Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- trainer_state.json +3054 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,13 +20,13 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"o_proj",
|
| 24 |
-
"gate_proj",
|
| 25 |
"down_proj",
|
| 26 |
"up_proj",
|
| 27 |
-
"
|
|
|
|
| 28 |
"v_proj",
|
| 29 |
-
"
|
|
|
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"down_proj",
|
| 24 |
"up_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
+
"k_proj",
|
| 27 |
"v_proj",
|
| 28 |
+
"gate_proj",
|
| 29 |
+
"q_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"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 78480072
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:16606fca4ea6f77bcf19caa4c75fbbf98d9032ce6a344b0a25e624916065b60e
|
| 3 |
size 78480072
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3054 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.994216310005784,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4320,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02313475997686524,
|
| 13 |
+
"grad_norm": 4.818889617919922,
|
| 14 |
+
"learning_rate": 0.00019999735576321776,
|
| 15 |
+
"loss": 2.7586,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04626951995373048,
|
| 20 |
+
"grad_norm": 1.2888654470443726,
|
| 21 |
+
"learning_rate": 0.00019998942319271077,
|
| 22 |
+
"loss": 2.4225,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06940427993059572,
|
| 27 |
+
"grad_norm": 0.9638609290122986,
|
| 28 |
+
"learning_rate": 0.00019997620270799092,
|
| 29 |
+
"loss": 2.2447,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.09253903990746096,
|
| 34 |
+
"grad_norm": 0.882834255695343,
|
| 35 |
+
"learning_rate": 0.0001999576950082201,
|
| 36 |
+
"loss": 2.2015,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1156737998843262,
|
| 41 |
+
"grad_norm": 1.0840070247650146,
|
| 42 |
+
"learning_rate": 0.00019993390107217302,
|
| 43 |
+
"loss": 2.1489,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.13880855986119145,
|
| 48 |
+
"grad_norm": 13.946427345275879,
|
| 49 |
+
"learning_rate": 0.0001999048221581858,
|
| 50 |
+
"loss": 2.1928,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.16194331983805668,
|
| 55 |
+
"grad_norm": 0.9393675327301025,
|
| 56 |
+
"learning_rate": 0.00019987045980408906,
|
| 57 |
+
"loss": 2.172,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.18507807981492191,
|
| 62 |
+
"grad_norm": 0.882853627204895,
|
| 63 |
+
"learning_rate": 0.00019983081582712685,
|
| 64 |
+
"loss": 2.1491,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.20821283979178715,
|
| 69 |
+
"grad_norm": 0.8838891386985779,
|
| 70 |
+
"learning_rate": 0.00019978589232386035,
|
| 71 |
+
"loss": 2.1147,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2313475997686524,
|
| 76 |
+
"grad_norm": 0.9649007320404053,
|
| 77 |
+
"learning_rate": 0.00019973569167005723,
|
| 78 |
+
"loss": 2.0984,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.25448235974551764,
|
| 83 |
+
"grad_norm": 0.9888186454772949,
|
| 84 |
+
"learning_rate": 0.00019968021652056576,
|
| 85 |
+
"loss": 2.0735,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2776171197223829,
|
| 90 |
+
"grad_norm": 0.8648116588592529,
|
| 91 |
+
"learning_rate": 0.00019961946980917456,
|
| 92 |
+
"loss": 2.0005,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3007518796992481,
|
| 97 |
+
"grad_norm": 0.932079553604126,
|
| 98 |
+
"learning_rate": 0.0001995534547484574,
|
| 99 |
+
"loss": 1.9938,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.32388663967611336,
|
| 104 |
+
"grad_norm": 0.9043927788734436,
|
| 105 |
+
"learning_rate": 0.0001994821748296033,
|
| 106 |
+
"loss": 1.9941,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3470213996529786,
|
| 111 |
+
"grad_norm": 0.9096814393997192,
|
| 112 |
+
"learning_rate": 0.00019940563382223197,
|
| 113 |
+
"loss": 2.0175,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.37015615962984383,
|
| 118 |
+
"grad_norm": 0.9436717629432678,
|
| 119 |
+
"learning_rate": 0.00019932383577419432,
|
| 120 |
+
"loss": 2.0761,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3932909196067091,
|
| 125 |
+
"grad_norm": 0.8598865270614624,
|
| 126 |
+
"learning_rate": 0.00019923678501135848,
|
| 127 |
+
"loss": 2.0289,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4164256795835743,
|
| 132 |
+
"grad_norm": 0.8160977363586426,
|
| 133 |
+
"learning_rate": 0.00019914448613738106,
|
| 134 |
+
"loss": 2.0161,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.43956043956043955,
|
| 139 |
+
"grad_norm": 0.8758726119995117,
|
| 140 |
+
"learning_rate": 0.0001990469440334636,
|
| 141 |
+
"loss": 1.95,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4626951995373048,
|
| 146 |
+
"grad_norm": 0.9540367126464844,
|
| 147 |
+
"learning_rate": 0.00019894416385809444,
|
| 148 |
+
"loss": 2.033,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.48582995951417,
|
| 153 |
+
"grad_norm": 0.9104657173156738,
|
| 154 |
+
"learning_rate": 0.00019883615104677608,
|
| 155 |
+
"loss": 1.965,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5089647194910353,
|
| 160 |
+
"grad_norm": 0.9348485469818115,
|
| 161 |
+
"learning_rate": 0.00019872291131173742,
|
| 162 |
+
"loss": 1.9257,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5320994794679005,
|
| 167 |
+
"grad_norm": 1.0083684921264648,
|
| 168 |
+
"learning_rate": 0.00019860445064163193,
|
| 169 |
+
"loss": 1.9851,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5552342394447658,
|
| 174 |
+
"grad_norm": 0.9179242849349976,
|
| 175 |
+
"learning_rate": 0.00019848077530122083,
|
| 176 |
+
"loss": 1.9656,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.578368999421631,
|
| 181 |
+
"grad_norm": 0.881529688835144,
|
| 182 |
+
"learning_rate": 0.00019835189183104178,
|
| 183 |
+
"loss": 1.9819,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6015037593984962,
|
| 188 |
+
"grad_norm": 0.931932270526886,
|
| 189 |
+
"learning_rate": 0.00019821780704706307,
|
| 190 |
+
"loss": 1.9791,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6246385193753615,
|
| 195 |
+
"grad_norm": 0.9711050391197205,
|
| 196 |
+
"learning_rate": 0.00019807852804032305,
|
| 197 |
+
"loss": 1.9672,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6477732793522267,
|
| 202 |
+
"grad_norm": 0.9100776314735413,
|
| 203 |
+
"learning_rate": 0.00019793406217655517,
|
| 204 |
+
"loss": 1.9403,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6709080393290919,
|
| 209 |
+
"grad_norm": 1.0750184059143066,
|
| 210 |
+
"learning_rate": 0.0001977844170957984,
|
| 211 |
+
"loss": 1.9222,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6940427993059572,
|
| 216 |
+
"grad_norm": 1.0510389804840088,
|
| 217 |
+
"learning_rate": 0.00019762960071199333,
|
| 218 |
+
"loss": 1.8928,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7171775592828225,
|
| 223 |
+
"grad_norm": 0.8637176156044006,
|
| 224 |
+
"learning_rate": 0.0001974696212125635,
|
| 225 |
+
"loss": 1.9064,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7403123192596877,
|
| 230 |
+
"grad_norm": 0.9364832043647766,
|
| 231 |
+
"learning_rate": 0.00019730448705798239,
|
| 232 |
+
"loss": 1.9649,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.763447079236553,
|
| 237 |
+
"grad_norm": 1.0437376499176025,
|
| 238 |
+
"learning_rate": 0.00019713420698132614,
|
| 239 |
+
"loss": 1.9472,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7865818392134182,
|
| 244 |
+
"grad_norm": 0.8804395198822021,
|
| 245 |
+
"learning_rate": 0.0001969587899878116,
|
| 246 |
+
"loss": 1.8961,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8097165991902834,
|
| 251 |
+
"grad_norm": 0.9558535814285278,
|
| 252 |
+
"learning_rate": 0.0001967782453543201,
|
| 253 |
+
"loss": 1.9841,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8328513591671486,
|
| 258 |
+
"grad_norm": 1.0408087968826294,
|
| 259 |
+
"learning_rate": 0.00019659258262890683,
|
| 260 |
+
"loss": 1.9095,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8559861191440139,
|
| 265 |
+
"grad_norm": 1.0767790079116821,
|
| 266 |
+
"learning_rate": 0.00019640181163029595,
|
| 267 |
+
"loss": 1.8774,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8791208791208791,
|
| 272 |
+
"grad_norm": 0.9003483057022095,
|
| 273 |
+
"learning_rate": 0.00019620594244736133,
|
| 274 |
+
"loss": 1.8791,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9022556390977443,
|
| 279 |
+
"grad_norm": 0.9258147478103638,
|
| 280 |
+
"learning_rate": 0.0001960049854385929,
|
| 281 |
+
"loss": 1.9243,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9253903990746096,
|
| 286 |
+
"grad_norm": 0.9662225842475891,
|
| 287 |
+
"learning_rate": 0.0001957989512315489,
|
| 288 |
+
"loss": 1.8853,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9485251590514748,
|
| 293 |
+
"grad_norm": 0.8478869795799255,
|
| 294 |
+
"learning_rate": 0.00019558785072229396,
|
| 295 |
+
"loss": 1.8573,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.97165991902834,
|
| 300 |
+
"grad_norm": 0.9442031979560852,
|
| 301 |
+
"learning_rate": 0.0001953716950748227,
|
| 302 |
+
"loss": 1.8542,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.9947946790052054,
|
| 307 |
+
"grad_norm": 0.970368504524231,
|
| 308 |
+
"learning_rate": 0.00019515049572046937,
|
| 309 |
+
"loss": 1.9055,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.0179294389820706,
|
| 314 |
+
"grad_norm": 0.9968038201332092,
|
| 315 |
+
"learning_rate": 0.0001949242643573034,
|
| 316 |
+
"loss": 1.7246,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.0410641989589358,
|
| 321 |
+
"grad_norm": 0.9842697978019714,
|
| 322 |
+
"learning_rate": 0.0001946930129495106,
|
| 323 |
+
"loss": 1.726,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.064198958935801,
|
| 328 |
+
"grad_norm": 1.029723882675171,
|
| 329 |
+
"learning_rate": 0.0001944567537267605,
|
| 330 |
+
"loss": 1.6401,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.0873337189126664,
|
| 335 |
+
"grad_norm": 0.95682692527771,
|
| 336 |
+
"learning_rate": 0.00019421549918355966,
|
| 337 |
+
"loss": 1.642,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.1104684788895316,
|
| 342 |
+
"grad_norm": 1.1155472993850708,
|
| 343 |
+
"learning_rate": 0.00019396926207859084,
|
| 344 |
+
"loss": 1.7003,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.1336032388663968,
|
| 349 |
+
"grad_norm": 1.0741841793060303,
|
| 350 |
+
"learning_rate": 0.00019371805543403825,
|
| 351 |
+
"loss": 1.6915,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.156737998843262,
|
| 356 |
+
"grad_norm": 1.0780987739562988,
|
| 357 |
+
"learning_rate": 0.00019346189253489885,
|
| 358 |
+
"loss": 1.5675,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.1798727588201272,
|
| 363 |
+
"grad_norm": 1.0540192127227783,
|
| 364 |
+
"learning_rate": 0.00019320078692827987,
|
| 365 |
+
"loss": 1.6459,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.2030075187969924,
|
| 370 |
+
"grad_norm": 1.0274903774261475,
|
| 371 |
+
"learning_rate": 0.00019293475242268223,
|
| 372 |
+
"loss": 1.6763,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.2261422787738576,
|
| 377 |
+
"grad_norm": 1.0954372882843018,
|
| 378 |
+
"learning_rate": 0.00019266380308727055,
|
| 379 |
+
"loss": 1.7276,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.249277038750723,
|
| 384 |
+
"grad_norm": 1.1079906225204468,
|
| 385 |
+
"learning_rate": 0.0001923879532511287,
|
| 386 |
+
"loss": 1.6581,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.2724117987275883,
|
| 391 |
+
"grad_norm": 1.053789734840393,
|
| 392 |
+
"learning_rate": 0.00019210721750250235,
|
| 393 |
+
"loss": 1.709,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.2955465587044535,
|
| 398 |
+
"grad_norm": 1.0135327577590942,
|
| 399 |
+
"learning_rate": 0.00019182161068802741,
|
| 400 |
+
"loss": 1.6606,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.3186813186813187,
|
| 405 |
+
"grad_norm": 1.0936933755874634,
|
| 406 |
+
"learning_rate": 0.00019153114791194473,
|
| 407 |
+
"loss": 1.6489,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.3418160786581839,
|
| 412 |
+
"grad_norm": 1.0684951543807983,
|
| 413 |
+
"learning_rate": 0.00019123584453530144,
|
| 414 |
+
"loss": 1.7086,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.3649508386350493,
|
| 419 |
+
"grad_norm": 1.1002228260040283,
|
| 420 |
+
"learning_rate": 0.00019093571617513852,
|
| 421 |
+
"loss": 1.6785,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.3880855986119145,
|
| 426 |
+
"grad_norm": 1.0296435356140137,
|
| 427 |
+
"learning_rate": 0.000190630778703665,
|
| 428 |
+
"loss": 1.6931,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.4112203585887797,
|
| 433 |
+
"grad_norm": 1.2503811120986938,
|
| 434 |
+
"learning_rate": 0.00019032104824741843,
|
| 435 |
+
"loss": 1.6651,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.434355118565645,
|
| 440 |
+
"grad_norm": 1.048912763595581,
|
| 441 |
+
"learning_rate": 0.00019000654118641211,
|
| 442 |
+
"loss": 1.7226,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.45748987854251,
|
| 447 |
+
"grad_norm": 1.1050567626953125,
|
| 448 |
+
"learning_rate": 0.00018968727415326884,
|
| 449 |
+
"loss": 1.6195,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.4806246385193753,
|
| 454 |
+
"grad_norm": 1.1447317600250244,
|
| 455 |
+
"learning_rate": 0.00018936326403234125,
|
| 456 |
+
"loss": 1.6719,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.5037593984962405,
|
| 461 |
+
"grad_norm": 1.2653794288635254,
|
| 462 |
+
"learning_rate": 0.0001890345279588189,
|
| 463 |
+
"loss": 1.6512,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.5268941584731057,
|
| 468 |
+
"grad_norm": 1.2562129497528076,
|
| 469 |
+
"learning_rate": 0.00018870108331782217,
|
| 470 |
+
"loss": 1.7135,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.550028918449971,
|
| 475 |
+
"grad_norm": 1.158982753753662,
|
| 476 |
+
"learning_rate": 0.00018836294774348278,
|
| 477 |
+
"loss": 1.6415,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.5731636784268364,
|
| 482 |
+
"grad_norm": 1.102187156677246,
|
| 483 |
+
"learning_rate": 0.00018802013911801112,
|
| 484 |
+
"loss": 1.589,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.5962984384037016,
|
| 489 |
+
"grad_norm": 1.126552939414978,
|
| 490 |
+
"learning_rate": 0.0001876726755707508,
|
| 491 |
+
"loss": 1.661,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.6194331983805668,
|
| 496 |
+
"grad_norm": 1.1704193353652954,
|
| 497 |
+
"learning_rate": 0.0001873205754772196,
|
| 498 |
+
"loss": 1.7446,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.6425679583574322,
|
| 503 |
+
"grad_norm": 1.1397557258605957,
|
| 504 |
+
"learning_rate": 0.00018696385745813792,
|
| 505 |
+
"loss": 1.6528,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6657027183342974,
|
| 510 |
+
"grad_norm": 1.212633728981018,
|
| 511 |
+
"learning_rate": 0.00018660254037844388,
|
| 512 |
+
"loss": 1.6315,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.6888374783111626,
|
| 517 |
+
"grad_norm": 1.124883770942688,
|
| 518 |
+
"learning_rate": 0.00018623664334629575,
|
| 519 |
+
"loss": 1.6961,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.7119722382880278,
|
| 524 |
+
"grad_norm": 1.1361215114593506,
|
| 525 |
+
"learning_rate": 0.00018586618571206134,
|
| 526 |
+
"loss": 1.7002,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.735106998264893,
|
| 531 |
+
"grad_norm": 1.3424755334854126,
|
| 532 |
+
"learning_rate": 0.00018549118706729468,
|
| 533 |
+
"loss": 1.6754,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.7582417582417582,
|
| 538 |
+
"grad_norm": 1.1060627698898315,
|
| 539 |
+
"learning_rate": 0.00018511166724369997,
|
| 540 |
+
"loss": 1.6363,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.7813765182186234,
|
| 545 |
+
"grad_norm": 1.1465959548950195,
|
| 546 |
+
"learning_rate": 0.0001847276463120828,
|
| 547 |
+
"loss": 1.5913,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.8045112781954886,
|
| 552 |
+
"grad_norm": 1.1735104322433472,
|
| 553 |
+
"learning_rate": 0.0001843391445812886,
|
| 554 |
+
"loss": 1.6201,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.8276460381723538,
|
| 559 |
+
"grad_norm": 1.0827069282531738,
|
| 560 |
+
"learning_rate": 0.00018394618259712865,
|
| 561 |
+
"loss": 1.6184,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.8507807981492193,
|
| 566 |
+
"grad_norm": 1.0566381216049194,
|
| 567 |
+
"learning_rate": 0.00018354878114129367,
|
| 568 |
+
"loss": 1.64,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.8739155581260845,
|
| 573 |
+
"grad_norm": 1.223117470741272,
|
| 574 |
+
"learning_rate": 0.00018314696123025454,
|
| 575 |
+
"loss": 1.651,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.8970503181029497,
|
| 580 |
+
"grad_norm": 1.2156977653503418,
|
| 581 |
+
"learning_rate": 0.00018274074411415105,
|
| 582 |
+
"loss": 1.6914,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.9201850780798149,
|
| 587 |
+
"grad_norm": 1.179998517036438,
|
| 588 |
+
"learning_rate": 0.00018233015127566807,
|
| 589 |
+
"loss": 1.6974,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.9433198380566803,
|
| 594 |
+
"grad_norm": 1.1092870235443115,
|
| 595 |
+
"learning_rate": 0.0001819152044288992,
|
| 596 |
+
"loss": 1.5961,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.9664545980335455,
|
| 601 |
+
"grad_norm": 1.0674152374267578,
|
| 602 |
+
"learning_rate": 0.0001814959255181988,
|
| 603 |
+
"loss": 1.6687,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.9895893580104107,
|
| 608 |
+
"grad_norm": 1.3172643184661865,
|
| 609 |
+
"learning_rate": 0.00018107233671702124,
|
| 610 |
+
"loss": 1.7101,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.012724117987276,
|
| 615 |
+
"grad_norm": 1.3380361795425415,
|
| 616 |
+
"learning_rate": 0.00018064446042674828,
|
| 617 |
+
"loss": 1.5151,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.035858877964141,
|
| 622 |
+
"grad_norm": 1.431959867477417,
|
| 623 |
+
"learning_rate": 0.0001802123192755044,
|
| 624 |
+
"loss": 1.3543,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.0589936379410063,
|
| 629 |
+
"grad_norm": 1.2023745775222778,
|
| 630 |
+
"learning_rate": 0.00017977593611696015,
|
| 631 |
+
"loss": 1.3866,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.0821283979178715,
|
| 636 |
+
"grad_norm": 1.4215834140777588,
|
| 637 |
+
"learning_rate": 0.00017933533402912354,
|
| 638 |
+
"loss": 1.3933,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.1052631578947367,
|
| 643 |
+
"grad_norm": 1.2749346494674683,
|
| 644 |
+
"learning_rate": 0.00017889053631311947,
|
| 645 |
+
"loss": 1.3135,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.128397917871602,
|
| 650 |
+
"grad_norm": 1.347468376159668,
|
| 651 |
+
"learning_rate": 0.00017844156649195759,
|
| 652 |
+
"loss": 1.3276,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.151532677848467,
|
| 657 |
+
"grad_norm": 1.3235530853271484,
|
| 658 |
+
"learning_rate": 0.00017798844830928817,
|
| 659 |
+
"loss": 1.3517,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.174667437825333,
|
| 664 |
+
"grad_norm": 1.4696698188781738,
|
| 665 |
+
"learning_rate": 0.0001775312057281466,
|
| 666 |
+
"loss": 1.3442,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.197802197802198,
|
| 671 |
+
"grad_norm": 1.3937824964523315,
|
| 672 |
+
"learning_rate": 0.00017706986292968582,
|
| 673 |
+
"loss": 1.3082,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.220936957779063,
|
| 678 |
+
"grad_norm": 1.4411293268203735,
|
| 679 |
+
"learning_rate": 0.0001766044443118978,
|
| 680 |
+
"loss": 1.313,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.2440717177559284,
|
| 685 |
+
"grad_norm": 1.4743627309799194,
|
| 686 |
+
"learning_rate": 0.00017613497448832312,
|
| 687 |
+
"loss": 1.3783,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.2672064777327936,
|
| 692 |
+
"grad_norm": 1.3222951889038086,
|
| 693 |
+
"learning_rate": 0.00017566147828674931,
|
| 694 |
+
"loss": 1.3253,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.290341237709659,
|
| 699 |
+
"grad_norm": 1.3810490369796753,
|
| 700 |
+
"learning_rate": 0.00017518398074789775,
|
| 701 |
+
"loss": 1.3191,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.313475997686524,
|
| 706 |
+
"grad_norm": 1.3942046165466309,
|
| 707 |
+
"learning_rate": 0.0001747025071240996,
|
| 708 |
+
"loss": 1.4389,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.336610757663389,
|
| 713 |
+
"grad_norm": 1.523859977722168,
|
| 714 |
+
"learning_rate": 0.00017421708287796017,
|
| 715 |
+
"loss": 1.4015,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.3597455176402544,
|
| 720 |
+
"grad_norm": 1.435693383216858,
|
| 721 |
+
"learning_rate": 0.0001737277336810124,
|
| 722 |
+
"loss": 1.3434,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.3828802776171196,
|
| 727 |
+
"grad_norm": 1.4890649318695068,
|
| 728 |
+
"learning_rate": 0.00017323448541235924,
|
| 729 |
+
"loss": 1.3337,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.406015037593985,
|
| 734 |
+
"grad_norm": 1.3789806365966797,
|
| 735 |
+
"learning_rate": 0.00017273736415730488,
|
| 736 |
+
"loss": 1.3136,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.42914979757085,
|
| 741 |
+
"grad_norm": 1.4059851169586182,
|
| 742 |
+
"learning_rate": 0.00017223639620597556,
|
| 743 |
+
"loss": 1.4041,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.4522845575477152,
|
| 748 |
+
"grad_norm": 1.4800046682357788,
|
| 749 |
+
"learning_rate": 0.00017173160805192893,
|
| 750 |
+
"loss": 1.3156,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.4754193175245804,
|
| 755 |
+
"grad_norm": 1.556971549987793,
|
| 756 |
+
"learning_rate": 0.0001712230263907531,
|
| 757 |
+
"loss": 1.394,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.498554077501446,
|
| 762 |
+
"grad_norm": 1.4921575784683228,
|
| 763 |
+
"learning_rate": 0.00017071067811865476,
|
| 764 |
+
"loss": 1.3528,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.5216888374783113,
|
| 769 |
+
"grad_norm": 1.53052818775177,
|
| 770 |
+
"learning_rate": 0.00017019459033103682,
|
| 771 |
+
"loss": 1.3099,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.5448235974551765,
|
| 776 |
+
"grad_norm": 1.4016938209533691,
|
| 777 |
+
"learning_rate": 0.0001696747903210655,
|
| 778 |
+
"loss": 1.4067,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.5679583574320417,
|
| 783 |
+
"grad_norm": 1.5849937200546265,
|
| 784 |
+
"learning_rate": 0.00016915130557822695,
|
| 785 |
+
"loss": 1.3557,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.591093117408907,
|
| 790 |
+
"grad_norm": 1.5975929498672485,
|
| 791 |
+
"learning_rate": 0.0001686241637868734,
|
| 792 |
+
"loss": 1.304,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.614227877385772,
|
| 797 |
+
"grad_norm": 1.3660000562667847,
|
| 798 |
+
"learning_rate": 0.00016809339282475904,
|
| 799 |
+
"loss": 1.3462,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.6373626373626373,
|
| 804 |
+
"grad_norm": 1.4659340381622314,
|
| 805 |
+
"learning_rate": 0.00016755902076156604,
|
| 806 |
+
"loss": 1.3905,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.6604973973395025,
|
| 811 |
+
"grad_norm": 1.398289680480957,
|
| 812 |
+
"learning_rate": 0.0001670210758574196,
|
| 813 |
+
"loss": 1.3498,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.6836321573163677,
|
| 818 |
+
"grad_norm": 1.4304271936416626,
|
| 819 |
+
"learning_rate": 0.00016647958656139378,
|
| 820 |
+
"loss": 1.3641,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.706766917293233,
|
| 825 |
+
"grad_norm": 1.455327033996582,
|
| 826 |
+
"learning_rate": 0.00016593458151000688,
|
| 827 |
+
"loss": 1.3586,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.7299016772700986,
|
| 832 |
+
"grad_norm": 1.4968023300170898,
|
| 833 |
+
"learning_rate": 0.00016538608952570698,
|
| 834 |
+
"loss": 1.3731,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.753036437246964,
|
| 839 |
+
"grad_norm": 1.5257724523544312,
|
| 840 |
+
"learning_rate": 0.00016483413961534762,
|
| 841 |
+
"loss": 1.3484,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.776171197223829,
|
| 846 |
+
"grad_norm": 1.5006804466247559,
|
| 847 |
+
"learning_rate": 0.00016427876096865394,
|
| 848 |
+
"loss": 1.3702,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.799305957200694,
|
| 853 |
+
"grad_norm": 1.376607060432434,
|
| 854 |
+
"learning_rate": 0.00016371998295667886,
|
| 855 |
+
"loss": 1.4002,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.8224407171775594,
|
| 860 |
+
"grad_norm": 1.5122621059417725,
|
| 861 |
+
"learning_rate": 0.00016315783513024977,
|
| 862 |
+
"loss": 1.368,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.8455754771544246,
|
| 867 |
+
"grad_norm": 1.5241082906723022,
|
| 868 |
+
"learning_rate": 0.00016259234721840591,
|
| 869 |
+
"loss": 1.3907,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.86871023713129,
|
| 874 |
+
"grad_norm": 1.5061618089675903,
|
| 875 |
+
"learning_rate": 0.000162023549126826,
|
| 876 |
+
"loss": 1.3166,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.891844997108155,
|
| 881 |
+
"grad_norm": 1.3175408840179443,
|
| 882 |
+
"learning_rate": 0.00016145147093624677,
|
| 883 |
+
"loss": 1.3919,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.91497975708502,
|
| 888 |
+
"grad_norm": 1.614513635635376,
|
| 889 |
+
"learning_rate": 0.00016087614290087208,
|
| 890 |
+
"loss": 1.3702,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.9381145170618854,
|
| 895 |
+
"grad_norm": 1.460629940032959,
|
| 896 |
+
"learning_rate": 0.00016029759544677297,
|
| 897 |
+
"loss": 1.3504,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.9612492770387506,
|
| 902 |
+
"grad_norm": 1.5735194683074951,
|
| 903 |
+
"learning_rate": 0.00015971585917027862,
|
| 904 |
+
"loss": 1.3726,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.984384037015616,
|
| 909 |
+
"grad_norm": 1.3974887132644653,
|
| 910 |
+
"learning_rate": 0.00015913096483635824,
|
| 911 |
+
"loss": 1.3811,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.007518796992481,
|
| 916 |
+
"grad_norm": 1.3209352493286133,
|
| 917 |
+
"learning_rate": 0.00015854294337699408,
|
| 918 |
+
"loss": 1.2467,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.0306535569693462,
|
| 923 |
+
"grad_norm": 1.64750337600708,
|
| 924 |
+
"learning_rate": 0.0001579518258895455,
|
| 925 |
+
"loss": 1.0457,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.053788316946212,
|
| 930 |
+
"grad_norm": 1.7243373394012451,
|
| 931 |
+
"learning_rate": 0.0001573576436351046,
|
| 932 |
+
"loss": 1.0609,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.076923076923077,
|
| 937 |
+
"grad_norm": 1.5499110221862793,
|
| 938 |
+
"learning_rate": 0.00015676042803684288,
|
| 939 |
+
"loss": 1.0522,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.1000578368999423,
|
| 944 |
+
"grad_norm": 1.7944879531860352,
|
| 945 |
+
"learning_rate": 0.0001561602106783493,
|
| 946 |
+
"loss": 1.0416,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.1231925968768075,
|
| 951 |
+
"grad_norm": 1.7069816589355469,
|
| 952 |
+
"learning_rate": 0.00015555702330196023,
|
| 953 |
+
"loss": 0.9983,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.1463273568536727,
|
| 958 |
+
"grad_norm": 1.7057902812957764,
|
| 959 |
+
"learning_rate": 0.0001549508978070806,
|
| 960 |
+
"loss": 1.0115,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.169462116830538,
|
| 965 |
+
"grad_norm": 1.6382006406784058,
|
| 966 |
+
"learning_rate": 0.000154341866248497,
|
| 967 |
+
"loss": 1.0349,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.192596876807403,
|
| 972 |
+
"grad_norm": 1.6651489734649658,
|
| 973 |
+
"learning_rate": 0.0001537299608346824,
|
| 974 |
+
"loss": 1.092,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.2157316367842683,
|
| 979 |
+
"grad_norm": 1.7190675735473633,
|
| 980 |
+
"learning_rate": 0.00015311521392609282,
|
| 981 |
+
"loss": 1.0611,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.2388663967611335,
|
| 986 |
+
"grad_norm": 1.6004880666732788,
|
| 987 |
+
"learning_rate": 0.000152497658033456,
|
| 988 |
+
"loss": 1.0455,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.2620011567379987,
|
| 993 |
+
"grad_norm": 1.870151162147522,
|
| 994 |
+
"learning_rate": 0.00015187732581605217,
|
| 995 |
+
"loss": 1.0883,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.285135916714864,
|
| 1000 |
+
"grad_norm": 1.7651612758636475,
|
| 1001 |
+
"learning_rate": 0.00015125425007998653,
|
| 1002 |
+
"loss": 1.0853,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.308270676691729,
|
| 1007 |
+
"grad_norm": 1.578336477279663,
|
| 1008 |
+
"learning_rate": 0.00015062846377645475,
|
| 1009 |
+
"loss": 1.0612,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.331405436668595,
|
| 1014 |
+
"grad_norm": 1.8268855810165405,
|
| 1015 |
+
"learning_rate": 0.00015000000000000001,
|
| 1016 |
+
"loss": 1.112,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.35454019664546,
|
| 1021 |
+
"grad_norm": 1.663640022277832,
|
| 1022 |
+
"learning_rate": 0.00014936889198676303,
|
| 1023 |
+
"loss": 1.0387,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.377674956622325,
|
| 1028 |
+
"grad_norm": 1.7626187801361084,
|
| 1029 |
+
"learning_rate": 0.00014873517311272423,
|
| 1030 |
+
"loss": 1.0766,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.4008097165991904,
|
| 1035 |
+
"grad_norm": 1.683388113975525,
|
| 1036 |
+
"learning_rate": 0.00014809887689193877,
|
| 1037 |
+
"loss": 1.0982,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.4239444765760556,
|
| 1042 |
+
"grad_norm": 1.6581867933273315,
|
| 1043 |
+
"learning_rate": 0.00014746003697476404,
|
| 1044 |
+
"loss": 1.1141,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.447079236552921,
|
| 1049 |
+
"grad_norm": 1.9846975803375244,
|
| 1050 |
+
"learning_rate": 0.0001468186871460802,
|
| 1051 |
+
"loss": 1.0853,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.470213996529786,
|
| 1056 |
+
"grad_norm": 1.7938522100448608,
|
| 1057 |
+
"learning_rate": 0.00014617486132350343,
|
| 1058 |
+
"loss": 1.1085,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.493348756506651,
|
| 1063 |
+
"grad_norm": 1.7541499137878418,
|
| 1064 |
+
"learning_rate": 0.00014552859355559204,
|
| 1065 |
+
"loss": 1.0778,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.5164835164835164,
|
| 1070 |
+
"grad_norm": 1.7515652179718018,
|
| 1071 |
+
"learning_rate": 0.00014487991802004623,
|
| 1072 |
+
"loss": 1.1077,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.5396182764603816,
|
| 1077 |
+
"grad_norm": 1.7582262754440308,
|
| 1078 |
+
"learning_rate": 0.00014422886902190014,
|
| 1079 |
+
"loss": 1.0703,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.562753036437247,
|
| 1084 |
+
"grad_norm": 1.6952829360961914,
|
| 1085 |
+
"learning_rate": 0.00014357548099170795,
|
| 1086 |
+
"loss": 1.0588,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.585887796414112,
|
| 1091 |
+
"grad_norm": 1.8166347742080688,
|
| 1092 |
+
"learning_rate": 0.00014291978848372293,
|
| 1093 |
+
"loss": 1.053,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.6090225563909772,
|
| 1098 |
+
"grad_norm": 1.8070217370986938,
|
| 1099 |
+
"learning_rate": 0.00014226182617406996,
|
| 1100 |
+
"loss": 1.1416,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.6321573163678424,
|
| 1105 |
+
"grad_norm": 1.9227118492126465,
|
| 1106 |
+
"learning_rate": 0.00014160162885891193,
|
| 1107 |
+
"loss": 1.0947,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.6552920763447077,
|
| 1112 |
+
"grad_norm": 1.8285157680511475,
|
| 1113 |
+
"learning_rate": 0.00014093923145260925,
|
| 1114 |
+
"loss": 1.062,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.6784268363215733,
|
| 1119 |
+
"grad_norm": 1.8435360193252563,
|
| 1120 |
+
"learning_rate": 0.00014027466898587374,
|
| 1121 |
+
"loss": 1.0761,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.7015615962984385,
|
| 1126 |
+
"grad_norm": 1.8980731964111328,
|
| 1127 |
+
"learning_rate": 0.0001396079766039157,
|
| 1128 |
+
"loss": 1.0855,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.7246963562753037,
|
| 1133 |
+
"grad_norm": 1.8872120380401611,
|
| 1134 |
+
"learning_rate": 0.00013893918956458552,
|
| 1135 |
+
"loss": 1.0775,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.747831116252169,
|
| 1140 |
+
"grad_norm": 1.8531919717788696,
|
| 1141 |
+
"learning_rate": 0.000138268343236509,
|
| 1142 |
+
"loss": 1.0411,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.770965876229034,
|
| 1147 |
+
"grad_norm": 1.6847087144851685,
|
| 1148 |
+
"learning_rate": 0.0001375954730972168,
|
| 1149 |
+
"loss": 1.0747,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.7941006362058993,
|
| 1154 |
+
"grad_norm": 1.818551778793335,
|
| 1155 |
+
"learning_rate": 0.00013692061473126845,
|
| 1156 |
+
"loss": 1.0527,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.8172353961827645,
|
| 1161 |
+
"grad_norm": 1.7148271799087524,
|
| 1162 |
+
"learning_rate": 0.00013624380382837016,
|
| 1163 |
+
"loss": 1.0999,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.8403701561596297,
|
| 1168 |
+
"grad_norm": 1.6660877466201782,
|
| 1169 |
+
"learning_rate": 0.0001355650761814877,
|
| 1170 |
+
"loss": 1.0868,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.863504916136495,
|
| 1175 |
+
"grad_norm": 1.787760615348816,
|
| 1176 |
+
"learning_rate": 0.0001348844676849531,
|
| 1177 |
+
"loss": 1.1218,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.8866396761133606,
|
| 1182 |
+
"grad_norm": 1.7259087562561035,
|
| 1183 |
+
"learning_rate": 0.00013420201433256689,
|
| 1184 |
+
"loss": 1.085,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.909774436090226,
|
| 1189 |
+
"grad_norm": 1.666365146636963,
|
| 1190 |
+
"learning_rate": 0.00013351775221569415,
|
| 1191 |
+
"loss": 1.1676,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.932909196067091,
|
| 1196 |
+
"grad_norm": 1.7131948471069336,
|
| 1197 |
+
"learning_rate": 0.00013283171752135614,
|
| 1198 |
+
"loss": 1.0713,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.956043956043956,
|
| 1203 |
+
"grad_norm": 1.7973147630691528,
|
| 1204 |
+
"learning_rate": 0.00013214394653031616,
|
| 1205 |
+
"loss": 1.0864,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.9791787160208214,
|
| 1210 |
+
"grad_norm": 1.7572144269943237,
|
| 1211 |
+
"learning_rate": 0.00013145447561516138,
|
| 1212 |
+
"loss": 1.0458,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.002313475997687,
|
| 1217 |
+
"grad_norm": 1.5910913944244385,
|
| 1218 |
+
"learning_rate": 0.00013076334123837883,
|
| 1219 |
+
"loss": 1.0368,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.025448235974552,
|
| 1224 |
+
"grad_norm": 1.9898087978363037,
|
| 1225 |
+
"learning_rate": 0.00013007057995042732,
|
| 1226 |
+
"loss": 0.7595,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.048582995951417,
|
| 1231 |
+
"grad_norm": 1.8028409481048584,
|
| 1232 |
+
"learning_rate": 0.00012937622838780444,
|
| 1233 |
+
"loss": 0.7896,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.071717755928282,
|
| 1238 |
+
"grad_norm": 2.1292295455932617,
|
| 1239 |
+
"learning_rate": 0.00012868032327110904,
|
| 1240 |
+
"loss": 0.748,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 4.094852515905147,
|
| 1245 |
+
"grad_norm": 2.0062098503112793,
|
| 1246 |
+
"learning_rate": 0.00012798290140309923,
|
| 1247 |
+
"loss": 0.7741,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.117987275882013,
|
| 1252 |
+
"grad_norm": 2.028233528137207,
|
| 1253 |
+
"learning_rate": 0.00012728399966674612,
|
| 1254 |
+
"loss": 0.8249,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.141122035858878,
|
| 1259 |
+
"grad_norm": 2.332066297531128,
|
| 1260 |
+
"learning_rate": 0.0001265836550232833,
|
| 1261 |
+
"loss": 0.8377,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.164256795835743,
|
| 1266 |
+
"grad_norm": 2.0353000164031982,
|
| 1267 |
+
"learning_rate": 0.00012588190451025207,
|
| 1268 |
+
"loss": 0.7875,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.187391555812608,
|
| 1273 |
+
"grad_norm": 2.017143964767456,
|
| 1274 |
+
"learning_rate": 0.00012517878523954286,
|
| 1275 |
+
"loss": 0.7837,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.2105263157894735,
|
| 1280 |
+
"grad_norm": 2.2068400382995605,
|
| 1281 |
+
"learning_rate": 0.0001244743343954324,
|
| 1282 |
+
"loss": 0.844,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 4.233661075766339,
|
| 1287 |
+
"grad_norm": 2.1417810916900635,
|
| 1288 |
+
"learning_rate": 0.00012376858923261733,
|
| 1289 |
+
"loss": 0.7947,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.256795835743204,
|
| 1294 |
+
"grad_norm": 2.1283600330352783,
|
| 1295 |
+
"learning_rate": 0.00012306158707424403,
|
| 1296 |
+
"loss": 0.846,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.279930595720069,
|
| 1301 |
+
"grad_norm": 1.9856690168380737,
|
| 1302 |
+
"learning_rate": 0.00012235336530993474,
|
| 1303 |
+
"loss": 0.8118,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.303065355696934,
|
| 1308 |
+
"grad_norm": 1.9329304695129395,
|
| 1309 |
+
"learning_rate": 0.00012164396139381029,
|
| 1310 |
+
"loss": 0.78,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.3262001156737995,
|
| 1315 |
+
"grad_norm": 1.9364598989486694,
|
| 1316 |
+
"learning_rate": 0.0001209334128425092,
|
| 1317 |
+
"loss": 0.8625,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 4.349334875650666,
|
| 1322 |
+
"grad_norm": 1.803899884223938,
|
| 1323 |
+
"learning_rate": 0.00012022175723320381,
|
| 1324 |
+
"loss": 0.7904,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 4.372469635627531,
|
| 1329 |
+
"grad_norm": 1.9305102825164795,
|
| 1330 |
+
"learning_rate": 0.00011950903220161285,
|
| 1331 |
+
"loss": 0.8376,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 4.395604395604396,
|
| 1336 |
+
"grad_norm": 2.184109687805176,
|
| 1337 |
+
"learning_rate": 0.00011879527544001119,
|
| 1338 |
+
"loss": 0.755,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 4.418739155581261,
|
| 1343 |
+
"grad_norm": 1.9741244316101074,
|
| 1344 |
+
"learning_rate": 0.00011808052469523654,
|
| 1345 |
+
"loss": 0.8244,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.441873915558126,
|
| 1350 |
+
"grad_norm": 1.954925537109375,
|
| 1351 |
+
"learning_rate": 0.00011736481776669306,
|
| 1352 |
+
"loss": 0.8061,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 4.465008675534992,
|
| 1357 |
+
"grad_norm": 1.891669750213623,
|
| 1358 |
+
"learning_rate": 0.00011664819250435246,
|
| 1359 |
+
"loss": 0.8089,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.488143435511857,
|
| 1364 |
+
"grad_norm": 2.0183587074279785,
|
| 1365 |
+
"learning_rate": 0.00011593068680675228,
|
| 1366 |
+
"loss": 0.8304,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 4.511278195488722,
|
| 1371 |
+
"grad_norm": 1.9433139562606812,
|
| 1372 |
+
"learning_rate": 0.00011521233861899167,
|
| 1373 |
+
"loss": 0.8356,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.534412955465587,
|
| 1378 |
+
"grad_norm": 2.0215744972229004,
|
| 1379 |
+
"learning_rate": 0.00011449318593072466,
|
| 1380 |
+
"loss": 0.8106,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 4.557547715442452,
|
| 1385 |
+
"grad_norm": 2.0268826484680176,
|
| 1386 |
+
"learning_rate": 0.00011377326677415108,
|
| 1387 |
+
"loss": 0.8186,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.580682475419318,
|
| 1392 |
+
"grad_norm": 2.270048141479492,
|
| 1393 |
+
"learning_rate": 0.00011305261922200519,
|
| 1394 |
+
"loss": 0.7984,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 4.603817235396183,
|
| 1399 |
+
"grad_norm": 1.9835270643234253,
|
| 1400 |
+
"learning_rate": 0.0001123312813855422,
|
| 1401 |
+
"loss": 0.8106,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 4.626951995373048,
|
| 1406 |
+
"grad_norm": 2.036339044570923,
|
| 1407 |
+
"learning_rate": 0.00011160929141252303,
|
| 1408 |
+
"loss": 0.8778,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 4.650086755349913,
|
| 1413 |
+
"grad_norm": 2.1386802196502686,
|
| 1414 |
+
"learning_rate": 0.00011088668748519647,
|
| 1415 |
+
"loss": 0.8408,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 4.673221515326778,
|
| 1420 |
+
"grad_norm": 2.199148416519165,
|
| 1421 |
+
"learning_rate": 0.00011016350781828019,
|
| 1422 |
+
"loss": 0.8356,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 4.696356275303644,
|
| 1427 |
+
"grad_norm": 1.943888783454895,
|
| 1428 |
+
"learning_rate": 0.00010943979065693975,
|
| 1429 |
+
"loss": 0.8521,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.719491035280509,
|
| 1434 |
+
"grad_norm": 2.2225332260131836,
|
| 1435 |
+
"learning_rate": 0.00010871557427476583,
|
| 1436 |
+
"loss": 0.8487,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 4.742625795257374,
|
| 1441 |
+
"grad_norm": 2.0183157920837402,
|
| 1442 |
+
"learning_rate": 0.0001079908969717504,
|
| 1443 |
+
"loss": 0.8258,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 4.765760555234239,
|
| 1448 |
+
"grad_norm": 2.1665942668914795,
|
| 1449 |
+
"learning_rate": 0.00010726579707226108,
|
| 1450 |
+
"loss": 0.8786,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.7888953152111045,
|
| 1455 |
+
"grad_norm": 2.0331172943115234,
|
| 1456 |
+
"learning_rate": 0.00010654031292301432,
|
| 1457 |
+
"loss": 0.8344,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 4.81203007518797,
|
| 1462 |
+
"grad_norm": 2.299880266189575,
|
| 1463 |
+
"learning_rate": 0.00010581448289104758,
|
| 1464 |
+
"loss": 0.8267,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 4.835164835164835,
|
| 1469 |
+
"grad_norm": 2.0382227897644043,
|
| 1470 |
+
"learning_rate": 0.00010508834536169028,
|
| 1471 |
+
"loss": 0.8533,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 4.8582995951417,
|
| 1476 |
+
"grad_norm": 1.9726015329360962,
|
| 1477 |
+
"learning_rate": 0.00010436193873653361,
|
| 1478 |
+
"loss": 0.8654,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 4.881434355118565,
|
| 1483 |
+
"grad_norm": 2.1659207344055176,
|
| 1484 |
+
"learning_rate": 0.00010363530143139999,
|
| 1485 |
+
"loss": 0.8723,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 4.9045691150954305,
|
| 1490 |
+
"grad_norm": 2.0480141639709473,
|
| 1491 |
+
"learning_rate": 0.00010290847187431113,
|
| 1492 |
+
"loss": 0.8272,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 4.927703875072297,
|
| 1497 |
+
"grad_norm": 2.1771929264068604,
|
| 1498 |
+
"learning_rate": 0.00010218148850345613,
|
| 1499 |
+
"loss": 0.8352,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 4.950838635049161,
|
| 1504 |
+
"grad_norm": 2.022383689880371,
|
| 1505 |
+
"learning_rate": 0.00010145438976515828,
|
| 1506 |
+
"loss": 0.9057,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 4.973973395026027,
|
| 1511 |
+
"grad_norm": 2.1294972896575928,
|
| 1512 |
+
"learning_rate": 0.00010072721411184219,
|
| 1513 |
+
"loss": 0.8775,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 4.997108155002892,
|
| 1518 |
+
"grad_norm": 2.14635968208313,
|
| 1519 |
+
"learning_rate": 0.0001,
|
| 1520 |
+
"loss": 0.8175,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.020242914979757,
|
| 1525 |
+
"grad_norm": 2.4558303356170654,
|
| 1526 |
+
"learning_rate": 9.927278588815786e-05,
|
| 1527 |
+
"loss": 0.6267,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.043377674956623,
|
| 1532 |
+
"grad_norm": 2.078815221786499,
|
| 1533 |
+
"learning_rate": 9.854561023484173e-05,
|
| 1534 |
+
"loss": 0.6033,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 5.066512434933488,
|
| 1539 |
+
"grad_norm": 2.428105354309082,
|
| 1540 |
+
"learning_rate": 9.78185114965439e-05,
|
| 1541 |
+
"loss": 0.6157,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 5.089647194910353,
|
| 1546 |
+
"grad_norm": 2.1209535598754883,
|
| 1547 |
+
"learning_rate": 9.709152812568886e-05,
|
| 1548 |
+
"loss": 0.5899,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 5.112781954887218,
|
| 1553 |
+
"grad_norm": 2.1761510372161865,
|
| 1554 |
+
"learning_rate": 9.636469856860005e-05,
|
| 1555 |
+
"loss": 0.5986,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 5.135916714864083,
|
| 1560 |
+
"grad_norm": 2.0283255577087402,
|
| 1561 |
+
"learning_rate": 9.563806126346642e-05,
|
| 1562 |
+
"loss": 0.5719,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 5.159051474840949,
|
| 1567 |
+
"grad_norm": 2.2018613815307617,
|
| 1568 |
+
"learning_rate": 9.491165463830975e-05,
|
| 1569 |
+
"loss": 0.5884,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 5.182186234817814,
|
| 1574 |
+
"grad_norm": 2.153411626815796,
|
| 1575 |
+
"learning_rate": 9.418551710895243e-05,
|
| 1576 |
+
"loss": 0.5664,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.205320994794679,
|
| 1581 |
+
"grad_norm": 2.133164405822754,
|
| 1582 |
+
"learning_rate": 9.345968707698569e-05,
|
| 1583 |
+
"loss": 0.5873,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 5.228455754771544,
|
| 1588 |
+
"grad_norm": 2.3184361457824707,
|
| 1589 |
+
"learning_rate": 9.273420292773894e-05,
|
| 1590 |
+
"loss": 0.6031,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 5.251590514748409,
|
| 1595 |
+
"grad_norm": 1.985695481300354,
|
| 1596 |
+
"learning_rate": 9.200910302824963e-05,
|
| 1597 |
+
"loss": 0.603,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 5.274725274725275,
|
| 1602 |
+
"grad_norm": 2.204653024673462,
|
| 1603 |
+
"learning_rate": 9.128442572523417e-05,
|
| 1604 |
+
"loss": 0.5943,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 5.29786003470214,
|
| 1609 |
+
"grad_norm": 2.212740659713745,
|
| 1610 |
+
"learning_rate": 9.056020934306032e-05,
|
| 1611 |
+
"loss": 0.6301,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 5.320994794679005,
|
| 1616 |
+
"grad_norm": 2.320549964904785,
|
| 1617 |
+
"learning_rate": 8.983649218171982e-05,
|
| 1618 |
+
"loss": 0.6231,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 5.34412955465587,
|
| 1623 |
+
"grad_norm": 2.21589732170105,
|
| 1624 |
+
"learning_rate": 8.911331251480357e-05,
|
| 1625 |
+
"loss": 0.5959,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 5.3672643146327355,
|
| 1630 |
+
"grad_norm": 2.234104633331299,
|
| 1631 |
+
"learning_rate": 8.839070858747697e-05,
|
| 1632 |
+
"loss": 0.6234,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 5.390399074609601,
|
| 1637 |
+
"grad_norm": 2.3640241622924805,
|
| 1638 |
+
"learning_rate": 8.76687186144578e-05,
|
| 1639 |
+
"loss": 0.6501,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 5.413533834586466,
|
| 1644 |
+
"grad_norm": 2.0853078365325928,
|
| 1645 |
+
"learning_rate": 8.694738077799488e-05,
|
| 1646 |
+
"loss": 0.5925,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 5.436668594563331,
|
| 1651 |
+
"grad_norm": 2.4165499210357666,
|
| 1652 |
+
"learning_rate": 8.622673322584893e-05,
|
| 1653 |
+
"loss": 0.6399,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 5.459803354540196,
|
| 1658 |
+
"grad_norm": 2.2116127014160156,
|
| 1659 |
+
"learning_rate": 8.550681406927535e-05,
|
| 1660 |
+
"loss": 0.6438,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 5.4829381145170615,
|
| 1665 |
+
"grad_norm": 2.056300163269043,
|
| 1666 |
+
"learning_rate": 8.478766138100834e-05,
|
| 1667 |
+
"loss": 0.5918,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 5.506072874493928,
|
| 1672 |
+
"grad_norm": 2.453059196472168,
|
| 1673 |
+
"learning_rate": 8.406931319324776e-05,
|
| 1674 |
+
"loss": 0.6414,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 5.529207634470792,
|
| 1679 |
+
"grad_norm": 2.425656318664551,
|
| 1680 |
+
"learning_rate": 8.33518074956476e-05,
|
| 1681 |
+
"loss": 0.6227,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 5.552342394447658,
|
| 1686 |
+
"grad_norm": 2.2804975509643555,
|
| 1687 |
+
"learning_rate": 8.263518223330697e-05,
|
| 1688 |
+
"loss": 0.6061,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 5.575477154424523,
|
| 1693 |
+
"grad_norm": 2.6980812549591064,
|
| 1694 |
+
"learning_rate": 8.19194753047635e-05,
|
| 1695 |
+
"loss": 0.5937,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 5.598611914401388,
|
| 1700 |
+
"grad_norm": 2.084348201751709,
|
| 1701 |
+
"learning_rate": 8.120472455998882e-05,
|
| 1702 |
+
"loss": 0.6146,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 5.621746674378254,
|
| 1707 |
+
"grad_norm": 2.330526113510132,
|
| 1708 |
+
"learning_rate": 8.049096779838719e-05,
|
| 1709 |
+
"loss": 0.6077,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 5.644881434355119,
|
| 1714 |
+
"grad_norm": 2.280693769454956,
|
| 1715 |
+
"learning_rate": 7.977824276679623e-05,
|
| 1716 |
+
"loss": 0.6267,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 5.668016194331984,
|
| 1721 |
+
"grad_norm": 2.3855202198028564,
|
| 1722 |
+
"learning_rate": 7.90665871574908e-05,
|
| 1723 |
+
"loss": 0.6045,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 5.691150954308849,
|
| 1728 |
+
"grad_norm": 2.3103318214416504,
|
| 1729 |
+
"learning_rate": 7.835603860618972e-05,
|
| 1730 |
+
"loss": 0.6166,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 5.714285714285714,
|
| 1735 |
+
"grad_norm": 2.136725425720215,
|
| 1736 |
+
"learning_rate": 7.764663469006526e-05,
|
| 1737 |
+
"loss": 0.6116,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 5.73742047426258,
|
| 1742 |
+
"grad_norm": 2.1863021850585938,
|
| 1743 |
+
"learning_rate": 7.693841292575598e-05,
|
| 1744 |
+
"loss": 0.6302,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 5.760555234239445,
|
| 1749 |
+
"grad_norm": 2.1635851860046387,
|
| 1750 |
+
"learning_rate": 7.623141076738271e-05,
|
| 1751 |
+
"loss": 0.6232,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 5.78368999421631,
|
| 1756 |
+
"grad_norm": 2.2908101081848145,
|
| 1757 |
+
"learning_rate": 7.552566560456762e-05,
|
| 1758 |
+
"loss": 0.6573,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 5.806824754193175,
|
| 1763 |
+
"grad_norm": 2.1210217475891113,
|
| 1764 |
+
"learning_rate": 7.482121476045716e-05,
|
| 1765 |
+
"loss": 0.6391,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 5.82995951417004,
|
| 1770 |
+
"grad_norm": 2.4276537895202637,
|
| 1771 |
+
"learning_rate": 7.411809548974792e-05,
|
| 1772 |
+
"loss": 0.6313,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 5.853094274146906,
|
| 1777 |
+
"grad_norm": 2.296722412109375,
|
| 1778 |
+
"learning_rate": 7.341634497671672e-05,
|
| 1779 |
+
"loss": 0.6218,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 5.876229034123771,
|
| 1784 |
+
"grad_norm": 2.1190712451934814,
|
| 1785 |
+
"learning_rate": 7.271600033325393e-05,
|
| 1786 |
+
"loss": 0.6521,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 5.899363794100636,
|
| 1791 |
+
"grad_norm": 2.1762571334838867,
|
| 1792 |
+
"learning_rate": 7.20170985969008e-05,
|
| 1793 |
+
"loss": 0.6107,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 5.922498554077501,
|
| 1798 |
+
"grad_norm": 2.5130181312561035,
|
| 1799 |
+
"learning_rate": 7.131967672889101e-05,
|
| 1800 |
+
"loss": 0.6325,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 5.9456333140543665,
|
| 1805 |
+
"grad_norm": 2.1171939373016357,
|
| 1806 |
+
"learning_rate": 7.062377161219556e-05,
|
| 1807 |
+
"loss": 0.6242,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 5.968768074031232,
|
| 1812 |
+
"grad_norm": 2.013845682144165,
|
| 1813 |
+
"learning_rate": 6.992942004957271e-05,
|
| 1814 |
+
"loss": 0.6266,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 5.991902834008097,
|
| 1819 |
+
"grad_norm": 2.2959377765655518,
|
| 1820 |
+
"learning_rate": 6.923665876162118e-05,
|
| 1821 |
+
"loss": 0.6191,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 6.015037593984962,
|
| 1826 |
+
"grad_norm": 2.0503013134002686,
|
| 1827 |
+
"learning_rate": 6.854552438483865e-05,
|
| 1828 |
+
"loss": 0.5465,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 6.038172353961827,
|
| 1833 |
+
"grad_norm": 2.3042502403259277,
|
| 1834 |
+
"learning_rate": 6.785605346968386e-05,
|
| 1835 |
+
"loss": 0.4176,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.0613071139386925,
|
| 1840 |
+
"grad_norm": 2.150449275970459,
|
| 1841 |
+
"learning_rate": 6.71682824786439e-05,
|
| 1842 |
+
"loss": 0.4201,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 6.084441873915559,
|
| 1847 |
+
"grad_norm": 2.2146551609039307,
|
| 1848 |
+
"learning_rate": 6.648224778430586e-05,
|
| 1849 |
+
"loss": 0.4305,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.107576633892424,
|
| 1854 |
+
"grad_norm": 2.163520097732544,
|
| 1855 |
+
"learning_rate": 6.579798566743314e-05,
|
| 1856 |
+
"loss": 0.4445,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 6.130711393869289,
|
| 1861 |
+
"grad_norm": 2.339061975479126,
|
| 1862 |
+
"learning_rate": 6.511553231504693e-05,
|
| 1863 |
+
"loss": 0.4453,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 6.153846153846154,
|
| 1868 |
+
"grad_norm": 2.3959970474243164,
|
| 1869 |
+
"learning_rate": 6.443492381851237e-05,
|
| 1870 |
+
"loss": 0.4061,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 6.176980913823019,
|
| 1875 |
+
"grad_norm": 2.2613327503204346,
|
| 1876 |
+
"learning_rate": 6.375619617162985e-05,
|
| 1877 |
+
"loss": 0.4074,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 6.200115673799885,
|
| 1882 |
+
"grad_norm": 2.2177650928497314,
|
| 1883 |
+
"learning_rate": 6.307938526873157e-05,
|
| 1884 |
+
"loss": 0.441,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 6.22325043377675,
|
| 1889 |
+
"grad_norm": 2.184535503387451,
|
| 1890 |
+
"learning_rate": 6.24045269027832e-05,
|
| 1891 |
+
"loss": 0.4518,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.246385193753615,
|
| 1896 |
+
"grad_norm": 2.2392778396606445,
|
| 1897 |
+
"learning_rate": 6.173165676349103e-05,
|
| 1898 |
+
"loss": 0.4613,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 6.26951995373048,
|
| 1903 |
+
"grad_norm": 2.2998814582824707,
|
| 1904 |
+
"learning_rate": 6.106081043541451e-05,
|
| 1905 |
+
"loss": 0.4466,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 6.292654713707345,
|
| 1910 |
+
"grad_norm": 2.1889827251434326,
|
| 1911 |
+
"learning_rate": 6.039202339608432e-05,
|
| 1912 |
+
"loss": 0.4625,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 6.315789473684211,
|
| 1917 |
+
"grad_norm": 2.439364194869995,
|
| 1918 |
+
"learning_rate": 5.9725331014126294e-05,
|
| 1919 |
+
"loss": 0.4722,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 6.338924233661076,
|
| 1924 |
+
"grad_norm": 2.3686563968658447,
|
| 1925 |
+
"learning_rate": 5.906076854739074e-05,
|
| 1926 |
+
"loss": 0.4256,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 6.362058993637941,
|
| 1931 |
+
"grad_norm": 2.355642795562744,
|
| 1932 |
+
"learning_rate": 5.83983711410881e-05,
|
| 1933 |
+
"loss": 0.4405,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.385193753614806,
|
| 1938 |
+
"grad_norm": 2.5149965286254883,
|
| 1939 |
+
"learning_rate": 5.773817382593008e-05,
|
| 1940 |
+
"loss": 0.4775,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 6.408328513591671,
|
| 1945 |
+
"grad_norm": 2.1884281635284424,
|
| 1946 |
+
"learning_rate": 5.708021151627712e-05,
|
| 1947 |
+
"loss": 0.4219,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 6.431463273568537,
|
| 1952 |
+
"grad_norm": 2.5396625995635986,
|
| 1953 |
+
"learning_rate": 5.642451900829209e-05,
|
| 1954 |
+
"loss": 0.4617,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 6.454598033545402,
|
| 1959 |
+
"grad_norm": 2.1632354259490967,
|
| 1960 |
+
"learning_rate": 5.577113097809989e-05,
|
| 1961 |
+
"loss": 0.4407,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 6.477732793522267,
|
| 1966 |
+
"grad_norm": 2.3358309268951416,
|
| 1967 |
+
"learning_rate": 5.5120081979953785e-05,
|
| 1968 |
+
"loss": 0.4526,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 6.500867553499132,
|
| 1973 |
+
"grad_norm": 2.1623499393463135,
|
| 1974 |
+
"learning_rate": 5.447140644440798e-05,
|
| 1975 |
+
"loss": 0.4408,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 6.5240023134759975,
|
| 1980 |
+
"grad_norm": 2.002521514892578,
|
| 1981 |
+
"learning_rate": 5.382513867649663e-05,
|
| 1982 |
+
"loss": 0.4536,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 6.547137073452863,
|
| 1987 |
+
"grad_norm": 2.0050201416015625,
|
| 1988 |
+
"learning_rate": 5.3181312853919804e-05,
|
| 1989 |
+
"loss": 0.4468,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 6.570271833429728,
|
| 1994 |
+
"grad_norm": 1.9836812019348145,
|
| 1995 |
+
"learning_rate": 5.253996302523596e-05,
|
| 1996 |
+
"loss": 0.4524,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 6.593406593406593,
|
| 2001 |
+
"grad_norm": 2.3806679248809814,
|
| 2002 |
+
"learning_rate": 5.190112310806126e-05,
|
| 2003 |
+
"loss": 0.4598,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 6.616541353383458,
|
| 2008 |
+
"grad_norm": 2.2022571563720703,
|
| 2009 |
+
"learning_rate": 5.1264826887275776e-05,
|
| 2010 |
+
"loss": 0.4798,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 6.6396761133603235,
|
| 2015 |
+
"grad_norm": 2.4624128341674805,
|
| 2016 |
+
"learning_rate": 5.063110801323697e-05,
|
| 2017 |
+
"loss": 0.4561,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 6.66281087333719,
|
| 2022 |
+
"grad_norm": 2.4618606567382812,
|
| 2023 |
+
"learning_rate": 5.000000000000002e-05,
|
| 2024 |
+
"loss": 0.4389,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 6.685945633314054,
|
| 2029 |
+
"grad_norm": 2.107267379760742,
|
| 2030 |
+
"learning_rate": 4.937153622354529e-05,
|
| 2031 |
+
"loss": 0.4696,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 6.70908039329092,
|
| 2036 |
+
"grad_norm": 2.5566253662109375,
|
| 2037 |
+
"learning_rate": 4.8745749920013486e-05,
|
| 2038 |
+
"loss": 0.4778,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 6.732215153267785,
|
| 2043 |
+
"grad_norm": 2.2925281524658203,
|
| 2044 |
+
"learning_rate": 4.8122674183947836e-05,
|
| 2045 |
+
"loss": 0.4536,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 6.75534991324465,
|
| 2050 |
+
"grad_norm": 2.449045419692993,
|
| 2051 |
+
"learning_rate": 4.7502341966544e-05,
|
| 2052 |
+
"loss": 0.4679,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 6.778484673221516,
|
| 2057 |
+
"grad_norm": 2.3320090770721436,
|
| 2058 |
+
"learning_rate": 4.688478607390723e-05,
|
| 2059 |
+
"loss": 0.4517,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 6.801619433198381,
|
| 2064 |
+
"grad_norm": 2.4661507606506348,
|
| 2065 |
+
"learning_rate": 4.6270039165317605e-05,
|
| 2066 |
+
"loss": 0.4699,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 6.824754193175246,
|
| 2071 |
+
"grad_norm": 2.398205280303955,
|
| 2072 |
+
"learning_rate": 4.565813375150302e-05,
|
| 2073 |
+
"loss": 0.4614,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 6.847888953152111,
|
| 2078 |
+
"grad_norm": 2.3216257095336914,
|
| 2079 |
+
"learning_rate": 4.50491021929194e-05,
|
| 2080 |
+
"loss": 0.465,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 6.871023713128976,
|
| 2085 |
+
"grad_norm": 2.2298338413238525,
|
| 2086 |
+
"learning_rate": 4.444297669803981e-05,
|
| 2087 |
+
"loss": 0.4679,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 6.894158473105842,
|
| 2092 |
+
"grad_norm": 2.1584606170654297,
|
| 2093 |
+
"learning_rate": 4.383978932165073e-05,
|
| 2094 |
+
"loss": 0.4532,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 6.917293233082707,
|
| 2099 |
+
"grad_norm": 2.500182628631592,
|
| 2100 |
+
"learning_rate": 4.3239571963157145e-05,
|
| 2101 |
+
"loss": 0.4784,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 6.940427993059572,
|
| 2106 |
+
"grad_norm": 2.2554585933685303,
|
| 2107 |
+
"learning_rate": 4.264235636489542e-05,
|
| 2108 |
+
"loss": 0.471,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 6.963562753036437,
|
| 2113 |
+
"grad_norm": 2.2849414348602295,
|
| 2114 |
+
"learning_rate": 4.2048174110454496e-05,
|
| 2115 |
+
"loss": 0.4595,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 6.986697513013302,
|
| 2120 |
+
"grad_norm": 2.273737668991089,
|
| 2121 |
+
"learning_rate": 4.145705662300595e-05,
|
| 2122 |
+
"loss": 0.4557,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.009832272990168,
|
| 2127 |
+
"grad_norm": 1.8555997610092163,
|
| 2128 |
+
"learning_rate": 4.086903516364179e-05,
|
| 2129 |
+
"loss": 0.4155,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.032967032967033,
|
| 2134 |
+
"grad_norm": 2.351282835006714,
|
| 2135 |
+
"learning_rate": 4.028414082972141e-05,
|
| 2136 |
+
"loss": 0.3444,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.056101792943898,
|
| 2141 |
+
"grad_norm": 1.9804246425628662,
|
| 2142 |
+
"learning_rate": 3.9702404553227046e-05,
|
| 2143 |
+
"loss": 0.3247,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.079236552920763,
|
| 2148 |
+
"grad_norm": 2.0093185901641846,
|
| 2149 |
+
"learning_rate": 3.9123857099127936e-05,
|
| 2150 |
+
"loss": 0.3455,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.1023713128976285,
|
| 2155 |
+
"grad_norm": 2.369689702987671,
|
| 2156 |
+
"learning_rate": 3.854852906375326e-05,
|
| 2157 |
+
"loss": 0.3474,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.125506072874494,
|
| 2162 |
+
"grad_norm": 2.10414981842041,
|
| 2163 |
+
"learning_rate": 3.7976450873174005e-05,
|
| 2164 |
+
"loss": 0.3179,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.148640832851359,
|
| 2169 |
+
"grad_norm": 2.3380303382873535,
|
| 2170 |
+
"learning_rate": 3.7407652781594095e-05,
|
| 2171 |
+
"loss": 0.3298,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.171775592828224,
|
| 2176 |
+
"grad_norm": 1.9055235385894775,
|
| 2177 |
+
"learning_rate": 3.684216486975026e-05,
|
| 2178 |
+
"loss": 0.3446,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.194910352805089,
|
| 2183 |
+
"grad_norm": 2.2698934078216553,
|
| 2184 |
+
"learning_rate": 3.628001704332118e-05,
|
| 2185 |
+
"loss": 0.3202,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 7.2180451127819545,
|
| 2190 |
+
"grad_norm": 2.1574339866638184,
|
| 2191 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 2192 |
+
"loss": 0.3299,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.241179872758821,
|
| 2197 |
+
"grad_norm": 2.068159818649292,
|
| 2198 |
+
"learning_rate": 3.516586038465237e-05,
|
| 2199 |
+
"loss": 0.3158,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 7.264314632735685,
|
| 2204 |
+
"grad_norm": 2.033571720123291,
|
| 2205 |
+
"learning_rate": 3.461391047429304e-05,
|
| 2206 |
+
"loss": 0.3511,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.287449392712551,
|
| 2211 |
+
"grad_norm": 2.3351962566375732,
|
| 2212 |
+
"learning_rate": 3.406541848999312e-05,
|
| 2213 |
+
"loss": 0.3495,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 7.310584152689416,
|
| 2218 |
+
"grad_norm": 1.9200048446655273,
|
| 2219 |
+
"learning_rate": 3.352041343860621e-05,
|
| 2220 |
+
"loss": 0.3329,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 7.333718912666281,
|
| 2225 |
+
"grad_norm": 2.1292202472686768,
|
| 2226 |
+
"learning_rate": 3.297892414258043e-05,
|
| 2227 |
+
"loss": 0.3233,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.356853672643147,
|
| 2232 |
+
"grad_norm": 2.009242057800293,
|
| 2233 |
+
"learning_rate": 3.244097923843398e-05,
|
| 2234 |
+
"loss": 0.3412,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.379988432620012,
|
| 2239 |
+
"grad_norm": 2.1473593711853027,
|
| 2240 |
+
"learning_rate": 3.190660717524094e-05,
|
| 2241 |
+
"loss": 0.3379,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 7.403123192596877,
|
| 2246 |
+
"grad_norm": 2.2157785892486572,
|
| 2247 |
+
"learning_rate": 3.137583621312665e-05,
|
| 2248 |
+
"loss": 0.3417,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.426257952573742,
|
| 2253 |
+
"grad_norm": 2.3066554069519043,
|
| 2254 |
+
"learning_rate": 3.0848694421773075e-05,
|
| 2255 |
+
"loss": 0.3319,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 7.449392712550607,
|
| 2260 |
+
"grad_norm": 2.0240354537963867,
|
| 2261 |
+
"learning_rate": 3.032520967893453e-05,
|
| 2262 |
+
"loss": 0.3343,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 7.472527472527473,
|
| 2267 |
+
"grad_norm": 2.417139768600464,
|
| 2268 |
+
"learning_rate": 2.9805409668963168e-05,
|
| 2269 |
+
"loss": 0.3236,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 7.495662232504338,
|
| 2274 |
+
"grad_norm": 2.220024347305298,
|
| 2275 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 2276 |
+
"loss": 0.3374,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 7.518796992481203,
|
| 2281 |
+
"grad_norm": 2.071851968765259,
|
| 2282 |
+
"learning_rate": 2.8776973609246928e-05,
|
| 2283 |
+
"loss": 0.3625,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 7.541931752458068,
|
| 2288 |
+
"grad_norm": 2.0256235599517822,
|
| 2289 |
+
"learning_rate": 2.826839194807105e-05,
|
| 2290 |
+
"loss": 0.346,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 7.565066512434933,
|
| 2295 |
+
"grad_norm": 2.0721919536590576,
|
| 2296 |
+
"learning_rate": 2.776360379402445e-05,
|
| 2297 |
+
"loss": 0.3444,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 7.588201272411799,
|
| 2302 |
+
"grad_norm": 2.2439305782318115,
|
| 2303 |
+
"learning_rate": 2.7262635842695127e-05,
|
| 2304 |
+
"loss": 0.3413,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 7.611336032388664,
|
| 2309 |
+
"grad_norm": 2.2420361042022705,
|
| 2310 |
+
"learning_rate": 2.6765514587640818e-05,
|
| 2311 |
+
"loss": 0.3413,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 7.634470792365529,
|
| 2316 |
+
"grad_norm": 2.33591628074646,
|
| 2317 |
+
"learning_rate": 2.6272266318987603e-05,
|
| 2318 |
+
"loss": 0.3309,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 7.657605552342394,
|
| 2323 |
+
"grad_norm": 2.183473825454712,
|
| 2324 |
+
"learning_rate": 2.578291712203983e-05,
|
| 2325 |
+
"loss": 0.3341,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 7.6807403123192595,
|
| 2330 |
+
"grad_norm": 2.333395481109619,
|
| 2331 |
+
"learning_rate": 2.529749287590042e-05,
|
| 2332 |
+
"loss": 0.3665,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 7.703875072296125,
|
| 2337 |
+
"grad_norm": 2.368133783340454,
|
| 2338 |
+
"learning_rate": 2.4816019252102273e-05,
|
| 2339 |
+
"loss": 0.3341,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 7.72700983227299,
|
| 2344 |
+
"grad_norm": 2.055673360824585,
|
| 2345 |
+
"learning_rate": 2.433852171325072e-05,
|
| 2346 |
+
"loss": 0.331,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 7.750144592249855,
|
| 2351 |
+
"grad_norm": 2.531468391418457,
|
| 2352 |
+
"learning_rate": 2.3865025511676897e-05,
|
| 2353 |
+
"loss": 0.3553,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 7.77327935222672,
|
| 2358 |
+
"grad_norm": 2.126619338989258,
|
| 2359 |
+
"learning_rate": 2.339555568810221e-05,
|
| 2360 |
+
"loss": 0.3332,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 7.7964141122035855,
|
| 2365 |
+
"grad_norm": 2.1222150325775146,
|
| 2366 |
+
"learning_rate": 2.2930137070314194e-05,
|
| 2367 |
+
"loss": 0.3428,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 7.819548872180452,
|
| 2372 |
+
"grad_norm": 2.4255363941192627,
|
| 2373 |
+
"learning_rate": 2.246879427185341e-05,
|
| 2374 |
+
"loss": 0.3372,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 7.842683632157316,
|
| 2379 |
+
"grad_norm": 2.0324771404266357,
|
| 2380 |
+
"learning_rate": 2.201155169071184e-05,
|
| 2381 |
+
"loss": 0.3334,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 7.865818392134182,
|
| 2386 |
+
"grad_norm": 2.5304958820343018,
|
| 2387 |
+
"learning_rate": 2.155843350804243e-05,
|
| 2388 |
+
"loss": 0.3539,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 7.888953152111047,
|
| 2393 |
+
"grad_norm": 2.394742965698242,
|
| 2394 |
+
"learning_rate": 2.110946368688055e-05,
|
| 2395 |
+
"loss": 0.3522,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 7.912087912087912,
|
| 2400 |
+
"grad_norm": 2.2429323196411133,
|
| 2401 |
+
"learning_rate": 2.0664665970876496e-05,
|
| 2402 |
+
"loss": 0.3564,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 7.935222672064778,
|
| 2407 |
+
"grad_norm": 2.269540309906006,
|
| 2408 |
+
"learning_rate": 2.0224063883039868e-05,
|
| 2409 |
+
"loss": 0.3414,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 7.958357432041643,
|
| 2414 |
+
"grad_norm": 2.2664902210235596,
|
| 2415 |
+
"learning_rate": 1.9787680724495617e-05,
|
| 2416 |
+
"loss": 0.3422,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 7.981492192018508,
|
| 2421 |
+
"grad_norm": 2.529334545135498,
|
| 2422 |
+
"learning_rate": 1.9355539573251734e-05,
|
| 2423 |
+
"loss": 0.3384,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 8.004626951995373,
|
| 2428 |
+
"grad_norm": 1.9292296171188354,
|
| 2429 |
+
"learning_rate": 1.892766328297878e-05,
|
| 2430 |
+
"loss": 0.3313,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 8.027761711972238,
|
| 2435 |
+
"grad_norm": 1.851493000984192,
|
| 2436 |
+
"learning_rate": 1.8504074481801238e-05,
|
| 2437 |
+
"loss": 0.2768,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 8.050896471949104,
|
| 2442 |
+
"grad_norm": 2.0147864818573,
|
| 2443 |
+
"learning_rate": 1.808479557110081e-05,
|
| 2444 |
+
"loss": 0.2539,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 8.074031231925968,
|
| 2449 |
+
"grad_norm": 2.2376036643981934,
|
| 2450 |
+
"learning_rate": 1.7669848724331984e-05,
|
| 2451 |
+
"loss": 0.2815,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 8.097165991902834,
|
| 2456 |
+
"grad_norm": 2.2430434226989746,
|
| 2457 |
+
"learning_rate": 1.7259255885848946e-05,
|
| 2458 |
+
"loss": 0.2603,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 8.1203007518797,
|
| 2463 |
+
"grad_norm": 1.880216121673584,
|
| 2464 |
+
"learning_rate": 1.6853038769745467e-05,
|
| 2465 |
+
"loss": 0.2647,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 8.143435511856564,
|
| 2470 |
+
"grad_norm": 2.154496192932129,
|
| 2471 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 2472 |
+
"loss": 0.2648,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 8.16657027183343,
|
| 2477 |
+
"grad_norm": 1.9910736083984375,
|
| 2478 |
+
"learning_rate": 1.6053817402871362e-05,
|
| 2479 |
+
"loss": 0.2737,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 8.189705031810295,
|
| 2484 |
+
"grad_norm": 1.7439430952072144,
|
| 2485 |
+
"learning_rate": 1.566085541871145e-05,
|
| 2486 |
+
"loss": 0.2728,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 8.212839791787161,
|
| 2491 |
+
"grad_norm": 1.8878188133239746,
|
| 2492 |
+
"learning_rate": 1.5272353687917196e-05,
|
| 2493 |
+
"loss": 0.2678,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 8.235974551764025,
|
| 2498 |
+
"grad_norm": 1.8868342638015747,
|
| 2499 |
+
"learning_rate": 1.4888332756300027e-05,
|
| 2500 |
+
"loss": 0.2629,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 8.259109311740891,
|
| 2505 |
+
"grad_norm": 2.0473947525024414,
|
| 2506 |
+
"learning_rate": 1.4508812932705363e-05,
|
| 2507 |
+
"loss": 0.2706,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 8.282244071717756,
|
| 2512 |
+
"grad_norm": 2.090017795562744,
|
| 2513 |
+
"learning_rate": 1.4133814287938706e-05,
|
| 2514 |
+
"loss": 0.2754,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 8.305378831694622,
|
| 2519 |
+
"grad_norm": 2.0764472484588623,
|
| 2520 |
+
"learning_rate": 1.3763356653704273e-05,
|
| 2521 |
+
"loss": 0.2801,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 8.328513591671486,
|
| 2526 |
+
"grad_norm": 1.978501796722412,
|
| 2527 |
+
"learning_rate": 1.339745962155613e-05,
|
| 2528 |
+
"loss": 0.2754,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 8.351648351648352,
|
| 2533 |
+
"grad_norm": 2.1843178272247314,
|
| 2534 |
+
"learning_rate": 1.3036142541862118e-05,
|
| 2535 |
+
"loss": 0.2734,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 8.374783111625216,
|
| 2540 |
+
"grad_norm": 2.240216016769409,
|
| 2541 |
+
"learning_rate": 1.2679424522780426e-05,
|
| 2542 |
+
"loss": 0.2819,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 8.397917871602083,
|
| 2547 |
+
"grad_norm": 2.18241548538208,
|
| 2548 |
+
"learning_rate": 1.2327324429249232e-05,
|
| 2549 |
+
"loss": 0.2744,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 8.421052631578947,
|
| 2554 |
+
"grad_norm": 2.1984431743621826,
|
| 2555 |
+
"learning_rate": 1.1979860881988902e-05,
|
| 2556 |
+
"loss": 0.2867,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 8.444187391555813,
|
| 2561 |
+
"grad_norm": 2.063441753387451,
|
| 2562 |
+
"learning_rate": 1.1637052256517244e-05,
|
| 2563 |
+
"loss": 0.2687,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 8.467322151532677,
|
| 2568 |
+
"grad_norm": 2.132188558578491,
|
| 2569 |
+
"learning_rate": 1.129891668217783e-05,
|
| 2570 |
+
"loss": 0.2542,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 8.490456911509543,
|
| 2575 |
+
"grad_norm": 2.227787971496582,
|
| 2576 |
+
"learning_rate": 1.0965472041181102e-05,
|
| 2577 |
+
"loss": 0.2849,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 8.513591671486408,
|
| 2582 |
+
"grad_norm": 1.9620667695999146,
|
| 2583 |
+
"learning_rate": 1.0636735967658784e-05,
|
| 2584 |
+
"loss": 0.2746,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 8.536726431463274,
|
| 2589 |
+
"grad_norm": 1.9825657606124878,
|
| 2590 |
+
"learning_rate": 1.0312725846731175e-05,
|
| 2591 |
+
"loss": 0.2665,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 8.559861191440138,
|
| 2596 |
+
"grad_norm": 1.9591861963272095,
|
| 2597 |
+
"learning_rate": 9.993458813587885e-06,
|
| 2598 |
+
"loss": 0.2598,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 8.582995951417004,
|
| 2603 |
+
"grad_norm": 2.2019407749176025,
|
| 2604 |
+
"learning_rate": 9.678951752581578e-06,
|
| 2605 |
+
"loss": 0.2801,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 8.606130711393869,
|
| 2610 |
+
"grad_norm": 2.131798267364502,
|
| 2611 |
+
"learning_rate": 9.369221296335006e-06,
|
| 2612 |
+
"loss": 0.2778,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 8.629265471370735,
|
| 2617 |
+
"grad_norm": 2.081733465194702,
|
| 2618 |
+
"learning_rate": 9.064283824861486e-06,
|
| 2619 |
+
"loss": 0.2774,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 8.652400231347599,
|
| 2624 |
+
"grad_norm": 2.1532142162323,
|
| 2625 |
+
"learning_rate": 8.764155464698597e-06,
|
| 2626 |
+
"loss": 0.2754,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 8.675534991324465,
|
| 2631 |
+
"grad_norm": 2.2483901977539062,
|
| 2632 |
+
"learning_rate": 8.46885208805529e-06,
|
| 2633 |
+
"loss": 0.2618,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 8.698669751301331,
|
| 2638 |
+
"grad_norm": 1.9993518590927124,
|
| 2639 |
+
"learning_rate": 8.178389311972612e-06,
|
| 2640 |
+
"loss": 0.2892,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 8.721804511278195,
|
| 2645 |
+
"grad_norm": 2.1293857097625732,
|
| 2646 |
+
"learning_rate": 7.892782497497642e-06,
|
| 2647 |
+
"loss": 0.2909,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 8.744939271255062,
|
| 2652 |
+
"grad_norm": 1.9576681852340698,
|
| 2653 |
+
"learning_rate": 7.612046748871327e-06,
|
| 2654 |
+
"loss": 0.2783,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 8.768074031231926,
|
| 2659 |
+
"grad_norm": 1.7659740447998047,
|
| 2660 |
+
"learning_rate": 7.336196912729487e-06,
|
| 2661 |
+
"loss": 0.2757,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 8.791208791208792,
|
| 2666 |
+
"grad_norm": 1.8570443391799927,
|
| 2667 |
+
"learning_rate": 7.0652475773177464e-06,
|
| 2668 |
+
"loss": 0.262,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 8.814343551185656,
|
| 2673 |
+
"grad_norm": 1.9978190660476685,
|
| 2674 |
+
"learning_rate": 6.7992130717201564e-06,
|
| 2675 |
+
"loss": 0.2686,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 8.837478311162522,
|
| 2680 |
+
"grad_norm": 2.242790460586548,
|
| 2681 |
+
"learning_rate": 6.538107465101162e-06,
|
| 2682 |
+
"loss": 0.2857,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 8.860613071139387,
|
| 2687 |
+
"grad_norm": 2.2894692420959473,
|
| 2688 |
+
"learning_rate": 6.281944565961773e-06,
|
| 2689 |
+
"loss": 0.2762,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 8.883747831116253,
|
| 2694 |
+
"grad_norm": 1.879543662071228,
|
| 2695 |
+
"learning_rate": 6.030737921409169e-06,
|
| 2696 |
+
"loss": 0.275,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 8.906882591093117,
|
| 2701 |
+
"grad_norm": 1.8957421779632568,
|
| 2702 |
+
"learning_rate": 5.784500816440353e-06,
|
| 2703 |
+
"loss": 0.2736,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 8.930017351069983,
|
| 2708 |
+
"grad_norm": 2.0215442180633545,
|
| 2709 |
+
"learning_rate": 5.543246273239533e-06,
|
| 2710 |
+
"loss": 0.2698,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 8.953152111046847,
|
| 2715 |
+
"grad_norm": 2.278769016265869,
|
| 2716 |
+
"learning_rate": 5.306987050489442e-06,
|
| 2717 |
+
"loss": 0.2832,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 8.976286871023714,
|
| 2722 |
+
"grad_norm": 2.370384931564331,
|
| 2723 |
+
"learning_rate": 5.075735642696611e-06,
|
| 2724 |
+
"loss": 0.2739,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 8.999421631000578,
|
| 2729 |
+
"grad_norm": 2.1495909690856934,
|
| 2730 |
+
"learning_rate": 4.849504279530636e-06,
|
| 2731 |
+
"loss": 0.2664,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 9.022556390977444,
|
| 2736 |
+
"grad_norm": 1.8931546211242676,
|
| 2737 |
+
"learning_rate": 4.628304925177318e-06,
|
| 2738 |
+
"loss": 0.251,
|
| 2739 |
+
"step": 3900
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 9.045691150954308,
|
| 2743 |
+
"grad_norm": 1.964045763015747,
|
| 2744 |
+
"learning_rate": 4.412149277706046e-06,
|
| 2745 |
+
"loss": 0.2394,
|
| 2746 |
+
"step": 3910
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 9.068825910931174,
|
| 2750 |
+
"grad_norm": 1.8425894975662231,
|
| 2751 |
+
"learning_rate": 4.20104876845111e-06,
|
| 2752 |
+
"loss": 0.2368,
|
| 2753 |
+
"step": 3920
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 9.091960670908039,
|
| 2757 |
+
"grad_norm": 1.8906612396240234,
|
| 2758 |
+
"learning_rate": 3.99501456140714e-06,
|
| 2759 |
+
"loss": 0.2517,
|
| 2760 |
+
"step": 3930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 9.115095430884905,
|
| 2764 |
+
"grad_norm": 1.9606223106384277,
|
| 2765 |
+
"learning_rate": 3.7940575526386857e-06,
|
| 2766 |
+
"loss": 0.2638,
|
| 2767 |
+
"step": 3940
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 9.13823019086177,
|
| 2771 |
+
"grad_norm": 2.0585832595825195,
|
| 2772 |
+
"learning_rate": 3.598188369704036e-06,
|
| 2773 |
+
"loss": 0.2465,
|
| 2774 |
+
"step": 3950
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 9.161364950838635,
|
| 2778 |
+
"grad_norm": 2.0932703018188477,
|
| 2779 |
+
"learning_rate": 3.40741737109318e-06,
|
| 2780 |
+
"loss": 0.2454,
|
| 2781 |
+
"step": 3960
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 9.1844997108155,
|
| 2785 |
+
"grad_norm": 1.9221563339233398,
|
| 2786 |
+
"learning_rate": 3.2217546456799086e-06,
|
| 2787 |
+
"loss": 0.2387,
|
| 2788 |
+
"step": 3970
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 9.207634470792366,
|
| 2792 |
+
"grad_norm": 1.973889946937561,
|
| 2793 |
+
"learning_rate": 3.0412100121884e-06,
|
| 2794 |
+
"loss": 0.2475,
|
| 2795 |
+
"step": 3980
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 9.23076923076923,
|
| 2799 |
+
"grad_norm": 1.8199151754379272,
|
| 2800 |
+
"learning_rate": 2.8657930186738567e-06,
|
| 2801 |
+
"loss": 0.2364,
|
| 2802 |
+
"step": 3990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 9.253903990746096,
|
| 2806 |
+
"grad_norm": 1.847708821296692,
|
| 2807 |
+
"learning_rate": 2.6955129420176196e-06,
|
| 2808 |
+
"loss": 0.2431,
|
| 2809 |
+
"step": 4000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 9.277038750722962,
|
| 2813 |
+
"grad_norm": 1.6919292211532593,
|
| 2814 |
+
"learning_rate": 2.530378787436527e-06,
|
| 2815 |
+
"loss": 0.2546,
|
| 2816 |
+
"step": 4010
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 9.300173510699826,
|
| 2820 |
+
"grad_norm": 1.7380826473236084,
|
| 2821 |
+
"learning_rate": 2.3703992880066638e-06,
|
| 2822 |
+
"loss": 0.2471,
|
| 2823 |
+
"step": 4020
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 9.323308270676693,
|
| 2827 |
+
"grad_norm": 2.1914947032928467,
|
| 2828 |
+
"learning_rate": 2.2155829042015962e-06,
|
| 2829 |
+
"loss": 0.2406,
|
| 2830 |
+
"step": 4030
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 9.346443030653557,
|
| 2834 |
+
"grad_norm": 1.788342833518982,
|
| 2835 |
+
"learning_rate": 2.0659378234448525e-06,
|
| 2836 |
+
"loss": 0.2292,
|
| 2837 |
+
"step": 4040
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 9.369577790630423,
|
| 2841 |
+
"grad_norm": 1.953272819519043,
|
| 2842 |
+
"learning_rate": 1.921471959676957e-06,
|
| 2843 |
+
"loss": 0.2393,
|
| 2844 |
+
"step": 4050
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 9.392712550607287,
|
| 2848 |
+
"grad_norm": 2.092925548553467,
|
| 2849 |
+
"learning_rate": 1.7821929529369342e-06,
|
| 2850 |
+
"loss": 0.2455,
|
| 2851 |
+
"step": 4060
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 9.415847310584153,
|
| 2855 |
+
"grad_norm": 1.312064528465271,
|
| 2856 |
+
"learning_rate": 1.648108168958229e-06,
|
| 2857 |
+
"loss": 0.2464,
|
| 2858 |
+
"step": 4070
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 9.438982070561018,
|
| 2862 |
+
"grad_norm": 1.8902608156204224,
|
| 2863 |
+
"learning_rate": 1.5192246987791981e-06,
|
| 2864 |
+
"loss": 0.2503,
|
| 2865 |
+
"step": 4080
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 9.462116830537884,
|
| 2869 |
+
"grad_norm": 2.0683326721191406,
|
| 2870 |
+
"learning_rate": 1.3955493583680868e-06,
|
| 2871 |
+
"loss": 0.2532,
|
| 2872 |
+
"step": 4090
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 9.485251590514748,
|
| 2876 |
+
"grad_norm": 2.037275552749634,
|
| 2877 |
+
"learning_rate": 1.2770886882625955e-06,
|
| 2878 |
+
"loss": 0.2494,
|
| 2879 |
+
"step": 4100
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 9.508386350491614,
|
| 2883 |
+
"grad_norm": 1.884734034538269,
|
| 2884 |
+
"learning_rate": 1.163848953223934e-06,
|
| 2885 |
+
"loss": 0.2427,
|
| 2886 |
+
"step": 4110
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 9.531521110468478,
|
| 2890 |
+
"grad_norm": 2.0720531940460205,
|
| 2891 |
+
"learning_rate": 1.055836141905553e-06,
|
| 2892 |
+
"loss": 0.2361,
|
| 2893 |
+
"step": 4120
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 9.554655870445345,
|
| 2897 |
+
"grad_norm": 1.7988377809524536,
|
| 2898 |
+
"learning_rate": 9.530559665364202e-07,
|
| 2899 |
+
"loss": 0.2524,
|
| 2900 |
+
"step": 4130
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 9.577790630422209,
|
| 2904 |
+
"grad_norm": 2.159759044647217,
|
| 2905 |
+
"learning_rate": 8.555138626189618e-07,
|
| 2906 |
+
"loss": 0.2562,
|
| 2907 |
+
"step": 4140
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 9.600925390399075,
|
| 2911 |
+
"grad_norm": 2.3228530883789062,
|
| 2912 |
+
"learning_rate": 7.632149886415363e-07,
|
| 2913 |
+
"loss": 0.2326,
|
| 2914 |
+
"step": 4150
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 9.62406015037594,
|
| 2918 |
+
"grad_norm": 2.210181713104248,
|
| 2919 |
+
"learning_rate": 6.761642258056978e-07,
|
| 2920 |
+
"loss": 0.2628,
|
| 2921 |
+
"step": 4160
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 9.647194910352805,
|
| 2925 |
+
"grad_norm": 1.6865817308425903,
|
| 2926 |
+
"learning_rate": 5.943661777680354e-07,
|
| 2927 |
+
"loss": 0.2409,
|
| 2928 |
+
"step": 4170
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 9.67032967032967,
|
| 2932 |
+
"grad_norm": 1.9101097583770752,
|
| 2933 |
+
"learning_rate": 5.178251703967018e-07,
|
| 2934 |
+
"loss": 0.2453,
|
| 2935 |
+
"step": 4180
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 9.693464430306536,
|
| 2939 |
+
"grad_norm": 2.2354347705841064,
|
| 2940 |
+
"learning_rate": 4.465452515426072e-07,
|
| 2941 |
+
"loss": 0.2339,
|
| 2942 |
+
"step": 4190
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 9.7165991902834,
|
| 2946 |
+
"grad_norm": 2.173842668533325,
|
| 2947 |
+
"learning_rate": 3.805301908254455e-07,
|
| 2948 |
+
"loss": 0.2369,
|
| 2949 |
+
"step": 4200
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 9.739733950260266,
|
| 2953 |
+
"grad_norm": 1.74326491355896,
|
| 2954 |
+
"learning_rate": 3.197834794342436e-07,
|
| 2955 |
+
"loss": 0.2431,
|
| 2956 |
+
"step": 4210
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 9.76286871023713,
|
| 2960 |
+
"grad_norm": 1.731026291847229,
|
| 2961 |
+
"learning_rate": 2.643083299427751e-07,
|
| 2962 |
+
"loss": 0.2587,
|
| 2963 |
+
"step": 4220
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 9.786003470213997,
|
| 2967 |
+
"grad_norm": 2.085442066192627,
|
| 2968 |
+
"learning_rate": 2.141076761396521e-07,
|
| 2969 |
+
"loss": 0.2525,
|
| 2970 |
+
"step": 4230
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 9.809138230190861,
|
| 2974 |
+
"grad_norm": 1.921111822128296,
|
| 2975 |
+
"learning_rate": 1.6918417287318245e-07,
|
| 2976 |
+
"loss": 0.2561,
|
| 2977 |
+
"step": 4240
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 9.832272990167727,
|
| 2981 |
+
"grad_norm": 1.7905348539352417,
|
| 2982 |
+
"learning_rate": 1.295401959109599e-07,
|
| 2983 |
+
"loss": 0.2425,
|
| 2984 |
+
"step": 4250
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 9.855407750144593,
|
| 2988 |
+
"grad_norm": 2.0938756465911865,
|
| 2989 |
+
"learning_rate": 9.517784181422019e-08,
|
| 2990 |
+
"loss": 0.2486,
|
| 2991 |
+
"step": 4260
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 9.878542510121457,
|
| 2995 |
+
"grad_norm": 2.0553646087646484,
|
| 2996 |
+
"learning_rate": 6.609892782699633e-08,
|
| 2997 |
+
"loss": 0.2397,
|
| 2998 |
+
"step": 4270
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 9.901677270098324,
|
| 3002 |
+
"grad_norm": 1.8413350582122803,
|
| 3003 |
+
"learning_rate": 4.230499177994007e-08,
|
| 3004 |
+
"loss": 0.2504,
|
| 3005 |
+
"step": 4280
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 9.924812030075188,
|
| 3009 |
+
"grad_norm": 1.7618731260299683,
|
| 3010 |
+
"learning_rate": 2.379729200908676e-08,
|
| 3011 |
+
"loss": 0.2491,
|
| 3012 |
+
"step": 4290
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 9.947946790052054,
|
| 3016 |
+
"grad_norm": 1.7408370971679688,
|
| 3017 |
+
"learning_rate": 1.0576807289253143e-08,
|
| 3018 |
+
"loss": 0.2448,
|
| 3019 |
+
"step": 4300
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 9.971081550028918,
|
| 3023 |
+
"grad_norm": 1.9657200574874878,
|
| 3024 |
+
"learning_rate": 2.6442367822565596e-09,
|
| 3025 |
+
"loss": 0.2623,
|
| 3026 |
+
"step": 4310
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 9.994216310005784,
|
| 3030 |
+
"grad_norm": 1.993733525276184,
|
| 3031 |
+
"learning_rate": 0.0,
|
| 3032 |
+
"loss": 0.2453,
|
| 3033 |
+
"step": 4320
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 9.994216310005784,
|
| 3037 |
+
"step": 4320,
|
| 3038 |
+
"total_flos": 2.1249547460149248e+17,
|
| 3039 |
+
"train_loss": 0.887424631913503,
|
| 3040 |
+
"train_runtime": 17181.5838,
|
| 3041 |
+
"train_samples_per_second": 1.006,
|
| 3042 |
+
"train_steps_per_second": 0.251
|
| 3043 |
+
}
|
| 3044 |
+
],
|
| 3045 |
+
"logging_steps": 10,
|
| 3046 |
+
"max_steps": 4320,
|
| 3047 |
+
"num_input_tokens_seen": 0,
|
| 3048 |
+
"num_train_epochs": 10,
|
| 3049 |
+
"save_steps": 500,
|
| 3050 |
+
"total_flos": 2.1249547460149248e+17,
|
| 3051 |
+
"train_batch_size": 1,
|
| 3052 |
+
"trial_name": null,
|
| 3053 |
+
"trial_params": null
|
| 3054 |
+
}
|
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:ae18098947d709c4a75c7f8f7e079e765707d3f3dc9c259d5b9c9b7edc83d2d7
|
| 3 |
size 5048
|