Training in progress, epoch 0
Browse files- adapter_config.json +4 -4
- trainer_state.json +2340 -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 +1,2340 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.992429977289932,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3300,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03028009084027252,
|
| 13 |
+
"grad_norm": NaN,
|
| 14 |
+
"learning_rate": 0.0001999954685354173,
|
| 15 |
+
"loss": 0.0,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06056018168054504,
|
| 20 |
+
"grad_norm": NaN,
|
| 21 |
+
"learning_rate": 0.0001999818745523526,
|
| 22 |
+
"loss": 0.0,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09084027252081757,
|
| 27 |
+
"grad_norm": NaN,
|
| 28 |
+
"learning_rate": 0.00019995921928281894,
|
| 29 |
+
"loss": 0.0,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12112036336109008,
|
| 34 |
+
"grad_norm": NaN,
|
| 35 |
+
"learning_rate": 0.00019992750478004738,
|
| 36 |
+
"loss": 0.0,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1514004542013626,
|
| 41 |
+
"grad_norm": NaN,
|
| 42 |
+
"learning_rate": 0.0001998867339183008,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18168054504163514,
|
| 48 |
+
"grad_norm": NaN,
|
| 49 |
+
"learning_rate": 0.00019983691039261357,
|
| 50 |
+
"loss": 0.0,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21196063588190764,
|
| 55 |
+
"grad_norm": NaN,
|
| 56 |
+
"learning_rate": 0.0001997780387184565,
|
| 57 |
+
"loss": 0.0,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24224072672218017,
|
| 62 |
+
"grad_norm": NaN,
|
| 63 |
+
"learning_rate": 0.00019971012423132775,
|
| 64 |
+
"loss": 0.0,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.27252081756245267,
|
| 69 |
+
"grad_norm": NaN,
|
| 70 |
+
"learning_rate": 0.00019963317308626914,
|
| 71 |
+
"loss": 0.0,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3028009084027252,
|
| 76 |
+
"grad_norm": NaN,
|
| 77 |
+
"learning_rate": 0.00019954719225730847,
|
| 78 |
+
"loss": 0.0,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3330809992429977,
|
| 83 |
+
"grad_norm": NaN,
|
| 84 |
+
"learning_rate": 0.00019945218953682734,
|
| 85 |
+
"loss": 0.0,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3633610900832703,
|
| 90 |
+
"grad_norm": NaN,
|
| 91 |
+
"learning_rate": 0.00019934817353485501,
|
| 92 |
+
"loss": 0.0,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3936411809235428,
|
| 97 |
+
"grad_norm": NaN,
|
| 98 |
+
"learning_rate": 0.0001992351536782881,
|
| 99 |
+
"loss": 0.0,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4239212717638153,
|
| 104 |
+
"grad_norm": NaN,
|
| 105 |
+
"learning_rate": 0.00019911314021003613,
|
| 106 |
+
"loss": 0.0,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.45420136260408783,
|
| 111 |
+
"grad_norm": NaN,
|
| 112 |
+
"learning_rate": 0.0001989821441880933,
|
| 113 |
+
"loss": 0.0,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.48448145344436033,
|
| 118 |
+
"grad_norm": NaN,
|
| 119 |
+
"learning_rate": 0.00019884217748453623,
|
| 120 |
+
"loss": 0.0,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5147615442846328,
|
| 125 |
+
"grad_norm": NaN,
|
| 126 |
+
"learning_rate": 0.00019869325278444824,
|
| 127 |
+
"loss": 0.0,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5450416351249053,
|
| 132 |
+
"grad_norm": NaN,
|
| 133 |
+
"learning_rate": 0.00019853538358476932,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5753217259651779,
|
| 139 |
+
"grad_norm": NaN,
|
| 140 |
+
"learning_rate": 0.00019836858419307324,
|
| 141 |
+
"loss": 0.0,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6056018168054504,
|
| 146 |
+
"grad_norm": NaN,
|
| 147 |
+
"learning_rate": 0.00019819286972627066,
|
| 148 |
+
"loss": 0.0,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6358819076457229,
|
| 153 |
+
"grad_norm": NaN,
|
| 154 |
+
"learning_rate": 0.00019800825610923934,
|
| 155 |
+
"loss": 0.0,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6661619984859954,
|
| 160 |
+
"grad_norm": NaN,
|
| 161 |
+
"learning_rate": 0.00019781476007338058,
|
| 162 |
+
"loss": 0.0,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6964420893262679,
|
| 167 |
+
"grad_norm": NaN,
|
| 168 |
+
"learning_rate": 0.00019761239915510302,
|
| 169 |
+
"loss": 0.0,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7267221801665406,
|
| 174 |
+
"grad_norm": NaN,
|
| 175 |
+
"learning_rate": 0.00019740119169423337,
|
| 176 |
+
"loss": 0.0,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.757002271006813,
|
| 181 |
+
"grad_norm": NaN,
|
| 182 |
+
"learning_rate": 0.00019718115683235417,
|
| 183 |
+
"loss": 0.0,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7872823618470856,
|
| 188 |
+
"grad_norm": NaN,
|
| 189 |
+
"learning_rate": 0.00019695231451106912,
|
| 190 |
+
"loss": 0.0,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.817562452687358,
|
| 195 |
+
"grad_norm": NaN,
|
| 196 |
+
"learning_rate": 0.00019671468547019573,
|
| 197 |
+
"loss": 0.0,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8478425435276306,
|
| 202 |
+
"grad_norm": NaN,
|
| 203 |
+
"learning_rate": 0.0001964682912458856,
|
| 204 |
+
"loss": 0.0,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.878122634367903,
|
| 209 |
+
"grad_norm": NaN,
|
| 210 |
+
"learning_rate": 0.00019621315416867274,
|
| 211 |
+
"loss": 0.0,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9084027252081757,
|
| 216 |
+
"grad_norm": NaN,
|
| 217 |
+
"learning_rate": 0.00019594929736144976,
|
| 218 |
+
"loss": 0.0,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9386828160484482,
|
| 223 |
+
"grad_norm": NaN,
|
| 224 |
+
"learning_rate": 0.00019567674473737218,
|
| 225 |
+
"loss": 0.0,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9689629068887207,
|
| 230 |
+
"grad_norm": NaN,
|
| 231 |
+
"learning_rate": 0.00019539552099769126,
|
| 232 |
+
"loss": 0.0,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9992429977289932,
|
| 237 |
+
"grad_norm": NaN,
|
| 238 |
+
"learning_rate": 0.00019510565162951537,
|
| 239 |
+
"loss": 0.0,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0295230885692657,
|
| 244 |
+
"grad_norm": NaN,
|
| 245 |
+
"learning_rate": 0.00019480716290349995,
|
| 246 |
+
"loss": 0.0,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0598031794095382,
|
| 251 |
+
"grad_norm": NaN,
|
| 252 |
+
"learning_rate": 0.00019450008187146684,
|
| 253 |
+
"loss": 0.0,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0900832702498107,
|
| 258 |
+
"grad_norm": NaN,
|
| 259 |
+
"learning_rate": 0.00019418443636395248,
|
| 260 |
+
"loss": 0.0,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1203633610900834,
|
| 265 |
+
"grad_norm": NaN,
|
| 266 |
+
"learning_rate": 0.00019386025498768558,
|
| 267 |
+
"loss": 0.0,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1506434519303559,
|
| 272 |
+
"grad_norm": NaN,
|
| 273 |
+
"learning_rate": 0.00019352756712299468,
|
| 274 |
+
"loss": 0.0,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1809235427706284,
|
| 279 |
+
"grad_norm": NaN,
|
| 280 |
+
"learning_rate": 0.00019318640292114524,
|
| 281 |
+
"loss": 0.0,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2112036336109009,
|
| 286 |
+
"grad_norm": NaN,
|
| 287 |
+
"learning_rate": 0.00019283679330160726,
|
| 288 |
+
"loss": 0.0,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2414837244511734,
|
| 293 |
+
"grad_norm": NaN,
|
| 294 |
+
"learning_rate": 0.00019247876994925292,
|
| 295 |
+
"loss": 0.0,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2717638152914459,
|
| 300 |
+
"grad_norm": NaN,
|
| 301 |
+
"learning_rate": 0.000192112365311485,
|
| 302 |
+
"loss": 0.0,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3020439061317184,
|
| 307 |
+
"grad_norm": NaN,
|
| 308 |
+
"learning_rate": 0.00019173761259529633,
|
| 309 |
+
"loss": 0.0,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3323239969719909,
|
| 314 |
+
"grad_norm": NaN,
|
| 315 |
+
"learning_rate": 0.0001913545457642601,
|
| 316 |
+
"loss": 0.0,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3626040878122634,
|
| 321 |
+
"grad_norm": NaN,
|
| 322 |
+
"learning_rate": 0.00019096319953545185,
|
| 323 |
+
"loss": 0.0,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.3928841786525359,
|
| 328 |
+
"grad_norm": NaN,
|
| 329 |
+
"learning_rate": 0.0001905636093763031,
|
| 330 |
+
"loss": 0.0,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4231642694928084,
|
| 335 |
+
"grad_norm": NaN,
|
| 336 |
+
"learning_rate": 0.00019015581150138693,
|
| 337 |
+
"loss": 0.0,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4534443603330809,
|
| 342 |
+
"grad_norm": NaN,
|
| 343 |
+
"learning_rate": 0.00018973984286913584,
|
| 344 |
+
"loss": 0.0,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4837244511733534,
|
| 349 |
+
"grad_norm": NaN,
|
| 350 |
+
"learning_rate": 0.0001893157411784924,
|
| 351 |
+
"loss": 0.0,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5140045420136259,
|
| 356 |
+
"grad_norm": NaN,
|
| 357 |
+
"learning_rate": 0.00018888354486549237,
|
| 358 |
+
"loss": 0.0,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5442846328538986,
|
| 363 |
+
"grad_norm": NaN,
|
| 364 |
+
"learning_rate": 0.00018844329309978145,
|
| 365 |
+
"loss": 0.0,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.574564723694171,
|
| 370 |
+
"grad_norm": NaN,
|
| 371 |
+
"learning_rate": 0.00018799502578106534,
|
| 372 |
+
"loss": 0.0,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6048448145344436,
|
| 377 |
+
"grad_norm": NaN,
|
| 378 |
+
"learning_rate": 0.00018753878353549357,
|
| 379 |
+
"loss": 0.0,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.635124905374716,
|
| 384 |
+
"grad_norm": NaN,
|
| 385 |
+
"learning_rate": 0.00018707460771197774,
|
| 386 |
+
"loss": 0.0,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6654049962149886,
|
| 391 |
+
"grad_norm": NaN,
|
| 392 |
+
"learning_rate": 0.00018660254037844388,
|
| 393 |
+
"loss": 0.0,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.6956850870552613,
|
| 398 |
+
"grad_norm": NaN,
|
| 399 |
+
"learning_rate": 0.00018612262431802007,
|
| 400 |
+
"loss": 0.0,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7259651778955338,
|
| 405 |
+
"grad_norm": NaN,
|
| 406 |
+
"learning_rate": 0.0001856349030251589,
|
| 407 |
+
"loss": 0.0,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7562452687358063,
|
| 412 |
+
"grad_norm": NaN,
|
| 413 |
+
"learning_rate": 0.0001851394207016957,
|
| 414 |
+
"loss": 0.0,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.7865253595760788,
|
| 419 |
+
"grad_norm": NaN,
|
| 420 |
+
"learning_rate": 0.00018463622225284242,
|
| 421 |
+
"loss": 0.0,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8168054504163513,
|
| 426 |
+
"grad_norm": NaN,
|
| 427 |
+
"learning_rate": 0.00018412535328311814,
|
| 428 |
+
"loss": 0.0,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.8470855412566238,
|
| 433 |
+
"grad_norm": NaN,
|
| 434 |
+
"learning_rate": 0.0001836068600922156,
|
| 435 |
+
"loss": 0.0,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.8773656320968963,
|
| 440 |
+
"grad_norm": NaN,
|
| 441 |
+
"learning_rate": 0.00018308078967080546,
|
| 442 |
+
"loss": 0.0,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9076457229371688,
|
| 447 |
+
"grad_norm": NaN,
|
| 448 |
+
"learning_rate": 0.0001825471896962774,
|
| 449 |
+
"loss": 0.0,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9379258137774413,
|
| 454 |
+
"grad_norm": NaN,
|
| 455 |
+
"learning_rate": 0.00018200610852841913,
|
| 456 |
+
"loss": 0.0,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9682059046177138,
|
| 461 |
+
"grad_norm": NaN,
|
| 462 |
+
"learning_rate": 0.00018145759520503358,
|
| 463 |
+
"loss": 0.0,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.9984859954579863,
|
| 468 |
+
"grad_norm": NaN,
|
| 469 |
+
"learning_rate": 0.00018090169943749476,
|
| 470 |
+
"loss": 0.0,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.028766086298259,
|
| 475 |
+
"grad_norm": NaN,
|
| 476 |
+
"learning_rate": 0.00018033847160624225,
|
| 477 |
+
"loss": 0.0,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0590461771385313,
|
| 482 |
+
"grad_norm": NaN,
|
| 483 |
+
"learning_rate": 0.00017976796275621555,
|
| 484 |
+
"loss": 0.0,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.089326267978804,
|
| 489 |
+
"grad_norm": NaN,
|
| 490 |
+
"learning_rate": 0.00017919022459222752,
|
| 491 |
+
"loss": 0.0,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1196063588190763,
|
| 496 |
+
"grad_norm": NaN,
|
| 497 |
+
"learning_rate": 0.00017860530947427875,
|
| 498 |
+
"loss": 0.0,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.149886449659349,
|
| 503 |
+
"grad_norm": NaN,
|
| 504 |
+
"learning_rate": 0.00017801327041281207,
|
| 505 |
+
"loss": 0.0,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.1801665404996213,
|
| 510 |
+
"grad_norm": NaN,
|
| 511 |
+
"learning_rate": 0.00017741416106390826,
|
| 512 |
+
"loss": 0.0,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.210446631339894,
|
| 517 |
+
"grad_norm": NaN,
|
| 518 |
+
"learning_rate": 0.00017680803572442318,
|
| 519 |
+
"loss": 0.0,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.2407267221801668,
|
| 524 |
+
"grad_norm": NaN,
|
| 525 |
+
"learning_rate": 0.0001761949493270671,
|
| 526 |
+
"loss": 0.0,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.2710068130204393,
|
| 531 |
+
"grad_norm": NaN,
|
| 532 |
+
"learning_rate": 0.00017557495743542585,
|
| 533 |
+
"loss": 0.0,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3012869038607118,
|
| 538 |
+
"grad_norm": NaN,
|
| 539 |
+
"learning_rate": 0.0001749481162389254,
|
| 540 |
+
"loss": 0.0,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3315669947009843,
|
| 545 |
+
"grad_norm": NaN,
|
| 546 |
+
"learning_rate": 0.00017431448254773944,
|
| 547 |
+
"loss": 0.0,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3618470855412568,
|
| 552 |
+
"grad_norm": NaN,
|
| 553 |
+
"learning_rate": 0.0001736741137876405,
|
| 554 |
+
"loss": 0.0,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.3921271763815293,
|
| 559 |
+
"grad_norm": NaN,
|
| 560 |
+
"learning_rate": 0.00017302706799479574,
|
| 561 |
+
"loss": 0.0,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4224072672218018,
|
| 566 |
+
"grad_norm": NaN,
|
| 567 |
+
"learning_rate": 0.00017237340381050703,
|
| 568 |
+
"loss": 0.0,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4526873580620743,
|
| 573 |
+
"grad_norm": NaN,
|
| 574 |
+
"learning_rate": 0.00017171318047589637,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.4829674489023468,
|
| 580 |
+
"grad_norm": NaN,
|
| 581 |
+
"learning_rate": 0.0001710464578265369,
|
| 582 |
+
"loss": 0.0,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5132475397426193,
|
| 587 |
+
"grad_norm": NaN,
|
| 588 |
+
"learning_rate": 0.00017037329628703004,
|
| 589 |
+
"loss": 0.0,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5435276305828918,
|
| 594 |
+
"grad_norm": NaN,
|
| 595 |
+
"learning_rate": 0.00016969375686552937,
|
| 596 |
+
"loss": 0.0,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.5738077214231643,
|
| 601 |
+
"grad_norm": NaN,
|
| 602 |
+
"learning_rate": 0.00016900790114821122,
|
| 603 |
+
"loss": 0.0,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6040878122634368,
|
| 608 |
+
"grad_norm": NaN,
|
| 609 |
+
"learning_rate": 0.00016831579129369346,
|
| 610 |
+
"loss": 0.0,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6343679031037093,
|
| 615 |
+
"grad_norm": NaN,
|
| 616 |
+
"learning_rate": 0.00016761749002740193,
|
| 617 |
+
"loss": 0.0,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.6646479939439818,
|
| 622 |
+
"grad_norm": NaN,
|
| 623 |
+
"learning_rate": 0.00016691306063588583,
|
| 624 |
+
"loss": 0.0,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.6949280847842543,
|
| 629 |
+
"grad_norm": NaN,
|
| 630 |
+
"learning_rate": 0.00016620256696108188,
|
| 631 |
+
"loss": 0.0,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.7252081756245268,
|
| 636 |
+
"grad_norm": NaN,
|
| 637 |
+
"learning_rate": 0.00016548607339452853,
|
| 638 |
+
"loss": 0.0,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.7554882664647993,
|
| 643 |
+
"grad_norm": NaN,
|
| 644 |
+
"learning_rate": 0.00016476364487153023,
|
| 645 |
+
"loss": 0.0,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.7857683573050718,
|
| 650 |
+
"grad_norm": NaN,
|
| 651 |
+
"learning_rate": 0.00016403534686527225,
|
| 652 |
+
"loss": 0.0,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8160484481453443,
|
| 657 |
+
"grad_norm": NaN,
|
| 658 |
+
"learning_rate": 0.00016330124538088705,
|
| 659 |
+
"loss": 0.0,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8463285389856168,
|
| 664 |
+
"grad_norm": NaN,
|
| 665 |
+
"learning_rate": 0.00016256140694947217,
|
| 666 |
+
"loss": 0.0,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.8766086298258893,
|
| 671 |
+
"grad_norm": NaN,
|
| 672 |
+
"learning_rate": 0.00016181589862206052,
|
| 673 |
+
"loss": 0.0,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9068887206661618,
|
| 678 |
+
"grad_norm": NaN,
|
| 679 |
+
"learning_rate": 0.00016106478796354382,
|
| 680 |
+
"loss": 0.0,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9371688115064343,
|
| 685 |
+
"grad_norm": NaN,
|
| 686 |
+
"learning_rate": 0.00016030814304654895,
|
| 687 |
+
"loss": 0.0,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.9674489023467068,
|
| 692 |
+
"grad_norm": NaN,
|
| 693 |
+
"learning_rate": 0.0001595460324452688,
|
| 694 |
+
"loss": 0.0,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.9977289931869797,
|
| 699 |
+
"grad_norm": NaN,
|
| 700 |
+
"learning_rate": 0.00015877852522924732,
|
| 701 |
+
"loss": 0.0,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.028009084027252,
|
| 706 |
+
"grad_norm": NaN,
|
| 707 |
+
"learning_rate": 0.00015800569095711982,
|
| 708 |
+
"loss": 0.0,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.0582891748675247,
|
| 713 |
+
"grad_norm": NaN,
|
| 714 |
+
"learning_rate": 0.00015722759967030898,
|
| 715 |
+
"loss": 0.0,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.088569265707797,
|
| 720 |
+
"grad_norm": NaN,
|
| 721 |
+
"learning_rate": 0.00015644432188667695,
|
| 722 |
+
"loss": 0.0,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.1188493565480697,
|
| 727 |
+
"grad_norm": NaN,
|
| 728 |
+
"learning_rate": 0.0001556559285941344,
|
| 729 |
+
"loss": 0.0,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.149129447388342,
|
| 734 |
+
"grad_norm": NaN,
|
| 735 |
+
"learning_rate": 0.000154862491244207,
|
| 736 |
+
"loss": 0.0,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.1794095382286147,
|
| 741 |
+
"grad_norm": NaN,
|
| 742 |
+
"learning_rate": 0.00015406408174555976,
|
| 743 |
+
"loss": 0.0,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.209689629068887,
|
| 748 |
+
"grad_norm": NaN,
|
| 749 |
+
"learning_rate": 0.00015326077245747999,
|
| 750 |
+
"loss": 0.0,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.2399697199091597,
|
| 755 |
+
"grad_norm": NaN,
|
| 756 |
+
"learning_rate": 0.00015245263618331945,
|
| 757 |
+
"loss": 0.0,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.270249810749432,
|
| 762 |
+
"grad_norm": NaN,
|
| 763 |
+
"learning_rate": 0.0001516397461638962,
|
| 764 |
+
"loss": 0.0,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.3005299015897047,
|
| 769 |
+
"grad_norm": NaN,
|
| 770 |
+
"learning_rate": 0.00015082217607085692,
|
| 771 |
+
"loss": 0.0,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.330809992429977,
|
| 776 |
+
"grad_norm": NaN,
|
| 777 |
+
"learning_rate": 0.00015000000000000001,
|
| 778 |
+
"loss": 0.0,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.3610900832702497,
|
| 783 |
+
"grad_norm": NaN,
|
| 784 |
+
"learning_rate": 0.0001491732924645604,
|
| 785 |
+
"loss": 0.0,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.391370174110522,
|
| 790 |
+
"grad_norm": NaN,
|
| 791 |
+
"learning_rate": 0.00014834212838845637,
|
| 792 |
+
"loss": 0.0,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.4216502649507947,
|
| 797 |
+
"grad_norm": NaN,
|
| 798 |
+
"learning_rate": 0.0001475065830994995,
|
| 799 |
+
"loss": 0.0,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.451930355791067,
|
| 804 |
+
"grad_norm": NaN,
|
| 805 |
+
"learning_rate": 0.00014666673232256738,
|
| 806 |
+
"loss": 0.0,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.4822104466313397,
|
| 811 |
+
"grad_norm": NaN,
|
| 812 |
+
"learning_rate": 0.00014582265217274104,
|
| 813 |
+
"loss": 0.0,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.5124905374716127,
|
| 818 |
+
"grad_norm": NaN,
|
| 819 |
+
"learning_rate": 0.0001449744191484066,
|
| 820 |
+
"loss": 0.0,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.542770628311885,
|
| 825 |
+
"grad_norm": NaN,
|
| 826 |
+
"learning_rate": 0.00014412211012432212,
|
| 827 |
+
"loss": 0.0,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.5730507191521577,
|
| 832 |
+
"grad_norm": NaN,
|
| 833 |
+
"learning_rate": 0.00014326580234465085,
|
| 834 |
+
"loss": 0.0,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.60333080999243,
|
| 839 |
+
"grad_norm": NaN,
|
| 840 |
+
"learning_rate": 0.00014240557341596018,
|
| 841 |
+
"loss": 0.0,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.6336109008327027,
|
| 846 |
+
"grad_norm": NaN,
|
| 847 |
+
"learning_rate": 0.00014154150130018866,
|
| 848 |
+
"loss": 0.0,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.663890991672975,
|
| 853 |
+
"grad_norm": NaN,
|
| 854 |
+
"learning_rate": 0.00014067366430758004,
|
| 855 |
+
"loss": 0.0,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.6941710825132477,
|
| 860 |
+
"grad_norm": NaN,
|
| 861 |
+
"learning_rate": 0.00013980214108958624,
|
| 862 |
+
"loss": 0.0,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.72445117335352,
|
| 867 |
+
"grad_norm": NaN,
|
| 868 |
+
"learning_rate": 0.00013892701063173918,
|
| 869 |
+
"loss": 0.0,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.7547312641937927,
|
| 874 |
+
"grad_norm": NaN,
|
| 875 |
+
"learning_rate": 0.0001380483522464923,
|
| 876 |
+
"loss": 0.0,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.785011355034065,
|
| 881 |
+
"grad_norm": NaN,
|
| 882 |
+
"learning_rate": 0.00013716624556603274,
|
| 883 |
+
"loss": 0.0,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.8152914458743377,
|
| 888 |
+
"grad_norm": NaN,
|
| 889 |
+
"learning_rate": 0.0001362807705350641,
|
| 890 |
+
"loss": 0.0,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.84557153671461,
|
| 895 |
+
"grad_norm": NaN,
|
| 896 |
+
"learning_rate": 0.00013539200740356118,
|
| 897 |
+
"loss": 0.0,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.8758516275548827,
|
| 902 |
+
"grad_norm": NaN,
|
| 903 |
+
"learning_rate": 0.00013450003671949706,
|
| 904 |
+
"loss": 0.0,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.906131718395155,
|
| 909 |
+
"grad_norm": NaN,
|
| 910 |
+
"learning_rate": 0.00013360493932154302,
|
| 911 |
+
"loss": 0.0,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.9364118092354277,
|
| 916 |
+
"grad_norm": NaN,
|
| 917 |
+
"learning_rate": 0.00013270679633174218,
|
| 918 |
+
"loss": 0.0,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.9666919000757,
|
| 923 |
+
"grad_norm": NaN,
|
| 924 |
+
"learning_rate": 0.00013180568914815752,
|
| 925 |
+
"loss": 0.0,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.9969719909159727,
|
| 930 |
+
"grad_norm": NaN,
|
| 931 |
+
"learning_rate": 0.00013090169943749476,
|
| 932 |
+
"loss": 0.0,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.027252081756245,
|
| 937 |
+
"grad_norm": NaN,
|
| 938 |
+
"learning_rate": 0.00012999490912770107,
|
| 939 |
+
"loss": 0.0,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.057532172596518,
|
| 944 |
+
"grad_norm": NaN,
|
| 945 |
+
"learning_rate": 0.0001290854004005399,
|
| 946 |
+
"loss": 0.0,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.08781226343679,
|
| 951 |
+
"grad_norm": NaN,
|
| 952 |
+
"learning_rate": 0.00012817325568414297,
|
| 953 |
+
"loss": 0.0,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.118092354277063,
|
| 958 |
+
"grad_norm": NaN,
|
| 959 |
+
"learning_rate": 0.0001272585576455398,
|
| 960 |
+
"loss": 0.0,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.148372445117335,
|
| 965 |
+
"grad_norm": NaN,
|
| 966 |
+
"learning_rate": 0.00012634138918316568,
|
| 967 |
+
"loss": 0.0,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.178652535957608,
|
| 972 |
+
"grad_norm": NaN,
|
| 973 |
+
"learning_rate": 0.00012542183341934872,
|
| 974 |
+
"loss": 0.0,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.20893262679788,
|
| 979 |
+
"grad_norm": NaN,
|
| 980 |
+
"learning_rate": 0.0001244999736927764,
|
| 981 |
+
"loss": 0.0,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.239212717638153,
|
| 986 |
+
"grad_norm": NaN,
|
| 987 |
+
"learning_rate": 0.00012357589355094275,
|
| 988 |
+
"loss": 0.0,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.269492808478425,
|
| 993 |
+
"grad_norm": NaN,
|
| 994 |
+
"learning_rate": 0.00012264967674257646,
|
| 995 |
+
"loss": 0.0,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.299772899318698,
|
| 1000 |
+
"grad_norm": NaN,
|
| 1001 |
+
"learning_rate": 0.00012172140721005079,
|
| 1002 |
+
"loss": 0.0,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.33005299015897,
|
| 1007 |
+
"grad_norm": NaN,
|
| 1008 |
+
"learning_rate": 0.00012079116908177593,
|
| 1009 |
+
"loss": 0.0,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.360333080999243,
|
| 1014 |
+
"grad_norm": NaN,
|
| 1015 |
+
"learning_rate": 0.00011985904666457455,
|
| 1016 |
+
"loss": 0.0,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.390613171839515,
|
| 1021 |
+
"grad_norm": NaN,
|
| 1022 |
+
"learning_rate": 0.00011892512443604102,
|
| 1023 |
+
"loss": 0.0,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.420893262679788,
|
| 1028 |
+
"grad_norm": NaN,
|
| 1029 |
+
"learning_rate": 0.00011798948703688539,
|
| 1030 |
+
"loss": 0.0,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.45117335352006,
|
| 1035 |
+
"grad_norm": NaN,
|
| 1036 |
+
"learning_rate": 0.0001170522192632624,
|
| 1037 |
+
"loss": 0.0,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.4814534443603335,
|
| 1042 |
+
"grad_norm": NaN,
|
| 1043 |
+
"learning_rate": 0.00011611340605908642,
|
| 1044 |
+
"loss": 0.0,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.511733535200605,
|
| 1049 |
+
"grad_norm": NaN,
|
| 1050 |
+
"learning_rate": 0.00011517313250833317,
|
| 1051 |
+
"loss": 0.0,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.5420136260408785,
|
| 1056 |
+
"grad_norm": NaN,
|
| 1057 |
+
"learning_rate": 0.00011423148382732853,
|
| 1058 |
+
"loss": 0.0,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.57229371688115,
|
| 1063 |
+
"grad_norm": NaN,
|
| 1064 |
+
"learning_rate": 0.00011328854535702543,
|
| 1065 |
+
"loss": 0.0,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.6025738077214235,
|
| 1070 |
+
"grad_norm": NaN,
|
| 1071 |
+
"learning_rate": 0.00011234440255526948,
|
| 1072 |
+
"loss": 0.0,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.632853898561696,
|
| 1077 |
+
"grad_norm": NaN,
|
| 1078 |
+
"learning_rate": 0.00011139914098905406,
|
| 1079 |
+
"loss": 0.0,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.6631339894019685,
|
| 1084 |
+
"grad_norm": NaN,
|
| 1085 |
+
"learning_rate": 0.00011045284632676536,
|
| 1086 |
+
"loss": 0.0,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.693414080242241,
|
| 1091 |
+
"grad_norm": NaN,
|
| 1092 |
+
"learning_rate": 0.00010950560433041826,
|
| 1093 |
+
"loss": 0.0,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.7236941710825135,
|
| 1098 |
+
"grad_norm": NaN,
|
| 1099 |
+
"learning_rate": 0.00010855750084788398,
|
| 1100 |
+
"loss": 0.0,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.753974261922786,
|
| 1105 |
+
"grad_norm": NaN,
|
| 1106 |
+
"learning_rate": 0.00010760862180510951,
|
| 1107 |
+
"loss": 0.0,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.7842543527630585,
|
| 1112 |
+
"grad_norm": NaN,
|
| 1113 |
+
"learning_rate": 0.00010665905319833041,
|
| 1114 |
+
"loss": 0.0,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.814534443603331,
|
| 1119 |
+
"grad_norm": NaN,
|
| 1120 |
+
"learning_rate": 0.00010570888108627681,
|
| 1121 |
+
"loss": 0.0,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.8448145344436035,
|
| 1126 |
+
"grad_norm": NaN,
|
| 1127 |
+
"learning_rate": 0.00010475819158237425,
|
| 1128 |
+
"loss": 0.0,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.875094625283876,
|
| 1133 |
+
"grad_norm": NaN,
|
| 1134 |
+
"learning_rate": 0.00010380707084693901,
|
| 1135 |
+
"loss": 0.0,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.9053747161241485,
|
| 1140 |
+
"grad_norm": NaN,
|
| 1141 |
+
"learning_rate": 0.00010285560507936961,
|
| 1142 |
+
"loss": 0.0,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.935654806964421,
|
| 1147 |
+
"grad_norm": NaN,
|
| 1148 |
+
"learning_rate": 0.00010190388051033466,
|
| 1149 |
+
"loss": 0.0,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.9659348978046935,
|
| 1154 |
+
"grad_norm": NaN,
|
| 1155 |
+
"learning_rate": 0.00010095198339395769,
|
| 1156 |
+
"loss": 0.0,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.996214988644966,
|
| 1161 |
+
"grad_norm": NaN,
|
| 1162 |
+
"learning_rate": 0.0001,
|
| 1163 |
+
"loss": 0.0,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.0264950794852385,
|
| 1168 |
+
"grad_norm": NaN,
|
| 1169 |
+
"learning_rate": 9.904801660604234e-05,
|
| 1170 |
+
"loss": 0.0,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.056775170325511,
|
| 1175 |
+
"grad_norm": NaN,
|
| 1176 |
+
"learning_rate": 9.809611948966533e-05,
|
| 1177 |
+
"loss": 0.0,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.0870552611657835,
|
| 1182 |
+
"grad_norm": NaN,
|
| 1183 |
+
"learning_rate": 9.71443949206304e-05,
|
| 1184 |
+
"loss": 0.0,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.117335352006056,
|
| 1189 |
+
"grad_norm": NaN,
|
| 1190 |
+
"learning_rate": 9.619292915306101e-05,
|
| 1191 |
+
"loss": 0.0,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.1476154428463285,
|
| 1196 |
+
"grad_norm": NaN,
|
| 1197 |
+
"learning_rate": 9.524180841762577e-05,
|
| 1198 |
+
"loss": 0.0,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.177895533686601,
|
| 1203 |
+
"grad_norm": NaN,
|
| 1204 |
+
"learning_rate": 9.42911189137232e-05,
|
| 1205 |
+
"loss": 0.0,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.2081756245268735,
|
| 1210 |
+
"grad_norm": NaN,
|
| 1211 |
+
"learning_rate": 9.334094680166962e-05,
|
| 1212 |
+
"loss": 0.0,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.238455715367146,
|
| 1217 |
+
"grad_norm": NaN,
|
| 1218 |
+
"learning_rate": 9.239137819489047e-05,
|
| 1219 |
+
"loss": 0.0,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.2687358062074185,
|
| 1224 |
+
"grad_norm": NaN,
|
| 1225 |
+
"learning_rate": 9.144249915211605e-05,
|
| 1226 |
+
"loss": 0.0,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.299015897047691,
|
| 1231 |
+
"grad_norm": NaN,
|
| 1232 |
+
"learning_rate": 9.049439566958175e-05,
|
| 1233 |
+
"loss": 0.0,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.3292959878879635,
|
| 1238 |
+
"grad_norm": NaN,
|
| 1239 |
+
"learning_rate": 8.954715367323468e-05,
|
| 1240 |
+
"loss": 0.0,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.359576078728236,
|
| 1245 |
+
"grad_norm": NaN,
|
| 1246 |
+
"learning_rate": 8.860085901094595e-05,
|
| 1247 |
+
"loss": 0.0,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.3898561695685085,
|
| 1252 |
+
"grad_norm": NaN,
|
| 1253 |
+
"learning_rate": 8.765559744473053e-05,
|
| 1254 |
+
"loss": 0.0,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.420136260408781,
|
| 1259 |
+
"grad_norm": NaN,
|
| 1260 |
+
"learning_rate": 8.67114546429746e-05,
|
| 1261 |
+
"loss": 0.0,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.4504163512490535,
|
| 1266 |
+
"grad_norm": NaN,
|
| 1267 |
+
"learning_rate": 8.57685161726715e-05,
|
| 1268 |
+
"loss": 0.0,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.480696442089326,
|
| 1273 |
+
"grad_norm": NaN,
|
| 1274 |
+
"learning_rate": 8.482686749166686e-05,
|
| 1275 |
+
"loss": 0.0,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.5109765329295985,
|
| 1280 |
+
"grad_norm": NaN,
|
| 1281 |
+
"learning_rate": 8.38865939409136e-05,
|
| 1282 |
+
"loss": 0.0,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.541256623769871,
|
| 1287 |
+
"grad_norm": NaN,
|
| 1288 |
+
"learning_rate": 8.294778073673762e-05,
|
| 1289 |
+
"loss": 0.0,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.5715367146101435,
|
| 1294 |
+
"grad_norm": NaN,
|
| 1295 |
+
"learning_rate": 8.201051296311462e-05,
|
| 1296 |
+
"loss": 0.0,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.601816805450416,
|
| 1301 |
+
"grad_norm": NaN,
|
| 1302 |
+
"learning_rate": 8.107487556395901e-05,
|
| 1303 |
+
"loss": 0.0,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.6320968962906885,
|
| 1308 |
+
"grad_norm": NaN,
|
| 1309 |
+
"learning_rate": 8.014095333542548e-05,
|
| 1310 |
+
"loss": 0.0,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.662376987130961,
|
| 1315 |
+
"grad_norm": NaN,
|
| 1316 |
+
"learning_rate": 7.920883091822408e-05,
|
| 1317 |
+
"loss": 0.0,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.6926570779712335,
|
| 1322 |
+
"grad_norm": NaN,
|
| 1323 |
+
"learning_rate": 7.827859278994925e-05,
|
| 1324 |
+
"loss": 0.0,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.722937168811506,
|
| 1329 |
+
"grad_norm": NaN,
|
| 1330 |
+
"learning_rate": 7.735032325742355e-05,
|
| 1331 |
+
"loss": 0.0,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.7532172596517785,
|
| 1336 |
+
"grad_norm": NaN,
|
| 1337 |
+
"learning_rate": 7.642410644905726e-05,
|
| 1338 |
+
"loss": 0.0,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.783497350492052,
|
| 1343 |
+
"grad_norm": NaN,
|
| 1344 |
+
"learning_rate": 7.550002630722366e-05,
|
| 1345 |
+
"loss": 0.0,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.8137774413323235,
|
| 1350 |
+
"grad_norm": NaN,
|
| 1351 |
+
"learning_rate": 7.457816658065134e-05,
|
| 1352 |
+
"loss": 0.0,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.844057532172597,
|
| 1357 |
+
"grad_norm": NaN,
|
| 1358 |
+
"learning_rate": 7.365861081683433e-05,
|
| 1359 |
+
"loss": 0.0,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.8743376230128685,
|
| 1364 |
+
"grad_norm": NaN,
|
| 1365 |
+
"learning_rate": 7.274144235446023e-05,
|
| 1366 |
+
"loss": 0.0,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.904617713853142,
|
| 1371 |
+
"grad_norm": NaN,
|
| 1372 |
+
"learning_rate": 7.182674431585704e-05,
|
| 1373 |
+
"loss": 0.0,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.934897804693414,
|
| 1378 |
+
"grad_norm": NaN,
|
| 1379 |
+
"learning_rate": 7.09145995994601e-05,
|
| 1380 |
+
"loss": 0.0,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.965177895533687,
|
| 1385 |
+
"grad_norm": NaN,
|
| 1386 |
+
"learning_rate": 7.000509087229895e-05,
|
| 1387 |
+
"loss": 0.0,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.995457986373959,
|
| 1392 |
+
"grad_norm": NaN,
|
| 1393 |
+
"learning_rate": 6.909830056250527e-05,
|
| 1394 |
+
"loss": 0.0,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.025738077214232,
|
| 1399 |
+
"grad_norm": NaN,
|
| 1400 |
+
"learning_rate": 6.819431085184251e-05,
|
| 1401 |
+
"loss": 0.0,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.056018168054504,
|
| 1406 |
+
"grad_norm": NaN,
|
| 1407 |
+
"learning_rate": 6.729320366825784e-05,
|
| 1408 |
+
"loss": 0.0,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.086298258894777,
|
| 1413 |
+
"grad_norm": NaN,
|
| 1414 |
+
"learning_rate": 6.639506067845697e-05,
|
| 1415 |
+
"loss": 0.0,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.116578349735049,
|
| 1420 |
+
"grad_norm": NaN,
|
| 1421 |
+
"learning_rate": 6.549996328050296e-05,
|
| 1422 |
+
"loss": 0.0,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.146858440575322,
|
| 1427 |
+
"grad_norm": NaN,
|
| 1428 |
+
"learning_rate": 6.460799259643884e-05,
|
| 1429 |
+
"loss": 0.0,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.177138531415594,
|
| 1434 |
+
"grad_norm": NaN,
|
| 1435 |
+
"learning_rate": 6.371922946493591e-05,
|
| 1436 |
+
"loss": 0.0,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.207418622255867,
|
| 1441 |
+
"grad_norm": NaN,
|
| 1442 |
+
"learning_rate": 6.283375443396726e-05,
|
| 1443 |
+
"loss": 0.0,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.237698713096139,
|
| 1448 |
+
"grad_norm": NaN,
|
| 1449 |
+
"learning_rate": 6.19516477535077e-05,
|
| 1450 |
+
"loss": 0.0,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.267978803936412,
|
| 1455 |
+
"grad_norm": NaN,
|
| 1456 |
+
"learning_rate": 6.107298936826086e-05,
|
| 1457 |
+
"loss": 0.0,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.298258894776684,
|
| 1462 |
+
"grad_norm": NaN,
|
| 1463 |
+
"learning_rate": 6.019785891041381e-05,
|
| 1464 |
+
"loss": 0.0,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.328538985616957,
|
| 1469 |
+
"grad_norm": NaN,
|
| 1470 |
+
"learning_rate": 5.9326335692419995e-05,
|
| 1471 |
+
"loss": 0.0,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.358819076457229,
|
| 1476 |
+
"grad_norm": NaN,
|
| 1477 |
+
"learning_rate": 5.845849869981137e-05,
|
| 1478 |
+
"loss": 0.0,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.389099167297502,
|
| 1483 |
+
"grad_norm": NaN,
|
| 1484 |
+
"learning_rate": 5.759442658403985e-05,
|
| 1485 |
+
"loss": 0.0,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.419379258137774,
|
| 1490 |
+
"grad_norm": NaN,
|
| 1491 |
+
"learning_rate": 5.6734197655349156e-05,
|
| 1492 |
+
"loss": 0.0,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.449659348978047,
|
| 1497 |
+
"grad_norm": NaN,
|
| 1498 |
+
"learning_rate": 5.5877889875677845e-05,
|
| 1499 |
+
"loss": 0.0,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.479939439818319,
|
| 1504 |
+
"grad_norm": NaN,
|
| 1505 |
+
"learning_rate": 5.5025580851593436e-05,
|
| 1506 |
+
"loss": 0.0,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.510219530658592,
|
| 1511 |
+
"grad_norm": NaN,
|
| 1512 |
+
"learning_rate": 5.417734782725896e-05,
|
| 1513 |
+
"loss": 0.0,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.540499621498864,
|
| 1518 |
+
"grad_norm": NaN,
|
| 1519 |
+
"learning_rate": 5.333326767743263e-05,
|
| 1520 |
+
"loss": 0.0,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.570779712339137,
|
| 1525 |
+
"grad_norm": NaN,
|
| 1526 |
+
"learning_rate": 5.249341690050051e-05,
|
| 1527 |
+
"loss": 0.0,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.601059803179409,
|
| 1532 |
+
"grad_norm": NaN,
|
| 1533 |
+
"learning_rate": 5.1657871611543605e-05,
|
| 1534 |
+
"loss": 0.0,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.631339894019682,
|
| 1539 |
+
"grad_norm": NaN,
|
| 1540 |
+
"learning_rate": 5.082670753543961e-05,
|
| 1541 |
+
"loss": 0.0,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.661619984859954,
|
| 1546 |
+
"grad_norm": NaN,
|
| 1547 |
+
"learning_rate": 5.000000000000002e-05,
|
| 1548 |
+
"loss": 0.0,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.691900075700227,
|
| 1553 |
+
"grad_norm": NaN,
|
| 1554 |
+
"learning_rate": 4.9177823929143106e-05,
|
| 1555 |
+
"loss": 0.0,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.722180166540499,
|
| 1560 |
+
"grad_norm": NaN,
|
| 1561 |
+
"learning_rate": 4.836025383610382e-05,
|
| 1562 |
+
"loss": 0.0,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.752460257380772,
|
| 1567 |
+
"grad_norm": NaN,
|
| 1568 |
+
"learning_rate": 4.754736381668057e-05,
|
| 1569 |
+
"loss": 0.0,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.782740348221044,
|
| 1574 |
+
"grad_norm": NaN,
|
| 1575 |
+
"learning_rate": 4.673922754252002e-05,
|
| 1576 |
+
"loss": 0.0,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.813020439061317,
|
| 1581 |
+
"grad_norm": NaN,
|
| 1582 |
+
"learning_rate": 4.593591825444028e-05,
|
| 1583 |
+
"loss": 0.0,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.843300529901589,
|
| 1588 |
+
"grad_norm": NaN,
|
| 1589 |
+
"learning_rate": 4.513750875579303e-05,
|
| 1590 |
+
"loss": 0.0,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.873580620741862,
|
| 1595 |
+
"grad_norm": NaN,
|
| 1596 |
+
"learning_rate": 4.434407140586565e-05,
|
| 1597 |
+
"loss": 0.0,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.903860711582134,
|
| 1602 |
+
"grad_norm": NaN,
|
| 1603 |
+
"learning_rate": 4.355567811332311e-05,
|
| 1604 |
+
"loss": 0.0,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.934140802422407,
|
| 1609 |
+
"grad_norm": NaN,
|
| 1610 |
+
"learning_rate": 4.277240032969105e-05,
|
| 1611 |
+
"loss": 0.0,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.964420893262679,
|
| 1616 |
+
"grad_norm": NaN,
|
| 1617 |
+
"learning_rate": 4.19943090428802e-05,
|
| 1618 |
+
"loss": 0.0,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 6.994700984102952,
|
| 1623 |
+
"grad_norm": NaN,
|
| 1624 |
+
"learning_rate": 4.12214747707527e-05,
|
| 1625 |
+
"loss": 0.0,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.024981074943224,
|
| 1630 |
+
"grad_norm": NaN,
|
| 1631 |
+
"learning_rate": 4.045396755473121e-05,
|
| 1632 |
+
"loss": 0.0,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.055261165783497,
|
| 1637 |
+
"grad_norm": NaN,
|
| 1638 |
+
"learning_rate": 3.969185695345105e-05,
|
| 1639 |
+
"loss": 0.0,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.085541256623769,
|
| 1644 |
+
"grad_norm": NaN,
|
| 1645 |
+
"learning_rate": 3.893521203645618e-05,
|
| 1646 |
+
"loss": 0.0,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.115821347464043,
|
| 1651 |
+
"grad_norm": NaN,
|
| 1652 |
+
"learning_rate": 3.8184101377939476e-05,
|
| 1653 |
+
"loss": 0.0,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.146101438304315,
|
| 1658 |
+
"grad_norm": NaN,
|
| 1659 |
+
"learning_rate": 3.7438593050527845e-05,
|
| 1660 |
+
"loss": 0.0,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.176381529144588,
|
| 1665 |
+
"grad_norm": NaN,
|
| 1666 |
+
"learning_rate": 3.669875461911297e-05,
|
| 1667 |
+
"loss": 0.0,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.20666161998486,
|
| 1672 |
+
"grad_norm": NaN,
|
| 1673 |
+
"learning_rate": 3.5964653134727776e-05,
|
| 1674 |
+
"loss": 0.0,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.236941710825133,
|
| 1679 |
+
"grad_norm": NaN,
|
| 1680 |
+
"learning_rate": 3.523635512846981e-05,
|
| 1681 |
+
"loss": 0.0,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.267221801665405,
|
| 1686 |
+
"grad_norm": NaN,
|
| 1687 |
+
"learning_rate": 3.45139266054715e-05,
|
| 1688 |
+
"loss": 0.0,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.297501892505678,
|
| 1693 |
+
"grad_norm": NaN,
|
| 1694 |
+
"learning_rate": 3.379743303891815e-05,
|
| 1695 |
+
"loss": 0.0,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.32778198334595,
|
| 1700 |
+
"grad_norm": NaN,
|
| 1701 |
+
"learning_rate": 3.308693936411421e-05,
|
| 1702 |
+
"loss": 0.0,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.358062074186223,
|
| 1707 |
+
"grad_norm": NaN,
|
| 1708 |
+
"learning_rate": 3.238250997259808e-05,
|
| 1709 |
+
"loss": 0.0,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.388342165026495,
|
| 1714 |
+
"grad_norm": NaN,
|
| 1715 |
+
"learning_rate": 3.1684208706306574e-05,
|
| 1716 |
+
"loss": 0.0,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.418622255866768,
|
| 1721 |
+
"grad_norm": NaN,
|
| 1722 |
+
"learning_rate": 3.099209885178882e-05,
|
| 1723 |
+
"loss": 0.0,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.44890234670704,
|
| 1728 |
+
"grad_norm": NaN,
|
| 1729 |
+
"learning_rate": 3.030624313447067e-05,
|
| 1730 |
+
"loss": 0.0,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.479182437547313,
|
| 1735 |
+
"grad_norm": NaN,
|
| 1736 |
+
"learning_rate": 2.962670371296996e-05,
|
| 1737 |
+
"loss": 0.0,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.509462528387585,
|
| 1742 |
+
"grad_norm": NaN,
|
| 1743 |
+
"learning_rate": 2.8953542173463133e-05,
|
| 1744 |
+
"loss": 0.0,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.539742619227858,
|
| 1749 |
+
"grad_norm": NaN,
|
| 1750 |
+
"learning_rate": 2.828681952410366e-05,
|
| 1751 |
+
"loss": 0.0,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.57002271006813,
|
| 1756 |
+
"grad_norm": NaN,
|
| 1757 |
+
"learning_rate": 2.7626596189492983e-05,
|
| 1758 |
+
"loss": 0.0,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.600302800908403,
|
| 1763 |
+
"grad_norm": NaN,
|
| 1764 |
+
"learning_rate": 2.6972932005204267e-05,
|
| 1765 |
+
"loss": 0.0,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.630582891748675,
|
| 1770 |
+
"grad_norm": NaN,
|
| 1771 |
+
"learning_rate": 2.6325886212359498e-05,
|
| 1772 |
+
"loss": 0.0,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.660862982588948,
|
| 1777 |
+
"grad_norm": NaN,
|
| 1778 |
+
"learning_rate": 2.5685517452260567e-05,
|
| 1779 |
+
"loss": 0.0,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.69114307342922,
|
| 1784 |
+
"grad_norm": NaN,
|
| 1785 |
+
"learning_rate": 2.5051883761074614e-05,
|
| 1786 |
+
"loss": 0.0,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.721423164269493,
|
| 1791 |
+
"grad_norm": NaN,
|
| 1792 |
+
"learning_rate": 2.4425042564574184e-05,
|
| 1793 |
+
"loss": 0.0,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.751703255109765,
|
| 1798 |
+
"grad_norm": NaN,
|
| 1799 |
+
"learning_rate": 2.3805050672932928e-05,
|
| 1800 |
+
"loss": 0.0,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.781983345950038,
|
| 1805 |
+
"grad_norm": NaN,
|
| 1806 |
+
"learning_rate": 2.3191964275576805e-05,
|
| 1807 |
+
"loss": 0.0,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.81226343679031,
|
| 1812 |
+
"grad_norm": NaN,
|
| 1813 |
+
"learning_rate": 2.2585838936091754e-05,
|
| 1814 |
+
"loss": 0.0,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.842543527630583,
|
| 1819 |
+
"grad_norm": NaN,
|
| 1820 |
+
"learning_rate": 2.198672958718796e-05,
|
| 1821 |
+
"loss": 0.0,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.872823618470855,
|
| 1826 |
+
"grad_norm": NaN,
|
| 1827 |
+
"learning_rate": 2.139469052572127e-05,
|
| 1828 |
+
"loss": 0.0,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 7.903103709311128,
|
| 1833 |
+
"grad_norm": NaN,
|
| 1834 |
+
"learning_rate": 2.0809775407772503e-05,
|
| 1835 |
+
"loss": 0.0,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 7.9333838001514,
|
| 1840 |
+
"grad_norm": NaN,
|
| 1841 |
+
"learning_rate": 2.0232037243784475e-05,
|
| 1842 |
+
"loss": 0.0,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 7.963663890991673,
|
| 1847 |
+
"grad_norm": NaN,
|
| 1848 |
+
"learning_rate": 1.9661528393757744e-05,
|
| 1849 |
+
"loss": 0.0,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 7.993943981831945,
|
| 1854 |
+
"grad_norm": NaN,
|
| 1855 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 1856 |
+
"loss": 0.0,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.024224072672219,
|
| 1861 |
+
"grad_norm": NaN,
|
| 1862 |
+
"learning_rate": 1.854240479496643e-05,
|
| 1863 |
+
"loss": 0.0,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.05450416351249,
|
| 1868 |
+
"grad_norm": NaN,
|
| 1869 |
+
"learning_rate": 1.7993891471580893e-05,
|
| 1870 |
+
"loss": 0.0,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.084784254352764,
|
| 1875 |
+
"grad_norm": NaN,
|
| 1876 |
+
"learning_rate": 1.74528103037226e-05,
|
| 1877 |
+
"loss": 0.0,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.115064345193035,
|
| 1882 |
+
"grad_norm": NaN,
|
| 1883 |
+
"learning_rate": 1.6919210329194533e-05,
|
| 1884 |
+
"loss": 0.0,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.145344436033309,
|
| 1889 |
+
"grad_norm": NaN,
|
| 1890 |
+
"learning_rate": 1.6393139907784404e-05,
|
| 1891 |
+
"loss": 0.0,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.17562452687358,
|
| 1896 |
+
"grad_norm": NaN,
|
| 1897 |
+
"learning_rate": 1.587464671688187e-05,
|
| 1898 |
+
"loss": 0.0,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.205904617713854,
|
| 1903 |
+
"grad_norm": NaN,
|
| 1904 |
+
"learning_rate": 1.5363777747157572e-05,
|
| 1905 |
+
"loss": 0.0,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.236184708554125,
|
| 1910 |
+
"grad_norm": NaN,
|
| 1911 |
+
"learning_rate": 1.4860579298304312e-05,
|
| 1912 |
+
"loss": 0.0,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.266464799394399,
|
| 1917 |
+
"grad_norm": NaN,
|
| 1918 |
+
"learning_rate": 1.4365096974841108e-05,
|
| 1919 |
+
"loss": 0.0,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.29674489023467,
|
| 1924 |
+
"grad_norm": NaN,
|
| 1925 |
+
"learning_rate": 1.3877375681979943e-05,
|
| 1926 |
+
"loss": 0.0,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.327024981074944,
|
| 1931 |
+
"grad_norm": NaN,
|
| 1932 |
+
"learning_rate": 1.339745962155613e-05,
|
| 1933 |
+
"loss": 0.0,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.357305071915215,
|
| 1938 |
+
"grad_norm": NaN,
|
| 1939 |
+
"learning_rate": 1.2925392288022298e-05,
|
| 1940 |
+
"loss": 0.0,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.387585162755489,
|
| 1945 |
+
"grad_norm": NaN,
|
| 1946 |
+
"learning_rate": 1.2461216464506454e-05,
|
| 1947 |
+
"loss": 0.0,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.41786525359576,
|
| 1952 |
+
"grad_norm": NaN,
|
| 1953 |
+
"learning_rate": 1.2004974218934695e-05,
|
| 1954 |
+
"loss": 0.0,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.448145344436034,
|
| 1959 |
+
"grad_norm": NaN,
|
| 1960 |
+
"learning_rate": 1.1556706900218572e-05,
|
| 1961 |
+
"loss": 0.0,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.478425435276305,
|
| 1966 |
+
"grad_norm": NaN,
|
| 1967 |
+
"learning_rate": 1.1116455134507664e-05,
|
| 1968 |
+
"loss": 0.0,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.508705526116579,
|
| 1973 |
+
"grad_norm": NaN,
|
| 1974 |
+
"learning_rate": 1.068425882150762e-05,
|
| 1975 |
+
"loss": 0.0,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.53898561695685,
|
| 1980 |
+
"grad_norm": NaN,
|
| 1981 |
+
"learning_rate": 1.026015713086418e-05,
|
| 1982 |
+
"loss": 0.0,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.569265707797124,
|
| 1987 |
+
"grad_norm": NaN,
|
| 1988 |
+
"learning_rate": 9.844188498613116e-06,
|
| 1989 |
+
"loss": 0.0,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.599545798637395,
|
| 1994 |
+
"grad_norm": NaN,
|
| 1995 |
+
"learning_rate": 9.436390623696911e-06,
|
| 1996 |
+
"loss": 0.0,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.629825889477669,
|
| 2001 |
+
"grad_norm": NaN,
|
| 2002 |
+
"learning_rate": 9.036800464548157e-06,
|
| 2003 |
+
"loss": 0.0,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.66010598031794,
|
| 2008 |
+
"grad_norm": NaN,
|
| 2009 |
+
"learning_rate": 8.645454235739903e-06,
|
| 2010 |
+
"loss": 0.0,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.690386071158214,
|
| 2015 |
+
"grad_norm": NaN,
|
| 2016 |
+
"learning_rate": 8.262387404703653e-06,
|
| 2017 |
+
"loss": 0.0,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.720666161998485,
|
| 2022 |
+
"grad_norm": NaN,
|
| 2023 |
+
"learning_rate": 7.887634688515e-06,
|
| 2024 |
+
"loss": 0.0,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 8.750946252838759,
|
| 2029 |
+
"grad_norm": NaN,
|
| 2030 |
+
"learning_rate": 7.521230050747086e-06,
|
| 2031 |
+
"loss": 0.0,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 8.78122634367903,
|
| 2036 |
+
"grad_norm": NaN,
|
| 2037 |
+
"learning_rate": 7.163206698392744e-06,
|
| 2038 |
+
"loss": 0.0,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 8.811506434519304,
|
| 2043 |
+
"grad_norm": NaN,
|
| 2044 |
+
"learning_rate": 6.813597078854772e-06,
|
| 2045 |
+
"loss": 0.0,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.841786525359575,
|
| 2050 |
+
"grad_norm": NaN,
|
| 2051 |
+
"learning_rate": 6.472432877005341e-06,
|
| 2052 |
+
"loss": 0.0,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 8.872066616199849,
|
| 2057 |
+
"grad_norm": NaN,
|
| 2058 |
+
"learning_rate": 6.139745012314424e-06,
|
| 2059 |
+
"loss": 0.0,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 8.90234670704012,
|
| 2064 |
+
"grad_norm": NaN,
|
| 2065 |
+
"learning_rate": 5.8155636360475385e-06,
|
| 2066 |
+
"loss": 0.0,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 8.932626797880394,
|
| 2071 |
+
"grad_norm": NaN,
|
| 2072 |
+
"learning_rate": 5.499918128533155e-06,
|
| 2073 |
+
"loss": 0.0,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 8.962906888720667,
|
| 2078 |
+
"grad_norm": NaN,
|
| 2079 |
+
"learning_rate": 5.192837096500058e-06,
|
| 2080 |
+
"loss": 0.0,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 8.993186979560939,
|
| 2085 |
+
"grad_norm": NaN,
|
| 2086 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 2087 |
+
"loss": 0.0,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.02346707040121,
|
| 2092 |
+
"grad_norm": NaN,
|
| 2093 |
+
"learning_rate": 4.604479002308737e-06,
|
| 2094 |
+
"loss": 0.0,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.053747161241484,
|
| 2099 |
+
"grad_norm": NaN,
|
| 2100 |
+
"learning_rate": 4.323255262627846e-06,
|
| 2101 |
+
"loss": 0.0,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.084027252081757,
|
| 2106 |
+
"grad_norm": NaN,
|
| 2107 |
+
"learning_rate": 4.050702638550275e-06,
|
| 2108 |
+
"loss": 0.0,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.114307342922029,
|
| 2113 |
+
"grad_norm": NaN,
|
| 2114 |
+
"learning_rate": 3.7868458313272904e-06,
|
| 2115 |
+
"loss": 0.0,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.144587433762302,
|
| 2120 |
+
"grad_norm": NaN,
|
| 2121 |
+
"learning_rate": 3.5317087541144377e-06,
|
| 2122 |
+
"loss": 0.0,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.174867524602574,
|
| 2127 |
+
"grad_norm": NaN,
|
| 2128 |
+
"learning_rate": 3.2853145298042953e-06,
|
| 2129 |
+
"loss": 0.0,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.205147615442847,
|
| 2134 |
+
"grad_norm": NaN,
|
| 2135 |
+
"learning_rate": 3.047685488930874e-06,
|
| 2136 |
+
"loss": 0.0,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.235427706283119,
|
| 2141 |
+
"grad_norm": NaN,
|
| 2142 |
+
"learning_rate": 2.818843167645835e-06,
|
| 2143 |
+
"loss": 0.0,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.265707797123392,
|
| 2148 |
+
"grad_norm": NaN,
|
| 2149 |
+
"learning_rate": 2.5988083057666533e-06,
|
| 2150 |
+
"loss": 0.0,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.295987887963664,
|
| 2155 |
+
"grad_norm": NaN,
|
| 2156 |
+
"learning_rate": 2.3876008448969976e-06,
|
| 2157 |
+
"loss": 0.0,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 9.326267978803937,
|
| 2162 |
+
"grad_norm": NaN,
|
| 2163 |
+
"learning_rate": 2.1852399266194314e-06,
|
| 2164 |
+
"loss": 0.0,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 9.356548069644209,
|
| 2169 |
+
"grad_norm": NaN,
|
| 2170 |
+
"learning_rate": 1.9917438907606556e-06,
|
| 2171 |
+
"loss": 0.0,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 9.386828160484482,
|
| 2176 |
+
"grad_norm": NaN,
|
| 2177 |
+
"learning_rate": 1.8071302737293295e-06,
|
| 2178 |
+
"loss": 0.0,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 9.417108251324754,
|
| 2183 |
+
"grad_norm": NaN,
|
| 2184 |
+
"learning_rate": 1.6314158069267948e-06,
|
| 2185 |
+
"loss": 0.0,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.447388342165027,
|
| 2190 |
+
"grad_norm": NaN,
|
| 2191 |
+
"learning_rate": 1.4646164152307018e-06,
|
| 2192 |
+
"loss": 0.0,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 9.477668433005299,
|
| 2197 |
+
"grad_norm": NaN,
|
| 2198 |
+
"learning_rate": 1.3067472155517735e-06,
|
| 2199 |
+
"loss": 0.0,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 9.507948523845572,
|
| 2204 |
+
"grad_norm": NaN,
|
| 2205 |
+
"learning_rate": 1.157822515463758e-06,
|
| 2206 |
+
"loss": 0.0,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 9.538228614685844,
|
| 2211 |
+
"grad_norm": NaN,
|
| 2212 |
+
"learning_rate": 1.0178558119067315e-06,
|
| 2213 |
+
"loss": 0.0,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 9.568508705526117,
|
| 2218 |
+
"grad_norm": NaN,
|
| 2219 |
+
"learning_rate": 8.868597899638898e-07,
|
| 2220 |
+
"loss": 0.0,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 9.598788796366389,
|
| 2225 |
+
"grad_norm": NaN,
|
| 2226 |
+
"learning_rate": 7.648463217118984e-07,
|
| 2227 |
+
"loss": 0.0,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.629068887206662,
|
| 2232 |
+
"grad_norm": NaN,
|
| 2233 |
+
"learning_rate": 6.518264651449779e-07,
|
| 2234 |
+
"loss": 0.0,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 9.659348978046934,
|
| 2239 |
+
"grad_norm": NaN,
|
| 2240 |
+
"learning_rate": 5.478104631726711e-07,
|
| 2241 |
+
"loss": 0.0,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 9.689629068887207,
|
| 2246 |
+
"grad_norm": NaN,
|
| 2247 |
+
"learning_rate": 4.5280774269154115e-07,
|
| 2248 |
+
"loss": 0.0,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 9.719909159727479,
|
| 2253 |
+
"grad_norm": NaN,
|
| 2254 |
+
"learning_rate": 3.6682691373086665e-07,
|
| 2255 |
+
"loss": 0.0,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 9.750189250567752,
|
| 2260 |
+
"grad_norm": NaN,
|
| 2261 |
+
"learning_rate": 2.898757686722542e-07,
|
| 2262 |
+
"loss": 0.0,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 9.780469341408024,
|
| 2267 |
+
"grad_norm": NaN,
|
| 2268 |
+
"learning_rate": 2.219612815434924e-07,
|
| 2269 |
+
"loss": 0.0,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 9.810749432248297,
|
| 2274 |
+
"grad_norm": NaN,
|
| 2275 |
+
"learning_rate": 1.630896073864352e-07,
|
| 2276 |
+
"loss": 0.0,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 9.841029523088569,
|
| 2281 |
+
"grad_norm": NaN,
|
| 2282 |
+
"learning_rate": 1.1326608169920372e-07,
|
| 2283 |
+
"loss": 0.0,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 9.871309613928842,
|
| 2288 |
+
"grad_norm": NaN,
|
| 2289 |
+
"learning_rate": 7.249521995263964e-08,
|
| 2290 |
+
"loss": 0.0,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 9.901589704769114,
|
| 2295 |
+
"grad_norm": NaN,
|
| 2296 |
+
"learning_rate": 4.078071718107701e-08,
|
| 2297 |
+
"loss": 0.0,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 9.931869795609387,
|
| 2302 |
+
"grad_norm": NaN,
|
| 2303 |
+
"learning_rate": 1.81254476474213e-08,
|
| 2304 |
+
"loss": 0.0,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 9.962149886449659,
|
| 2309 |
+
"grad_norm": NaN,
|
| 2310 |
+
"learning_rate": 4.531464582713252e-09,
|
| 2311 |
+
"loss": 0.0,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 9.992429977289932,
|
| 2316 |
+
"grad_norm": NaN,
|
| 2317 |
+
"learning_rate": 0.0,
|
| 2318 |
+
"loss": 0.0,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 9.992429977289932,
|
| 2323 |
+
"step": 3300,
|
| 2324 |
+
"total_flos": 1.623229319872512e+17,
|
| 2325 |
+
"train_loss": 0.0,
|
| 2326 |
+
"train_runtime": 3221.4406,
|
| 2327 |
+
"train_samples_per_second": 4.101,
|
| 2328 |
+
"train_steps_per_second": 1.024
|
| 2329 |
+
}
|
| 2330 |
+
],
|
| 2331 |
+
"logging_steps": 10,
|
| 2332 |
+
"max_steps": 3300,
|
| 2333 |
+
"num_input_tokens_seen": 0,
|
| 2334 |
+
"num_train_epochs": 10,
|
| 2335 |
+
"save_steps": 500,
|
| 2336 |
+
"total_flos": 1.623229319872512e+17,
|
| 2337 |
+
"train_batch_size": 1,
|
| 2338 |
+
"trial_name": null,
|
| 2339 |
+
"trial_params": null
|
| 2340 |
+
}
|
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 sha256:44d636e57f19c91fd4f5087f4884d8ce9c533ae1bb1da60ec048d48942764c3b
|
| 3 |
size 5048
|