Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- trainer_state.json +2403 -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 sha256:2259e193d8da80e789418114dc6629e6079cba681e1e01213e56754a4d4fe031
|
3 |
size 47236352
|
trainer_state.json
ADDED
@@ -0,0 +1,2403 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 9.985272459499264,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 3390,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.029455081001472753,
|
13 |
+
"grad_norm": 10.109375,
|
14 |
+
"learning_rate": 0.00019999570594853575,
|
15 |
+
"loss": 2.6151,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.05891016200294551,
|
20 |
+
"grad_norm": 9.828125,
|
21 |
+
"learning_rate": 0.00019998282416292055,
|
22 |
+
"loss": 2.4649,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.08836524300441827,
|
27 |
+
"grad_norm": 8.59375,
|
28 |
+
"learning_rate": 0.00019996135574945544,
|
29 |
+
"loss": 2.4015,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.11782032400589101,
|
34 |
+
"grad_norm": 7.8359375,
|
35 |
+
"learning_rate": 0.00019993130255186977,
|
36 |
+
"loss": 2.4925,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.14727540500736377,
|
41 |
+
"grad_norm": 7.76953125,
|
42 |
+
"learning_rate": 0.00019989266715116316,
|
43 |
+
"loss": 2.4309,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.17673048600883653,
|
48 |
+
"grad_norm": 10.2265625,
|
49 |
+
"learning_rate": 0.0001998454528653836,
|
50 |
+
"loss": 2.3814,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.20618556701030927,
|
55 |
+
"grad_norm": 10.375,
|
56 |
+
"learning_rate": 0.00019978966374934254,
|
57 |
+
"loss": 2.3852,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.23564064801178203,
|
62 |
+
"grad_norm": 9.4453125,
|
63 |
+
"learning_rate": 0.00019972530459426663,
|
64 |
+
"loss": 2.396,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.2650957290132548,
|
69 |
+
"grad_norm": 9.21875,
|
70 |
+
"learning_rate": 0.00019965238092738643,
|
71 |
+
"loss": 2.386,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.29455081001472755,
|
76 |
+
"grad_norm": 10.1328125,
|
77 |
+
"learning_rate": 0.00019957089901146148,
|
78 |
+
"loss": 2.3984,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.3240058910162003,
|
83 |
+
"grad_norm": 8.484375,
|
84 |
+
"learning_rate": 0.00019948086584424256,
|
85 |
+
"loss": 2.3893,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.35346097201767307,
|
90 |
+
"grad_norm": 7.83203125,
|
91 |
+
"learning_rate": 0.0001993822891578708,
|
92 |
+
"loss": 2.4262,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.38291605301914583,
|
97 |
+
"grad_norm": 6.61328125,
|
98 |
+
"learning_rate": 0.00019927517741821343,
|
99 |
+
"loss": 2.3399,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.41237113402061853,
|
104 |
+
"grad_norm": 7.91015625,
|
105 |
+
"learning_rate": 0.0001991595398241369,
|
106 |
+
"loss": 2.4667,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.4418262150220913,
|
111 |
+
"grad_norm": 5.46484375,
|
112 |
+
"learning_rate": 0.0001990353863067169,
|
113 |
+
"loss": 2.4316,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.47128129602356406,
|
118 |
+
"grad_norm": 7.76171875,
|
119 |
+
"learning_rate": 0.00019890272752838518,
|
120 |
+
"loss": 2.4068,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.5007363770250368,
|
125 |
+
"grad_norm": 6.78515625,
|
126 |
+
"learning_rate": 0.00019876157488201424,
|
127 |
+
"loss": 2.4295,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.5301914580265096,
|
132 |
+
"grad_norm": 6.67578125,
|
133 |
+
"learning_rate": 0.00019861194048993863,
|
134 |
+
"loss": 2.4124,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.5596465390279823,
|
139 |
+
"grad_norm": 4.140625,
|
140 |
+
"learning_rate": 0.00019845383720291392,
|
141 |
+
"loss": 2.3651,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.5891016200294551,
|
146 |
+
"grad_norm": 3.337890625,
|
147 |
+
"learning_rate": 0.00019828727859901317,
|
148 |
+
"loss": 2.4137,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.6185567010309279,
|
153 |
+
"grad_norm": 4.01953125,
|
154 |
+
"learning_rate": 0.0001981122789824607,
|
155 |
+
"loss": 2.3851,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.6480117820324006,
|
160 |
+
"grad_norm": 4.94140625,
|
161 |
+
"learning_rate": 0.00019792885338240374,
|
162 |
+
"loss": 2.3647,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.6774668630338734,
|
167 |
+
"grad_norm": 4.5234375,
|
168 |
+
"learning_rate": 0.00019773701755162158,
|
169 |
+
"loss": 2.2891,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.7069219440353461,
|
174 |
+
"grad_norm": 3.619140625,
|
175 |
+
"learning_rate": 0.00019753678796517282,
|
176 |
+
"loss": 2.33,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.7363770250368189,
|
181 |
+
"grad_norm": 3.244140625,
|
182 |
+
"learning_rate": 0.00019732818181898045,
|
183 |
+
"loss": 2.3588,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.7658321060382917,
|
188 |
+
"grad_norm": 5.95703125,
|
189 |
+
"learning_rate": 0.00019711121702835504,
|
190 |
+
"loss": 2.4006,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.7952871870397643,
|
195 |
+
"grad_norm": 3.873046875,
|
196 |
+
"learning_rate": 0.00019688591222645607,
|
197 |
+
"loss": 2.3874,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.8247422680412371,
|
202 |
+
"grad_norm": 3.833984375,
|
203 |
+
"learning_rate": 0.0001966522867626919,
|
204 |
+
"loss": 2.3104,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.8541973490427098,
|
209 |
+
"grad_norm": 4.55078125,
|
210 |
+
"learning_rate": 0.00019641036070105778,
|
211 |
+
"loss": 2.3146,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.8836524300441826,
|
216 |
+
"grad_norm": 5.9765625,
|
217 |
+
"learning_rate": 0.0001961601548184129,
|
218 |
+
"loss": 2.3112,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.9131075110456554,
|
223 |
+
"grad_norm": 5.18359375,
|
224 |
+
"learning_rate": 0.00019590169060269602,
|
225 |
+
"loss": 2.2807,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.9425625920471281,
|
230 |
+
"grad_norm": 8.359375,
|
231 |
+
"learning_rate": 0.00019563499025107998,
|
232 |
+
"loss": 2.3536,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.9720176730486009,
|
237 |
+
"grad_norm": 5.60546875,
|
238 |
+
"learning_rate": 0.00019536007666806556,
|
239 |
+
"loss": 2.3461,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 1.0014727540500736,
|
244 |
+
"grad_norm": 4.32421875,
|
245 |
+
"learning_rate": 0.00019507697346351414,
|
246 |
+
"loss": 2.3032,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.0309278350515463,
|
251 |
+
"grad_norm": 3.591796875,
|
252 |
+
"learning_rate": 0.00019478570495062037,
|
253 |
+
"loss": 2.2973,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.0603829160530192,
|
258 |
+
"grad_norm": 5.73046875,
|
259 |
+
"learning_rate": 0.0001944862961438239,
|
260 |
+
"loss": 2.3204,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 1.0898379970544918,
|
265 |
+
"grad_norm": 3.8125,
|
266 |
+
"learning_rate": 0.0001941787727566613,
|
267 |
+
"loss": 2.3208,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 1.1192930780559647,
|
272 |
+
"grad_norm": 4.33984375,
|
273 |
+
"learning_rate": 0.00019386316119955756,
|
274 |
+
"loss": 2.4171,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.1487481590574373,
|
279 |
+
"grad_norm": 3.208984375,
|
280 |
+
"learning_rate": 0.00019353948857755803,
|
281 |
+
"loss": 2.2795,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 1.1782032400589102,
|
286 |
+
"grad_norm": 4.07421875,
|
287 |
+
"learning_rate": 0.00019320778268800066,
|
288 |
+
"loss": 2.2644,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 1.2076583210603828,
|
293 |
+
"grad_norm": 4.53125,
|
294 |
+
"learning_rate": 0.00019286807201812867,
|
295 |
+
"loss": 2.2787,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.2371134020618557,
|
300 |
+
"grad_norm": 3.912109375,
|
301 |
+
"learning_rate": 0.00019252038574264405,
|
302 |
+
"loss": 2.2143,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 1.2665684830633284,
|
307 |
+
"grad_norm": 3.607421875,
|
308 |
+
"learning_rate": 0.00019216475372120197,
|
309 |
+
"loss": 2.2773,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 1.2960235640648012,
|
314 |
+
"grad_norm": 2.80859375,
|
315 |
+
"learning_rate": 0.00019180120649584653,
|
316 |
+
"loss": 2.2923,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 1.3254786450662739,
|
321 |
+
"grad_norm": 2.791015625,
|
322 |
+
"learning_rate": 0.00019142977528838762,
|
323 |
+
"loss": 2.2583,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.3549337260677468,
|
328 |
+
"grad_norm": 2.541015625,
|
329 |
+
"learning_rate": 0.00019105049199771962,
|
330 |
+
"loss": 2.2466,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.3843888070692194,
|
335 |
+
"grad_norm": 2.541015625,
|
336 |
+
"learning_rate": 0.00019066338919708197,
|
337 |
+
"loss": 2.2686,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.413843888070692,
|
342 |
+
"grad_norm": 3.03125,
|
343 |
+
"learning_rate": 0.00019026850013126157,
|
344 |
+
"loss": 2.246,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.443298969072165,
|
349 |
+
"grad_norm": 2.46484375,
|
350 |
+
"learning_rate": 0.0001898658587137379,
|
351 |
+
"loss": 2.2931,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.4727540500736378,
|
356 |
+
"grad_norm": 2.740234375,
|
357 |
+
"learning_rate": 0.0001894554995237703,
|
358 |
+
"loss": 2.255,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.5022091310751104,
|
363 |
+
"grad_norm": 2.658203125,
|
364 |
+
"learning_rate": 0.00018903745780342839,
|
365 |
+
"loss": 2.2697,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.531664212076583,
|
370 |
+
"grad_norm": 2.486328125,
|
371 |
+
"learning_rate": 0.0001886117694545654,
|
372 |
+
"loss": 2.2511,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.561119293078056,
|
377 |
+
"grad_norm": 3.216796875,
|
378 |
+
"learning_rate": 0.00018817847103573486,
|
379 |
+
"loss": 2.2199,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.5905743740795288,
|
384 |
+
"grad_norm": 2.8984375,
|
385 |
+
"learning_rate": 0.00018773759975905098,
|
386 |
+
"loss": 2.2668,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.6200294550810015,
|
391 |
+
"grad_norm": 3.052734375,
|
392 |
+
"learning_rate": 0.00018728919348699283,
|
393 |
+
"loss": 2.283,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.6494845360824741,
|
398 |
+
"grad_norm": 2.8515625,
|
399 |
+
"learning_rate": 0.00018683329072915252,
|
400 |
+
"loss": 2.2239,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.678939617083947,
|
405 |
+
"grad_norm": 2.607421875,
|
406 |
+
"learning_rate": 0.0001863699306389282,
|
407 |
+
"loss": 2.1978,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.7083946980854199,
|
412 |
+
"grad_norm": 2.978515625,
|
413 |
+
"learning_rate": 0.0001858991530101613,
|
414 |
+
"loss": 2.2468,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.7378497790868925,
|
419 |
+
"grad_norm": 2.78125,
|
420 |
+
"learning_rate": 0.0001854209982737192,
|
421 |
+
"loss": 2.2494,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.7673048600883652,
|
426 |
+
"grad_norm": 2.75390625,
|
427 |
+
"learning_rate": 0.00018493550749402278,
|
428 |
+
"loss": 2.2163,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.7967599410898378,
|
433 |
+
"grad_norm": 2.908203125,
|
434 |
+
"learning_rate": 0.0001844427223655199,
|
435 |
+
"loss": 2.2156,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.8262150220913107,
|
440 |
+
"grad_norm": 3.236328125,
|
441 |
+
"learning_rate": 0.00018394268520910466,
|
442 |
+
"loss": 2.2757,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.8556701030927836,
|
447 |
+
"grad_norm": 2.51953125,
|
448 |
+
"learning_rate": 0.00018343543896848273,
|
449 |
+
"loss": 2.1927,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.8851251840942562,
|
454 |
+
"grad_norm": 3.451171875,
|
455 |
+
"learning_rate": 0.00018292102720648333,
|
456 |
+
"loss": 2.1343,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.9145802650957289,
|
461 |
+
"grad_norm": 2.0625,
|
462 |
+
"learning_rate": 0.00018239949410131802,
|
463 |
+
"loss": 2.2521,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.9440353460972017,
|
468 |
+
"grad_norm": 3.177734375,
|
469 |
+
"learning_rate": 0.00018187088444278674,
|
470 |
+
"loss": 2.2107,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.9734904270986746,
|
475 |
+
"grad_norm": 2.892578125,
|
476 |
+
"learning_rate": 0.00018133524362843104,
|
477 |
+
"loss": 2.1933,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 2.0029455081001473,
|
482 |
+
"grad_norm": 2.4296875,
|
483 |
+
"learning_rate": 0.00018079261765963537,
|
484 |
+
"loss": 2.1848,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 2.03240058910162,
|
489 |
+
"grad_norm": 2.189453125,
|
490 |
+
"learning_rate": 0.00018024305313767646,
|
491 |
+
"loss": 2.2219,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 2.0618556701030926,
|
496 |
+
"grad_norm": 2.189453125,
|
497 |
+
"learning_rate": 0.00017968659725972112,
|
498 |
+
"loss": 2.1635,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 2.0913107511045657,
|
503 |
+
"grad_norm": 2.607421875,
|
504 |
+
"learning_rate": 0.00017912329781477287,
|
505 |
+
"loss": 2.2145,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 2.1207658321060383,
|
510 |
+
"grad_norm": 1.8662109375,
|
511 |
+
"learning_rate": 0.00017855320317956784,
|
512 |
+
"loss": 2.178,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 2.150220913107511,
|
517 |
+
"grad_norm": 2.13671875,
|
518 |
+
"learning_rate": 0.00017797636231442016,
|
519 |
+
"loss": 2.1433,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 2.1796759941089836,
|
524 |
+
"grad_norm": 2.98828125,
|
525 |
+
"learning_rate": 0.000177392824759017,
|
526 |
+
"loss": 2.1158,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 2.2091310751104567,
|
531 |
+
"grad_norm": 2.25,
|
532 |
+
"learning_rate": 0.0001768026406281642,
|
533 |
+
"loss": 2.2127,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 2.2385861561119293,
|
538 |
+
"grad_norm": 2.484375,
|
539 |
+
"learning_rate": 0.00017620586060748252,
|
540 |
+
"loss": 2.1268,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 2.268041237113402,
|
545 |
+
"grad_norm": 1.9482421875,
|
546 |
+
"learning_rate": 0.00017560253594905425,
|
547 |
+
"loss": 2.1628,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.2974963181148746,
|
552 |
+
"grad_norm": 1.94140625,
|
553 |
+
"learning_rate": 0.00017499271846702213,
|
554 |
+
"loss": 2.1489,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 2.3269513991163477,
|
559 |
+
"grad_norm": 1.6005859375,
|
560 |
+
"learning_rate": 0.0001743764605331392,
|
561 |
+
"loss": 2.1078,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 2.3564064801178204,
|
566 |
+
"grad_norm": 1.677734375,
|
567 |
+
"learning_rate": 0.00017375381507227108,
|
568 |
+
"loss": 2.1365,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 2.385861561119293,
|
573 |
+
"grad_norm": 1.537109375,
|
574 |
+
"learning_rate": 0.00017312483555785086,
|
575 |
+
"loss": 2.1179,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 2.4153166421207657,
|
580 |
+
"grad_norm": 1.8291015625,
|
581 |
+
"learning_rate": 0.00017248957600728664,
|
582 |
+
"loss": 2.0896,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 2.444771723122239,
|
587 |
+
"grad_norm": 1.9423828125,
|
588 |
+
"learning_rate": 0.00017184809097732246,
|
589 |
+
"loss": 2.1057,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 2.4742268041237114,
|
594 |
+
"grad_norm": 1.443359375,
|
595 |
+
"learning_rate": 0.00017120043555935298,
|
596 |
+
"loss": 2.1325,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 2.503681885125184,
|
601 |
+
"grad_norm": 2.630859375,
|
602 |
+
"learning_rate": 0.00017054666537469213,
|
603 |
+
"loss": 2.0701,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 2.5331369661266567,
|
608 |
+
"grad_norm": 1.95703125,
|
609 |
+
"learning_rate": 0.00016988683656979624,
|
610 |
+
"loss": 2.1342,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 2.5625920471281294,
|
615 |
+
"grad_norm": 1.50390625,
|
616 |
+
"learning_rate": 0.00016922100581144228,
|
617 |
+
"loss": 2.1223,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 2.5920471281296025,
|
622 |
+
"grad_norm": 2.083984375,
|
623 |
+
"learning_rate": 0.00016854923028186111,
|
624 |
+
"loss": 2.1597,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 2.621502209131075,
|
629 |
+
"grad_norm": 1.7548828125,
|
630 |
+
"learning_rate": 0.00016787156767382659,
|
631 |
+
"loss": 2.1915,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 2.6509572901325478,
|
636 |
+
"grad_norm": 1.4912109375,
|
637 |
+
"learning_rate": 0.00016718807618570106,
|
638 |
+
"loss": 2.1678,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 2.680412371134021,
|
643 |
+
"grad_norm": 3.720703125,
|
644 |
+
"learning_rate": 0.00016649881451643705,
|
645 |
+
"loss": 2.1261,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 2.7098674521354935,
|
650 |
+
"grad_norm": 1.833984375,
|
651 |
+
"learning_rate": 0.0001658038418605361,
|
652 |
+
"loss": 2.0961,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 2.739322533136966,
|
657 |
+
"grad_norm": 1.80859375,
|
658 |
+
"learning_rate": 0.00016510321790296525,
|
659 |
+
"loss": 2.106,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 2.768777614138439,
|
664 |
+
"grad_norm": 1.5380859375,
|
665 |
+
"learning_rate": 0.00016439700281403114,
|
666 |
+
"loss": 2.112,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 2.7982326951399115,
|
671 |
+
"grad_norm": 1.341796875,
|
672 |
+
"learning_rate": 0.00016368525724421248,
|
673 |
+
"loss": 2.1424,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.827687776141384,
|
678 |
+
"grad_norm": 1.8505859375,
|
679 |
+
"learning_rate": 0.00016296804231895142,
|
680 |
+
"loss": 2.0889,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.857142857142857,
|
685 |
+
"grad_norm": 1.58203125,
|
686 |
+
"learning_rate": 0.00016224541963340391,
|
687 |
+
"loss": 2.0933,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.88659793814433,
|
692 |
+
"grad_norm": 1.447265625,
|
693 |
+
"learning_rate": 0.00016151745124715002,
|
694 |
+
"loss": 2.0716,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 2.9160530191458025,
|
699 |
+
"grad_norm": 1.2646484375,
|
700 |
+
"learning_rate": 0.00016078419967886402,
|
701 |
+
"loss": 2.0821,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 2.9455081001472756,
|
706 |
+
"grad_norm": 1.1318359375,
|
707 |
+
"learning_rate": 0.00016004572790094535,
|
708 |
+
"loss": 2.1024,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 2.9749631811487482,
|
713 |
+
"grad_norm": 1.330078125,
|
714 |
+
"learning_rate": 0.00015930209933411036,
|
715 |
+
"loss": 2.0347,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 3.004418262150221,
|
720 |
+
"grad_norm": 1.2509765625,
|
721 |
+
"learning_rate": 0.00015855337784194577,
|
722 |
+
"loss": 2.0021,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 3.0338733431516935,
|
727 |
+
"grad_norm": 1.4970703125,
|
728 |
+
"learning_rate": 0.00015779962772542402,
|
729 |
+
"loss": 2.0147,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 3.063328424153166,
|
734 |
+
"grad_norm": 1.4833984375,
|
735 |
+
"learning_rate": 0.0001570409137173809,
|
736 |
+
"loss": 2.078,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 3.0927835051546393,
|
741 |
+
"grad_norm": 1.5166015625,
|
742 |
+
"learning_rate": 0.00015627730097695638,
|
743 |
+
"loss": 2.0246,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 3.122238586156112,
|
748 |
+
"grad_norm": 1.5107421875,
|
749 |
+
"learning_rate": 0.00015550885508399856,
|
750 |
+
"loss": 2.0394,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 3.1516936671575846,
|
755 |
+
"grad_norm": 1.375,
|
756 |
+
"learning_rate": 0.00015473564203343174,
|
757 |
+
"loss": 2.0159,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 3.1811487481590572,
|
762 |
+
"grad_norm": 1.2431640625,
|
763 |
+
"learning_rate": 0.00015395772822958845,
|
764 |
+
"loss": 2.0241,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 3.2106038291605303,
|
769 |
+
"grad_norm": 1.3994140625,
|
770 |
+
"learning_rate": 0.00015317518048050697,
|
771 |
+
"loss": 1.9786,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 3.240058910162003,
|
776 |
+
"grad_norm": 1.34375,
|
777 |
+
"learning_rate": 0.00015238806599219336,
|
778 |
+
"loss": 2.0276,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 3.2695139911634756,
|
783 |
+
"grad_norm": 1.251953125,
|
784 |
+
"learning_rate": 0.0001515964523628501,
|
785 |
+
"loss": 1.9723,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 3.2989690721649483,
|
790 |
+
"grad_norm": 1.392578125,
|
791 |
+
"learning_rate": 0.00015080040757707046,
|
792 |
+
"loss": 1.951,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 3.3284241531664214,
|
797 |
+
"grad_norm": 1.380859375,
|
798 |
+
"learning_rate": 0.00015000000000000001,
|
799 |
+
"loss": 1.989,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 3.357879234167894,
|
804 |
+
"grad_norm": 1.5498046875,
|
805 |
+
"learning_rate": 0.00014919529837146528,
|
806 |
+
"loss": 2.0103,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 3.3873343151693667,
|
811 |
+
"grad_norm": 1.5712890625,
|
812 |
+
"learning_rate": 0.00014838637180007047,
|
813 |
+
"loss": 1.9914,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 3.4167893961708393,
|
818 |
+
"grad_norm": 1.8037109375,
|
819 |
+
"learning_rate": 0.00014757328975726207,
|
820 |
+
"loss": 1.9981,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 3.4462444771723124,
|
825 |
+
"grad_norm": 1.3818359375,
|
826 |
+
"learning_rate": 0.0001467561220713628,
|
827 |
+
"loss": 2.0569,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 3.475699558173785,
|
832 |
+
"grad_norm": 1.263671875,
|
833 |
+
"learning_rate": 0.00014593493892157473,
|
834 |
+
"loss": 2.0572,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 3.5051546391752577,
|
839 |
+
"grad_norm": 1.322265625,
|
840 |
+
"learning_rate": 0.00014510981083195188,
|
841 |
+
"loss": 2.0181,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 3.5346097201767304,
|
846 |
+
"grad_norm": 1.2841796875,
|
847 |
+
"learning_rate": 0.00014428080866534396,
|
848 |
+
"loss": 1.9768,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 3.564064801178203,
|
853 |
+
"grad_norm": 1.0927734375,
|
854 |
+
"learning_rate": 0.00014344800361731027,
|
855 |
+
"loss": 1.9746,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 3.593519882179676,
|
860 |
+
"grad_norm": 1.3828125,
|
861 |
+
"learning_rate": 0.00014261146721000553,
|
862 |
+
"loss": 1.9891,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 3.6229749631811488,
|
867 |
+
"grad_norm": 1.3505859375,
|
868 |
+
"learning_rate": 0.00014177127128603745,
|
869 |
+
"loss": 1.9369,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 3.6524300441826214,
|
874 |
+
"grad_norm": 1.2265625,
|
875 |
+
"learning_rate": 0.00014092748800229683,
|
876 |
+
"loss": 1.963,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 3.6818851251840945,
|
881 |
+
"grad_norm": 1.123046875,
|
882 |
+
"learning_rate": 0.00014008018982376044,
|
883 |
+
"loss": 1.9741,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 3.711340206185567,
|
888 |
+
"grad_norm": 1.1748046875,
|
889 |
+
"learning_rate": 0.0001392294495172681,
|
890 |
+
"loss": 1.9275,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 3.74079528718704,
|
895 |
+
"grad_norm": 1.369140625,
|
896 |
+
"learning_rate": 0.0001383753401452729,
|
897 |
+
"loss": 1.9831,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 3.7702503681885124,
|
902 |
+
"grad_norm": 1.2236328125,
|
903 |
+
"learning_rate": 0.0001375179350595669,
|
904 |
+
"loss": 1.9445,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 3.799705449189985,
|
909 |
+
"grad_norm": 1.341796875,
|
910 |
+
"learning_rate": 0.0001366573078949813,
|
911 |
+
"loss": 2.014,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 3.829160530191458,
|
916 |
+
"grad_norm": 1.1904296875,
|
917 |
+
"learning_rate": 0.00013579353256306287,
|
918 |
+
"loss": 2.006,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 3.858615611192931,
|
923 |
+
"grad_norm": 1.2060546875,
|
924 |
+
"learning_rate": 0.00013492668324572614,
|
925 |
+
"loss": 2.007,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 3.8880706921944035,
|
930 |
+
"grad_norm": 1.2109375,
|
931 |
+
"learning_rate": 0.00013405683438888282,
|
932 |
+
"loss": 1.9583,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 3.917525773195876,
|
937 |
+
"grad_norm": 1.216796875,
|
938 |
+
"learning_rate": 0.00013318406069604794,
|
939 |
+
"loss": 2.0087,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 3.9469808541973492,
|
944 |
+
"grad_norm": 1.2822265625,
|
945 |
+
"learning_rate": 0.00013230843712192463,
|
946 |
+
"loss": 1.993,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 3.976435935198822,
|
951 |
+
"grad_norm": 1.3349609375,
|
952 |
+
"learning_rate": 0.00013143003886596669,
|
953 |
+
"loss": 1.9554,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 4.0058910162002945,
|
958 |
+
"grad_norm": 1.1884765625,
|
959 |
+
"learning_rate": 0.00013054894136592052,
|
960 |
+
"loss": 1.8819,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 4.035346097201767,
|
965 |
+
"grad_norm": 1.3740234375,
|
966 |
+
"learning_rate": 0.00012966522029134623,
|
967 |
+
"loss": 1.8809,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 4.06480117820324,
|
972 |
+
"grad_norm": 1.3447265625,
|
973 |
+
"learning_rate": 0.00012877895153711935,
|
974 |
+
"loss": 1.8892,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 4.0942562592047125,
|
979 |
+
"grad_norm": 1.462890625,
|
980 |
+
"learning_rate": 0.00012789021121691274,
|
981 |
+
"loss": 1.892,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 4.123711340206185,
|
986 |
+
"grad_norm": 1.5107421875,
|
987 |
+
"learning_rate": 0.00012699907565665982,
|
988 |
+
"loss": 1.8828,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 4.153166421207659,
|
993 |
+
"grad_norm": 1.376953125,
|
994 |
+
"learning_rate": 0.00012610562138799978,
|
995 |
+
"loss": 1.9302,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 4.182621502209131,
|
1000 |
+
"grad_norm": 1.4072265625,
|
1001 |
+
"learning_rate": 0.0001252099251417048,
|
1002 |
+
"loss": 1.8978,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 4.212076583210604,
|
1007 |
+
"grad_norm": 1.3017578125,
|
1008 |
+
"learning_rate": 0.00012431206384109044,
|
1009 |
+
"loss": 1.8759,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 4.241531664212077,
|
1014 |
+
"grad_norm": 1.302734375,
|
1015 |
+
"learning_rate": 0.0001234121145954094,
|
1016 |
+
"loss": 1.83,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 4.270986745213549,
|
1021 |
+
"grad_norm": 1.2998046875,
|
1022 |
+
"learning_rate": 0.00012251015469322916,
|
1023 |
+
"loss": 1.9297,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 4.300441826215022,
|
1028 |
+
"grad_norm": 1.1630859375,
|
1029 |
+
"learning_rate": 0.00012160626159579447,
|
1030 |
+
"loss": 1.8802,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 4.329896907216495,
|
1035 |
+
"grad_norm": 1.2919921875,
|
1036 |
+
"learning_rate": 0.00012070051293037492,
|
1037 |
+
"loss": 1.8425,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 4.359351988217967,
|
1042 |
+
"grad_norm": 1.2509765625,
|
1043 |
+
"learning_rate": 0.00011979298648359823,
|
1044 |
+
"loss": 1.9012,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 4.388807069219441,
|
1049 |
+
"grad_norm": 1.1875,
|
1050 |
+
"learning_rate": 0.00011888376019476966,
|
1051 |
+
"loss": 1.8615,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 4.418262150220913,
|
1056 |
+
"grad_norm": 1.423828125,
|
1057 |
+
"learning_rate": 0.00011797291214917881,
|
1058 |
+
"loss": 1.8457,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 4.447717231222386,
|
1063 |
+
"grad_norm": 1.0859375,
|
1064 |
+
"learning_rate": 0.00011706052057139335,
|
1065 |
+
"loss": 1.8759,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 4.477172312223859,
|
1070 |
+
"grad_norm": 1.2333984375,
|
1071 |
+
"learning_rate": 0.00011614666381854107,
|
1072 |
+
"loss": 1.8354,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 4.506627393225331,
|
1077 |
+
"grad_norm": 1.4306640625,
|
1078 |
+
"learning_rate": 0.0001152314203735805,
|
1079 |
+
"loss": 1.9068,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 4.536082474226804,
|
1084 |
+
"grad_norm": 1.1806640625,
|
1085 |
+
"learning_rate": 0.00011431486883856082,
|
1086 |
+
"loss": 1.8969,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 4.565537555228277,
|
1091 |
+
"grad_norm": 1.11328125,
|
1092 |
+
"learning_rate": 0.00011339708792787119,
|
1093 |
+
"loss": 1.8408,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 4.594992636229749,
|
1098 |
+
"grad_norm": 1.2548828125,
|
1099 |
+
"learning_rate": 0.00011247815646148087,
|
1100 |
+
"loss": 1.9036,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 4.624447717231222,
|
1105 |
+
"grad_norm": 1.2119140625,
|
1106 |
+
"learning_rate": 0.00011155815335817011,
|
1107 |
+
"loss": 1.8163,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 4.6539027982326955,
|
1112 |
+
"grad_norm": 1.2255859375,
|
1113 |
+
"learning_rate": 0.00011063715762875225,
|
1114 |
+
"loss": 1.8382,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 4.683357879234168,
|
1119 |
+
"grad_norm": 1.3173828125,
|
1120 |
+
"learning_rate": 0.0001097152483692886,
|
1121 |
+
"loss": 1.8436,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 4.712812960235641,
|
1126 |
+
"grad_norm": 1.17578125,
|
1127 |
+
"learning_rate": 0.00010879250475429523,
|
1128 |
+
"loss": 1.8737,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 4.742268041237113,
|
1133 |
+
"grad_norm": 1.134765625,
|
1134 |
+
"learning_rate": 0.00010786900602994359,
|
1135 |
+
"loss": 1.7844,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 4.771723122238586,
|
1140 |
+
"grad_norm": 1.3564453125,
|
1141 |
+
"learning_rate": 0.00010694483150725458,
|
1142 |
+
"loss": 1.8593,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 4.801178203240059,
|
1147 |
+
"grad_norm": 1.0625,
|
1148 |
+
"learning_rate": 0.0001060200605552876,
|
1149 |
+
"loss": 1.9182,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 4.830633284241531,
|
1154 |
+
"grad_norm": 1.2099609375,
|
1155 |
+
"learning_rate": 0.00010509477259432372,
|
1156 |
+
"loss": 1.8436,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 4.860088365243005,
|
1161 |
+
"grad_norm": 1.2333984375,
|
1162 |
+
"learning_rate": 0.00010416904708904548,
|
1163 |
+
"loss": 1.8418,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 4.889543446244478,
|
1168 |
+
"grad_norm": 1.240234375,
|
1169 |
+
"learning_rate": 0.00010324296354171207,
|
1170 |
+
"loss": 1.8142,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 4.91899852724595,
|
1175 |
+
"grad_norm": 1.169921875,
|
1176 |
+
"learning_rate": 0.00010231660148533183,
|
1177 |
+
"loss": 1.8852,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 4.948453608247423,
|
1182 |
+
"grad_norm": 1.142578125,
|
1183 |
+
"learning_rate": 0.00010139004047683151,
|
1184 |
+
"loss": 1.8006,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 4.9779086892488955,
|
1189 |
+
"grad_norm": 1.1513671875,
|
1190 |
+
"learning_rate": 0.00010046336009022435,
|
1191 |
+
"loss": 1.8859,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 5.007363770250368,
|
1196 |
+
"grad_norm": 1.189453125,
|
1197 |
+
"learning_rate": 9.953663990977568e-05,
|
1198 |
+
"loss": 1.8146,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 5.036818851251841,
|
1203 |
+
"grad_norm": 1.2392578125,
|
1204 |
+
"learning_rate": 9.860995952316851e-05,
|
1205 |
+
"loss": 1.7467,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 5.0662739322533135,
|
1210 |
+
"grad_norm": 1.302734375,
|
1211 |
+
"learning_rate": 9.768339851466818e-05,
|
1212 |
+
"loss": 1.8023,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 5.095729013254786,
|
1217 |
+
"grad_norm": 1.3076171875,
|
1218 |
+
"learning_rate": 9.675703645828794e-05,
|
1219 |
+
"loss": 1.746,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 5.125184094256259,
|
1224 |
+
"grad_norm": 1.2001953125,
|
1225 |
+
"learning_rate": 9.583095291095453e-05,
|
1226 |
+
"loss": 1.7308,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 5.154639175257732,
|
1231 |
+
"grad_norm": 1.3076171875,
|
1232 |
+
"learning_rate": 9.490522740567633e-05,
|
1233 |
+
"loss": 1.7452,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 5.184094256259205,
|
1238 |
+
"grad_norm": 1.201171875,
|
1239 |
+
"learning_rate": 9.397993944471244e-05,
|
1240 |
+
"loss": 1.7385,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 5.213549337260678,
|
1245 |
+
"grad_norm": 1.3212890625,
|
1246 |
+
"learning_rate": 9.305516849274541e-05,
|
1247 |
+
"loss": 1.7481,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 5.24300441826215,
|
1252 |
+
"grad_norm": 1.4599609375,
|
1253 |
+
"learning_rate": 9.213099397005646e-05,
|
1254 |
+
"loss": 1.7526,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 5.272459499263623,
|
1259 |
+
"grad_norm": 1.2783203125,
|
1260 |
+
"learning_rate": 9.12074952457048e-05,
|
1261 |
+
"loss": 1.7547,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 5.3019145802650955,
|
1266 |
+
"grad_norm": 1.203125,
|
1267 |
+
"learning_rate": 9.028475163071141e-05,
|
1268 |
+
"loss": 1.7631,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 5.331369661266568,
|
1273 |
+
"grad_norm": 1.4521484375,
|
1274 |
+
"learning_rate": 8.936284237124778e-05,
|
1275 |
+
"loss": 1.8386,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 5.360824742268041,
|
1280 |
+
"grad_norm": 1.1865234375,
|
1281 |
+
"learning_rate": 8.844184664182993e-05,
|
1282 |
+
"loss": 1.8157,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 5.390279823269514,
|
1287 |
+
"grad_norm": 1.2294921875,
|
1288 |
+
"learning_rate": 8.752184353851916e-05,
|
1289 |
+
"loss": 1.7475,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 5.419734904270987,
|
1294 |
+
"grad_norm": 1.2275390625,
|
1295 |
+
"learning_rate": 8.660291207212882e-05,
|
1296 |
+
"loss": 1.7475,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 5.44918998527246,
|
1301 |
+
"grad_norm": 1.263671875,
|
1302 |
+
"learning_rate": 8.568513116143919e-05,
|
1303 |
+
"loss": 1.7088,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 5.478645066273932,
|
1308 |
+
"grad_norm": 1.23046875,
|
1309 |
+
"learning_rate": 8.47685796264195e-05,
|
1310 |
+
"loss": 1.7893,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 5.508100147275405,
|
1315 |
+
"grad_norm": 1.427734375,
|
1316 |
+
"learning_rate": 8.385333618145896e-05,
|
1317 |
+
"loss": 1.7206,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 5.537555228276878,
|
1322 |
+
"grad_norm": 1.1630859375,
|
1323 |
+
"learning_rate": 8.293947942860666e-05,
|
1324 |
+
"loss": 1.7444,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 5.56701030927835,
|
1329 |
+
"grad_norm": 1.1826171875,
|
1330 |
+
"learning_rate": 8.202708785082121e-05,
|
1331 |
+
"loss": 1.7685,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 5.596465390279823,
|
1336 |
+
"grad_norm": 1.2373046875,
|
1337 |
+
"learning_rate": 8.111623980523035e-05,
|
1338 |
+
"loss": 1.7494,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 5.625920471281296,
|
1343 |
+
"grad_norm": 1.248046875,
|
1344 |
+
"learning_rate": 8.020701351640182e-05,
|
1345 |
+
"loss": 1.6737,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 5.655375552282769,
|
1350 |
+
"grad_norm": 1.361328125,
|
1351 |
+
"learning_rate": 7.929948706962508e-05,
|
1352 |
+
"loss": 1.7643,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 5.684830633284242,
|
1357 |
+
"grad_norm": 1.1435546875,
|
1358 |
+
"learning_rate": 7.839373840420554e-05,
|
1359 |
+
"loss": 1.7467,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 5.714285714285714,
|
1364 |
+
"grad_norm": 1.111328125,
|
1365 |
+
"learning_rate": 7.748984530677089e-05,
|
1366 |
+
"loss": 1.7421,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 5.743740795287187,
|
1371 |
+
"grad_norm": 1.3056640625,
|
1372 |
+
"learning_rate": 7.658788540459062e-05,
|
1373 |
+
"loss": 1.7451,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 5.77319587628866,
|
1378 |
+
"grad_norm": 1.18359375,
|
1379 |
+
"learning_rate": 7.568793615890954e-05,
|
1380 |
+
"loss": 1.7545,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 5.802650957290132,
|
1385 |
+
"grad_norm": 1.115234375,
|
1386 |
+
"learning_rate": 7.479007485829523e-05,
|
1387 |
+
"loss": 1.7657,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 5.832106038291605,
|
1392 |
+
"grad_norm": 1.2001953125,
|
1393 |
+
"learning_rate": 7.389437861200024e-05,
|
1394 |
+
"loss": 1.8164,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 5.8615611192930785,
|
1399 |
+
"grad_norm": 1.1943359375,
|
1400 |
+
"learning_rate": 7.30009243433402e-05,
|
1401 |
+
"loss": 1.748,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 5.891016200294551,
|
1406 |
+
"grad_norm": 1.2333984375,
|
1407 |
+
"learning_rate": 7.210978878308729e-05,
|
1408 |
+
"loss": 1.7395,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 5.920471281296024,
|
1413 |
+
"grad_norm": 1.2294921875,
|
1414 |
+
"learning_rate": 7.122104846288064e-05,
|
1415 |
+
"loss": 1.7774,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 5.9499263622974965,
|
1420 |
+
"grad_norm": 1.2802734375,
|
1421 |
+
"learning_rate": 7.033477970865381e-05,
|
1422 |
+
"loss": 1.7505,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 5.979381443298969,
|
1427 |
+
"grad_norm": 1.2568359375,
|
1428 |
+
"learning_rate": 6.945105863407951e-05,
|
1429 |
+
"loss": 1.7314,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 6.008836524300442,
|
1434 |
+
"grad_norm": 1.240234375,
|
1435 |
+
"learning_rate": 6.85699611340333e-05,
|
1436 |
+
"loss": 1.7157,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 6.0382916053019144,
|
1441 |
+
"grad_norm": 1.3017578125,
|
1442 |
+
"learning_rate": 6.76915628780754e-05,
|
1443 |
+
"loss": 1.6333,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 6.067746686303387,
|
1448 |
+
"grad_norm": 1.3681640625,
|
1449 |
+
"learning_rate": 6.681593930395209e-05,
|
1450 |
+
"loss": 1.678,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 6.09720176730486,
|
1455 |
+
"grad_norm": 1.2626953125,
|
1456 |
+
"learning_rate": 6.594316561111724e-05,
|
1457 |
+
"loss": 1.7078,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 6.126656848306332,
|
1462 |
+
"grad_norm": 1.228515625,
|
1463 |
+
"learning_rate": 6.507331675427387e-05,
|
1464 |
+
"loss": 1.7065,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 6.156111929307806,
|
1469 |
+
"grad_norm": 1.267578125,
|
1470 |
+
"learning_rate": 6.420646743693714e-05,
|
1471 |
+
"loss": 1.6533,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 6.185567010309279,
|
1476 |
+
"grad_norm": 1.2919921875,
|
1477 |
+
"learning_rate": 6.334269210501875e-05,
|
1478 |
+
"loss": 1.6894,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 6.215022091310751,
|
1483 |
+
"grad_norm": 1.2294921875,
|
1484 |
+
"learning_rate": 6.248206494043313e-05,
|
1485 |
+
"loss": 1.6709,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 6.244477172312224,
|
1490 |
+
"grad_norm": 1.2490234375,
|
1491 |
+
"learning_rate": 6.16246598547271e-05,
|
1492 |
+
"loss": 1.6944,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 6.2739322533136965,
|
1497 |
+
"grad_norm": 1.373046875,
|
1498 |
+
"learning_rate": 6.0770550482731924e-05,
|
1499 |
+
"loss": 1.6468,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 6.303387334315169,
|
1504 |
+
"grad_norm": 1.3154296875,
|
1505 |
+
"learning_rate": 5.991981017623955e-05,
|
1506 |
+
"loss": 1.6622,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 6.332842415316642,
|
1511 |
+
"grad_norm": 1.2998046875,
|
1512 |
+
"learning_rate": 5.9072511997703226e-05,
|
1513 |
+
"loss": 1.7171,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 6.3622974963181145,
|
1518 |
+
"grad_norm": 1.1591796875,
|
1519 |
+
"learning_rate": 5.8228728713962543e-05,
|
1520 |
+
"loss": 1.6501,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 6.391752577319588,
|
1525 |
+
"grad_norm": 1.2763671875,
|
1526 |
+
"learning_rate": 5.7388532789994476e-05,
|
1527 |
+
"loss": 1.5946,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 6.421207658321061,
|
1532 |
+
"grad_norm": 1.1728515625,
|
1533 |
+
"learning_rate": 5.6551996382689776e-05,
|
1534 |
+
"loss": 1.6424,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 6.450662739322533,
|
1539 |
+
"grad_norm": 1.279296875,
|
1540 |
+
"learning_rate": 5.571919133465605e-05,
|
1541 |
+
"loss": 1.6173,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 6.480117820324006,
|
1546 |
+
"grad_norm": 1.26953125,
|
1547 |
+
"learning_rate": 5.489018916804813e-05,
|
1548 |
+
"loss": 1.6357,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 6.509572901325479,
|
1553 |
+
"grad_norm": 3.232421875,
|
1554 |
+
"learning_rate": 5.4065061078425315e-05,
|
1555 |
+
"loss": 1.6616,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 6.539027982326951,
|
1560 |
+
"grad_norm": 1.2939453125,
|
1561 |
+
"learning_rate": 5.324387792863719e-05,
|
1562 |
+
"loss": 1.7015,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 6.568483063328424,
|
1567 |
+
"grad_norm": 1.271484375,
|
1568 |
+
"learning_rate": 5.242671024273798e-05,
|
1569 |
+
"loss": 1.7161,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 6.597938144329897,
|
1574 |
+
"grad_norm": 1.2939453125,
|
1575 |
+
"learning_rate": 5.1613628199929544e-05,
|
1576 |
+
"loss": 1.6494,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 6.627393225331369,
|
1581 |
+
"grad_norm": 1.21484375,
|
1582 |
+
"learning_rate": 5.080470162853472e-05,
|
1583 |
+
"loss": 1.6409,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 6.656848306332843,
|
1588 |
+
"grad_norm": 1.359375,
|
1589 |
+
"learning_rate": 5.000000000000002e-05,
|
1590 |
+
"loss": 1.648,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 6.686303387334315,
|
1595 |
+
"grad_norm": 2.451171875,
|
1596 |
+
"learning_rate": 4.919959242292954e-05,
|
1597 |
+
"loss": 1.7565,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 6.715758468335788,
|
1602 |
+
"grad_norm": 1.3623046875,
|
1603 |
+
"learning_rate": 4.840354763714991e-05,
|
1604 |
+
"loss": 1.6658,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 6.745213549337261,
|
1609 |
+
"grad_norm": 1.388671875,
|
1610 |
+
"learning_rate": 4.7611934007806666e-05,
|
1611 |
+
"loss": 1.5883,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 6.774668630338733,
|
1616 |
+
"grad_norm": 1.3837890625,
|
1617 |
+
"learning_rate": 4.6824819519493057e-05,
|
1618 |
+
"loss": 1.6502,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 6.804123711340206,
|
1623 |
+
"grad_norm": 1.333984375,
|
1624 |
+
"learning_rate": 4.604227177041156e-05,
|
1625 |
+
"loss": 1.632,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 6.833578792341679,
|
1630 |
+
"grad_norm": 1.3525390625,
|
1631 |
+
"learning_rate": 4.5264357966568306e-05,
|
1632 |
+
"loss": 1.6804,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 6.863033873343152,
|
1637 |
+
"grad_norm": 1.2587890625,
|
1638 |
+
"learning_rate": 4.4491144916001425e-05,
|
1639 |
+
"loss": 1.6897,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 6.892488954344625,
|
1644 |
+
"grad_norm": 1.2490234375,
|
1645 |
+
"learning_rate": 4.372269902304363e-05,
|
1646 |
+
"loss": 1.6649,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 6.9219440353460975,
|
1651 |
+
"grad_norm": 1.2705078125,
|
1652 |
+
"learning_rate": 4.29590862826191e-05,
|
1653 |
+
"loss": 1.7002,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 6.95139911634757,
|
1658 |
+
"grad_norm": 1.240234375,
|
1659 |
+
"learning_rate": 4.2200372274576e-05,
|
1660 |
+
"loss": 1.6725,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 6.980854197349043,
|
1665 |
+
"grad_norm": 1.26953125,
|
1666 |
+
"learning_rate": 4.144662215805426e-05,
|
1667 |
+
"loss": 1.6438,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 7.010309278350515,
|
1672 |
+
"grad_norm": 1.30859375,
|
1673 |
+
"learning_rate": 4.069790066588967e-05,
|
1674 |
+
"loss": 1.6172,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 7.039764359351988,
|
1679 |
+
"grad_norm": 1.58203125,
|
1680 |
+
"learning_rate": 3.995427209905469e-05,
|
1681 |
+
"loss": 1.6116,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 7.069219440353461,
|
1686 |
+
"grad_norm": 1.31640625,
|
1687 |
+
"learning_rate": 3.921580032113602e-05,
|
1688 |
+
"loss": 1.5673,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 7.098674521354933,
|
1693 |
+
"grad_norm": 1.275390625,
|
1694 |
+
"learning_rate": 3.848254875285e-05,
|
1695 |
+
"loss": 1.5971,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 7.128129602356406,
|
1700 |
+
"grad_norm": 1.521484375,
|
1701 |
+
"learning_rate": 3.7754580366596115e-05,
|
1702 |
+
"loss": 1.6331,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 7.15758468335788,
|
1707 |
+
"grad_norm": 1.4345703125,
|
1708 |
+
"learning_rate": 3.7031957681048604e-05,
|
1709 |
+
"loss": 1.5961,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 7.187039764359352,
|
1714 |
+
"grad_norm": 1.3564453125,
|
1715 |
+
"learning_rate": 3.631474275578754e-05,
|
1716 |
+
"loss": 1.6064,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 7.216494845360825,
|
1721 |
+
"grad_norm": 1.3017578125,
|
1722 |
+
"learning_rate": 3.560299718596889e-05,
|
1723 |
+
"loss": 1.5493,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 7.2459499263622975,
|
1728 |
+
"grad_norm": 1.2578125,
|
1729 |
+
"learning_rate": 3.489678209703475e-05,
|
1730 |
+
"loss": 1.6051,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 7.27540500736377,
|
1735 |
+
"grad_norm": 1.3056640625,
|
1736 |
+
"learning_rate": 3.4196158139463915e-05,
|
1737 |
+
"loss": 1.6425,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 7.304860088365243,
|
1742 |
+
"grad_norm": 1.423828125,
|
1743 |
+
"learning_rate": 3.3501185483562994e-05,
|
1744 |
+
"loss": 1.68,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 7.3343151693667155,
|
1749 |
+
"grad_norm": 1.1806640625,
|
1750 |
+
"learning_rate": 3.281192381429894e-05,
|
1751 |
+
"loss": 1.6,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 7.363770250368188,
|
1756 |
+
"grad_norm": 1.3212890625,
|
1757 |
+
"learning_rate": 3.212843232617343e-05,
|
1758 |
+
"loss": 1.5919,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 7.393225331369662,
|
1763 |
+
"grad_norm": 1.2099609375,
|
1764 |
+
"learning_rate": 3.145076971813891e-05,
|
1765 |
+
"loss": 1.6371,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 7.422680412371134,
|
1770 |
+
"grad_norm": 1.6533203125,
|
1771 |
+
"learning_rate": 3.077899418855772e-05,
|
1772 |
+
"loss": 1.6582,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 7.452135493372607,
|
1777 |
+
"grad_norm": 1.3759765625,
|
1778 |
+
"learning_rate": 3.0113163430203772e-05,
|
1779 |
+
"loss": 1.642,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 7.48159057437408,
|
1784 |
+
"grad_norm": 1.2509765625,
|
1785 |
+
"learning_rate": 2.945333462530788e-05,
|
1786 |
+
"loss": 1.5679,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 7.511045655375552,
|
1791 |
+
"grad_norm": 1.3740234375,
|
1792 |
+
"learning_rate": 2.879956444064703e-05,
|
1793 |
+
"loss": 1.5998,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 7.540500736377025,
|
1798 |
+
"grad_norm": 1.3916015625,
|
1799 |
+
"learning_rate": 2.815190902267757e-05,
|
1800 |
+
"loss": 1.5767,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 7.5699558173784975,
|
1805 |
+
"grad_norm": 1.3115234375,
|
1806 |
+
"learning_rate": 2.7510423992713374e-05,
|
1807 |
+
"loss": 1.597,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 7.59941089837997,
|
1812 |
+
"grad_norm": 1.3876953125,
|
1813 |
+
"learning_rate": 2.6875164442149147e-05,
|
1814 |
+
"loss": 1.559,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 7.628865979381443,
|
1819 |
+
"grad_norm": 1.330078125,
|
1820 |
+
"learning_rate": 2.624618492772891e-05,
|
1821 |
+
"loss": 1.6197,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 7.658321060382916,
|
1826 |
+
"grad_norm": 1.478515625,
|
1827 |
+
"learning_rate": 2.5623539466860813e-05,
|
1828 |
+
"loss": 1.6207,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 7.687776141384389,
|
1833 |
+
"grad_norm": 1.421875,
|
1834 |
+
"learning_rate": 2.500728153297788e-05,
|
1835 |
+
"loss": 1.6783,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 7.717231222385862,
|
1840 |
+
"grad_norm": 1.3701171875,
|
1841 |
+
"learning_rate": 2.439746405094575e-05,
|
1842 |
+
"loss": 1.6265,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 7.746686303387334,
|
1847 |
+
"grad_norm": 1.3115234375,
|
1848 |
+
"learning_rate": 2.379413939251751e-05,
|
1849 |
+
"loss": 1.6028,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 7.776141384388807,
|
1854 |
+
"grad_norm": 1.70703125,
|
1855 |
+
"learning_rate": 2.3197359371835802e-05,
|
1856 |
+
"loss": 1.6263,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 7.80559646539028,
|
1861 |
+
"grad_norm": 1.318359375,
|
1862 |
+
"learning_rate": 2.2607175240983026e-05,
|
1863 |
+
"loss": 1.6417,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 7.835051546391752,
|
1868 |
+
"grad_norm": 1.3759765625,
|
1869 |
+
"learning_rate": 2.2023637685579856e-05,
|
1870 |
+
"loss": 1.5317,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 7.864506627393226,
|
1875 |
+
"grad_norm": 1.453125,
|
1876 |
+
"learning_rate": 2.1446796820432167e-05,
|
1877 |
+
"loss": 1.5853,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 7.8939617083946985,
|
1882 |
+
"grad_norm": 1.3369140625,
|
1883 |
+
"learning_rate": 2.0876702185227137e-05,
|
1884 |
+
"loss": 1.5672,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 7.923416789396171,
|
1889 |
+
"grad_norm": 1.3466796875,
|
1890 |
+
"learning_rate": 2.0313402740278908e-05,
|
1891 |
+
"loss": 1.5869,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 7.952871870397644,
|
1896 |
+
"grad_norm": 1.3984375,
|
1897 |
+
"learning_rate": 1.9756946862323535e-05,
|
1898 |
+
"loss": 1.5456,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 7.982326951399116,
|
1903 |
+
"grad_norm": 1.310546875,
|
1904 |
+
"learning_rate": 1.9207382340364634e-05,
|
1905 |
+
"loss": 1.5585,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 8.011782032400589,
|
1910 |
+
"grad_norm": 1.47265625,
|
1911 |
+
"learning_rate": 1.866475637156898e-05,
|
1912 |
+
"loss": 1.6146,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 8.041237113402062,
|
1917 |
+
"grad_norm": 1.3291015625,
|
1918 |
+
"learning_rate": 1.8129115557213262e-05,
|
1919 |
+
"loss": 1.5909,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 8.070692194403534,
|
1924 |
+
"grad_norm": 1.4443359375,
|
1925 |
+
"learning_rate": 1.7600505898681997e-05,
|
1926 |
+
"loss": 1.5275,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 8.100147275405007,
|
1931 |
+
"grad_norm": 1.3486328125,
|
1932 |
+
"learning_rate": 1.707897279351671e-05,
|
1933 |
+
"loss": 1.4641,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 8.12960235640648,
|
1938 |
+
"grad_norm": 1.3115234375,
|
1939 |
+
"learning_rate": 1.656456103151728e-05,
|
1940 |
+
"loss": 1.61,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 8.159057437407952,
|
1945 |
+
"grad_norm": 1.3779296875,
|
1946 |
+
"learning_rate": 1.605731479089534e-05,
|
1947 |
+
"loss": 1.5912,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 8.188512518409425,
|
1952 |
+
"grad_norm": 1.3515625,
|
1953 |
+
"learning_rate": 1.5557277634480083e-05,
|
1954 |
+
"loss": 1.5664,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 8.217967599410898,
|
1959 |
+
"grad_norm": 1.458984375,
|
1960 |
+
"learning_rate": 1.5064492505977234e-05,
|
1961 |
+
"loss": 1.5658,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 8.24742268041237,
|
1966 |
+
"grad_norm": 1.3330078125,
|
1967 |
+
"learning_rate": 1.4579001726280828e-05,
|
1968 |
+
"loss": 1.6019,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 8.276877761413845,
|
1973 |
+
"grad_norm": 1.4013671875,
|
1974 |
+
"learning_rate": 1.41008469898387e-05,
|
1975 |
+
"loss": 1.6153,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 8.306332842415317,
|
1980 |
+
"grad_norm": 1.6611328125,
|
1981 |
+
"learning_rate": 1.363006936107183e-05,
|
1982 |
+
"loss": 1.6229,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 8.33578792341679,
|
1987 |
+
"grad_norm": 1.3984375,
|
1988 |
+
"learning_rate": 1.3166709270847511e-05,
|
1989 |
+
"loss": 1.5794,
|
1990 |
+
"step": 2830
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 8.365243004418263,
|
1994 |
+
"grad_norm": 1.4560546875,
|
1995 |
+
"learning_rate": 1.271080651300719e-05,
|
1996 |
+
"loss": 1.5704,
|
1997 |
+
"step": 2840
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 8.394698085419735,
|
2001 |
+
"grad_norm": 1.3349609375,
|
2002 |
+
"learning_rate": 1.2262400240949023e-05,
|
2003 |
+
"loss": 1.5124,
|
2004 |
+
"step": 2850
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 8.424153166421208,
|
2008 |
+
"grad_norm": 1.5546875,
|
2009 |
+
"learning_rate": 1.182152896426515e-05,
|
2010 |
+
"loss": 1.5907,
|
2011 |
+
"step": 2860
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 8.45360824742268,
|
2015 |
+
"grad_norm": 1.2841796875,
|
2016 |
+
"learning_rate": 1.1388230545434653e-05,
|
2017 |
+
"loss": 1.5517,
|
2018 |
+
"step": 2870
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 8.483063328424153,
|
2022 |
+
"grad_norm": 1.3115234375,
|
2023 |
+
"learning_rate": 1.0962542196571634e-05,
|
2024 |
+
"loss": 1.567,
|
2025 |
+
"step": 2880
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 8.512518409425626,
|
2029 |
+
"grad_norm": 1.41796875,
|
2030 |
+
"learning_rate": 1.0544500476229713e-05,
|
2031 |
+
"loss": 1.4715,
|
2032 |
+
"step": 2890
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 8.541973490427099,
|
2036 |
+
"grad_norm": 1.5263671875,
|
2037 |
+
"learning_rate": 1.013414128626211e-05,
|
2038 |
+
"loss": 1.5595,
|
2039 |
+
"step": 2900
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 8.571428571428571,
|
2043 |
+
"grad_norm": 1.44140625,
|
2044 |
+
"learning_rate": 9.731499868738447e-06,
|
2045 |
+
"loss": 1.5258,
|
2046 |
+
"step": 2910
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 8.600883652430044,
|
2050 |
+
"grad_norm": 1.2353515625,
|
2051 |
+
"learning_rate": 9.336610802918044e-06,
|
2052 |
+
"loss": 1.5395,
|
2053 |
+
"step": 2920
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 8.630338733431516,
|
2057 |
+
"grad_norm": 1.3125,
|
2058 |
+
"learning_rate": 8.949508002280382e-06,
|
2059 |
+
"loss": 1.571,
|
2060 |
+
"step": 2930
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 8.65979381443299,
|
2064 |
+
"grad_norm": 1.3330078125,
|
2065 |
+
"learning_rate": 8.570224711612385e-06,
|
2066 |
+
"loss": 1.6215,
|
2067 |
+
"step": 2940
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 8.689248895434462,
|
2071 |
+
"grad_norm": 1.404296875,
|
2072 |
+
"learning_rate": 8.19879350415349e-06,
|
2073 |
+
"loss": 1.5996,
|
2074 |
+
"step": 2950
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 8.718703976435934,
|
2078 |
+
"grad_norm": 1.5576171875,
|
2079 |
+
"learning_rate": 7.835246278798037e-06,
|
2080 |
+
"loss": 1.5238,
|
2081 |
+
"step": 2960
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 8.748159057437409,
|
2085 |
+
"grad_norm": 1.501953125,
|
2086 |
+
"learning_rate": 7.479614257355971e-06,
|
2087 |
+
"loss": 1.593,
|
2088 |
+
"step": 2970
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 8.777614138438881,
|
2092 |
+
"grad_norm": 1.5166015625,
|
2093 |
+
"learning_rate": 7.1319279818713445e-06,
|
2094 |
+
"loss": 1.5612,
|
2095 |
+
"step": 2980
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 8.807069219440354,
|
2099 |
+
"grad_norm": 1.7763671875,
|
2100 |
+
"learning_rate": 6.7922173119993606e-06,
|
2101 |
+
"loss": 1.5534,
|
2102 |
+
"step": 2990
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 8.836524300441827,
|
2106 |
+
"grad_norm": 1.388671875,
|
2107 |
+
"learning_rate": 6.460511422441984e-06,
|
2108 |
+
"loss": 1.5691,
|
2109 |
+
"step": 3000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 8.8659793814433,
|
2113 |
+
"grad_norm": 1.412109375,
|
2114 |
+
"learning_rate": 6.136838800442457e-06,
|
2115 |
+
"loss": 1.6044,
|
2116 |
+
"step": 3010
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 8.895434462444772,
|
2120 |
+
"grad_norm": 1.36328125,
|
2121 |
+
"learning_rate": 5.821227243338712e-06,
|
2122 |
+
"loss": 1.6178,
|
2123 |
+
"step": 3020
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 8.924889543446245,
|
2127 |
+
"grad_norm": 1.41796875,
|
2128 |
+
"learning_rate": 5.5137038561761115e-06,
|
2129 |
+
"loss": 1.6223,
|
2130 |
+
"step": 3030
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 8.954344624447717,
|
2134 |
+
"grad_norm": 1.44140625,
|
2135 |
+
"learning_rate": 5.214295049379658e-06,
|
2136 |
+
"loss": 1.5837,
|
2137 |
+
"step": 3040
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 8.98379970544919,
|
2141 |
+
"grad_norm": 1.3671875,
|
2142 |
+
"learning_rate": 4.923026536485875e-06,
|
2143 |
+
"loss": 1.523,
|
2144 |
+
"step": 3050
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 9.013254786450663,
|
2148 |
+
"grad_norm": 1.33203125,
|
2149 |
+
"learning_rate": 4.639923331934471e-06,
|
2150 |
+
"loss": 1.5245,
|
2151 |
+
"step": 3060
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 9.042709867452135,
|
2155 |
+
"grad_norm": 1.40625,
|
2156 |
+
"learning_rate": 4.365009748920012e-06,
|
2157 |
+
"loss": 1.5816,
|
2158 |
+
"step": 3070
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 9.072164948453608,
|
2162 |
+
"grad_norm": 1.255859375,
|
2163 |
+
"learning_rate": 4.098309397303978e-06,
|
2164 |
+
"loss": 1.5324,
|
2165 |
+
"step": 3080
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 9.10162002945508,
|
2169 |
+
"grad_norm": 1.5517578125,
|
2170 |
+
"learning_rate": 3.839845181587098e-06,
|
2171 |
+
"loss": 1.6157,
|
2172 |
+
"step": 3090
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 9.131075110456553,
|
2176 |
+
"grad_norm": 1.2978515625,
|
2177 |
+
"learning_rate": 3.5896392989422377e-06,
|
2178 |
+
"loss": 1.5071,
|
2179 |
+
"step": 3100
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 9.160530191458026,
|
2183 |
+
"grad_norm": 1.537109375,
|
2184 |
+
"learning_rate": 3.3477132373081254e-06,
|
2185 |
+
"loss": 1.646,
|
2186 |
+
"step": 3110
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 9.189985272459499,
|
2190 |
+
"grad_norm": 1.373046875,
|
2191 |
+
"learning_rate": 3.1140877735439387e-06,
|
2192 |
+
"loss": 1.5886,
|
2193 |
+
"step": 3120
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 9.219440353460971,
|
2197 |
+
"grad_norm": 1.314453125,
|
2198 |
+
"learning_rate": 2.8887829716449876e-06,
|
2199 |
+
"loss": 1.5618,
|
2200 |
+
"step": 3130
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 9.248895434462444,
|
2204 |
+
"grad_norm": 1.326171875,
|
2205 |
+
"learning_rate": 2.6718181810195696e-06,
|
2206 |
+
"loss": 1.5741,
|
2207 |
+
"step": 3140
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 9.278350515463918,
|
2211 |
+
"grad_norm": 1.3994140625,
|
2212 |
+
"learning_rate": 2.4632120348272003e-06,
|
2213 |
+
"loss": 1.5673,
|
2214 |
+
"step": 3150
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 9.307805596465391,
|
2218 |
+
"grad_norm": 1.4306640625,
|
2219 |
+
"learning_rate": 2.2629824483784366e-06,
|
2220 |
+
"loss": 1.5312,
|
2221 |
+
"step": 3160
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 9.337260677466864,
|
2225 |
+
"grad_norm": 1.4052734375,
|
2226 |
+
"learning_rate": 2.0711466175962756e-06,
|
2227 |
+
"loss": 1.5329,
|
2228 |
+
"step": 3170
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 9.366715758468336,
|
2232 |
+
"grad_norm": 1.3193359375,
|
2233 |
+
"learning_rate": 1.88772101753929e-06,
|
2234 |
+
"loss": 1.5319,
|
2235 |
+
"step": 3180
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 9.396170839469809,
|
2239 |
+
"grad_norm": 1.3486328125,
|
2240 |
+
"learning_rate": 1.7127214009868385e-06,
|
2241 |
+
"loss": 1.4908,
|
2242 |
+
"step": 3190
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 9.425625920471282,
|
2246 |
+
"grad_norm": 1.4267578125,
|
2247 |
+
"learning_rate": 1.5461627970860814e-06,
|
2248 |
+
"loss": 1.6365,
|
2249 |
+
"step": 3200
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 9.455081001472754,
|
2253 |
+
"grad_norm": 1.4658203125,
|
2254 |
+
"learning_rate": 1.3880595100613792e-06,
|
2255 |
+
"loss": 1.5504,
|
2256 |
+
"step": 3210
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 9.484536082474227,
|
2260 |
+
"grad_norm": 1.685546875,
|
2261 |
+
"learning_rate": 1.2384251179857643e-06,
|
2262 |
+
"loss": 1.5451,
|
2263 |
+
"step": 3220
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 9.5139911634757,
|
2267 |
+
"grad_norm": 1.4287109375,
|
2268 |
+
"learning_rate": 1.0972724716148187e-06,
|
2269 |
+
"loss": 1.5691,
|
2270 |
+
"step": 3230
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 9.543446244477172,
|
2274 |
+
"grad_norm": 1.345703125,
|
2275 |
+
"learning_rate": 9.64613693283123e-07,
|
2276 |
+
"loss": 1.5342,
|
2277 |
+
"step": 3240
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 9.572901325478645,
|
2281 |
+
"grad_norm": 1.3359375,
|
2282 |
+
"learning_rate": 8.404601758630892e-07,
|
2283 |
+
"loss": 1.5575,
|
2284 |
+
"step": 3250
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 9.602356406480117,
|
2288 |
+
"grad_norm": 1.3681640625,
|
2289 |
+
"learning_rate": 7.248225817865884e-07,
|
2290 |
+
"loss": 1.5327,
|
2291 |
+
"step": 3260
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 9.63181148748159,
|
2295 |
+
"grad_norm": 1.37890625,
|
2296 |
+
"learning_rate": 6.177108421292266e-07,
|
2297 |
+
"loss": 1.5548,
|
2298 |
+
"step": 3270
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 9.661266568483063,
|
2302 |
+
"grad_norm": 1.353515625,
|
2303 |
+
"learning_rate": 5.191341557574392e-07,
|
2304 |
+
"loss": 1.6446,
|
2305 |
+
"step": 3280
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 9.690721649484535,
|
2309 |
+
"grad_norm": 1.3310546875,
|
2310 |
+
"learning_rate": 4.291009885385333e-07,
|
2311 |
+
"loss": 1.4653,
|
2312 |
+
"step": 3290
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 9.720176730486008,
|
2316 |
+
"grad_norm": 1.380859375,
|
2317 |
+
"learning_rate": 3.4761907261356976e-07,
|
2318 |
+
"loss": 1.5371,
|
2319 |
+
"step": 3300
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 9.749631811487482,
|
2323 |
+
"grad_norm": 1.4384765625,
|
2324 |
+
"learning_rate": 2.746954057333606e-07,
|
2325 |
+
"loss": 1.5145,
|
2326 |
+
"step": 3310
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 9.779086892488955,
|
2330 |
+
"grad_norm": 1.3583984375,
|
2331 |
+
"learning_rate": 2.1033625065747242e-07,
|
2332 |
+
"loss": 1.6238,
|
2333 |
+
"step": 3320
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 9.808541973490428,
|
2337 |
+
"grad_norm": 1.3671875,
|
2338 |
+
"learning_rate": 1.545471346164007e-07,
|
2339 |
+
"loss": 1.5669,
|
2340 |
+
"step": 3330
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 9.8379970544919,
|
2344 |
+
"grad_norm": 1.34765625,
|
2345 |
+
"learning_rate": 1.0733284883682749e-07,
|
2346 |
+
"loss": 1.5792,
|
2347 |
+
"step": 3340
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 9.867452135493373,
|
2351 |
+
"grad_norm": 1.4501953125,
|
2352 |
+
"learning_rate": 6.869744813023937e-08,
|
2353 |
+
"loss": 1.6,
|
2354 |
+
"step": 3350
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 9.896907216494846,
|
2358 |
+
"grad_norm": 1.443359375,
|
2359 |
+
"learning_rate": 3.8644250544594975e-08,
|
2360 |
+
"loss": 1.5093,
|
2361 |
+
"step": 3360
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 9.926362297496318,
|
2365 |
+
"grad_norm": 1.51171875,
|
2366 |
+
"learning_rate": 1.7175837079452804e-08,
|
2367 |
+
"loss": 1.5322,
|
2368 |
+
"step": 3370
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 9.955817378497791,
|
2372 |
+
"grad_norm": 1.3876953125,
|
2373 |
+
"learning_rate": 4.2940514642597626e-09,
|
2374 |
+
"loss": 1.5623,
|
2375 |
+
"step": 3380
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 9.985272459499264,
|
2379 |
+
"grad_norm": 1.30078125,
|
2380 |
+
"learning_rate": 0.0,
|
2381 |
+
"loss": 1.5123,
|
2382 |
+
"step": 3390
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 9.985272459499264,
|
2386 |
+
"step": 3390,
|
2387 |
+
"total_flos": 2.22628436508672e+17,
|
2388 |
+
"train_loss": 1.877983266273431,
|
2389 |
+
"train_runtime": 3392.4525,
|
2390 |
+
"train_samples_per_second": 4.003,
|
2391 |
+
"train_steps_per_second": 0.999
|
2392 |
+
}
|
2393 |
+
],
|
2394 |
+
"logging_steps": 10,
|
2395 |
+
"max_steps": 3390,
|
2396 |
+
"num_input_tokens_seen": 0,
|
2397 |
+
"num_train_epochs": 10,
|
2398 |
+
"save_steps": 500,
|
2399 |
+
"total_flos": 2.22628436508672e+17,
|
2400 |
+
"train_batch_size": 1,
|
2401 |
+
"trial_name": null,
|
2402 |
+
"trial_params": null
|
2403 |
+
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5112
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:24850aff70c0816e287c31070a300308cfb0ebf456a8f1d7fbf9718ee683ebea
|
3 |
size 5112
|