Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- trainer_state.json +2368 -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,
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2368 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.992520568436799,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3340,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.029917726252804786,
|
| 13 |
+
"grad_norm": NaN,
|
| 14 |
+
"learning_rate": 0.00019999557642276714,
|
| 15 |
+
"loss": 0.0,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05983545250560957,
|
| 20 |
+
"grad_norm": NaN,
|
| 21 |
+
"learning_rate": 0.00019998230608242915,
|
| 22 |
+
"loss": 0.0,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08975317875841436,
|
| 27 |
+
"grad_norm": NaN,
|
| 28 |
+
"learning_rate": 0.00019996019015303368,
|
| 29 |
+
"loss": 0.0,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11967090501121914,
|
| 34 |
+
"grad_norm": NaN,
|
| 35 |
+
"learning_rate": 0.00019992923059121106,
|
| 36 |
+
"loss": 0.0,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.14958863126402394,
|
| 41 |
+
"grad_norm": NaN,
|
| 42 |
+
"learning_rate": 0.00019988943013600157,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.17950635751682872,
|
| 48 |
+
"grad_norm": NaN,
|
| 49 |
+
"learning_rate": 0.00019984079230861296,
|
| 50 |
+
"loss": 0.0,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2094240837696335,
|
| 55 |
+
"grad_norm": NaN,
|
| 56 |
+
"learning_rate": 0.00019978332141210893,
|
| 57 |
+
"loss": 0.0,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2393418100224383,
|
| 62 |
+
"grad_norm": NaN,
|
| 63 |
+
"learning_rate": 0.00019971702253102856,
|
| 64 |
+
"loss": 0.0,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.26925953627524307,
|
| 69 |
+
"grad_norm": NaN,
|
| 70 |
+
"learning_rate": 0.00019964190153093613,
|
| 71 |
+
"loss": 0.0,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2991772625280479,
|
| 76 |
+
"grad_norm": NaN,
|
| 77 |
+
"learning_rate": 0.00019955796505790268,
|
| 78 |
+
"loss": 0.0,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.32909498878085264,
|
| 83 |
+
"grad_norm": NaN,
|
| 84 |
+
"learning_rate": 0.00019946522053791753,
|
| 85 |
+
"loss": 0.0,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.35901271503365745,
|
| 90 |
+
"grad_norm": NaN,
|
| 91 |
+
"learning_rate": 0.00019936367617623171,
|
| 92 |
+
"loss": 0.0,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3889304412864622,
|
| 97 |
+
"grad_norm": NaN,
|
| 98 |
+
"learning_rate": 0.00019925334095663167,
|
| 99 |
+
"loss": 0.0,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.418848167539267,
|
| 104 |
+
"grad_norm": NaN,
|
| 105 |
+
"learning_rate": 0.0001991342246406448,
|
| 106 |
+
"loss": 0.0,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.4487658937920718,
|
| 111 |
+
"grad_norm": NaN,
|
| 112 |
+
"learning_rate": 0.0001990063377666755,
|
| 113 |
+
"loss": 0.0,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4786836200448766,
|
| 118 |
+
"grad_norm": NaN,
|
| 119 |
+
"learning_rate": 0.00019886969164907307,
|
| 120 |
+
"loss": 0.0,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5086013462976814,
|
| 125 |
+
"grad_norm": NaN,
|
| 126 |
+
"learning_rate": 0.0001987242983771306,
|
| 127 |
+
"loss": 0.0,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5385190725504861,
|
| 132 |
+
"grad_norm": NaN,
|
| 133 |
+
"learning_rate": 0.00019857017081401548,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5684367988032909,
|
| 139 |
+
"grad_norm": NaN,
|
| 140 |
+
"learning_rate": 0.00019840732259563128,
|
| 141 |
+
"loss": 0.0,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5983545250560958,
|
| 146 |
+
"grad_norm": NaN,
|
| 147 |
+
"learning_rate": 0.00019823576812941137,
|
| 148 |
+
"loss": 0.0,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6282722513089005,
|
| 153 |
+
"grad_norm": NaN,
|
| 154 |
+
"learning_rate": 0.00019805552259304445,
|
| 155 |
+
"loss": 0.0,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6581899775617053,
|
| 160 |
+
"grad_norm": NaN,
|
| 161 |
+
"learning_rate": 0.00019786660193313148,
|
| 162 |
+
"loss": 0.0,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6881077038145101,
|
| 167 |
+
"grad_norm": NaN,
|
| 168 |
+
"learning_rate": 0.00019766902286377506,
|
| 169 |
+
"loss": 0.0,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7180254300673149,
|
| 174 |
+
"grad_norm": NaN,
|
| 175 |
+
"learning_rate": 0.0001974628028651007,
|
| 176 |
+
"loss": 0.0,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7479431563201197,
|
| 181 |
+
"grad_norm": NaN,
|
| 182 |
+
"learning_rate": 0.00019724796018171021,
|
| 183 |
+
"loss": 0.0,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7778608825729244,
|
| 188 |
+
"grad_norm": NaN,
|
| 189 |
+
"learning_rate": 0.0001970245138210676,
|
| 190 |
+
"loss": 0.0,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8077786088257293,
|
| 195 |
+
"grad_norm": NaN,
|
| 196 |
+
"learning_rate": 0.0001967924835518176,
|
| 197 |
+
"loss": 0.0,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.837696335078534,
|
| 202 |
+
"grad_norm": NaN,
|
| 203 |
+
"learning_rate": 0.00019655188990203647,
|
| 204 |
+
"loss": 0.0,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8676140613313388,
|
| 209 |
+
"grad_norm": NaN,
|
| 210 |
+
"learning_rate": 0.0001963027541574161,
|
| 211 |
+
"loss": 0.0,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.8975317875841436,
|
| 216 |
+
"grad_norm": NaN,
|
| 217 |
+
"learning_rate": 0.00019604509835938064,
|
| 218 |
+
"loss": 0.0,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9274495138369484,
|
| 223 |
+
"grad_norm": NaN,
|
| 224 |
+
"learning_rate": 0.00019577894530313654,
|
| 225 |
+
"loss": 0.0,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9573672400897532,
|
| 230 |
+
"grad_norm": NaN,
|
| 231 |
+
"learning_rate": 0.00019550431853565577,
|
| 232 |
+
"loss": 0.0,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9872849663425579,
|
| 237 |
+
"grad_norm": NaN,
|
| 238 |
+
"learning_rate": 0.00019522124235359263,
|
| 239 |
+
"loss": 0.0,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0172026925953628,
|
| 244 |
+
"grad_norm": NaN,
|
| 245 |
+
"learning_rate": 0.00019492974180113426,
|
| 246 |
+
"loss": 0.0,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0471204188481675,
|
| 251 |
+
"grad_norm": NaN,
|
| 252 |
+
"learning_rate": 0.0001946298426677848,
|
| 253 |
+
"loss": 0.0,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0770381451009723,
|
| 258 |
+
"grad_norm": NaN,
|
| 259 |
+
"learning_rate": 0.00019432157148608378,
|
| 260 |
+
"loss": 0.0,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.106955871353777,
|
| 265 |
+
"grad_norm": NaN,
|
| 266 |
+
"learning_rate": 0.00019400495552925886,
|
| 267 |
+
"loss": 0.0,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.136873597606582,
|
| 272 |
+
"grad_norm": NaN,
|
| 273 |
+
"learning_rate": 0.00019368002280881278,
|
| 274 |
+
"loss": 0.0,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1667913238593868,
|
| 279 |
+
"grad_norm": NaN,
|
| 280 |
+
"learning_rate": 0.00019334680207204523,
|
| 281 |
+
"loss": 0.0,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.1967090501121915,
|
| 286 |
+
"grad_norm": NaN,
|
| 287 |
+
"learning_rate": 0.00019300532279950948,
|
| 288 |
+
"loss": 0.0,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2266267763649963,
|
| 293 |
+
"grad_norm": NaN,
|
| 294 |
+
"learning_rate": 0.0001926556152024043,
|
| 295 |
+
"loss": 0.0,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.256544502617801,
|
| 300 |
+
"grad_norm": NaN,
|
| 301 |
+
"learning_rate": 0.00019229771021990095,
|
| 302 |
+
"loss": 0.0,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.2864622288706058,
|
| 307 |
+
"grad_norm": NaN,
|
| 308 |
+
"learning_rate": 0.00019193163951640606,
|
| 309 |
+
"loss": 0.0,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3163799551234106,
|
| 314 |
+
"grad_norm": NaN,
|
| 315 |
+
"learning_rate": 0.00019155743547876023,
|
| 316 |
+
"loss": 0.0,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3462976813762153,
|
| 321 |
+
"grad_norm": NaN,
|
| 322 |
+
"learning_rate": 0.0001911751312133727,
|
| 323 |
+
"loss": 0.0,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.37621540762902,
|
| 328 |
+
"grad_norm": NaN,
|
| 329 |
+
"learning_rate": 0.00019078476054329237,
|
| 330 |
+
"loss": 0.0,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.406133133881825,
|
| 335 |
+
"grad_norm": NaN,
|
| 336 |
+
"learning_rate": 0.0001903863580052154,
|
| 337 |
+
"loss": 0.0,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4360508601346298,
|
| 342 |
+
"grad_norm": NaN,
|
| 343 |
+
"learning_rate": 0.00018997995884642967,
|
| 344 |
+
"loss": 0.0,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4659685863874345,
|
| 349 |
+
"grad_norm": NaN,
|
| 350 |
+
"learning_rate": 0.0001895655990216966,
|
| 351 |
+
"loss": 0.0,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.4958863126402393,
|
| 356 |
+
"grad_norm": NaN,
|
| 357 |
+
"learning_rate": 0.00018914331519006986,
|
| 358 |
+
"loss": 0.0,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5258040388930443,
|
| 363 |
+
"grad_norm": NaN,
|
| 364 |
+
"learning_rate": 0.00018871314471165232,
|
| 365 |
+
"loss": 0.0,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.555721765145849,
|
| 370 |
+
"grad_norm": NaN,
|
| 371 |
+
"learning_rate": 0.00018827512564429074,
|
| 372 |
+
"loss": 0.0,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.5856394913986538,
|
| 377 |
+
"grad_norm": NaN,
|
| 378 |
+
"learning_rate": 0.00018782929674020856,
|
| 379 |
+
"loss": 0.0,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6155572176514585,
|
| 384 |
+
"grad_norm": NaN,
|
| 385 |
+
"learning_rate": 0.00018737569744257756,
|
| 386 |
+
"loss": 0.0,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6454749439042633,
|
| 391 |
+
"grad_norm": NaN,
|
| 392 |
+
"learning_rate": 0.00018691436788202822,
|
| 393 |
+
"loss": 0.0,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.675392670157068,
|
| 398 |
+
"grad_norm": NaN,
|
| 399 |
+
"learning_rate": 0.00018644534887309944,
|
| 400 |
+
"loss": 0.0,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7053103964098728,
|
| 405 |
+
"grad_norm": NaN,
|
| 406 |
+
"learning_rate": 0.0001859686819106274,
|
| 407 |
+
"loss": 0.0,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7352281226626776,
|
| 412 |
+
"grad_norm": NaN,
|
| 413 |
+
"learning_rate": 0.00018548440916607453,
|
| 414 |
+
"loss": 0.0,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.7651458489154823,
|
| 419 |
+
"grad_norm": NaN,
|
| 420 |
+
"learning_rate": 0.00018499257348379854,
|
| 421 |
+
"loss": 0.0,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.795063575168287,
|
| 426 |
+
"grad_norm": NaN,
|
| 427 |
+
"learning_rate": 0.00018449321837726208,
|
| 428 |
+
"loss": 0.0,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.824981301421092,
|
| 433 |
+
"grad_norm": NaN,
|
| 434 |
+
"learning_rate": 0.00018398638802518262,
|
| 435 |
+
"loss": 0.0,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.8548990276738968,
|
| 440 |
+
"grad_norm": NaN,
|
| 441 |
+
"learning_rate": 0.00018347212726762437,
|
| 442 |
+
"loss": 0.0,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.8848167539267016,
|
| 447 |
+
"grad_norm": NaN,
|
| 448 |
+
"learning_rate": 0.00018295048160203088,
|
| 449 |
+
"loss": 0.0,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9147344801795063,
|
| 454 |
+
"grad_norm": NaN,
|
| 455 |
+
"learning_rate": 0.00018242149717919993,
|
| 456 |
+
"loss": 0.0,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9446522064323113,
|
| 461 |
+
"grad_norm": NaN,
|
| 462 |
+
"learning_rate": 0.00018188522079920055,
|
| 463 |
+
"loss": 0.0,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.974569932685116,
|
| 468 |
+
"grad_norm": NaN,
|
| 469 |
+
"learning_rate": 0.00018134169990723247,
|
| 470 |
+
"loss": 0.0,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.004487658937921,
|
| 475 |
+
"grad_norm": NaN,
|
| 476 |
+
"learning_rate": 0.00018079098258942845,
|
| 477 |
+
"loss": 0.0,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0344053851907256,
|
| 482 |
+
"grad_norm": NaN,
|
| 483 |
+
"learning_rate": 0.00018023311756860038,
|
| 484 |
+
"loss": 0.0,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.0643231114435303,
|
| 489 |
+
"grad_norm": NaN,
|
| 490 |
+
"learning_rate": 0.00017966815419992829,
|
| 491 |
+
"loss": 0.0,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.094240837696335,
|
| 496 |
+
"grad_norm": NaN,
|
| 497 |
+
"learning_rate": 0.00017909614246659409,
|
| 498 |
+
"loss": 0.0,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.12415856394914,
|
| 503 |
+
"grad_norm": NaN,
|
| 504 |
+
"learning_rate": 0.0001785171329753594,
|
| 505 |
+
"loss": 0.0,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.1540762902019446,
|
| 510 |
+
"grad_norm": NaN,
|
| 511 |
+
"learning_rate": 0.00017793117695208828,
|
| 512 |
+
"loss": 0.0,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.1839940164547493,
|
| 517 |
+
"grad_norm": NaN,
|
| 518 |
+
"learning_rate": 0.0001773383262372152,
|
| 519 |
+
"loss": 0.0,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.213911742707554,
|
| 524 |
+
"grad_norm": NaN,
|
| 525 |
+
"learning_rate": 0.0001767386332811587,
|
| 526 |
+
"loss": 0.0,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.243829468960359,
|
| 531 |
+
"grad_norm": NaN,
|
| 532 |
+
"learning_rate": 0.0001761321511396809,
|
| 533 |
+
"loss": 0.0,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.273747195213164,
|
| 538 |
+
"grad_norm": NaN,
|
| 539 |
+
"learning_rate": 0.00017551893346919362,
|
| 540 |
+
"loss": 0.0,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.303664921465969,
|
| 545 |
+
"grad_norm": NaN,
|
| 546 |
+
"learning_rate": 0.00017489903452201145,
|
| 547 |
+
"loss": 0.0,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3335826477187736,
|
| 552 |
+
"grad_norm": NaN,
|
| 553 |
+
"learning_rate": 0.0001742725091415517,
|
| 554 |
+
"loss": 0.0,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.3635003739715783,
|
| 559 |
+
"grad_norm": NaN,
|
| 560 |
+
"learning_rate": 0.00017363941275748263,
|
| 561 |
+
"loss": 0.0,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.393418100224383,
|
| 566 |
+
"grad_norm": NaN,
|
| 567 |
+
"learning_rate": 0.00017299980138081922,
|
| 568 |
+
"loss": 0.0,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.423335826477188,
|
| 573 |
+
"grad_norm": NaN,
|
| 574 |
+
"learning_rate": 0.00017235373159896794,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.4532535527299926,
|
| 580 |
+
"grad_norm": NaN,
|
| 581 |
+
"learning_rate": 0.00017170126057072032,
|
| 582 |
+
"loss": 0.0,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.4831712789827973,
|
| 587 |
+
"grad_norm": NaN,
|
| 588 |
+
"learning_rate": 0.00017104244602119613,
|
| 589 |
+
"loss": 0.0,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.513089005235602,
|
| 594 |
+
"grad_norm": NaN,
|
| 595 |
+
"learning_rate": 0.00017037734623673615,
|
| 596 |
+
"loss": 0.0,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.543006731488407,
|
| 601 |
+
"grad_norm": NaN,
|
| 602 |
+
"learning_rate": 0.00016970602005974568,
|
| 603 |
+
"loss": 0.0,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.5729244577412116,
|
| 608 |
+
"grad_norm": NaN,
|
| 609 |
+
"learning_rate": 0.00016902852688348866,
|
| 610 |
+
"loss": 0.0,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6028421839940163,
|
| 615 |
+
"grad_norm": NaN,
|
| 616 |
+
"learning_rate": 0.00016834492664683274,
|
| 617 |
+
"loss": 0.0,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.632759910246821,
|
| 622 |
+
"grad_norm": NaN,
|
| 623 |
+
"learning_rate": 0.00016765527982894689,
|
| 624 |
+
"loss": 0.0,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.662677636499626,
|
| 629 |
+
"grad_norm": NaN,
|
| 630 |
+
"learning_rate": 0.00016695964744395028,
|
| 631 |
+
"loss": 0.0,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.6925953627524306,
|
| 636 |
+
"grad_norm": NaN,
|
| 637 |
+
"learning_rate": 0.00016625809103551462,
|
| 638 |
+
"loss": 0.0,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.7225130890052354,
|
| 643 |
+
"grad_norm": NaN,
|
| 644 |
+
"learning_rate": 0.00016555067267141896,
|
| 645 |
+
"loss": 0.0,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.75243081525804,
|
| 650 |
+
"grad_norm": NaN,
|
| 651 |
+
"learning_rate": 0.0001648374549380587,
|
| 652 |
+
"loss": 0.0,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.7823485415108453,
|
| 657 |
+
"grad_norm": NaN,
|
| 658 |
+
"learning_rate": 0.00016411850093490837,
|
| 659 |
+
"loss": 0.0,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.81226626776365,
|
| 664 |
+
"grad_norm": NaN,
|
| 665 |
+
"learning_rate": 0.00016339387426893918,
|
| 666 |
+
"loss": 0.0,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.842183994016455,
|
| 671 |
+
"grad_norm": NaN,
|
| 672 |
+
"learning_rate": 0.00016266363904899163,
|
| 673 |
+
"loss": 0.0,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.8721017202692596,
|
| 678 |
+
"grad_norm": NaN,
|
| 679 |
+
"learning_rate": 0.00016192785988010353,
|
| 680 |
+
"loss": 0.0,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9020194465220643,
|
| 685 |
+
"grad_norm": NaN,
|
| 686 |
+
"learning_rate": 0.00016118660185779446,
|
| 687 |
+
"loss": 0.0,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.931937172774869,
|
| 692 |
+
"grad_norm": NaN,
|
| 693 |
+
"learning_rate": 0.0001604399305623067,
|
| 694 |
+
"loss": 0.0,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.961854899027674,
|
| 699 |
+
"grad_norm": NaN,
|
| 700 |
+
"learning_rate": 0.00015968791205280306,
|
| 701 |
+
"loss": 0.0,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.9917726252804786,
|
| 706 |
+
"grad_norm": NaN,
|
| 707 |
+
"learning_rate": 0.00015893061286152276,
|
| 708 |
+
"loss": 0.0,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.0216903515332834,
|
| 713 |
+
"grad_norm": NaN,
|
| 714 |
+
"learning_rate": 0.00015816809998789492,
|
| 715 |
+
"loss": 0.0,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.051608077786088,
|
| 720 |
+
"grad_norm": NaN,
|
| 721 |
+
"learning_rate": 0.00015740044089261138,
|
| 722 |
+
"loss": 0.0,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.081525804038893,
|
| 727 |
+
"grad_norm": NaN,
|
| 728 |
+
"learning_rate": 0.00015662770349165802,
|
| 729 |
+
"loss": 0.0,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.1114435302916976,
|
| 734 |
+
"grad_norm": NaN,
|
| 735 |
+
"learning_rate": 0.00015584995615030634,
|
| 736 |
+
"loss": 0.0,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.141361256544503,
|
| 741 |
+
"grad_norm": NaN,
|
| 742 |
+
"learning_rate": 0.00015506726767706497,
|
| 743 |
+
"loss": 0.0,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.1712789827973076,
|
| 748 |
+
"grad_norm": NaN,
|
| 749 |
+
"learning_rate": 0.00015427970731759212,
|
| 750 |
+
"loss": 0.0,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.2011967090501123,
|
| 755 |
+
"grad_norm": NaN,
|
| 756 |
+
"learning_rate": 0.00015348734474856932,
|
| 757 |
+
"loss": 0.0,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.231114435302917,
|
| 762 |
+
"grad_norm": NaN,
|
| 763 |
+
"learning_rate": 0.000152690250071537,
|
| 764 |
+
"loss": 0.0,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.261032161555722,
|
| 769 |
+
"grad_norm": NaN,
|
| 770 |
+
"learning_rate": 0.0001518884938066924,
|
| 771 |
+
"loss": 0.0,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.2909498878085266,
|
| 776 |
+
"grad_norm": NaN,
|
| 777 |
+
"learning_rate": 0.0001510821468866508,
|
| 778 |
+
"loss": 0.0,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.3208676140613314,
|
| 783 |
+
"grad_norm": NaN,
|
| 784 |
+
"learning_rate": 0.0001502712806501697,
|
| 785 |
+
"loss": 0.0,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.350785340314136,
|
| 790 |
+
"grad_norm": NaN,
|
| 791 |
+
"learning_rate": 0.00014945596683583754,
|
| 792 |
+
"loss": 0.0,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.380703066566941,
|
| 797 |
+
"grad_norm": NaN,
|
| 798 |
+
"learning_rate": 0.00014863627757572687,
|
| 799 |
+
"loss": 0.0,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.4106207928197456,
|
| 804 |
+
"grad_norm": NaN,
|
| 805 |
+
"learning_rate": 0.00014781228538901267,
|
| 806 |
+
"loss": 0.0,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.4405385190725504,
|
| 811 |
+
"grad_norm": NaN,
|
| 812 |
+
"learning_rate": 0.00014698406317555652,
|
| 813 |
+
"loss": 0.0,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.470456245325355,
|
| 818 |
+
"grad_norm": NaN,
|
| 819 |
+
"learning_rate": 0.00014615168420945688,
|
| 820 |
+
"loss": 0.0,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.50037397157816,
|
| 825 |
+
"grad_norm": NaN,
|
| 826 |
+
"learning_rate": 0.00014531522213256674,
|
| 827 |
+
"loss": 0.0,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.5302916978309646,
|
| 832 |
+
"grad_norm": NaN,
|
| 833 |
+
"learning_rate": 0.00014447475094797797,
|
| 834 |
+
"loss": 0.0,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.5602094240837694,
|
| 839 |
+
"grad_norm": NaN,
|
| 840 |
+
"learning_rate": 0.00014363034501347456,
|
| 841 |
+
"loss": 0.0,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.590127150336574,
|
| 846 |
+
"grad_norm": NaN,
|
| 847 |
+
"learning_rate": 0.0001427820790349539,
|
| 848 |
+
"loss": 0.0,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.6200448765893793,
|
| 853 |
+
"grad_norm": NaN,
|
| 854 |
+
"learning_rate": 0.0001419300280598173,
|
| 855 |
+
"loss": 0.0,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.649962602842184,
|
| 860 |
+
"grad_norm": NaN,
|
| 861 |
+
"learning_rate": 0.00014107426747033075,
|
| 862 |
+
"loss": 0.0,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.679880329094989,
|
| 867 |
+
"grad_norm": NaN,
|
| 868 |
+
"learning_rate": 0.0001402148729769554,
|
| 869 |
+
"loss": 0.0,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.7097980553477936,
|
| 874 |
+
"grad_norm": NaN,
|
| 875 |
+
"learning_rate": 0.00013935192061164956,
|
| 876 |
+
"loss": 0.0,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.7397157816005984,
|
| 881 |
+
"grad_norm": NaN,
|
| 882 |
+
"learning_rate": 0.00013848548672114197,
|
| 883 |
+
"loss": 0.0,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.769633507853403,
|
| 888 |
+
"grad_norm": NaN,
|
| 889 |
+
"learning_rate": 0.00013761564796017724,
|
| 890 |
+
"loss": 0.0,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.799551234106208,
|
| 895 |
+
"grad_norm": NaN,
|
| 896 |
+
"learning_rate": 0.00013674248128473418,
|
| 897 |
+
"loss": 0.0,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.8294689603590126,
|
| 902 |
+
"grad_norm": NaN,
|
| 903 |
+
"learning_rate": 0.0001358660639452173,
|
| 904 |
+
"loss": 0.0,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.8593866866118174,
|
| 909 |
+
"grad_norm": NaN,
|
| 910 |
+
"learning_rate": 0.0001349864734796224,
|
| 911 |
+
"loss": 0.0,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.889304412864622,
|
| 916 |
+
"grad_norm": NaN,
|
| 917 |
+
"learning_rate": 0.0001341037877066766,
|
| 918 |
+
"loss": 0.0,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.9192221391174273,
|
| 923 |
+
"grad_norm": NaN,
|
| 924 |
+
"learning_rate": 0.00013321808471895375,
|
| 925 |
+
"loss": 0.0,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.949139865370232,
|
| 930 |
+
"grad_norm": NaN,
|
| 931 |
+
"learning_rate": 0.00013232944287596522,
|
| 932 |
+
"loss": 0.0,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.979057591623037,
|
| 937 |
+
"grad_norm": NaN,
|
| 938 |
+
"learning_rate": 0.00013143794079722754,
|
| 939 |
+
"loss": 0.0,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.008975317875842,
|
| 944 |
+
"grad_norm": NaN,
|
| 945 |
+
"learning_rate": 0.00013054365735530664,
|
| 946 |
+
"loss": 0.0,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.038893044128646,
|
| 951 |
+
"grad_norm": NaN,
|
| 952 |
+
"learning_rate": 0.00012964667166884004,
|
| 953 |
+
"loss": 0.0,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.068810770381451,
|
| 958 |
+
"grad_norm": NaN,
|
| 959 |
+
"learning_rate": 0.00012874706309553698,
|
| 960 |
+
"loss": 0.0,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.098728496634256,
|
| 965 |
+
"grad_norm": NaN,
|
| 966 |
+
"learning_rate": 0.00012784491122515743,
|
| 967 |
+
"loss": 0.0,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.128646222887061,
|
| 972 |
+
"grad_norm": NaN,
|
| 973 |
+
"learning_rate": 0.00012694029587247095,
|
| 974 |
+
"loss": 0.0,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.158563949139865,
|
| 979 |
+
"grad_norm": NaN,
|
| 980 |
+
"learning_rate": 0.00012603329707019508,
|
| 981 |
+
"loss": 0.0,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.18848167539267,
|
| 986 |
+
"grad_norm": NaN,
|
| 987 |
+
"learning_rate": 0.0001251239950619149,
|
| 988 |
+
"loss": 0.0,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.218399401645475,
|
| 993 |
+
"grad_norm": NaN,
|
| 994 |
+
"learning_rate": 0.0001242124702949836,
|
| 995 |
+
"loss": 0.0,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.24831712789828,
|
| 1000 |
+
"grad_norm": NaN,
|
| 1001 |
+
"learning_rate": 0.0001232988034134053,
|
| 1002 |
+
"loss": 0.0,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.278234854151084,
|
| 1007 |
+
"grad_norm": NaN,
|
| 1008 |
+
"learning_rate": 0.00012238307525070042,
|
| 1009 |
+
"loss": 0.0,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.308152580403889,
|
| 1014 |
+
"grad_norm": NaN,
|
| 1015 |
+
"learning_rate": 0.00012146536682275387,
|
| 1016 |
+
"loss": 0.0,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.338070306656694,
|
| 1021 |
+
"grad_norm": NaN,
|
| 1022 |
+
"learning_rate": 0.00012054575932064787,
|
| 1023 |
+
"loss": 0.0,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.367988032909499,
|
| 1028 |
+
"grad_norm": NaN,
|
| 1029 |
+
"learning_rate": 0.0001196243341034786,
|
| 1030 |
+
"loss": 0.0,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.397905759162303,
|
| 1035 |
+
"grad_norm": NaN,
|
| 1036 |
+
"learning_rate": 0.00011870117269115833,
|
| 1037 |
+
"loss": 0.0,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.427823485415108,
|
| 1042 |
+
"grad_norm": NaN,
|
| 1043 |
+
"learning_rate": 0.00011777635675720313,
|
| 1044 |
+
"loss": 0.0,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.457741211667913,
|
| 1049 |
+
"grad_norm": NaN,
|
| 1050 |
+
"learning_rate": 0.00011684996812150726,
|
| 1051 |
+
"loss": 0.0,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.487658937920718,
|
| 1056 |
+
"grad_norm": NaN,
|
| 1057 |
+
"learning_rate": 0.00011592208874310426,
|
| 1058 |
+
"loss": 0.0,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.517576664173522,
|
| 1063 |
+
"grad_norm": NaN,
|
| 1064 |
+
"learning_rate": 0.00011499280071291596,
|
| 1065 |
+
"loss": 0.0,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.547494390426328,
|
| 1070 |
+
"grad_norm": NaN,
|
| 1071 |
+
"learning_rate": 0.00011406218624648985,
|
| 1072 |
+
"loss": 0.0,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.577412116679133,
|
| 1077 |
+
"grad_norm": NaN,
|
| 1078 |
+
"learning_rate": 0.00011313032767672524,
|
| 1079 |
+
"loss": 0.0,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.607329842931938,
|
| 1084 |
+
"grad_norm": NaN,
|
| 1085 |
+
"learning_rate": 0.00011219730744658921,
|
| 1086 |
+
"loss": 0.0,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.637247569184742,
|
| 1091 |
+
"grad_norm": NaN,
|
| 1092 |
+
"learning_rate": 0.00011126320810182272,
|
| 1093 |
+
"loss": 0.0,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.667165295437547,
|
| 1098 |
+
"grad_norm": NaN,
|
| 1099 |
+
"learning_rate": 0.00011032811228363766,
|
| 1100 |
+
"loss": 0.0,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.697083021690352,
|
| 1105 |
+
"grad_norm": NaN,
|
| 1106 |
+
"learning_rate": 0.00010939210272140545,
|
| 1107 |
+
"loss": 0.0,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.727000747943157,
|
| 1112 |
+
"grad_norm": NaN,
|
| 1113 |
+
"learning_rate": 0.00010845526222533791,
|
| 1114 |
+
"loss": 0.0,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.756918474195961,
|
| 1119 |
+
"grad_norm": NaN,
|
| 1120 |
+
"learning_rate": 0.00010751767367916082,
|
| 1121 |
+
"loss": 0.0,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.786836200448766,
|
| 1126 |
+
"grad_norm": NaN,
|
| 1127 |
+
"learning_rate": 0.00010657942003278107,
|
| 1128 |
+
"loss": 0.0,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.816753926701571,
|
| 1133 |
+
"grad_norm": NaN,
|
| 1134 |
+
"learning_rate": 0.00010564058429494808,
|
| 1135 |
+
"loss": 0.0,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.846671652954376,
|
| 1140 |
+
"grad_norm": NaN,
|
| 1141 |
+
"learning_rate": 0.00010470124952590977,
|
| 1142 |
+
"loss": 0.0,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.87658937920718,
|
| 1147 |
+
"grad_norm": NaN,
|
| 1148 |
+
"learning_rate": 0.00010376149883006406,
|
| 1149 |
+
"loss": 0.0,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.906507105459985,
|
| 1154 |
+
"grad_norm": NaN,
|
| 1155 |
+
"learning_rate": 0.0001028214153486066,
|
| 1156 |
+
"loss": 0.0,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.93642483171279,
|
| 1161 |
+
"grad_norm": NaN,
|
| 1162 |
+
"learning_rate": 0.00010188108225217516,
|
| 1163 |
+
"loss": 0.0,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.966342557965595,
|
| 1168 |
+
"grad_norm": NaN,
|
| 1169 |
+
"learning_rate": 0.00010094058273349125,
|
| 1170 |
+
"loss": 0.0,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.996260284218399,
|
| 1175 |
+
"grad_norm": NaN,
|
| 1176 |
+
"learning_rate": 0.0001,
|
| 1177 |
+
"loss": 0.0,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.026178010471204,
|
| 1182 |
+
"grad_norm": NaN,
|
| 1183 |
+
"learning_rate": 9.90594172665088e-05,
|
| 1184 |
+
"loss": 0.0,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.056095736724009,
|
| 1189 |
+
"grad_norm": NaN,
|
| 1190 |
+
"learning_rate": 9.811891774782484e-05,
|
| 1191 |
+
"loss": 0.0,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.086013462976814,
|
| 1196 |
+
"grad_norm": NaN,
|
| 1197 |
+
"learning_rate": 9.717858465139342e-05,
|
| 1198 |
+
"loss": 0.0,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.115931189229618,
|
| 1203 |
+
"grad_norm": NaN,
|
| 1204 |
+
"learning_rate": 9.623850116993596e-05,
|
| 1205 |
+
"loss": 0.0,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.145848915482423,
|
| 1210 |
+
"grad_norm": NaN,
|
| 1211 |
+
"learning_rate": 9.529875047409027e-05,
|
| 1212 |
+
"loss": 0.0,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.175766641735228,
|
| 1217 |
+
"grad_norm": NaN,
|
| 1218 |
+
"learning_rate": 9.43594157050519e-05,
|
| 1219 |
+
"loss": 0.0,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.205684367988033,
|
| 1224 |
+
"grad_norm": NaN,
|
| 1225 |
+
"learning_rate": 9.342057996721894e-05,
|
| 1226 |
+
"loss": 0.0,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.2356020942408374,
|
| 1231 |
+
"grad_norm": NaN,
|
| 1232 |
+
"learning_rate": 9.248232632083923e-05,
|
| 1233 |
+
"loss": 0.0,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.265519820493642,
|
| 1238 |
+
"grad_norm": NaN,
|
| 1239 |
+
"learning_rate": 9.15447377746621e-05,
|
| 1240 |
+
"loss": 0.0,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.295437546746447,
|
| 1245 |
+
"grad_norm": NaN,
|
| 1246 |
+
"learning_rate": 9.060789727859457e-05,
|
| 1247 |
+
"loss": 0.0,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.325355272999252,
|
| 1252 |
+
"grad_norm": NaN,
|
| 1253 |
+
"learning_rate": 8.967188771636236e-05,
|
| 1254 |
+
"loss": 0.0,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.3552729992520565,
|
| 1259 |
+
"grad_norm": NaN,
|
| 1260 |
+
"learning_rate": 8.873679189817732e-05,
|
| 1261 |
+
"loss": 0.0,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.385190725504861,
|
| 1266 |
+
"grad_norm": NaN,
|
| 1267 |
+
"learning_rate": 8.78026925534108e-05,
|
| 1268 |
+
"loss": 0.0,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.415108451757666,
|
| 1273 |
+
"grad_norm": NaN,
|
| 1274 |
+
"learning_rate": 8.686967232327478e-05,
|
| 1275 |
+
"loss": 0.0,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.445026178010472,
|
| 1280 |
+
"grad_norm": NaN,
|
| 1281 |
+
"learning_rate": 8.59378137535102e-05,
|
| 1282 |
+
"loss": 0.0,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.474943904263276,
|
| 1287 |
+
"grad_norm": NaN,
|
| 1288 |
+
"learning_rate": 8.500719928708405e-05,
|
| 1289 |
+
"loss": 0.0,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.504861630516081,
|
| 1294 |
+
"grad_norm": NaN,
|
| 1295 |
+
"learning_rate": 8.407791125689578e-05,
|
| 1296 |
+
"loss": 0.0,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.534779356768886,
|
| 1301 |
+
"grad_norm": NaN,
|
| 1302 |
+
"learning_rate": 8.315003187849277e-05,
|
| 1303 |
+
"loss": 0.0,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.564697083021691,
|
| 1308 |
+
"grad_norm": NaN,
|
| 1309 |
+
"learning_rate": 8.222364324279689e-05,
|
| 1310 |
+
"loss": 0.0,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.594614809274495,
|
| 1315 |
+
"grad_norm": NaN,
|
| 1316 |
+
"learning_rate": 8.129882730884168e-05,
|
| 1317 |
+
"loss": 0.0,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.6245325355273,
|
| 1322 |
+
"grad_norm": NaN,
|
| 1323 |
+
"learning_rate": 8.037566589652141e-05,
|
| 1324 |
+
"loss": 0.0,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.654450261780105,
|
| 1329 |
+
"grad_norm": NaN,
|
| 1330 |
+
"learning_rate": 7.945424067935217e-05,
|
| 1331 |
+
"loss": 0.0,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.68436798803291,
|
| 1336 |
+
"grad_norm": NaN,
|
| 1337 |
+
"learning_rate": 7.853463317724614e-05,
|
| 1338 |
+
"loss": 0.0,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.714285714285714,
|
| 1343 |
+
"grad_norm": NaN,
|
| 1344 |
+
"learning_rate": 7.761692474929961e-05,
|
| 1345 |
+
"loss": 0.0,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.744203440538519,
|
| 1350 |
+
"grad_norm": NaN,
|
| 1351 |
+
"learning_rate": 7.67011965865947e-05,
|
| 1352 |
+
"loss": 0.0,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.774121166791324,
|
| 1357 |
+
"grad_norm": NaN,
|
| 1358 |
+
"learning_rate": 7.578752970501642e-05,
|
| 1359 |
+
"loss": 0.0,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.804038893044129,
|
| 1364 |
+
"grad_norm": NaN,
|
| 1365 |
+
"learning_rate": 7.487600493808513e-05,
|
| 1366 |
+
"loss": 0.0,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.833956619296933,
|
| 1371 |
+
"grad_norm": NaN,
|
| 1372 |
+
"learning_rate": 7.396670292980492e-05,
|
| 1373 |
+
"loss": 0.0,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.863874345549738,
|
| 1378 |
+
"grad_norm": NaN,
|
| 1379 |
+
"learning_rate": 7.305970412752909e-05,
|
| 1380 |
+
"loss": 0.0,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.893792071802543,
|
| 1385 |
+
"grad_norm": NaN,
|
| 1386 |
+
"learning_rate": 7.215508877484258e-05,
|
| 1387 |
+
"loss": 0.0,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.923709798055348,
|
| 1392 |
+
"grad_norm": NaN,
|
| 1393 |
+
"learning_rate": 7.125293690446306e-05,
|
| 1394 |
+
"loss": 0.0,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.9536275243081525,
|
| 1399 |
+
"grad_norm": NaN,
|
| 1400 |
+
"learning_rate": 7.035332833115997e-05,
|
| 1401 |
+
"loss": 0.0,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.983545250560957,
|
| 1406 |
+
"grad_norm": NaN,
|
| 1407 |
+
"learning_rate": 6.945634264469339e-05,
|
| 1408 |
+
"loss": 0.0,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.013462976813762,
|
| 1413 |
+
"grad_norm": NaN,
|
| 1414 |
+
"learning_rate": 6.85620592027725e-05,
|
| 1415 |
+
"loss": 0.0,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.043380703066567,
|
| 1420 |
+
"grad_norm": NaN,
|
| 1421 |
+
"learning_rate": 6.76705571240348e-05,
|
| 1422 |
+
"loss": 0.0,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.0732984293193715,
|
| 1427 |
+
"grad_norm": NaN,
|
| 1428 |
+
"learning_rate": 6.678191528104629e-05,
|
| 1429 |
+
"loss": 0.0,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.103216155572176,
|
| 1434 |
+
"grad_norm": NaN,
|
| 1435 |
+
"learning_rate": 6.58962122933234e-05,
|
| 1436 |
+
"loss": 0.0,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.133133881824981,
|
| 1441 |
+
"grad_norm": NaN,
|
| 1442 |
+
"learning_rate": 6.501352652037764e-05,
|
| 1443 |
+
"loss": 0.0,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.163051608077786,
|
| 1448 |
+
"grad_norm": NaN,
|
| 1449 |
+
"learning_rate": 6.413393605478275e-05,
|
| 1450 |
+
"loss": 0.0,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.1929693343305905,
|
| 1455 |
+
"grad_norm": NaN,
|
| 1456 |
+
"learning_rate": 6.325751871526588e-05,
|
| 1457 |
+
"loss": 0.0,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.222887060583395,
|
| 1462 |
+
"grad_norm": NaN,
|
| 1463 |
+
"learning_rate": 6.238435203982278e-05,
|
| 1464 |
+
"loss": 0.0,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.252804786836201,
|
| 1469 |
+
"grad_norm": NaN,
|
| 1470 |
+
"learning_rate": 6.151451327885805e-05,
|
| 1471 |
+
"loss": 0.0,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.282722513089006,
|
| 1476 |
+
"grad_norm": NaN,
|
| 1477 |
+
"learning_rate": 6.0648079388350466e-05,
|
| 1478 |
+
"loss": 0.0,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.31264023934181,
|
| 1483 |
+
"grad_norm": NaN,
|
| 1484 |
+
"learning_rate": 5.978512702304461e-05,
|
| 1485 |
+
"loss": 0.0,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.342557965594615,
|
| 1490 |
+
"grad_norm": NaN,
|
| 1491 |
+
"learning_rate": 5.892573252966926e-05,
|
| 1492 |
+
"loss": 0.0,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.37247569184742,
|
| 1497 |
+
"grad_norm": NaN,
|
| 1498 |
+
"learning_rate": 5.806997194018271e-05,
|
| 1499 |
+
"loss": 0.0,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.402393418100225,
|
| 1504 |
+
"grad_norm": NaN,
|
| 1505 |
+
"learning_rate": 5.721792096504611e-05,
|
| 1506 |
+
"loss": 0.0,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.432311144353029,
|
| 1511 |
+
"grad_norm": NaN,
|
| 1512 |
+
"learning_rate": 5.636965498652544e-05,
|
| 1513 |
+
"loss": 0.0,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.462228870605834,
|
| 1518 |
+
"grad_norm": NaN,
|
| 1519 |
+
"learning_rate": 5.5525249052022076e-05,
|
| 1520 |
+
"loss": 0.0,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.492146596858639,
|
| 1525 |
+
"grad_norm": NaN,
|
| 1526 |
+
"learning_rate": 5.4684777867433335e-05,
|
| 1527 |
+
"loss": 0.0,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.522064323111444,
|
| 1532 |
+
"grad_norm": NaN,
|
| 1533 |
+
"learning_rate": 5.3848315790543126e-05,
|
| 1534 |
+
"loss": 0.0,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.551982049364248,
|
| 1539 |
+
"grad_norm": NaN,
|
| 1540 |
+
"learning_rate": 5.301593682444352e-05,
|
| 1541 |
+
"loss": 0.0,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.581899775617053,
|
| 1546 |
+
"grad_norm": NaN,
|
| 1547 |
+
"learning_rate": 5.218771461098733e-05,
|
| 1548 |
+
"loss": 0.0,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.611817501869858,
|
| 1553 |
+
"grad_norm": NaN,
|
| 1554 |
+
"learning_rate": 5.1363722424273156e-05,
|
| 1555 |
+
"loss": 0.0,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.641735228122663,
|
| 1560 |
+
"grad_norm": NaN,
|
| 1561 |
+
"learning_rate": 5.054403316416247e-05,
|
| 1562 |
+
"loss": 0.0,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.6716529543754675,
|
| 1567 |
+
"grad_norm": NaN,
|
| 1568 |
+
"learning_rate": 4.972871934983032e-05,
|
| 1569 |
+
"loss": 0.0,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.701570680628272,
|
| 1574 |
+
"grad_norm": NaN,
|
| 1575 |
+
"learning_rate": 4.891785311334923e-05,
|
| 1576 |
+
"loss": 0.0,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.731488406881077,
|
| 1581 |
+
"grad_norm": NaN,
|
| 1582 |
+
"learning_rate": 4.811150619330759e-05,
|
| 1583 |
+
"loss": 0.0,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.761406133133882,
|
| 1588 |
+
"grad_norm": NaN,
|
| 1589 |
+
"learning_rate": 4.7309749928463035e-05,
|
| 1590 |
+
"loss": 0.0,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.7913238593866865,
|
| 1595 |
+
"grad_norm": NaN,
|
| 1596 |
+
"learning_rate": 4.6512655251430695e-05,
|
| 1597 |
+
"loss": 0.0,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.821241585639491,
|
| 1602 |
+
"grad_norm": NaN,
|
| 1603 |
+
"learning_rate": 4.5720292682407874e-05,
|
| 1604 |
+
"loss": 0.0,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.851159311892296,
|
| 1609 |
+
"grad_norm": NaN,
|
| 1610 |
+
"learning_rate": 4.493273232293506e-05,
|
| 1611 |
+
"loss": 0.0,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.881077038145101,
|
| 1616 |
+
"grad_norm": NaN,
|
| 1617 |
+
"learning_rate": 4.41500438496937e-05,
|
| 1618 |
+
"loss": 0.0,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 6.9109947643979055,
|
| 1623 |
+
"grad_norm": NaN,
|
| 1624 |
+
"learning_rate": 4.3372296508342026e-05,
|
| 1625 |
+
"loss": 0.0,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 6.94091249065071,
|
| 1630 |
+
"grad_norm": NaN,
|
| 1631 |
+
"learning_rate": 4.2599559107388645e-05,
|
| 1632 |
+
"loss": 0.0,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 6.970830216903515,
|
| 1637 |
+
"grad_norm": NaN,
|
| 1638 |
+
"learning_rate": 4.1831900012105105e-05,
|
| 1639 |
+
"loss": 0.0,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.00074794315632,
|
| 1644 |
+
"grad_norm": NaN,
|
| 1645 |
+
"learning_rate": 4.10693871384773e-05,
|
| 1646 |
+
"loss": 0.0,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.0306656694091245,
|
| 1651 |
+
"grad_norm": NaN,
|
| 1652 |
+
"learning_rate": 4.031208794719695e-05,
|
| 1653 |
+
"loss": 0.0,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.060583395661929,
|
| 1658 |
+
"grad_norm": NaN,
|
| 1659 |
+
"learning_rate": 3.95600694376933e-05,
|
| 1660 |
+
"loss": 0.0,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.090501121914735,
|
| 1665 |
+
"grad_norm": NaN,
|
| 1666 |
+
"learning_rate": 3.881339814220555e-05,
|
| 1667 |
+
"loss": 0.0,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.12041884816754,
|
| 1672 |
+
"grad_norm": NaN,
|
| 1673 |
+
"learning_rate": 3.8072140119896504e-05,
|
| 1674 |
+
"loss": 0.0,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.150336574420344,
|
| 1679 |
+
"grad_norm": NaN,
|
| 1680 |
+
"learning_rate": 3.733636095100838e-05,
|
| 1681 |
+
"loss": 0.0,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.180254300673149,
|
| 1686 |
+
"grad_norm": NaN,
|
| 1687 |
+
"learning_rate": 3.660612573106081e-05,
|
| 1688 |
+
"loss": 0.0,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.210172026925954,
|
| 1693 |
+
"grad_norm": NaN,
|
| 1694 |
+
"learning_rate": 3.588149906509166e-05,
|
| 1695 |
+
"loss": 0.0,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.240089753178759,
|
| 1700 |
+
"grad_norm": NaN,
|
| 1701 |
+
"learning_rate": 3.5162545061941335e-05,
|
| 1702 |
+
"loss": 0.0,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.2700074794315634,
|
| 1707 |
+
"grad_norm": NaN,
|
| 1708 |
+
"learning_rate": 3.444932732858107e-05,
|
| 1709 |
+
"loss": 0.0,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.299925205684368,
|
| 1714 |
+
"grad_norm": NaN,
|
| 1715 |
+
"learning_rate": 3.3741908964485414e-05,
|
| 1716 |
+
"loss": 0.0,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.329842931937173,
|
| 1721 |
+
"grad_norm": NaN,
|
| 1722 |
+
"learning_rate": 3.304035255604974e-05,
|
| 1723 |
+
"loss": 0.0,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.359760658189978,
|
| 1728 |
+
"grad_norm": NaN,
|
| 1729 |
+
"learning_rate": 3.234472017105313e-05,
|
| 1730 |
+
"loss": 0.0,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.3896783844427825,
|
| 1735 |
+
"grad_norm": NaN,
|
| 1736 |
+
"learning_rate": 3.1655073353167285e-05,
|
| 1737 |
+
"loss": 0.0,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.419596110695587,
|
| 1742 |
+
"grad_norm": NaN,
|
| 1743 |
+
"learning_rate": 3.0971473116511394e-05,
|
| 1744 |
+
"loss": 0.0,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.449513836948392,
|
| 1749 |
+
"grad_norm": NaN,
|
| 1750 |
+
"learning_rate": 3.0293979940254314e-05,
|
| 1751 |
+
"loss": 0.0,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.479431563201197,
|
| 1756 |
+
"grad_norm": NaN,
|
| 1757 |
+
"learning_rate": 2.9622653763263874e-05,
|
| 1758 |
+
"loss": 0.0,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.5093492894540015,
|
| 1763 |
+
"grad_norm": NaN,
|
| 1764 |
+
"learning_rate": 2.8957553978803898e-05,
|
| 1765 |
+
"loss": 0.0,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.539267015706806,
|
| 1770 |
+
"grad_norm": NaN,
|
| 1771 |
+
"learning_rate": 2.8298739429279707e-05,
|
| 1772 |
+
"loss": 0.0,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.569184741959611,
|
| 1777 |
+
"grad_norm": NaN,
|
| 1778 |
+
"learning_rate": 2.764626840103208e-05,
|
| 1779 |
+
"loss": 0.0,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.599102468212416,
|
| 1784 |
+
"grad_norm": NaN,
|
| 1785 |
+
"learning_rate": 2.7000198619180794e-05,
|
| 1786 |
+
"loss": 0.0,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.6290201944652205,
|
| 1791 |
+
"grad_norm": NaN,
|
| 1792 |
+
"learning_rate": 2.636058724251739e-05,
|
| 1793 |
+
"loss": 0.0,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.658937920718025,
|
| 1798 |
+
"grad_norm": NaN,
|
| 1799 |
+
"learning_rate": 2.5727490858448288e-05,
|
| 1800 |
+
"loss": 0.0,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.68885564697083,
|
| 1805 |
+
"grad_norm": NaN,
|
| 1806 |
+
"learning_rate": 2.510096547798857e-05,
|
| 1807 |
+
"loss": 0.0,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.718773373223635,
|
| 1812 |
+
"grad_norm": NaN,
|
| 1813 |
+
"learning_rate": 2.4481066530806395e-05,
|
| 1814 |
+
"loss": 0.0,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.7486910994764395,
|
| 1819 |
+
"grad_norm": NaN,
|
| 1820 |
+
"learning_rate": 2.3867848860319144e-05,
|
| 1821 |
+
"loss": 0.0,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.778608825729244,
|
| 1826 |
+
"grad_norm": NaN,
|
| 1827 |
+
"learning_rate": 2.3261366718841305e-05,
|
| 1828 |
+
"loss": 0.0,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 7.808526551982049,
|
| 1833 |
+
"grad_norm": NaN,
|
| 1834 |
+
"learning_rate": 2.2661673762784807e-05,
|
| 1835 |
+
"loss": 0.0,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 7.838444278234854,
|
| 1840 |
+
"grad_norm": NaN,
|
| 1841 |
+
"learning_rate": 2.206882304791176e-05,
|
| 1842 |
+
"loss": 0.0,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 7.868362004487659,
|
| 1847 |
+
"grad_norm": NaN,
|
| 1848 |
+
"learning_rate": 2.148286702464063e-05,
|
| 1849 |
+
"loss": 0.0,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 7.898279730740464,
|
| 1854 |
+
"grad_norm": NaN,
|
| 1855 |
+
"learning_rate": 2.0903857533405958e-05,
|
| 1856 |
+
"loss": 0.0,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 7.928197456993269,
|
| 1861 |
+
"grad_norm": NaN,
|
| 1862 |
+
"learning_rate": 2.033184580007177e-05,
|
| 1863 |
+
"loss": 0.0,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 7.958115183246074,
|
| 1868 |
+
"grad_norm": NaN,
|
| 1869 |
+
"learning_rate": 1.9766882431399646e-05,
|
| 1870 |
+
"loss": 0.0,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 7.9880329094988785,
|
| 1875 |
+
"grad_norm": NaN,
|
| 1876 |
+
"learning_rate": 1.920901741057153e-05,
|
| 1877 |
+
"loss": 0.0,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.017950635751683,
|
| 1882 |
+
"grad_norm": NaN,
|
| 1883 |
+
"learning_rate": 1.8658300092767544e-05,
|
| 1884 |
+
"loss": 0.0,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.047868362004488,
|
| 1889 |
+
"grad_norm": NaN,
|
| 1890 |
+
"learning_rate": 1.8114779200799437e-05,
|
| 1891 |
+
"loss": 0.0,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.077786088257293,
|
| 1896 |
+
"grad_norm": NaN,
|
| 1897 |
+
"learning_rate": 1.7578502820800045e-05,
|
| 1898 |
+
"loss": 0.0,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.107703814510097,
|
| 1903 |
+
"grad_norm": NaN,
|
| 1904 |
+
"learning_rate": 1.7049518397969132e-05,
|
| 1905 |
+
"loss": 0.0,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.137621540762902,
|
| 1910 |
+
"grad_norm": NaN,
|
| 1911 |
+
"learning_rate": 1.652787273237565e-05,
|
| 1912 |
+
"loss": 0.0,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.167539267015707,
|
| 1917 |
+
"grad_norm": NaN,
|
| 1918 |
+
"learning_rate": 1.601361197481741e-05,
|
| 1919 |
+
"loss": 0.0,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.197456993268512,
|
| 1924 |
+
"grad_norm": NaN,
|
| 1925 |
+
"learning_rate": 1.5506781622737942e-05,
|
| 1926 |
+
"loss": 0.0,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.227374719521316,
|
| 1931 |
+
"grad_norm": NaN,
|
| 1932 |
+
"learning_rate": 1.500742651620145e-05,
|
| 1933 |
+
"loss": 0.0,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.257292445774121,
|
| 1938 |
+
"grad_norm": NaN,
|
| 1939 |
+
"learning_rate": 1.4515590833925507e-05,
|
| 1940 |
+
"loss": 0.0,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.287210172026926,
|
| 1945 |
+
"grad_norm": NaN,
|
| 1946 |
+
"learning_rate": 1.4031318089372615e-05,
|
| 1947 |
+
"loss": 0.0,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.31712789827973,
|
| 1952 |
+
"grad_norm": NaN,
|
| 1953 |
+
"learning_rate": 1.3554651126900564e-05,
|
| 1954 |
+
"loss": 0.0,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.347045624532536,
|
| 1959 |
+
"grad_norm": NaN,
|
| 1960 |
+
"learning_rate": 1.308563211797179e-05,
|
| 1961 |
+
"loss": 0.0,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.37696335078534,
|
| 1966 |
+
"grad_norm": NaN,
|
| 1967 |
+
"learning_rate": 1.2624302557422473e-05,
|
| 1968 |
+
"loss": 0.0,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.406881077038145,
|
| 1973 |
+
"grad_norm": NaN,
|
| 1974 |
+
"learning_rate": 1.2170703259791471e-05,
|
| 1975 |
+
"loss": 0.0,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.43679880329095,
|
| 1980 |
+
"grad_norm": NaN,
|
| 1981 |
+
"learning_rate": 1.172487435570926e-05,
|
| 1982 |
+
"loss": 0.0,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.466716529543755,
|
| 1987 |
+
"grad_norm": NaN,
|
| 1988 |
+
"learning_rate": 1.1286855288347675e-05,
|
| 1989 |
+
"loss": 0.0,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.49663425579656,
|
| 1994 |
+
"grad_norm": NaN,
|
| 1995 |
+
"learning_rate": 1.0856684809930151e-05,
|
| 1996 |
+
"loss": 0.0,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.526551982049364,
|
| 2001 |
+
"grad_norm": NaN,
|
| 2002 |
+
"learning_rate": 1.0434400978303426e-05,
|
| 2003 |
+
"loss": 0.0,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.556469708302169,
|
| 2008 |
+
"grad_norm": NaN,
|
| 2009 |
+
"learning_rate": 1.0020041153570347e-05,
|
| 2010 |
+
"loss": 0.0,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.586387434554974,
|
| 2015 |
+
"grad_norm": NaN,
|
| 2016 |
+
"learning_rate": 9.613641994784651e-06,
|
| 2017 |
+
"loss": 0.0,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.616305160807778,
|
| 2022 |
+
"grad_norm": NaN,
|
| 2023 |
+
"learning_rate": 9.215239456707635e-06,
|
| 2024 |
+
"loss": 0.0,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 8.646222887060583,
|
| 2029 |
+
"grad_norm": NaN,
|
| 2030 |
+
"learning_rate": 8.824868786627304e-06,
|
| 2031 |
+
"loss": 0.0,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 8.676140613313388,
|
| 2036 |
+
"grad_norm": NaN,
|
| 2037 |
+
"learning_rate": 8.442564521239782e-06,
|
| 2038 |
+
"loss": 0.0,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 8.706058339566193,
|
| 2043 |
+
"grad_norm": NaN,
|
| 2044 |
+
"learning_rate": 8.06836048359394e-06,
|
| 2045 |
+
"loss": 0.0,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.735976065818997,
|
| 2050 |
+
"grad_norm": NaN,
|
| 2051 |
+
"learning_rate": 7.70228978009907e-06,
|
| 2052 |
+
"loss": 0.0,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 8.765893792071802,
|
| 2057 |
+
"grad_norm": NaN,
|
| 2058 |
+
"learning_rate": 7.344384797595716e-06,
|
| 2059 |
+
"loss": 0.0,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 8.795811518324607,
|
| 2064 |
+
"grad_norm": NaN,
|
| 2065 |
+
"learning_rate": 6.994677200490507e-06,
|
| 2066 |
+
"loss": 0.0,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 8.825729244577412,
|
| 2071 |
+
"grad_norm": NaN,
|
| 2072 |
+
"learning_rate": 6.653197927954791e-06,
|
| 2073 |
+
"loss": 0.0,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 8.855646970830216,
|
| 2078 |
+
"grad_norm": NaN,
|
| 2079 |
+
"learning_rate": 6.319977191187232e-06,
|
| 2080 |
+
"loss": 0.0,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 8.885564697083021,
|
| 2085 |
+
"grad_norm": NaN,
|
| 2086 |
+
"learning_rate": 5.995044470741151e-06,
|
| 2087 |
+
"loss": 0.0,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 8.915482423335826,
|
| 2092 |
+
"grad_norm": NaN,
|
| 2093 |
+
"learning_rate": 5.678428513916212e-06,
|
| 2094 |
+
"loss": 0.0,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 8.94540014958863,
|
| 2099 |
+
"grad_norm": NaN,
|
| 2100 |
+
"learning_rate": 5.370157332215209e-06,
|
| 2101 |
+
"loss": 0.0,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 8.975317875841435,
|
| 2106 |
+
"grad_norm": NaN,
|
| 2107 |
+
"learning_rate": 5.07025819886574e-06,
|
| 2108 |
+
"loss": 0.0,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.00523560209424,
|
| 2113 |
+
"grad_norm": NaN,
|
| 2114 |
+
"learning_rate": 4.778757646407362e-06,
|
| 2115 |
+
"loss": 0.0,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.035153328347045,
|
| 2120 |
+
"grad_norm": NaN,
|
| 2121 |
+
"learning_rate": 4.495681464344259e-06,
|
| 2122 |
+
"loss": 0.0,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.06507105459985,
|
| 2127 |
+
"grad_norm": NaN,
|
| 2128 |
+
"learning_rate": 4.22105469686348e-06,
|
| 2129 |
+
"loss": 0.0,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.094988780852654,
|
| 2134 |
+
"grad_norm": NaN,
|
| 2135 |
+
"learning_rate": 3.954901640619368e-06,
|
| 2136 |
+
"loss": 0.0,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.12490650710546,
|
| 2141 |
+
"grad_norm": NaN,
|
| 2142 |
+
"learning_rate": 3.6972458425838962e-06,
|
| 2143 |
+
"loss": 0.0,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.154824233358266,
|
| 2148 |
+
"grad_norm": NaN,
|
| 2149 |
+
"learning_rate": 3.4481100979635306e-06,
|
| 2150 |
+
"loss": 0.0,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.18474195961107,
|
| 2155 |
+
"grad_norm": NaN,
|
| 2156 |
+
"learning_rate": 3.207516448182435e-06,
|
| 2157 |
+
"loss": 0.0,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 9.214659685863875,
|
| 2162 |
+
"grad_norm": NaN,
|
| 2163 |
+
"learning_rate": 2.9754861789324073e-06,
|
| 2164 |
+
"loss": 0.0,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 9.24457741211668,
|
| 2169 |
+
"grad_norm": NaN,
|
| 2170 |
+
"learning_rate": 2.752039818289809e-06,
|
| 2171 |
+
"loss": 0.0,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 9.274495138369485,
|
| 2176 |
+
"grad_norm": NaN,
|
| 2177 |
+
"learning_rate": 2.537197134899294e-06,
|
| 2178 |
+
"loss": 0.0,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 9.30441286462229,
|
| 2183 |
+
"grad_norm": NaN,
|
| 2184 |
+
"learning_rate": 2.330977136224932e-06,
|
| 2185 |
+
"loss": 0.0,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.334330590875094,
|
| 2190 |
+
"grad_norm": NaN,
|
| 2191 |
+
"learning_rate": 2.1333980668685414e-06,
|
| 2192 |
+
"loss": 0.0,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 9.364248317127899,
|
| 2197 |
+
"grad_norm": NaN,
|
| 2198 |
+
"learning_rate": 1.9444774069555694e-06,
|
| 2199 |
+
"loss": 0.0,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 9.394166043380704,
|
| 2204 |
+
"grad_norm": NaN,
|
| 2205 |
+
"learning_rate": 1.7642318705886286e-06,
|
| 2206 |
+
"loss": 0.0,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 9.424083769633508,
|
| 2211 |
+
"grad_norm": NaN,
|
| 2212 |
+
"learning_rate": 1.5926774043687365e-06,
|
| 2213 |
+
"loss": 0.0,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 9.454001495886313,
|
| 2218 |
+
"grad_norm": NaN,
|
| 2219 |
+
"learning_rate": 1.4298291859845214e-06,
|
| 2220 |
+
"loss": 0.0,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 9.483919222139118,
|
| 2225 |
+
"grad_norm": NaN,
|
| 2226 |
+
"learning_rate": 1.2757016228693964e-06,
|
| 2227 |
+
"loss": 0.0,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.513836948391923,
|
| 2232 |
+
"grad_norm": NaN,
|
| 2233 |
+
"learning_rate": 1.1303083509269452e-06,
|
| 2234 |
+
"loss": 0.0,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 9.543754674644727,
|
| 2239 |
+
"grad_norm": NaN,
|
| 2240 |
+
"learning_rate": 9.936622333245104e-07,
|
| 2241 |
+
"loss": 0.0,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 9.573672400897532,
|
| 2246 |
+
"grad_norm": NaN,
|
| 2247 |
+
"learning_rate": 8.657753593552143e-07,
|
| 2248 |
+
"loss": 0.0,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 9.603590127150337,
|
| 2253 |
+
"grad_norm": NaN,
|
| 2254 |
+
"learning_rate": 7.466590433683251e-07,
|
| 2255 |
+
"loss": 0.0,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 9.633507853403142,
|
| 2260 |
+
"grad_norm": NaN,
|
| 2261 |
+
"learning_rate": 6.363238237683033e-07,
|
| 2262 |
+
"loss": 0.0,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 9.663425579655947,
|
| 2267 |
+
"grad_norm": NaN,
|
| 2268 |
+
"learning_rate": 5.347794620824576e-07,
|
| 2269 |
+
"loss": 0.0,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 9.693343305908751,
|
| 2274 |
+
"grad_norm": NaN,
|
| 2275 |
+
"learning_rate": 4.4203494209733576e-07,
|
| 2276 |
+
"loss": 0.0,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 9.723261032161556,
|
| 2281 |
+
"grad_norm": NaN,
|
| 2282 |
+
"learning_rate": 3.580984690638611e-07,
|
| 2283 |
+
"loss": 0.0,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 9.75317875841436,
|
| 2288 |
+
"grad_norm": NaN,
|
| 2289 |
+
"learning_rate": 2.8297746897146816e-07,
|
| 2290 |
+
"loss": 0.0,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 9.783096484667166,
|
| 2295 |
+
"grad_norm": NaN,
|
| 2296 |
+
"learning_rate": 2.1667858789105043e-07,
|
| 2297 |
+
"loss": 0.0,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 9.81301421091997,
|
| 2302 |
+
"grad_norm": NaN,
|
| 2303 |
+
"learning_rate": 1.5920769138706438e-07,
|
| 2304 |
+
"loss": 0.0,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 9.842931937172775,
|
| 2309 |
+
"grad_norm": NaN,
|
| 2310 |
+
"learning_rate": 1.1056986399845537e-07,
|
| 2311 |
+
"loss": 0.0,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 9.87284966342558,
|
| 2316 |
+
"grad_norm": NaN,
|
| 2317 |
+
"learning_rate": 7.07694087889621e-08,
|
| 2318 |
+
"loss": 0.0,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 9.902767389678385,
|
| 2323 |
+
"grad_norm": NaN,
|
| 2324 |
+
"learning_rate": 3.980984696634327e-08,
|
| 2325 |
+
"loss": 0.0,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 9.93268511593119,
|
| 2330 |
+
"grad_norm": NaN,
|
| 2331 |
+
"learning_rate": 1.7693917570837936e-08,
|
| 2332 |
+
"loss": 0.0,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 9.962602842183994,
|
| 2337 |
+
"grad_norm": NaN,
|
| 2338 |
+
"learning_rate": 4.42357723288156e-09,
|
| 2339 |
+
"loss": 0.0,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 9.992520568436799,
|
| 2344 |
+
"grad_norm": NaN,
|
| 2345 |
+
"learning_rate": 0.0,
|
| 2346 |
+
"loss": 0.0,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 9.992520568436799,
|
| 2351 |
+
"step": 3340,
|
| 2352 |
+
"total_flos": 1.6429048267800576e+17,
|
| 2353 |
+
"train_loss": 0.0,
|
| 2354 |
+
"train_runtime": 3248.6622,
|
| 2355 |
+
"train_samples_per_second": 4.116,
|
| 2356 |
+
"train_steps_per_second": 1.028
|
| 2357 |
+
}
|
| 2358 |
+
],
|
| 2359 |
+
"logging_steps": 10,
|
| 2360 |
+
"max_steps": 3340,
|
| 2361 |
+
"num_input_tokens_seen": 0,
|
| 2362 |
+
"num_train_epochs": 10,
|
| 2363 |
+
"save_steps": 500,
|
| 2364 |
+
"total_flos": 1.6429048267800576e+17,
|
| 2365 |
+
"train_batch_size": 1,
|
| 2366 |
+
"trial_name": null,
|
| 2367 |
+
"trial_params": null
|
| 2368 |
+
}
|
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 sha256:2f5cefe777ff3b9b4ebca31da3adce5e545eba944835c44abd83b753168ff016
|
| 3 |
size 5048
|