
|
| 2218 |
+
"grad_norm": 1.5751953125,
|
| 2219 |
+
"learning_rate": 5.1174083162680465e-06,
|
| 2220 |
+
"loss": 1.5525,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 9.005681818181818,
|
| 2225 |
+
"grad_norm": 1.46875,
|
| 2226 |
+
"learning_rate": 4.839340508520562e-06,
|
| 2227 |
+
"loss": 1.5572,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.034090909090908,
|
| 2232 |
+
"grad_norm": 1.6171875,
|
| 2233 |
+
"learning_rate": 4.568852696813619e-06,
|
| 2234 |
+
"loss": 1.5289,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 9.0625,
|
| 2239 |
+
"grad_norm": 1.5341796875,
|
| 2240 |
+
"learning_rate": 4.305966426779118e-06,
|
| 2241 |
+
"loss": 1.5175,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 9.090909090909092,
|
| 2246 |
+
"grad_norm": 1.6708984375,
|
| 2247 |
+
"learning_rate": 4.050702638550275e-06,
|
| 2248 |
+
"loss": 1.5103,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 9.119318181818182,
|
| 2253 |
+
"grad_norm": 1.697265625,
|
| 2254 |
+
"learning_rate": 3.8030816650935776e-06,
|
| 2255 |
+
"loss": 1.5117,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 9.147727272727273,
|
| 2260 |
+
"grad_norm": 1.548828125,
|
| 2261 |
+
"learning_rate": 3.5631232305893046e-06,
|
| 2262 |
+
"loss": 1.4877,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 9.176136363636363,
|
| 2267 |
+
"grad_norm": 1.6552734375,
|
| 2268 |
+
"learning_rate": 3.330846448860259e-06,
|
| 2269 |
+
"loss": 1.5132,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 9.204545454545455,
|
| 2274 |
+
"grad_norm": 1.5751953125,
|
| 2275 |
+
"learning_rate": 3.1062698218492724e-06,
|
| 2276 |
+
"loss": 1.488,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 9.232954545454545,
|
| 2281 |
+
"grad_norm": 1.9609375,
|
| 2282 |
+
"learning_rate": 2.889411238145545e-06,
|
| 2283 |
+
"loss": 1.5202,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 9.261363636363637,
|
| 2288 |
+
"grad_norm": 1.51953125,
|
| 2289 |
+
"learning_rate": 2.6802879715596585e-06,
|
| 2290 |
+
"loss": 1.5516,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 9.289772727272727,
|
| 2295 |
+
"grad_norm": 1.626953125,
|
| 2296 |
+
"learning_rate": 2.4789166797476228e-06,
|
| 2297 |
+
"loss": 1.4951,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 9.318181818181818,
|
| 2302 |
+
"grad_norm": 1.8134765625,
|
| 2303 |
+
"learning_rate": 2.2853134028840594e-06,
|
| 2304 |
+
"loss": 1.5397,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 9.346590909090908,
|
| 2309 |
+
"grad_norm": 1.62109375,
|
| 2310 |
+
"learning_rate": 2.0994935623844692e-06,
|
| 2311 |
+
"loss": 1.5236,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 9.375,
|
| 2316 |
+
"grad_norm": 1.3828125,
|
| 2317 |
+
"learning_rate": 1.921471959676957e-06,
|
| 2318 |
+
"loss": 1.4827,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 9.403409090909092,
|
| 2323 |
+
"grad_norm": 1.5693359375,
|
| 2324 |
+
"learning_rate": 1.751262775023077e-06,
|
| 2325 |
+
"loss": 1.5021,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 9.431818181818182,
|
| 2330 |
+
"grad_norm": 1.46484375,
|
| 2331 |
+
"learning_rate": 1.5888795663883904e-06,
|
| 2332 |
+
"loss": 1.4966,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 9.460227272727273,
|
| 2337 |
+
"grad_norm": 1.6455078125,
|
| 2338 |
+
"learning_rate": 1.4343352683625411e-06,
|
| 2339 |
+
"loss": 1.5455,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 9.488636363636363,
|
| 2344 |
+
"grad_norm": 1.666015625,
|
| 2345 |
+
"learning_rate": 1.2876421911288905e-06,
|
| 2346 |
+
"loss": 1.4876,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 9.517045454545455,
|
| 2351 |
+
"grad_norm": 1.6396484375,
|
| 2352 |
+
"learning_rate": 1.148812019483958e-06,
|
| 2353 |
+
"loss": 1.5912,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 9.545454545454545,
|
| 2358 |
+
"grad_norm": 2.1328125,
|
| 2359 |
+
"learning_rate": 1.0178558119067315e-06,
|
| 2360 |
+
"loss": 1.5355,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 9.573863636363637,
|
| 2365 |
+
"grad_norm": 2.03515625,
|
| 2366 |
+
"learning_rate": 8.947839996777285e-07,
|
| 2367 |
+
"loss": 1.5147,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 9.602272727272727,
|
| 2372 |
+
"grad_norm": 1.806640625,
|
| 2373 |
+
"learning_rate": 7.796063860481595e-07,
|
| 2374 |
+
"loss": 1.53,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 9.630681818181818,
|
| 2379 |
+
"grad_norm": 1.5048828125,
|
| 2380 |
+
"learning_rate": 6.723321454590092e-07,
|
| 2381 |
+
"loss": 1.5051,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 9.659090909090908,
|
| 2386 |
+
"grad_norm": 1.701171875,
|
| 2387 |
+
"learning_rate": 5.729698228102653e-07,
|
| 2388 |
+
"loss": 1.4818,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 9.6875,
|
| 2393 |
+
"grad_norm": 1.6064453125,
|
| 2394 |
+
"learning_rate": 4.815273327803182e-07,
|
| 2395 |
+
"loss": 1.5438,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 9.715909090909092,
|
| 2400 |
+
"grad_norm": 2.04296875,
|
| 2401 |
+
"learning_rate": 3.9801195919541014e-07,
|
| 2402 |
+
"loss": 1.551,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 9.744318181818182,
|
| 2407 |
+
"grad_norm": 1.8115234375,
|
| 2408 |
+
"learning_rate": 3.224303544495766e-07,
|
| 2409 |
+
"loss": 1.4888,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 9.772727272727273,
|
| 2414 |
+
"grad_norm": 1.923828125,
|
| 2415 |
+
"learning_rate": 2.547885389746485e-07,
|
| 2416 |
+
"loss": 1.5491,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 9.801136363636363,
|
| 2421 |
+
"grad_norm": 2.728515625,
|
| 2422 |
+
"learning_rate": 1.9509190076074657e-07,
|
| 2423 |
+
"loss": 1.5374,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 9.829545454545455,
|
| 2428 |
+
"grad_norm": 1.5556640625,
|
| 2429 |
+
"learning_rate": 1.4334519492711362e-07,
|
| 2430 |
+
"loss": 1.5385,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 9.857954545454545,
|
| 2435 |
+
"grad_norm": 1.5185546875,
|
| 2436 |
+
"learning_rate": 9.955254334328423e-08,
|
| 2437 |
+
"loss": 1.4884,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 9.886363636363637,
|
| 2442 |
+
"grad_norm": 1.48046875,
|
| 2443 |
+
"learning_rate": 6.37174343008251e-08,
|
| 2444 |
+
"loss": 1.5733,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 9.914772727272727,
|
| 2449 |
+
"grad_norm": 1.8115234375,
|
| 2450 |
+
"learning_rate": 3.584272223546847e-08,
|
| 2451 |
+
"loss": 1.5452,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 9.943181818181818,
|
| 2456 |
+
"grad_norm": 1.5380859375,
|
| 2457 |
+
"learning_rate": 1.593062749967178e-08,
|
| 2458 |
+
"loss": 1.5122,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 9.971590909090908,
|
| 2463 |
+
"grad_norm": 1.63671875,
|
| 2464 |
+
"learning_rate": 3.982736185859093e-09,
|
| 2465 |
+
"loss": 1.5273,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 10.0,
|
| 2470 |
+
"grad_norm": 1.5048828125,
|
| 2471 |
+
"learning_rate": 0.0,
|
| 2472 |
+
"loss": 1.5284,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 10.0,
|
| 2477 |
+
"step": 3520,
|
| 2478 |
+
"total_flos": 2.31165810180096e+17,
|
| 2479 |
+
"train_loss": 1.80265548852357,
|
| 2480 |
+
"train_runtime": 3529.4078,
|
| 2481 |
+
"train_samples_per_second": 3.989,
|
| 2482 |
+
"train_steps_per_second": 0.997
|
| 2483 |
+
}
|
| 2484 |
+
],
|
| 2485 |
+
"logging_steps": 10,
|
| 2486 |
+
"max_steps": 3520,
|
| 2487 |
+
"num_input_tokens_seen": 0,
|
| 2488 |
+
"num_train_epochs": 10,
|
| 2489 |
+
"save_steps": 500,
|
| 2490 |
+
"total_flos": 2.31165810180096e+17,
|
| 2491 |
+
"train_batch_size": 1,
|
| 2492 |
+
"trial_name": null,
|
| 2493 |
+
"trial_params": null
|
| 2494 |
+
}
|
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:b44ba095725710dadf4f5b58049664f468efcd16bf0f82f1eb07cc31fb95038e
|
| 3 |
size 5048
|