Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- trainer_state.json +2578 -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:ebed29c250b665fc721dda6b3fe56b828d995edb5051f585c9f46f6734d72291
|
| 3 |
size 47236352
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2578 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.986282578875171,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3640,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.027434842249657063,
|
| 13 |
+
"grad_norm": 9.328125,
|
| 14 |
+
"learning_rate": 0.00019999627553166294,
|
| 15 |
+
"loss": 2.6306,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05486968449931413,
|
| 20 |
+
"grad_norm": 5.1328125,
|
| 21 |
+
"learning_rate": 0.00019998510240408496,
|
| 22 |
+
"loss": 2.4194,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0823045267489712,
|
| 27 |
+
"grad_norm": 5.58203125,
|
| 28 |
+
"learning_rate": 0.0001999664814495453,
|
| 29 |
+
"loss": 2.336,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10973936899862825,
|
| 34 |
+
"grad_norm": 3.78515625,
|
| 35 |
+
"learning_rate": 0.00019994041405510705,
|
| 36 |
+
"loss": 2.4327,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.13717421124828533,
|
| 41 |
+
"grad_norm": 3.041015625,
|
| 42 |
+
"learning_rate": 0.00019990690216251396,
|
| 43 |
+
"loss": 2.3063,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1646090534979424,
|
| 48 |
+
"grad_norm": 3.302734375,
|
| 49 |
+
"learning_rate": 0.0001998659482680456,
|
| 50 |
+
"loss": 2.3151,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19204389574759945,
|
| 55 |
+
"grad_norm": 3.869140625,
|
| 56 |
+
"learning_rate": 0.00019981755542233177,
|
| 57 |
+
"loss": 2.3566,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2194787379972565,
|
| 62 |
+
"grad_norm": 5.70703125,
|
| 63 |
+
"learning_rate": 0.0001997617272301248,
|
| 64 |
+
"loss": 2.3368,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.24691358024691357,
|
| 69 |
+
"grad_norm": 4.81640625,
|
| 70 |
+
"learning_rate": 0.00019969846785003134,
|
| 71 |
+
"loss": 2.3303,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.27434842249657065,
|
| 76 |
+
"grad_norm": 4.6015625,
|
| 77 |
+
"learning_rate": 0.00019962778199420265,
|
| 78 |
+
"loss": 2.4144,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3017832647462277,
|
| 83 |
+
"grad_norm": 4.52734375,
|
| 84 |
+
"learning_rate": 0.00019954967492798333,
|
| 85 |
+
"loss": 2.4014,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3292181069958848,
|
| 90 |
+
"grad_norm": 4.68359375,
|
| 91 |
+
"learning_rate": 0.0001994641524695193,
|
| 92 |
+
"loss": 2.4312,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.35665294924554186,
|
| 97 |
+
"grad_norm": 3.36328125,
|
| 98 |
+
"learning_rate": 0.00019937122098932428,
|
| 99 |
+
"loss": 2.3563,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3840877914951989,
|
| 104 |
+
"grad_norm": 3.3984375,
|
| 105 |
+
"learning_rate": 0.0001992708874098054,
|
| 106 |
+
"loss": 2.351,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.411522633744856,
|
| 111 |
+
"grad_norm": 3.103515625,
|
| 112 |
+
"learning_rate": 0.0001991631592047475,
|
| 113 |
+
"loss": 2.2869,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.438957475994513,
|
| 118 |
+
"grad_norm": 2.904296875,
|
| 119 |
+
"learning_rate": 0.00019904804439875633,
|
| 120 |
+
"loss": 2.364,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4663923182441701,
|
| 125 |
+
"grad_norm": 4.171875,
|
| 126 |
+
"learning_rate": 0.00019892555156666089,
|
| 127 |
+
"loss": 2.362,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.49382716049382713,
|
| 132 |
+
"grad_norm": 3.611328125,
|
| 133 |
+
"learning_rate": 0.00019879568983287467,
|
| 134 |
+
"loss": 2.2994,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5212620027434842,
|
| 139 |
+
"grad_norm": 2.470703125,
|
| 140 |
+
"learning_rate": 0.00019865846887071596,
|
| 141 |
+
"loss": 2.2469,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5486968449931413,
|
| 146 |
+
"grad_norm": 2.314453125,
|
| 147 |
+
"learning_rate": 0.0001985138989016874,
|
| 148 |
+
"loss": 2.2429,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5761316872427984,
|
| 153 |
+
"grad_norm": 2.720703125,
|
| 154 |
+
"learning_rate": 0.00019836199069471437,
|
| 155 |
+
"loss": 2.3157,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6035665294924554,
|
| 160 |
+
"grad_norm": 2.998046875,
|
| 161 |
+
"learning_rate": 0.00019820275556534304,
|
| 162 |
+
"loss": 2.2214,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6310013717421125,
|
| 167 |
+
"grad_norm": 3.96484375,
|
| 168 |
+
"learning_rate": 0.00019803620537489736,
|
| 169 |
+
"loss": 2.3018,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6584362139917695,
|
| 174 |
+
"grad_norm": 2.982421875,
|
| 175 |
+
"learning_rate": 0.00019786235252959553,
|
| 176 |
+
"loss": 2.2603,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6858710562414266,
|
| 181 |
+
"grad_norm": 2.01953125,
|
| 182 |
+
"learning_rate": 0.00019768120997962592,
|
| 183 |
+
"loss": 2.3007,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7133058984910837,
|
| 188 |
+
"grad_norm": 2.552734375,
|
| 189 |
+
"learning_rate": 0.00019749279121818235,
|
| 190 |
+
"loss": 2.281,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7407407407407407,
|
| 195 |
+
"grad_norm": 2.98046875,
|
| 196 |
+
"learning_rate": 0.00019729711028045909,
|
| 197 |
+
"loss": 2.2611,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7681755829903978,
|
| 202 |
+
"grad_norm": 2.716796875,
|
| 203 |
+
"learning_rate": 0.0001970941817426052,
|
| 204 |
+
"loss": 2.1961,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7956104252400549,
|
| 209 |
+
"grad_norm": 2.318359375,
|
| 210 |
+
"learning_rate": 0.00019688402072063903,
|
| 211 |
+
"loss": 2.2652,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.823045267489712,
|
| 216 |
+
"grad_norm": 2.53515625,
|
| 217 |
+
"learning_rate": 0.00019666664286932198,
|
| 218 |
+
"loss": 2.1895,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.850480109739369,
|
| 223 |
+
"grad_norm": 2.451171875,
|
| 224 |
+
"learning_rate": 0.0001964420643809925,
|
| 225 |
+
"loss": 2.252,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.877914951989026,
|
| 230 |
+
"grad_norm": 1.8544921875,
|
| 231 |
+
"learning_rate": 0.00019621030198436006,
|
| 232 |
+
"loss": 2.1616,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9053497942386831,
|
| 237 |
+
"grad_norm": 2.611328125,
|
| 238 |
+
"learning_rate": 0.00019597137294325877,
|
| 239 |
+
"loss": 2.1698,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.9327846364883402,
|
| 244 |
+
"grad_norm": 2.349609375,
|
| 245 |
+
"learning_rate": 0.0001957252950553616,
|
| 246 |
+
"loss": 2.2043,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9602194787379973,
|
| 251 |
+
"grad_norm": 2.169921875,
|
| 252 |
+
"learning_rate": 0.00019547208665085457,
|
| 253 |
+
"loss": 2.1506,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9876543209876543,
|
| 258 |
+
"grad_norm": 1.8349609375,
|
| 259 |
+
"learning_rate": 0.00019521176659107142,
|
| 260 |
+
"loss": 2.1987,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.0150891632373114,
|
| 265 |
+
"grad_norm": 1.9765625,
|
| 266 |
+
"learning_rate": 0.00019494435426708855,
|
| 267 |
+
"loss": 2.1909,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.0425240054869684,
|
| 272 |
+
"grad_norm": 1.65234375,
|
| 273 |
+
"learning_rate": 0.0001946698695982806,
|
| 274 |
+
"loss": 2.1928,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.0699588477366255,
|
| 279 |
+
"grad_norm": 1.677734375,
|
| 280 |
+
"learning_rate": 0.00019438833303083678,
|
| 281 |
+
"loss": 2.1761,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0973936899862826,
|
| 286 |
+
"grad_norm": 1.94140625,
|
| 287 |
+
"learning_rate": 0.00019409976553623766,
|
| 288 |
+
"loss": 2.1634,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.1248285322359397,
|
| 293 |
+
"grad_norm": 1.8623046875,
|
| 294 |
+
"learning_rate": 0.00019380418860969322,
|
| 295 |
+
"loss": 2.2044,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.1522633744855968,
|
| 300 |
+
"grad_norm": 1.7763671875,
|
| 301 |
+
"learning_rate": 0.0001935016242685415,
|
| 302 |
+
"loss": 2.1264,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.1796982167352539,
|
| 307 |
+
"grad_norm": 2.4609375,
|
| 308 |
+
"learning_rate": 0.0001931920950506087,
|
| 309 |
+
"loss": 2.1819,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.2071330589849107,
|
| 314 |
+
"grad_norm": 2.990234375,
|
| 315 |
+
"learning_rate": 0.00019287562401253022,
|
| 316 |
+
"loss": 2.1799,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.2345679012345678,
|
| 321 |
+
"grad_norm": 4.86328125,
|
| 322 |
+
"learning_rate": 0.00019255223472803334,
|
| 323 |
+
"loss": 2.1497,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.262002743484225,
|
| 328 |
+
"grad_norm": 2.998046875,
|
| 329 |
+
"learning_rate": 0.00019222195128618106,
|
| 330 |
+
"loss": 2.0783,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.289437585733882,
|
| 335 |
+
"grad_norm": 2.033203125,
|
| 336 |
+
"learning_rate": 0.00019188479828957772,
|
| 337 |
+
"loss": 2.1195,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.316872427983539,
|
| 342 |
+
"grad_norm": 1.9560546875,
|
| 343 |
+
"learning_rate": 0.00019154080085253666,
|
| 344 |
+
"loss": 2.0549,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.3443072702331962,
|
| 349 |
+
"grad_norm": 2.150390625,
|
| 350 |
+
"learning_rate": 0.00019118998459920902,
|
| 351 |
+
"loss": 2.2041,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.3717421124828533,
|
| 356 |
+
"grad_norm": 1.9931640625,
|
| 357 |
+
"learning_rate": 0.0001908323756616754,
|
| 358 |
+
"loss": 2.1735,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3991769547325104,
|
| 363 |
+
"grad_norm": 1.732421875,
|
| 364 |
+
"learning_rate": 0.0001904680006779991,
|
| 365 |
+
"loss": 2.1329,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.4266117969821672,
|
| 370 |
+
"grad_norm": 1.703125,
|
| 371 |
+
"learning_rate": 0.0001900968867902419,
|
| 372 |
+
"loss": 2.075,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.4540466392318243,
|
| 377 |
+
"grad_norm": 2.724609375,
|
| 378 |
+
"learning_rate": 0.00018971906164244232,
|
| 379 |
+
"loss": 2.1452,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.4814814814814814,
|
| 384 |
+
"grad_norm": 3.53125,
|
| 385 |
+
"learning_rate": 0.00018933455337855632,
|
| 386 |
+
"loss": 2.081,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.5089163237311385,
|
| 391 |
+
"grad_norm": 2.15234375,
|
| 392 |
+
"learning_rate": 0.000188943390640361,
|
| 393 |
+
"loss": 2.1311,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.5363511659807956,
|
| 398 |
+
"grad_norm": 2.5078125,
|
| 399 |
+
"learning_rate": 0.000188545602565321,
|
| 400 |
+
"loss": 2.1465,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.5637860082304527,
|
| 405 |
+
"grad_norm": 2.421875,
|
| 406 |
+
"learning_rate": 0.00018814121878441814,
|
| 407 |
+
"loss": 2.242,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.5912208504801097,
|
| 412 |
+
"grad_norm": 2.095703125,
|
| 413 |
+
"learning_rate": 0.0001877302694199442,
|
| 414 |
+
"loss": 2.1339,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.6186556927297668,
|
| 419 |
+
"grad_norm": 1.5791015625,
|
| 420 |
+
"learning_rate": 0.00018731278508325708,
|
| 421 |
+
"loss": 2.1318,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.646090534979424,
|
| 426 |
+
"grad_norm": 1.8525390625,
|
| 427 |
+
"learning_rate": 0.00018688879687250067,
|
| 428 |
+
"loss": 2.108,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.673525377229081,
|
| 433 |
+
"grad_norm": 1.310546875,
|
| 434 |
+
"learning_rate": 0.00018645833637028825,
|
| 435 |
+
"loss": 2.2039,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.700960219478738,
|
| 440 |
+
"grad_norm": 1.50390625,
|
| 441 |
+
"learning_rate": 0.0001860214356413501,
|
| 442 |
+
"loss": 2.1096,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.7283950617283952,
|
| 447 |
+
"grad_norm": 1.6025390625,
|
| 448 |
+
"learning_rate": 0.00018557812723014476,
|
| 449 |
+
"loss": 2.0806,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.7558299039780523,
|
| 454 |
+
"grad_norm": 2.197265625,
|
| 455 |
+
"learning_rate": 0.00018512844415843514,
|
| 456 |
+
"loss": 2.0545,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.7832647462277091,
|
| 461 |
+
"grad_norm": 1.8056640625,
|
| 462 |
+
"learning_rate": 0.00018467241992282843,
|
| 463 |
+
"loss": 2.0552,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.8106995884773662,
|
| 468 |
+
"grad_norm": 1.40234375,
|
| 469 |
+
"learning_rate": 0.00018421008849228118,
|
| 470 |
+
"loss": 2.0906,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.8381344307270233,
|
| 475 |
+
"grad_norm": 1.7080078125,
|
| 476 |
+
"learning_rate": 0.0001837414843055689,
|
| 477 |
+
"loss": 2.0512,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.8655692729766804,
|
| 482 |
+
"grad_norm": 1.6162109375,
|
| 483 |
+
"learning_rate": 0.00018326664226872065,
|
| 484 |
+
"loss": 2.0945,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.8930041152263375,
|
| 489 |
+
"grad_norm": 1.603515625,
|
| 490 |
+
"learning_rate": 0.0001827855977524191,
|
| 491 |
+
"loss": 2.0553,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.9204389574759944,
|
| 496 |
+
"grad_norm": 1.5771484375,
|
| 497 |
+
"learning_rate": 0.00018229838658936564,
|
| 498 |
+
"loss": 2.0876,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.9478737997256514,
|
| 503 |
+
"grad_norm": 1.400390625,
|
| 504 |
+
"learning_rate": 0.0001818050450716113,
|
| 505 |
+
"loss": 2.0565,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.9753086419753085,
|
| 510 |
+
"grad_norm": 2.265625,
|
| 511 |
+
"learning_rate": 0.00018130560994785325,
|
| 512 |
+
"loss": 2.0883,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.0027434842249656,
|
| 517 |
+
"grad_norm": 3.23046875,
|
| 518 |
+
"learning_rate": 0.00018080011842069765,
|
| 519 |
+
"loss": 2.0904,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.0301783264746227,
|
| 524 |
+
"grad_norm": 1.611328125,
|
| 525 |
+
"learning_rate": 0.00018028860814388827,
|
| 526 |
+
"loss": 1.9926,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.05761316872428,
|
| 531 |
+
"grad_norm": 1.638671875,
|
| 532 |
+
"learning_rate": 0.00017977111721950164,
|
| 533 |
+
"loss": 1.9994,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.085048010973937,
|
| 538 |
+
"grad_norm": 1.7724609375,
|
| 539 |
+
"learning_rate": 0.00017924768419510904,
|
| 540 |
+
"loss": 2.0709,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.112482853223594,
|
| 545 |
+
"grad_norm": 1.9208984375,
|
| 546 |
+
"learning_rate": 0.00017871834806090501,
|
| 547 |
+
"loss": 1.9618,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.139917695473251,
|
| 552 |
+
"grad_norm": 2.603515625,
|
| 553 |
+
"learning_rate": 0.000178183148246803,
|
| 554 |
+
"loss": 1.9531,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.167352537722908,
|
| 559 |
+
"grad_norm": 2.236328125,
|
| 560 |
+
"learning_rate": 0.0001776421246194982,
|
| 561 |
+
"loss": 2.0776,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.1947873799725652,
|
| 566 |
+
"grad_norm": 2.79296875,
|
| 567 |
+
"learning_rate": 0.00017709531747949796,
|
| 568 |
+
"loss": 2.0563,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.2222222222222223,
|
| 573 |
+
"grad_norm": 2.55078125,
|
| 574 |
+
"learning_rate": 0.00017654276755811997,
|
| 575 |
+
"loss": 2.052,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.2496570644718794,
|
| 580 |
+
"grad_norm": 2.2421875,
|
| 581 |
+
"learning_rate": 0.0001759845160144579,
|
| 582 |
+
"loss": 2.0051,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.2770919067215365,
|
| 587 |
+
"grad_norm": 2.35546875,
|
| 588 |
+
"learning_rate": 0.00017542060443231572,
|
| 589 |
+
"loss": 2.0448,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.3045267489711936,
|
| 594 |
+
"grad_norm": 1.953125,
|
| 595 |
+
"learning_rate": 0.00017485107481711012,
|
| 596 |
+
"loss": 2.068,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.3319615912208507,
|
| 601 |
+
"grad_norm": 1.9912109375,
|
| 602 |
+
"learning_rate": 0.00017427596959274143,
|
| 603 |
+
"loss": 2.0173,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.3593964334705078,
|
| 608 |
+
"grad_norm": 1.841796875,
|
| 609 |
+
"learning_rate": 0.00017369533159843369,
|
| 610 |
+
"loss": 1.9539,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.386831275720165,
|
| 615 |
+
"grad_norm": 1.9970703125,
|
| 616 |
+
"learning_rate": 0.00017310920408554332,
|
| 617 |
+
"loss": 1.9894,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.4142661179698215,
|
| 622 |
+
"grad_norm": 1.80078125,
|
| 623 |
+
"learning_rate": 0.00017251763071433765,
|
| 624 |
+
"loss": 2.0438,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.4417009602194786,
|
| 629 |
+
"grad_norm": 1.8701171875,
|
| 630 |
+
"learning_rate": 0.00017192065555074245,
|
| 631 |
+
"loss": 2.0079,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.4691358024691357,
|
| 636 |
+
"grad_norm": 1.9775390625,
|
| 637 |
+
"learning_rate": 0.00017131832306305965,
|
| 638 |
+
"loss": 2.0738,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.4965706447187928,
|
| 643 |
+
"grad_norm": 2.087890625,
|
| 644 |
+
"learning_rate": 0.00017071067811865476,
|
| 645 |
+
"loss": 2.0805,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.52400548696845,
|
| 650 |
+
"grad_norm": 2.212890625,
|
| 651 |
+
"learning_rate": 0.00017009776598061495,
|
| 652 |
+
"loss": 2.0563,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.551440329218107,
|
| 657 |
+
"grad_norm": 2.291015625,
|
| 658 |
+
"learning_rate": 0.00016947963230437725,
|
| 659 |
+
"loss": 2.0289,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.578875171467764,
|
| 664 |
+
"grad_norm": 2.0625,
|
| 665 |
+
"learning_rate": 0.0001688563231343277,
|
| 666 |
+
"loss": 2.0648,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.606310013717421,
|
| 671 |
+
"grad_norm": 2.197265625,
|
| 672 |
+
"learning_rate": 0.00016822788490037177,
|
| 673 |
+
"loss": 1.9541,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.633744855967078,
|
| 678 |
+
"grad_norm": 1.5,
|
| 679 |
+
"learning_rate": 0.00016759436441447545,
|
| 680 |
+
"loss": 2.0415,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.6611796982167353,
|
| 685 |
+
"grad_norm": 1.7353515625,
|
| 686 |
+
"learning_rate": 0.00016695580886717858,
|
| 687 |
+
"loss": 2.0242,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.6886145404663924,
|
| 692 |
+
"grad_norm": 2.115234375,
|
| 693 |
+
"learning_rate": 0.00016631226582407952,
|
| 694 |
+
"loss": 2.0,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.7160493827160495,
|
| 699 |
+
"grad_norm": 1.78515625,
|
| 700 |
+
"learning_rate": 0.00016566378322229204,
|
| 701 |
+
"loss": 2.0559,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.7434842249657065,
|
| 706 |
+
"grad_norm": 1.6708984375,
|
| 707 |
+
"learning_rate": 0.00016501040936687443,
|
| 708 |
+
"loss": 2.0658,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.7709190672153636,
|
| 713 |
+
"grad_norm": 2.06640625,
|
| 714 |
+
"learning_rate": 0.00016435219292723147,
|
| 715 |
+
"loss": 2.0381,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.7983539094650207,
|
| 720 |
+
"grad_norm": 1.99609375,
|
| 721 |
+
"learning_rate": 0.00016368918293348892,
|
| 722 |
+
"loss": 1.9942,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.825788751714678,
|
| 727 |
+
"grad_norm": 2.3125,
|
| 728 |
+
"learning_rate": 0.00016302142877284138,
|
| 729 |
+
"loss": 2.0459,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.8532235939643344,
|
| 734 |
+
"grad_norm": 2.109375,
|
| 735 |
+
"learning_rate": 0.00016234898018587337,
|
| 736 |
+
"loss": 1.9964,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.8806584362139915,
|
| 741 |
+
"grad_norm": 1.828125,
|
| 742 |
+
"learning_rate": 0.00016167188726285434,
|
| 743 |
+
"loss": 1.9702,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.9080932784636486,
|
| 748 |
+
"grad_norm": 1.609375,
|
| 749 |
+
"learning_rate": 0.00016099020044000727,
|
| 750 |
+
"loss": 1.971,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.9355281207133057,
|
| 755 |
+
"grad_norm": 1.662109375,
|
| 756 |
+
"learning_rate": 0.00016030397049575203,
|
| 757 |
+
"loss": 2.0445,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.962962962962963,
|
| 762 |
+
"grad_norm": 2.14453125,
|
| 763 |
+
"learning_rate": 0.00015961324854692254,
|
| 764 |
+
"loss": 1.9905,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.99039780521262,
|
| 769 |
+
"grad_norm": 1.71875,
|
| 770 |
+
"learning_rate": 0.00015891808604495938,
|
| 771 |
+
"loss": 2.0048,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.017832647462277,
|
| 776 |
+
"grad_norm": 1.599609375,
|
| 777 |
+
"learning_rate": 0.00015821853477207708,
|
| 778 |
+
"loss": 2.0107,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.045267489711934,
|
| 783 |
+
"grad_norm": 2.4765625,
|
| 784 |
+
"learning_rate": 0.00015751464683740697,
|
| 785 |
+
"loss": 1.9425,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.072702331961591,
|
| 790 |
+
"grad_norm": 3.009765625,
|
| 791 |
+
"learning_rate": 0.00015680647467311557,
|
| 792 |
+
"loss": 1.9891,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.1001371742112482,
|
| 797 |
+
"grad_norm": 2.1328125,
|
| 798 |
+
"learning_rate": 0.00015609407103049896,
|
| 799 |
+
"loss": 1.9283,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.1275720164609053,
|
| 804 |
+
"grad_norm": 3.130859375,
|
| 805 |
+
"learning_rate": 0.0001553774889760533,
|
| 806 |
+
"loss": 1.9353,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.1550068587105624,
|
| 811 |
+
"grad_norm": 2.798828125,
|
| 812 |
+
"learning_rate": 0.0001546567818875221,
|
| 813 |
+
"loss": 1.9945,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.1824417009602195,
|
| 818 |
+
"grad_norm": 2.287109375,
|
| 819 |
+
"learning_rate": 0.00015393200344991995,
|
| 820 |
+
"loss": 1.9199,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.2098765432098766,
|
| 825 |
+
"grad_norm": 2.748046875,
|
| 826 |
+
"learning_rate": 0.00015320320765153367,
|
| 827 |
+
"loss": 1.9415,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.2373113854595337,
|
| 832 |
+
"grad_norm": 2.10546875,
|
| 833 |
+
"learning_rate": 0.0001524704487799008,
|
| 834 |
+
"loss": 1.9417,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.2647462277091908,
|
| 839 |
+
"grad_norm": 2.1015625,
|
| 840 |
+
"learning_rate": 0.00015173378141776568,
|
| 841 |
+
"loss": 1.9477,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.292181069958848,
|
| 846 |
+
"grad_norm": 1.8896484375,
|
| 847 |
+
"learning_rate": 0.0001509932604390136,
|
| 848 |
+
"loss": 1.8957,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.319615912208505,
|
| 853 |
+
"grad_norm": 2.36328125,
|
| 854 |
+
"learning_rate": 0.0001502489410045833,
|
| 855 |
+
"loss": 1.9313,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.347050754458162,
|
| 860 |
+
"grad_norm": 2.107421875,
|
| 861 |
+
"learning_rate": 0.00014950087855835815,
|
| 862 |
+
"loss": 1.902,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.374485596707819,
|
| 867 |
+
"grad_norm": 1.919921875,
|
| 868 |
+
"learning_rate": 0.000148749128823036,
|
| 869 |
+
"loss": 1.8492,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.401920438957476,
|
| 874 |
+
"grad_norm": 2.07421875,
|
| 875 |
+
"learning_rate": 0.00014799374779597867,
|
| 876 |
+
"loss": 1.9576,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.4293552812071333,
|
| 881 |
+
"grad_norm": 2.115234375,
|
| 882 |
+
"learning_rate": 0.00014723479174504037,
|
| 883 |
+
"loss": 1.9472,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.45679012345679,
|
| 888 |
+
"grad_norm": 2.748046875,
|
| 889 |
+
"learning_rate": 0.00014647231720437686,
|
| 890 |
+
"loss": 1.969,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.484224965706447,
|
| 895 |
+
"grad_norm": 2.20703125,
|
| 896 |
+
"learning_rate": 0.0001457063809702338,
|
| 897 |
+
"loss": 1.9607,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.511659807956104,
|
| 902 |
+
"grad_norm": 1.8447265625,
|
| 903 |
+
"learning_rate": 0.00014493704009671613,
|
| 904 |
+
"loss": 1.9347,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.539094650205761,
|
| 909 |
+
"grad_norm": 1.857421875,
|
| 910 |
+
"learning_rate": 0.00014416435189153846,
|
| 911 |
+
"loss": 1.9848,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.5665294924554183,
|
| 916 |
+
"grad_norm": 2.009765625,
|
| 917 |
+
"learning_rate": 0.00014338837391175582,
|
| 918 |
+
"loss": 1.9784,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.5939643347050754,
|
| 923 |
+
"grad_norm": 2.373046875,
|
| 924 |
+
"learning_rate": 0.00014260916395947656,
|
| 925 |
+
"loss": 1.9356,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.6213991769547325,
|
| 930 |
+
"grad_norm": 2.474609375,
|
| 931 |
+
"learning_rate": 0.0001418267800775565,
|
| 932 |
+
"loss": 1.9703,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.6488340192043895,
|
| 937 |
+
"grad_norm": 1.7783203125,
|
| 938 |
+
"learning_rate": 0.0001410412805452757,
|
| 939 |
+
"loss": 1.9149,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.6762688614540466,
|
| 944 |
+
"grad_norm": 2.7421875,
|
| 945 |
+
"learning_rate": 0.00014025272387399674,
|
| 946 |
+
"loss": 1.948,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.7037037037037037,
|
| 951 |
+
"grad_norm": 2.64453125,
|
| 952 |
+
"learning_rate": 0.00013946116880280681,
|
| 953 |
+
"loss": 1.9427,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.731138545953361,
|
| 958 |
+
"grad_norm": 3.169921875,
|
| 959 |
+
"learning_rate": 0.0001386666742941419,
|
| 960 |
+
"loss": 1.8966,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.758573388203018,
|
| 965 |
+
"grad_norm": 2.00390625,
|
| 966 |
+
"learning_rate": 0.00013786929952939477,
|
| 967 |
+
"loss": 1.9682,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.786008230452675,
|
| 972 |
+
"grad_norm": 1.7138671875,
|
| 973 |
+
"learning_rate": 0.00013706910390450677,
|
| 974 |
+
"loss": 1.9255,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.813443072702332,
|
| 979 |
+
"grad_norm": 1.9404296875,
|
| 980 |
+
"learning_rate": 0.0001362661470255432,
|
| 981 |
+
"loss": 1.8883,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.840877914951989,
|
| 986 |
+
"grad_norm": 2.662109375,
|
| 987 |
+
"learning_rate": 0.00013546048870425356,
|
| 988 |
+
"loss": 1.9409,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.8683127572016462,
|
| 993 |
+
"grad_norm": 1.763671875,
|
| 994 |
+
"learning_rate": 0.000134652188953616,
|
| 995 |
+
"loss": 1.9766,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.895747599451303,
|
| 1000 |
+
"grad_norm": 1.7626953125,
|
| 1001 |
+
"learning_rate": 0.00013384130798336705,
|
| 1002 |
+
"loss": 1.9428,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.92318244170096,
|
| 1007 |
+
"grad_norm": 1.8583984375,
|
| 1008 |
+
"learning_rate": 0.00013302790619551674,
|
| 1009 |
+
"loss": 1.9664,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.950617283950617,
|
| 1014 |
+
"grad_norm": 1.7919921875,
|
| 1015 |
+
"learning_rate": 0.00013221204417984908,
|
| 1016 |
+
"loss": 1.9387,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.978052126200274,
|
| 1021 |
+
"grad_norm": 2.16015625,
|
| 1022 |
+
"learning_rate": 0.000131393782709409,
|
| 1023 |
+
"loss": 1.8872,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.005486968449931,
|
| 1028 |
+
"grad_norm": 1.833984375,
|
| 1029 |
+
"learning_rate": 0.0001305731827359753,
|
| 1030 |
+
"loss": 1.9299,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.032921810699588,
|
| 1035 |
+
"grad_norm": 2.3359375,
|
| 1036 |
+
"learning_rate": 0.00012975030538552032,
|
| 1037 |
+
"loss": 1.8399,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.060356652949245,
|
| 1042 |
+
"grad_norm": 1.4580078125,
|
| 1043 |
+
"learning_rate": 0.00012892521195365678,
|
| 1044 |
+
"loss": 1.9137,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.0877914951989025,
|
| 1049 |
+
"grad_norm": 1.7353515625,
|
| 1050 |
+
"learning_rate": 0.00012809796390107195,
|
| 1051 |
+
"loss": 1.8806,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.11522633744856,
|
| 1056 |
+
"grad_norm": 2.189453125,
|
| 1057 |
+
"learning_rate": 0.00012726862284894938,
|
| 1058 |
+
"loss": 1.9019,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.142661179698217,
|
| 1063 |
+
"grad_norm": 1.490234375,
|
| 1064 |
+
"learning_rate": 0.0001264372505743789,
|
| 1065 |
+
"loss": 1.8454,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.170096021947874,
|
| 1070 |
+
"grad_norm": 1.6123046875,
|
| 1071 |
+
"learning_rate": 0.0001256039090057547,
|
| 1072 |
+
"loss": 1.9057,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.197530864197531,
|
| 1077 |
+
"grad_norm": 1.3447265625,
|
| 1078 |
+
"learning_rate": 0.0001247686602181626,
|
| 1079 |
+
"loss": 1.8994,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.224965706447188,
|
| 1084 |
+
"grad_norm": 1.7265625,
|
| 1085 |
+
"learning_rate": 0.0001239315664287558,
|
| 1086 |
+
"loss": 1.8779,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.252400548696845,
|
| 1091 |
+
"grad_norm": 2.07421875,
|
| 1092 |
+
"learning_rate": 0.0001230926899921206,
|
| 1093 |
+
"loss": 1.8386,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.279835390946502,
|
| 1098 |
+
"grad_norm": 1.7724609375,
|
| 1099 |
+
"learning_rate": 0.00012225209339563145,
|
| 1100 |
+
"loss": 1.8646,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.307270233196159,
|
| 1105 |
+
"grad_norm": 1.8017578125,
|
| 1106 |
+
"learning_rate": 0.00012140983925479662,
|
| 1107 |
+
"loss": 1.8488,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.334705075445816,
|
| 1112 |
+
"grad_norm": 2.1171875,
|
| 1113 |
+
"learning_rate": 0.00012056599030859366,
|
| 1114 |
+
"loss": 1.8531,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.362139917695473,
|
| 1119 |
+
"grad_norm": 1.9462890625,
|
| 1120 |
+
"learning_rate": 0.00011972060941479621,
|
| 1121 |
+
"loss": 1.8437,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.3895747599451305,
|
| 1126 |
+
"grad_norm": 1.5908203125,
|
| 1127 |
+
"learning_rate": 0.00011887375954529168,
|
| 1128 |
+
"loss": 1.8201,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.4170096021947876,
|
| 1133 |
+
"grad_norm": 5.125,
|
| 1134 |
+
"learning_rate": 0.0001180255037813906,
|
| 1135 |
+
"loss": 1.865,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.444444444444445,
|
| 1140 |
+
"grad_norm": 1.919921875,
|
| 1141 |
+
"learning_rate": 0.00011717590530912763,
|
| 1142 |
+
"loss": 1.8605,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.471879286694102,
|
| 1147 |
+
"grad_norm": 1.666015625,
|
| 1148 |
+
"learning_rate": 0.00011632502741455496,
|
| 1149 |
+
"loss": 1.8294,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.499314128943759,
|
| 1154 |
+
"grad_norm": 1.6572265625,
|
| 1155 |
+
"learning_rate": 0.00011547293347902812,
|
| 1156 |
+
"loss": 1.8254,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.526748971193416,
|
| 1161 |
+
"grad_norm": 1.4775390625,
|
| 1162 |
+
"learning_rate": 0.00011461968697448485,
|
| 1163 |
+
"loss": 1.8534,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.554183813443073,
|
| 1168 |
+
"grad_norm": 1.791015625,
|
| 1169 |
+
"learning_rate": 0.00011376535145871684,
|
| 1170 |
+
"loss": 1.8151,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.58161865569273,
|
| 1175 |
+
"grad_norm": 1.9013671875,
|
| 1176 |
+
"learning_rate": 0.00011290999057063569,
|
| 1177 |
+
"loss": 1.875,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.609053497942387,
|
| 1182 |
+
"grad_norm": 2.322265625,
|
| 1183 |
+
"learning_rate": 0.0001120536680255323,
|
| 1184 |
+
"loss": 1.9154,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.636488340192044,
|
| 1189 |
+
"grad_norm": 1.814453125,
|
| 1190 |
+
"learning_rate": 0.00011119644761033078,
|
| 1191 |
+
"loss": 1.898,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.663923182441701,
|
| 1196 |
+
"grad_norm": 2.25390625,
|
| 1197 |
+
"learning_rate": 0.00011033839317883701,
|
| 1198 |
+
"loss": 1.852,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.6913580246913575,
|
| 1203 |
+
"grad_norm": 2.240234375,
|
| 1204 |
+
"learning_rate": 0.00010947956864698223,
|
| 1205 |
+
"loss": 1.8394,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.7187928669410155,
|
| 1210 |
+
"grad_norm": 2.544921875,
|
| 1211 |
+
"learning_rate": 0.00010862003798806196,
|
| 1212 |
+
"loss": 1.84,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.746227709190672,
|
| 1217 |
+
"grad_norm": 1.7255859375,
|
| 1218 |
+
"learning_rate": 0.00010775986522797063,
|
| 1219 |
+
"loss": 1.8682,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.77366255144033,
|
| 1224 |
+
"grad_norm": 2.705078125,
|
| 1225 |
+
"learning_rate": 0.00010689911444043248,
|
| 1226 |
+
"loss": 1.8197,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.801097393689986,
|
| 1231 |
+
"grad_norm": 1.7275390625,
|
| 1232 |
+
"learning_rate": 0.00010603784974222861,
|
| 1233 |
+
"loss": 1.868,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.828532235939643,
|
| 1238 |
+
"grad_norm": 1.7900390625,
|
| 1239 |
+
"learning_rate": 0.00010517613528842097,
|
| 1240 |
+
"loss": 1.8828,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 4.8559670781893,
|
| 1245 |
+
"grad_norm": 1.603515625,
|
| 1246 |
+
"learning_rate": 0.00010431403526757347,
|
| 1247 |
+
"loss": 1.8683,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.883401920438957,
|
| 1252 |
+
"grad_norm": 1.958984375,
|
| 1253 |
+
"learning_rate": 0.00010345161389697082,
|
| 1254 |
+
"loss": 1.8725,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.910836762688614,
|
| 1259 |
+
"grad_norm": 1.857421875,
|
| 1260 |
+
"learning_rate": 0.00010258893541783476,
|
| 1261 |
+
"loss": 1.8893,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.938271604938271,
|
| 1266 |
+
"grad_norm": 1.755859375,
|
| 1267 |
+
"learning_rate": 0.00010172606409053886,
|
| 1268 |
+
"loss": 1.892,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.965706447187928,
|
| 1273 |
+
"grad_norm": 1.76171875,
|
| 1274 |
+
"learning_rate": 0.0001008630641898219,
|
| 1275 |
+
"loss": 1.8952,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.9931412894375855,
|
| 1280 |
+
"grad_norm": 1.767578125,
|
| 1281 |
+
"learning_rate": 0.0001,
|
| 1282 |
+
"loss": 1.8623,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.020576131687243,
|
| 1287 |
+
"grad_norm": 2.0234375,
|
| 1288 |
+
"learning_rate": 9.913693581017812e-05,
|
| 1289 |
+
"loss": 1.7838,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.0480109739369,
|
| 1294 |
+
"grad_norm": 1.9765625,
|
| 1295 |
+
"learning_rate": 9.827393590946116e-05,
|
| 1296 |
+
"loss": 1.7935,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.075445816186557,
|
| 1301 |
+
"grad_norm": 2.12109375,
|
| 1302 |
+
"learning_rate": 9.741106458216528e-05,
|
| 1303 |
+
"loss": 1.838,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.102880658436214,
|
| 1308 |
+
"grad_norm": 1.63671875,
|
| 1309 |
+
"learning_rate": 9.654838610302923e-05,
|
| 1310 |
+
"loss": 1.8097,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.130315500685871,
|
| 1315 |
+
"grad_norm": 2.62109375,
|
| 1316 |
+
"learning_rate": 9.568596473242654e-05,
|
| 1317 |
+
"loss": 1.7773,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.157750342935528,
|
| 1322 |
+
"grad_norm": 2.1015625,
|
| 1323 |
+
"learning_rate": 9.482386471157904e-05,
|
| 1324 |
+
"loss": 1.8083,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.185185185185185,
|
| 1329 |
+
"grad_norm": 1.8603515625,
|
| 1330 |
+
"learning_rate": 9.396215025777139e-05,
|
| 1331 |
+
"loss": 1.7376,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.212620027434842,
|
| 1336 |
+
"grad_norm": 1.7890625,
|
| 1337 |
+
"learning_rate": 9.31008855595675e-05,
|
| 1338 |
+
"loss": 1.8674,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.240054869684499,
|
| 1343 |
+
"grad_norm": 2.1796875,
|
| 1344 |
+
"learning_rate": 9.224013477202939e-05,
|
| 1345 |
+
"loss": 1.8239,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.267489711934156,
|
| 1350 |
+
"grad_norm": 2.208984375,
|
| 1351 |
+
"learning_rate": 9.137996201193805e-05,
|
| 1352 |
+
"loss": 1.811,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.2949245541838135,
|
| 1357 |
+
"grad_norm": 2.001953125,
|
| 1358 |
+
"learning_rate": 9.052043135301779e-05,
|
| 1359 |
+
"loss": 1.7938,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.322359396433471,
|
| 1364 |
+
"grad_norm": 2.083984375,
|
| 1365 |
+
"learning_rate": 8.9661606821163e-05,
|
| 1366 |
+
"loss": 1.8577,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.349794238683128,
|
| 1371 |
+
"grad_norm": 1.939453125,
|
| 1372 |
+
"learning_rate": 8.880355238966923e-05,
|
| 1373 |
+
"loss": 1.8207,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.377229080932785,
|
| 1378 |
+
"grad_norm": 1.873046875,
|
| 1379 |
+
"learning_rate": 8.79463319744677e-05,
|
| 1380 |
+
"loss": 1.7565,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.404663923182442,
|
| 1385 |
+
"grad_norm": 1.646484375,
|
| 1386 |
+
"learning_rate": 8.709000942936433e-05,
|
| 1387 |
+
"loss": 1.8572,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.432098765432099,
|
| 1392 |
+
"grad_norm": 1.5400390625,
|
| 1393 |
+
"learning_rate": 8.62346485412832e-05,
|
| 1394 |
+
"loss": 1.8169,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.459533607681756,
|
| 1399 |
+
"grad_norm": 1.8388671875,
|
| 1400 |
+
"learning_rate": 8.538031302551522e-05,
|
| 1401 |
+
"loss": 1.8642,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.486968449931413,
|
| 1406 |
+
"grad_norm": 2.240234375,
|
| 1407 |
+
"learning_rate": 8.452706652097186e-05,
|
| 1408 |
+
"loss": 1.7803,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.51440329218107,
|
| 1413 |
+
"grad_norm": 1.8984375,
|
| 1414 |
+
"learning_rate": 8.367497258544507e-05,
|
| 1415 |
+
"loss": 1.7859,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.541838134430727,
|
| 1420 |
+
"grad_norm": 1.880859375,
|
| 1421 |
+
"learning_rate": 8.282409469087239e-05,
|
| 1422 |
+
"loss": 1.8381,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.569272976680384,
|
| 1427 |
+
"grad_norm": 1.865234375,
|
| 1428 |
+
"learning_rate": 8.197449621860943e-05,
|
| 1429 |
+
"loss": 1.7921,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.596707818930041,
|
| 1434 |
+
"grad_norm": 1.671875,
|
| 1435 |
+
"learning_rate": 8.112624045470835e-05,
|
| 1436 |
+
"loss": 1.781,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.6241426611796985,
|
| 1441 |
+
"grad_norm": 1.7802734375,
|
| 1442 |
+
"learning_rate": 8.027939058520381e-05,
|
| 1443 |
+
"loss": 1.7988,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.651577503429356,
|
| 1448 |
+
"grad_norm": 1.72265625,
|
| 1449 |
+
"learning_rate": 7.943400969140635e-05,
|
| 1450 |
+
"loss": 1.7888,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.679012345679013,
|
| 1455 |
+
"grad_norm": 1.8701171875,
|
| 1456 |
+
"learning_rate": 7.85901607452034e-05,
|
| 1457 |
+
"loss": 1.7995,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.70644718792867,
|
| 1462 |
+
"grad_norm": 1.9873046875,
|
| 1463 |
+
"learning_rate": 7.774790660436858e-05,
|
| 1464 |
+
"loss": 1.7753,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.733882030178327,
|
| 1469 |
+
"grad_norm": 2.19921875,
|
| 1470 |
+
"learning_rate": 7.690731000787948e-05,
|
| 1471 |
+
"loss": 1.7387,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.761316872427983,
|
| 1476 |
+
"grad_norm": 1.85546875,
|
| 1477 |
+
"learning_rate": 7.606843357124426e-05,
|
| 1478 |
+
"loss": 1.8478,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.788751714677641,
|
| 1483 |
+
"grad_norm": 1.8662109375,
|
| 1484 |
+
"learning_rate": 7.52313397818374e-05,
|
| 1485 |
+
"loss": 1.8373,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.816186556927297,
|
| 1490 |
+
"grad_norm": 1.8505859375,
|
| 1491 |
+
"learning_rate": 7.43960909942453e-05,
|
| 1492 |
+
"loss": 1.8703,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.843621399176955,
|
| 1497 |
+
"grad_norm": 1.591796875,
|
| 1498 |
+
"learning_rate": 7.356274942562111e-05,
|
| 1499 |
+
"loss": 1.7647,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.871056241426611,
|
| 1504 |
+
"grad_norm": 1.8681640625,
|
| 1505 |
+
"learning_rate": 7.273137715105063e-05,
|
| 1506 |
+
"loss": 1.7957,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.8984910836762685,
|
| 1511 |
+
"grad_norm": 1.7666015625,
|
| 1512 |
+
"learning_rate": 7.190203609892808e-05,
|
| 1513 |
+
"loss": 1.7988,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.925925925925926,
|
| 1518 |
+
"grad_norm": 1.6123046875,
|
| 1519 |
+
"learning_rate": 7.107478804634325e-05,
|
| 1520 |
+
"loss": 1.7388,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.953360768175583,
|
| 1525 |
+
"grad_norm": 1.7646484375,
|
| 1526 |
+
"learning_rate": 7.024969461447972e-05,
|
| 1527 |
+
"loss": 1.7994,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.98079561042524,
|
| 1532 |
+
"grad_norm": 1.96484375,
|
| 1533 |
+
"learning_rate": 6.942681726402473e-05,
|
| 1534 |
+
"loss": 1.7937,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.008230452674897,
|
| 1539 |
+
"grad_norm": 2.099609375,
|
| 1540 |
+
"learning_rate": 6.8606217290591e-05,
|
| 1541 |
+
"loss": 1.7915,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.035665294924554,
|
| 1546 |
+
"grad_norm": 1.8466796875,
|
| 1547 |
+
"learning_rate": 6.778795582015097e-05,
|
| 1548 |
+
"loss": 1.8106,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.063100137174211,
|
| 1553 |
+
"grad_norm": 1.9228515625,
|
| 1554 |
+
"learning_rate": 6.697209380448333e-05,
|
| 1555 |
+
"loss": 1.7836,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.090534979423868,
|
| 1560 |
+
"grad_norm": 2.74609375,
|
| 1561 |
+
"learning_rate": 6.615869201663296e-05,
|
| 1562 |
+
"loss": 1.7202,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.117969821673525,
|
| 1567 |
+
"grad_norm": 1.708984375,
|
| 1568 |
+
"learning_rate": 6.534781104638399e-05,
|
| 1569 |
+
"loss": 1.7432,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.145404663923182,
|
| 1574 |
+
"grad_norm": 1.921875,
|
| 1575 |
+
"learning_rate": 6.453951129574644e-05,
|
| 1576 |
+
"loss": 1.719,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.172839506172839,
|
| 1581 |
+
"grad_norm": 1.7333984375,
|
| 1582 |
+
"learning_rate": 6.37338529744568e-05,
|
| 1583 |
+
"loss": 1.778,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.2002743484224965,
|
| 1588 |
+
"grad_norm": 1.7802734375,
|
| 1589 |
+
"learning_rate": 6.293089609549325e-05,
|
| 1590 |
+
"loss": 1.7294,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.227709190672154,
|
| 1595 |
+
"grad_norm": 2.220703125,
|
| 1596 |
+
"learning_rate": 6.213070047060524e-05,
|
| 1597 |
+
"loss": 1.6875,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.255144032921811,
|
| 1602 |
+
"grad_norm": 1.8193359375,
|
| 1603 |
+
"learning_rate": 6.133332570585812e-05,
|
| 1604 |
+
"loss": 1.8336,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.282578875171468,
|
| 1609 |
+
"grad_norm": 1.9384765625,
|
| 1610 |
+
"learning_rate": 6.05388311971932e-05,
|
| 1611 |
+
"loss": 1.7279,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.310013717421125,
|
| 1616 |
+
"grad_norm": 2.046875,
|
| 1617 |
+
"learning_rate": 5.9747276126003257e-05,
|
| 1618 |
+
"loss": 1.753,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 6.337448559670782,
|
| 1623 |
+
"grad_norm": 1.7490234375,
|
| 1624 |
+
"learning_rate": 5.8958719454724346e-05,
|
| 1625 |
+
"loss": 1.7593,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 6.364883401920439,
|
| 1630 |
+
"grad_norm": 1.9990234375,
|
| 1631 |
+
"learning_rate": 5.817321992244351e-05,
|
| 1632 |
+
"loss": 1.7361,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 6.392318244170096,
|
| 1637 |
+
"grad_norm": 2.55078125,
|
| 1638 |
+
"learning_rate": 5.739083604052351e-05,
|
| 1639 |
+
"loss": 1.7527,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 6.419753086419753,
|
| 1644 |
+
"grad_norm": 1.8408203125,
|
| 1645 |
+
"learning_rate": 5.6611626088244194e-05,
|
| 1646 |
+
"loss": 1.7893,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 6.44718792866941,
|
| 1651 |
+
"grad_norm": 1.8388671875,
|
| 1652 |
+
"learning_rate": 5.583564810846157e-05,
|
| 1653 |
+
"loss": 1.744,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 6.474622770919067,
|
| 1658 |
+
"grad_norm": 1.947265625,
|
| 1659 |
+
"learning_rate": 5.506295990328385e-05,
|
| 1660 |
+
"loss": 1.7609,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 6.502057613168724,
|
| 1665 |
+
"grad_norm": 1.947265625,
|
| 1666 |
+
"learning_rate": 5.429361902976624e-05,
|
| 1667 |
+
"loss": 1.7273,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 6.5294924554183815,
|
| 1672 |
+
"grad_norm": 1.9033203125,
|
| 1673 |
+
"learning_rate": 5.3527682795623146e-05,
|
| 1674 |
+
"loss": 1.7782,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 6.556927297668039,
|
| 1679 |
+
"grad_norm": 1.75390625,
|
| 1680 |
+
"learning_rate": 5.276520825495963e-05,
|
| 1681 |
+
"loss": 1.7612,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 6.584362139917696,
|
| 1686 |
+
"grad_norm": 1.8037109375,
|
| 1687 |
+
"learning_rate": 5.200625220402139e-05,
|
| 1688 |
+
"loss": 1.7672,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 6.611796982167353,
|
| 1693 |
+
"grad_norm": 1.7421875,
|
| 1694 |
+
"learning_rate": 5.1250871176964036e-05,
|
| 1695 |
+
"loss": 1.7832,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 6.63923182441701,
|
| 1700 |
+
"grad_norm": 1.609375,
|
| 1701 |
+
"learning_rate": 5.0499121441641864e-05,
|
| 1702 |
+
"loss": 1.7438,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 6.666666666666667,
|
| 1707 |
+
"grad_norm": 1.865234375,
|
| 1708 |
+
"learning_rate": 4.975105899541671e-05,
|
| 1709 |
+
"loss": 1.7172,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 6.694101508916324,
|
| 1714 |
+
"grad_norm": 1.8056640625,
|
| 1715 |
+
"learning_rate": 4.900673956098644e-05,
|
| 1716 |
+
"loss": 1.7476,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 6.721536351165981,
|
| 1721 |
+
"grad_norm": 1.75390625,
|
| 1722 |
+
"learning_rate": 4.826621858223431e-05,
|
| 1723 |
+
"loss": 1.7547,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.748971193415638,
|
| 1728 |
+
"grad_norm": 1.841796875,
|
| 1729 |
+
"learning_rate": 4.75295512200992e-05,
|
| 1730 |
+
"loss": 1.7363,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 6.776406035665294,
|
| 1735 |
+
"grad_norm": 1.8349609375,
|
| 1736 |
+
"learning_rate": 4.6796792348466356e-05,
|
| 1737 |
+
"loss": 1.7725,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 6.803840877914952,
|
| 1742 |
+
"grad_norm": 1.7373046875,
|
| 1743 |
+
"learning_rate": 4.606799655008009e-05,
|
| 1744 |
+
"loss": 1.6962,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 6.831275720164609,
|
| 1749 |
+
"grad_norm": 2.0234375,
|
| 1750 |
+
"learning_rate": 4.5343218112477904e-05,
|
| 1751 |
+
"loss": 1.6918,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 6.858710562414267,
|
| 1756 |
+
"grad_norm": 1.73828125,
|
| 1757 |
+
"learning_rate": 4.462251102394669e-05,
|
| 1758 |
+
"loss": 1.7336,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 6.886145404663923,
|
| 1763 |
+
"grad_norm": 1.974609375,
|
| 1764 |
+
"learning_rate": 4.3905928969501056e-05,
|
| 1765 |
+
"loss": 1.7502,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.91358024691358,
|
| 1770 |
+
"grad_norm": 1.8828125,
|
| 1771 |
+
"learning_rate": 4.3193525326884435e-05,
|
| 1772 |
+
"loss": 1.8341,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 6.941015089163237,
|
| 1777 |
+
"grad_norm": 1.7939453125,
|
| 1778 |
+
"learning_rate": 4.248535316259305e-05,
|
| 1779 |
+
"loss": 1.7671,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 6.968449931412894,
|
| 1784 |
+
"grad_norm": 1.5888671875,
|
| 1785 |
+
"learning_rate": 4.1781465227922957e-05,
|
| 1786 |
+
"loss": 1.7457,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 6.995884773662551,
|
| 1791 |
+
"grad_norm": 1.7177734375,
|
| 1792 |
+
"learning_rate": 4.108191395504064e-05,
|
| 1793 |
+
"loss": 1.7736,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.023319615912208,
|
| 1798 |
+
"grad_norm": 1.5703125,
|
| 1799 |
+
"learning_rate": 4.038675145307747e-05,
|
| 1800 |
+
"loss": 1.6925,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.050754458161865,
|
| 1805 |
+
"grad_norm": 2.091796875,
|
| 1806 |
+
"learning_rate": 3.9696029504247956e-05,
|
| 1807 |
+
"loss": 1.7176,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.078189300411522,
|
| 1812 |
+
"grad_norm": 1.919921875,
|
| 1813 |
+
"learning_rate": 3.900979955999271e-05,
|
| 1814 |
+
"loss": 1.7825,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.1056241426611795,
|
| 1819 |
+
"grad_norm": 1.943359375,
|
| 1820 |
+
"learning_rate": 3.832811273714569e-05,
|
| 1821 |
+
"loss": 1.7257,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.133058984910837,
|
| 1826 |
+
"grad_norm": 2.052734375,
|
| 1827 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 1828 |
+
"loss": 1.7665,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 7.160493827160494,
|
| 1833 |
+
"grad_norm": 1.6552734375,
|
| 1834 |
+
"learning_rate": 3.697857122715865e-05,
|
| 1835 |
+
"loss": 1.7373,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 7.187928669410151,
|
| 1840 |
+
"grad_norm": 1.8623046875,
|
| 1841 |
+
"learning_rate": 3.6310817066511105e-05,
|
| 1842 |
+
"loss": 1.6965,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 7.215363511659808,
|
| 1847 |
+
"grad_norm": 1.48828125,
|
| 1848 |
+
"learning_rate": 3.5647807072768526e-05,
|
| 1849 |
+
"loss": 1.7009,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 7.242798353909465,
|
| 1854 |
+
"grad_norm": 1.5126953125,
|
| 1855 |
+
"learning_rate": 3.498959063312558e-05,
|
| 1856 |
+
"loss": 1.7086,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 7.270233196159122,
|
| 1861 |
+
"grad_norm": 1.6748046875,
|
| 1862 |
+
"learning_rate": 3.4336216777708e-05,
|
| 1863 |
+
"loss": 1.6847,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 7.297668038408779,
|
| 1868 |
+
"grad_norm": 1.4794921875,
|
| 1869 |
+
"learning_rate": 3.36877341759205e-05,
|
| 1870 |
+
"loss": 1.756,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 7.325102880658436,
|
| 1875 |
+
"grad_norm": 1.6240234375,
|
| 1876 |
+
"learning_rate": 3.3044191132821454e-05,
|
| 1877 |
+
"loss": 1.6819,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 7.352537722908093,
|
| 1882 |
+
"grad_norm": 1.623046875,
|
| 1883 |
+
"learning_rate": 3.2405635585524565e-05,
|
| 1884 |
+
"loss": 1.7719,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 7.37997256515775,
|
| 1889 |
+
"grad_norm": 1.775390625,
|
| 1890 |
+
"learning_rate": 3.177211509962826e-05,
|
| 1891 |
+
"loss": 1.7437,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 7.407407407407407,
|
| 1896 |
+
"grad_norm": 1.880859375,
|
| 1897 |
+
"learning_rate": 3.114367686567228e-05,
|
| 1898 |
+
"loss": 1.702,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 7.4348422496570645,
|
| 1903 |
+
"grad_norm": 1.6015625,
|
| 1904 |
+
"learning_rate": 3.052036769562276e-05,
|
| 1905 |
+
"loss": 1.6124,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 7.462277091906722,
|
| 1910 |
+
"grad_norm": 1.8525390625,
|
| 1911 |
+
"learning_rate": 2.9902234019385057e-05,
|
| 1912 |
+
"loss": 1.6915,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 7.489711934156379,
|
| 1917 |
+
"grad_norm": 1.998046875,
|
| 1918 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 1919 |
+
"loss": 1.7637,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 7.517146776406036,
|
| 1924 |
+
"grad_norm": 1.744140625,
|
| 1925 |
+
"learning_rate": 2.8681676936940393e-05,
|
| 1926 |
+
"loss": 1.7212,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 7.544581618655693,
|
| 1931 |
+
"grad_norm": 1.486328125,
|
| 1932 |
+
"learning_rate": 2.8079344449257572e-05,
|
| 1933 |
+
"loss": 1.7415,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 7.57201646090535,
|
| 1938 |
+
"grad_norm": 2.044921875,
|
| 1939 |
+
"learning_rate": 2.7482369285662378e-05,
|
| 1940 |
+
"loss": 1.7262,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 7.599451303155007,
|
| 1945 |
+
"grad_norm": 1.6982421875,
|
| 1946 |
+
"learning_rate": 2.6890795914456678e-05,
|
| 1947 |
+
"loss": 1.6766,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 7.626886145404664,
|
| 1952 |
+
"grad_norm": 1.8671875,
|
| 1953 |
+
"learning_rate": 2.6304668401566335e-05,
|
| 1954 |
+
"loss": 1.7522,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 7.654320987654321,
|
| 1959 |
+
"grad_norm": 1.8017578125,
|
| 1960 |
+
"learning_rate": 2.572403040725855e-05,
|
| 1961 |
+
"loss": 1.7095,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 7.681755829903978,
|
| 1966 |
+
"grad_norm": 2.005859375,
|
| 1967 |
+
"learning_rate": 2.514892518288988e-05,
|
| 1968 |
+
"loss": 1.7196,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 7.709190672153635,
|
| 1973 |
+
"grad_norm": 1.7734375,
|
| 1974 |
+
"learning_rate": 2.4579395567684283e-05,
|
| 1975 |
+
"loss": 1.7174,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 7.7366255144032925,
|
| 1980 |
+
"grad_norm": 2.1015625,
|
| 1981 |
+
"learning_rate": 2.401548398554213e-05,
|
| 1982 |
+
"loss": 1.745,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 7.76406035665295,
|
| 1987 |
+
"grad_norm": 1.7236328125,
|
| 1988 |
+
"learning_rate": 2.345723244188006e-05,
|
| 1989 |
+
"loss": 1.7127,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.791495198902607,
|
| 1994 |
+
"grad_norm": 2.4296875,
|
| 1995 |
+
"learning_rate": 2.290468252050204e-05,
|
| 1996 |
+
"loss": 1.6945,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.818930041152264,
|
| 2001 |
+
"grad_norm": 1.67578125,
|
| 2002 |
+
"learning_rate": 2.2357875380501836e-05,
|
| 2003 |
+
"loss": 1.7206,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.84636488340192,
|
| 2008 |
+
"grad_norm": 2.359375,
|
| 2009 |
+
"learning_rate": 2.181685175319702e-05,
|
| 2010 |
+
"loss": 1.683,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.873799725651578,
|
| 2015 |
+
"grad_norm": 1.7236328125,
|
| 2016 |
+
"learning_rate": 2.1281651939094992e-05,
|
| 2017 |
+
"loss": 1.7218,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.901234567901234,
|
| 2022 |
+
"grad_norm": 1.82421875,
|
| 2023 |
+
"learning_rate": 2.0752315804890977e-05,
|
| 2024 |
+
"loss": 1.7274,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.928669410150892,
|
| 2029 |
+
"grad_norm": 3.5078125,
|
| 2030 |
+
"learning_rate": 2.0228882780498404e-05,
|
| 2031 |
+
"loss": 1.6874,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.956104252400548,
|
| 2036 |
+
"grad_norm": 2.08984375,
|
| 2037 |
+
"learning_rate": 1.971139185611176e-05,
|
| 2038 |
+
"loss": 1.7064,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.983539094650205,
|
| 2043 |
+
"grad_norm": 2.240234375,
|
| 2044 |
+
"learning_rate": 1.919988157930236e-05,
|
| 2045 |
+
"loss": 1.6922,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.010973936899862,
|
| 2050 |
+
"grad_norm": 2.25,
|
| 2051 |
+
"learning_rate": 1.8694390052146737e-05,
|
| 2052 |
+
"loss": 1.73,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 8.03840877914952,
|
| 2057 |
+
"grad_norm": 1.62890625,
|
| 2058 |
+
"learning_rate": 1.819495492838872e-05,
|
| 2059 |
+
"loss": 1.6912,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 8.065843621399177,
|
| 2064 |
+
"grad_norm": 1.755859375,
|
| 2065 |
+
"learning_rate": 1.7701613410634365e-05,
|
| 2066 |
+
"loss": 1.6858,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 8.093278463648835,
|
| 2071 |
+
"grad_norm": 1.5791015625,
|
| 2072 |
+
"learning_rate": 1.7214402247580918e-05,
|
| 2073 |
+
"loss": 1.6634,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 8.12071330589849,
|
| 2078 |
+
"grad_norm": 1.63671875,
|
| 2079 |
+
"learning_rate": 1.6733357731279377e-05,
|
| 2080 |
+
"loss": 1.6908,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 8.148148148148149,
|
| 2085 |
+
"grad_norm": 1.8701171875,
|
| 2086 |
+
"learning_rate": 1.6258515694431144e-05,
|
| 2087 |
+
"loss": 1.7138,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 8.175582990397805,
|
| 2092 |
+
"grad_norm": 1.6357421875,
|
| 2093 |
+
"learning_rate": 1.5789911507718826e-05,
|
| 2094 |
+
"loss": 1.7258,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 8.203017832647463,
|
| 2099 |
+
"grad_norm": 1.734375,
|
| 2100 |
+
"learning_rate": 1.5327580077171587e-05,
|
| 2101 |
+
"loss": 1.7178,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 8.23045267489712,
|
| 2106 |
+
"grad_norm": 1.787109375,
|
| 2107 |
+
"learning_rate": 1.4871555841564887e-05,
|
| 2108 |
+
"loss": 1.6809,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 8.257887517146777,
|
| 2113 |
+
"grad_norm": 1.7177734375,
|
| 2114 |
+
"learning_rate": 1.442187276985526e-05,
|
| 2115 |
+
"loss": 1.6501,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 8.285322359396433,
|
| 2120 |
+
"grad_norm": 1.822265625,
|
| 2121 |
+
"learning_rate": 1.3978564358649927e-05,
|
| 2122 |
+
"loss": 1.7259,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 8.312757201646091,
|
| 2127 |
+
"grad_norm": 1.7109375,
|
| 2128 |
+
"learning_rate": 1.3541663629711766e-05,
|
| 2129 |
+
"loss": 1.752,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 8.340192043895748,
|
| 2134 |
+
"grad_norm": 1.7744140625,
|
| 2135 |
+
"learning_rate": 1.311120312749935e-05,
|
| 2136 |
+
"loss": 1.6563,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 8.367626886145406,
|
| 2141 |
+
"grad_norm": 1.8427734375,
|
| 2142 |
+
"learning_rate": 1.2687214916742918e-05,
|
| 2143 |
+
"loss": 1.7103,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 8.395061728395062,
|
| 2148 |
+
"grad_norm": 2.611328125,
|
| 2149 |
+
"learning_rate": 1.2269730580055805e-05,
|
| 2150 |
+
"loss": 1.6951,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 8.422496570644718,
|
| 2155 |
+
"grad_norm": 1.763671875,
|
| 2156 |
+
"learning_rate": 1.185878121558186e-05,
|
| 2157 |
+
"loss": 1.6747,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 8.449931412894376,
|
| 2162 |
+
"grad_norm": 1.7568359375,
|
| 2163 |
+
"learning_rate": 1.1454397434679021e-05,
|
| 2164 |
+
"loss": 1.7284,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 8.477366255144032,
|
| 2169 |
+
"grad_norm": 1.8837890625,
|
| 2170 |
+
"learning_rate": 1.1056609359639025e-05,
|
| 2171 |
+
"loss": 1.6907,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 8.50480109739369,
|
| 2176 |
+
"grad_norm": 1.677734375,
|
| 2177 |
+
"learning_rate": 1.0665446621443708e-05,
|
| 2178 |
+
"loss": 1.652,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 8.532235939643346,
|
| 2183 |
+
"grad_norm": 1.7236328125,
|
| 2184 |
+
"learning_rate": 1.028093835755769e-05,
|
| 2185 |
+
"loss": 1.6751,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 8.559670781893004,
|
| 2190 |
+
"grad_norm": 1.716796875,
|
| 2191 |
+
"learning_rate": 9.903113209758096e-06,
|
| 2192 |
+
"loss": 1.6465,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 8.58710562414266,
|
| 2197 |
+
"grad_norm": 1.6328125,
|
| 2198 |
+
"learning_rate": 9.531999322000885e-06,
|
| 2199 |
+
"loss": 1.7407,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 8.614540466392318,
|
| 2204 |
+
"grad_norm": 1.681640625,
|
| 2205 |
+
"learning_rate": 9.1676243383246e-06,
|
| 2206 |
+
"loss": 1.7061,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 8.641975308641975,
|
| 2211 |
+
"grad_norm": 2.072265625,
|
| 2212 |
+
"learning_rate": 8.810015400790994e-06,
|
| 2213 |
+
"loss": 1.6604,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 8.669410150891633,
|
| 2218 |
+
"grad_norm": 1.7685546875,
|
| 2219 |
+
"learning_rate": 8.45919914746337e-06,
|
| 2220 |
+
"loss": 1.7301,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 8.696844993141289,
|
| 2225 |
+
"grad_norm": 2.193359375,
|
| 2226 |
+
"learning_rate": 8.115201710422282e-06,
|
| 2227 |
+
"loss": 1.6973,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 8.724279835390947,
|
| 2232 |
+
"grad_norm": 1.677734375,
|
| 2233 |
+
"learning_rate": 7.778048713818975e-06,
|
| 2234 |
+
"loss": 1.7165,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 8.751714677640603,
|
| 2239 |
+
"grad_norm": 2.1171875,
|
| 2240 |
+
"learning_rate": 7.447765271966656e-06,
|
| 2241 |
+
"loss": 1.6841,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 8.779149519890261,
|
| 2246 |
+
"grad_norm": 1.6279296875,
|
| 2247 |
+
"learning_rate": 7.124375987469767e-06,
|
| 2248 |
+
"loss": 1.7142,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 8.806584362139917,
|
| 2253 |
+
"grad_norm": 1.93359375,
|
| 2254 |
+
"learning_rate": 6.80790494939132e-06,
|
| 2255 |
+
"loss": 1.7619,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 8.834019204389575,
|
| 2260 |
+
"grad_norm": 2.171875,
|
| 2261 |
+
"learning_rate": 6.498375731458528e-06,
|
| 2262 |
+
"loss": 1.7146,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 8.861454046639231,
|
| 2267 |
+
"grad_norm": 1.59765625,
|
| 2268 |
+
"learning_rate": 6.195811390306816e-06,
|
| 2269 |
+
"loss": 1.6753,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 8.88888888888889,
|
| 2274 |
+
"grad_norm": 1.6484375,
|
| 2275 |
+
"learning_rate": 5.900234463762366e-06,
|
| 2276 |
+
"loss": 1.662,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 8.916323731138545,
|
| 2281 |
+
"grad_norm": 1.70703125,
|
| 2282 |
+
"learning_rate": 5.611666969163243e-06,
|
| 2283 |
+
"loss": 1.6781,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 8.943758573388203,
|
| 2288 |
+
"grad_norm": 1.8642578125,
|
| 2289 |
+
"learning_rate": 5.3301304017194135e-06,
|
| 2290 |
+
"loss": 1.6446,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 8.97119341563786,
|
| 2295 |
+
"grad_norm": 1.96484375,
|
| 2296 |
+
"learning_rate": 5.055645732911462e-06,
|
| 2297 |
+
"loss": 1.6632,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 8.998628257887518,
|
| 2302 |
+
"grad_norm": 2.181640625,
|
| 2303 |
+
"learning_rate": 4.788233408928589e-06,
|
| 2304 |
+
"loss": 1.707,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 9.026063100137174,
|
| 2309 |
+
"grad_norm": 1.75390625,
|
| 2310 |
+
"learning_rate": 4.527913349145441e-06,
|
| 2311 |
+
"loss": 1.7361,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 9.053497942386832,
|
| 2316 |
+
"grad_norm": 2.705078125,
|
| 2317 |
+
"learning_rate": 4.27470494463843e-06,
|
| 2318 |
+
"loss": 1.7412,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 9.080932784636488,
|
| 2323 |
+
"grad_norm": 2.5,
|
| 2324 |
+
"learning_rate": 4.028627056741252e-06,
|
| 2325 |
+
"loss": 1.6508,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 9.108367626886146,
|
| 2330 |
+
"grad_norm": 1.7490234375,
|
| 2331 |
+
"learning_rate": 3.789698015639953e-06,
|
| 2332 |
+
"loss": 1.7095,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 9.135802469135802,
|
| 2337 |
+
"grad_norm": 1.7919921875,
|
| 2338 |
+
"learning_rate": 3.5579356190074907e-06,
|
| 2339 |
+
"loss": 1.6629,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 9.16323731138546,
|
| 2344 |
+
"grad_norm": 1.662109375,
|
| 2345 |
+
"learning_rate": 3.3333571306780497e-06,
|
| 2346 |
+
"loss": 1.6755,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 9.190672153635116,
|
| 2351 |
+
"grad_norm": 5.0703125,
|
| 2352 |
+
"learning_rate": 3.115979279360992e-06,
|
| 2353 |
+
"loss": 1.6963,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 9.218106995884774,
|
| 2358 |
+
"grad_norm": 1.662109375,
|
| 2359 |
+
"learning_rate": 2.905818257394799e-06,
|
| 2360 |
+
"loss": 1.694,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 9.24554183813443,
|
| 2365 |
+
"grad_norm": 1.55078125,
|
| 2366 |
+
"learning_rate": 2.702889719540924e-06,
|
| 2367 |
+
"loss": 1.6488,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 9.272976680384089,
|
| 2372 |
+
"grad_norm": 2.107421875,
|
| 2373 |
+
"learning_rate": 2.5072087818176382e-06,
|
| 2374 |
+
"loss": 1.6729,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 9.300411522633745,
|
| 2379 |
+
"grad_norm": 1.5791015625,
|
| 2380 |
+
"learning_rate": 2.3187900203740844e-06,
|
| 2381 |
+
"loss": 1.6518,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 9.327846364883403,
|
| 2386 |
+
"grad_norm": 2.0234375,
|
| 2387 |
+
"learning_rate": 2.137647470404469e-06,
|
| 2388 |
+
"loss": 1.7342,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 9.355281207133059,
|
| 2393 |
+
"grad_norm": 1.9423828125,
|
| 2394 |
+
"learning_rate": 1.963794625102655e-06,
|
| 2395 |
+
"loss": 1.7223,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 9.382716049382717,
|
| 2400 |
+
"grad_norm": 2.234375,
|
| 2401 |
+
"learning_rate": 1.797244434656975e-06,
|
| 2402 |
+
"loss": 1.7224,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 9.410150891632373,
|
| 2407 |
+
"grad_norm": 1.6806640625,
|
| 2408 |
+
"learning_rate": 1.6380093052856483e-06,
|
| 2409 |
+
"loss": 1.6956,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 9.437585733882031,
|
| 2414 |
+
"grad_norm": 1.693359375,
|
| 2415 |
+
"learning_rate": 1.48610109831262e-06,
|
| 2416 |
+
"loss": 1.6918,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 9.465020576131687,
|
| 2421 |
+
"grad_norm": 1.7080078125,
|
| 2422 |
+
"learning_rate": 1.341531129284046e-06,
|
| 2423 |
+
"loss": 1.6376,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 9.492455418381343,
|
| 2428 |
+
"grad_norm": 2.220703125,
|
| 2429 |
+
"learning_rate": 1.2043101671253554e-06,
|
| 2430 |
+
"loss": 1.6831,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 9.519890260631001,
|
| 2435 |
+
"grad_norm": 2.203125,
|
| 2436 |
+
"learning_rate": 1.0744484333391368e-06,
|
| 2437 |
+
"loss": 1.6559,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 9.547325102880658,
|
| 2442 |
+
"grad_norm": 1.982421875,
|
| 2443 |
+
"learning_rate": 9.519556012436815e-07,
|
| 2444 |
+
"loss": 1.6864,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 9.574759945130316,
|
| 2449 |
+
"grad_norm": 2.025390625,
|
| 2450 |
+
"learning_rate": 8.368407952525026e-07,
|
| 2451 |
+
"loss": 1.6973,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 9.602194787379972,
|
| 2456 |
+
"grad_norm": 1.6865234375,
|
| 2457 |
+
"learning_rate": 7.291125901946027e-07,
|
| 2458 |
+
"loss": 1.7048,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 9.62962962962963,
|
| 2463 |
+
"grad_norm": 1.9140625,
|
| 2464 |
+
"learning_rate": 6.287790106757396e-07,
|
| 2465 |
+
"loss": 1.6939,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 9.657064471879286,
|
| 2470 |
+
"grad_norm": 1.5322265625,
|
| 2471 |
+
"learning_rate": 5.358475304807375e-07,
|
| 2472 |
+
"loss": 1.7541,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 9.684499314128944,
|
| 2477 |
+
"grad_norm": 2.29296875,
|
| 2478 |
+
"learning_rate": 4.503250720166774e-07,
|
| 2479 |
+
"loss": 1.6679,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 9.7119341563786,
|
| 2484 |
+
"grad_norm": 1.5634765625,
|
| 2485 |
+
"learning_rate": 3.7221800579735346e-07,
|
| 2486 |
+
"loss": 1.685,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 9.739368998628258,
|
| 2491 |
+
"grad_norm": 1.6884765625,
|
| 2492 |
+
"learning_rate": 3.0153214996866406e-07,
|
| 2493 |
+
"loss": 1.6628,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 9.766803840877914,
|
| 2498 |
+
"grad_norm": 1.8408203125,
|
| 2499 |
+
"learning_rate": 2.382727698752474e-07,
|
| 2500 |
+
"loss": 1.6632,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 9.794238683127572,
|
| 2505 |
+
"grad_norm": 1.9072265625,
|
| 2506 |
+
"learning_rate": 1.824445776682504e-07,
|
| 2507 |
+
"loss": 1.7231,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 9.821673525377228,
|
| 2512 |
+
"grad_norm": 1.791015625,
|
| 2513 |
+
"learning_rate": 1.340517319543877e-07,
|
| 2514 |
+
"loss": 1.6758,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 9.849108367626886,
|
| 2519 |
+
"grad_norm": 1.7236328125,
|
| 2520 |
+
"learning_rate": 9.309783748606693e-08,
|
| 2521 |
+
"loss": 1.6677,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 9.876543209876543,
|
| 2526 |
+
"grad_norm": 1.5498046875,
|
| 2527 |
+
"learning_rate": 5.958594489295921e-08,
|
| 2528 |
+
"loss": 1.6355,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 9.9039780521262,
|
| 2533 |
+
"grad_norm": 1.802734375,
|
| 2534 |
+
"learning_rate": 3.351855045471419e-08,
|
| 2535 |
+
"loss": 1.6854,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 9.931412894375857,
|
| 2540 |
+
"grad_norm": 1.9306640625,
|
| 2541 |
+
"learning_rate": 1.4897595915053242e-08,
|
| 2542 |
+
"loss": 1.6643,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 9.958847736625515,
|
| 2547 |
+
"grad_norm": 1.8544921875,
|
| 2548 |
+
"learning_rate": 3.724468337085174e-09,
|
| 2549 |
+
"loss": 1.7076,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 9.986282578875171,
|
| 2554 |
+
"grad_norm": 2.060546875,
|
| 2555 |
+
"learning_rate": 0.0,
|
| 2556 |
+
"loss": 1.6734,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 9.986282578875171,
|
| 2561 |
+
"step": 3640,
|
| 2562 |
+
"total_flos": 2.39046462799872e+17,
|
| 2563 |
+
"train_loss": 1.891263527398581,
|
| 2564 |
+
"train_runtime": 3639.5664,
|
| 2565 |
+
"train_samples_per_second": 4.006,
|
| 2566 |
+
"train_steps_per_second": 1.0
|
| 2567 |
+
}
|
| 2568 |
+
],
|
| 2569 |
+
"logging_steps": 10,
|
| 2570 |
+
"max_steps": 3640,
|
| 2571 |
+
"num_input_tokens_seen": 0,
|
| 2572 |
+
"num_train_epochs": 10,
|
| 2573 |
+
"save_steps": 500,
|
| 2574 |
+
"total_flos": 2.39046462799872e+17,
|
| 2575 |
+
"train_batch_size": 1,
|
| 2576 |
+
"trial_name": null,
|
| 2577 |
+
"trial_params": null
|
| 2578 |
+
}
|
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 sha256:5628f86206f80e318b56154c305eb33555c5ada950c5d867228b6f23b3fb4984
|
| 3 |
size 5048
|