Commit
•
a66cf45
1
Parent(s):
d63cffc
End of training
Browse files- README.md +3 -1
- trainer_state.json +1053 -59
README.md
CHANGED
|
@@ -3,6 +3,8 @@ base_model: facebook/musicgen-melody
|
|
| 3 |
library_name: peft
|
| 4 |
license: cc-by-nc-4.0
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
model-index:
|
| 8 |
- name: Classical-Data
|
|
@@ -14,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# Classical-Data
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 3 |
library_name: peft
|
| 4 |
license: cc-by-nc-4.0
|
| 5 |
tags:
|
| 6 |
+
- text-to-audio
|
| 7 |
+
- Punk
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: Classical-Data
|
|
|
|
| 16 |
|
| 17 |
# Classical-Data
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on the PlutoG99001/Classical-clean dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
trainer_state.json
CHANGED
|
@@ -1,109 +1,1103 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 3.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 0.
|
| 15 |
-
"loss":
|
| 16 |
"step": 2
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 0.
|
| 22 |
-
"loss":
|
| 23 |
"step": 4
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 0.
|
| 29 |
-
"loss": 8.
|
| 30 |
"step": 6
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch":
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 0.
|
| 36 |
-
"loss":
|
| 37 |
"step": 8
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch":
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 0.
|
| 43 |
-
"loss": 7.
|
| 44 |
"step": 10
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch":
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss":
|
| 51 |
"step": 12
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch":
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss":
|
| 58 |
"step": 14
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch":
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss":
|
| 65 |
"step": 16
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch":
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss":
|
| 72 |
"step": 18
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch":
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss":
|
| 79 |
"step": 20
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"grad_norm": 1.
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss":
|
| 86 |
"step": 22
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch":
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss":
|
| 93 |
"step": 24
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"
|
| 98 |
-
"
|
| 99 |
-
"
|
| 100 |
-
"
|
| 101 |
-
|
| 102 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
}
|
| 104 |
],
|
| 105 |
"logging_steps": 2,
|
| 106 |
-
"max_steps":
|
| 107 |
"num_input_tokens_seen": 0,
|
| 108 |
"num_train_epochs": 4,
|
| 109 |
"save_steps": 500,
|
|
@@ -119,7 +1113,7 @@
|
|
| 119 |
"attributes": {}
|
| 120 |
}
|
| 121 |
},
|
| 122 |
-
"total_flos":
|
| 123 |
"train_batch_size": 2,
|
| 124 |
"trial_name": null,
|
| 125 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.9741935483870967,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 308,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.025806451612903226,
|
| 13 |
+
"grad_norm": 3.5142855644226074,
|
| 14 |
+
"learning_rate": 0.00019870129870129872,
|
| 15 |
+
"loss": 8.9214,
|
| 16 |
"step": 2
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.05161290322580645,
|
| 20 |
+
"grad_norm": 3.772717237472534,
|
| 21 |
+
"learning_rate": 0.00019740259740259742,
|
| 22 |
+
"loss": 8.3405,
|
| 23 |
"step": 4
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.07741935483870968,
|
| 27 |
+
"grad_norm": 6.7029924392700195,
|
| 28 |
+
"learning_rate": 0.00019610389610389613,
|
| 29 |
+
"loss": 8.1379,
|
| 30 |
"step": 6
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.1032258064516129,
|
| 34 |
+
"grad_norm": 3.7924954891204834,
|
| 35 |
+
"learning_rate": 0.0001948051948051948,
|
| 36 |
+
"loss": 5.913,
|
| 37 |
"step": 8
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.12903225806451613,
|
| 41 |
+
"grad_norm": 4.214364051818848,
|
| 42 |
+
"learning_rate": 0.00019350649350649354,
|
| 43 |
+
"loss": 7.0624,
|
| 44 |
"step": 10
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.15483870967741936,
|
| 48 |
+
"grad_norm": 3.0830295085906982,
|
| 49 |
+
"learning_rate": 0.00019220779220779222,
|
| 50 |
+
"loss": 6.6031,
|
| 51 |
"step": 12
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.18064516129032257,
|
| 55 |
+
"grad_norm": 2.293950080871582,
|
| 56 |
+
"learning_rate": 0.00019090909090909092,
|
| 57 |
+
"loss": 6.1543,
|
| 58 |
"step": 14
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.2064516129032258,
|
| 62 |
+
"grad_norm": 2.8121275901794434,
|
| 63 |
+
"learning_rate": 0.00018961038961038963,
|
| 64 |
+
"loss": 5.6952,
|
| 65 |
"step": 16
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.23225806451612904,
|
| 69 |
+
"grad_norm": 3.2042553424835205,
|
| 70 |
+
"learning_rate": 0.00018831168831168833,
|
| 71 |
+
"loss": 5.6916,
|
| 72 |
"step": 18
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.25806451612903225,
|
| 76 |
+
"grad_norm": 3.4788341522216797,
|
| 77 |
+
"learning_rate": 0.000187012987012987,
|
| 78 |
+
"loss": 5.575,
|
| 79 |
"step": 20
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.2838709677419355,
|
| 83 |
+
"grad_norm": 1.1461117267608643,
|
| 84 |
+
"learning_rate": 0.00018571428571428572,
|
| 85 |
+
"loss": 5.7922,
|
| 86 |
"step": 22
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.3096774193548387,
|
| 90 |
+
"grad_norm": 1.7745267152786255,
|
| 91 |
+
"learning_rate": 0.00018441558441558442,
|
| 92 |
+
"loss": 5.5945,
|
| 93 |
"step": 24
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.33548387096774196,
|
| 97 |
+
"grad_norm": 1.0893975496292114,
|
| 98 |
+
"learning_rate": 0.00018311688311688313,
|
| 99 |
+
"loss": 6.1593,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.36129032258064514,
|
| 104 |
+
"grad_norm": 2.169384241104126,
|
| 105 |
+
"learning_rate": 0.00018181818181818183,
|
| 106 |
+
"loss": 5.8243,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3870967741935484,
|
| 111 |
+
"grad_norm": 1.8578031063079834,
|
| 112 |
+
"learning_rate": 0.00018051948051948054,
|
| 113 |
+
"loss": 5.5041,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4129032258064516,
|
| 118 |
+
"grad_norm": 1.448111653327942,
|
| 119 |
+
"learning_rate": 0.00017922077922077922,
|
| 120 |
+
"loss": 5.3471,
|
| 121 |
+
"step": 32
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43870967741935485,
|
| 125 |
+
"grad_norm": 1.828591227531433,
|
| 126 |
+
"learning_rate": 0.00017792207792207792,
|
| 127 |
+
"loss": 5.0863,
|
| 128 |
+
"step": 34
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4645161290322581,
|
| 132 |
+
"grad_norm": 4.83806848526001,
|
| 133 |
+
"learning_rate": 0.00017662337662337663,
|
| 134 |
+
"loss": 6.5823,
|
| 135 |
+
"step": 36
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.49032258064516127,
|
| 139 |
+
"grad_norm": 2.1270275115966797,
|
| 140 |
+
"learning_rate": 0.00017532467532467534,
|
| 141 |
+
"loss": 5.3268,
|
| 142 |
+
"step": 38
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5161290322580645,
|
| 146 |
+
"grad_norm": 1.42393958568573,
|
| 147 |
+
"learning_rate": 0.00017402597402597401,
|
| 148 |
+
"loss": 5.4203,
|
| 149 |
+
"step": 40
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5419354838709678,
|
| 153 |
+
"grad_norm": 2.768852472305298,
|
| 154 |
+
"learning_rate": 0.00017272727272727275,
|
| 155 |
+
"loss": 5.3403,
|
| 156 |
+
"step": 42
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.567741935483871,
|
| 160 |
+
"grad_norm": 0.7816085815429688,
|
| 161 |
+
"learning_rate": 0.00017142857142857143,
|
| 162 |
+
"loss": 5.6736,
|
| 163 |
+
"step": 44
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5935483870967742,
|
| 167 |
+
"grad_norm": 0.8368009924888611,
|
| 168 |
+
"learning_rate": 0.00017012987012987013,
|
| 169 |
+
"loss": 5.4177,
|
| 170 |
+
"step": 46
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6193548387096774,
|
| 174 |
+
"grad_norm": 1.7966578006744385,
|
| 175 |
+
"learning_rate": 0.00016883116883116884,
|
| 176 |
+
"loss": 5.8706,
|
| 177 |
+
"step": 48
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6451612903225806,
|
| 181 |
+
"grad_norm": 0.694948673248291,
|
| 182 |
+
"learning_rate": 0.00016753246753246754,
|
| 183 |
+
"loss": 5.2186,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6709677419354839,
|
| 188 |
+
"grad_norm": 2.0851364135742188,
|
| 189 |
+
"learning_rate": 0.00016623376623376625,
|
| 190 |
+
"loss": 5.452,
|
| 191 |
+
"step": 52
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6967741935483871,
|
| 195 |
+
"grad_norm": 1.705665946006775,
|
| 196 |
+
"learning_rate": 0.00016493506493506495,
|
| 197 |
+
"loss": 5.5373,
|
| 198 |
+
"step": 54
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7225806451612903,
|
| 202 |
+
"grad_norm": 1.5386658906936646,
|
| 203 |
+
"learning_rate": 0.00016363636363636366,
|
| 204 |
+
"loss": 5.8638,
|
| 205 |
+
"step": 56
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7483870967741936,
|
| 209 |
+
"grad_norm": 1.3625737428665161,
|
| 210 |
+
"learning_rate": 0.00016233766233766234,
|
| 211 |
+
"loss": 5.2913,
|
| 212 |
+
"step": 58
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7741935483870968,
|
| 216 |
+
"grad_norm": 1.357426643371582,
|
| 217 |
+
"learning_rate": 0.00016103896103896104,
|
| 218 |
+
"loss": 5.7317,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8,
|
| 223 |
+
"grad_norm": 1.1882686614990234,
|
| 224 |
+
"learning_rate": 0.00015974025974025975,
|
| 225 |
+
"loss": 5.9749,
|
| 226 |
+
"step": 62
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8258064516129032,
|
| 230 |
+
"grad_norm": 1.0777816772460938,
|
| 231 |
+
"learning_rate": 0.00015844155844155845,
|
| 232 |
+
"loss": 5.4791,
|
| 233 |
+
"step": 64
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8516129032258064,
|
| 237 |
+
"grad_norm": 0.7732660174369812,
|
| 238 |
+
"learning_rate": 0.00015714285714285716,
|
| 239 |
+
"loss": 5.9732,
|
| 240 |
+
"step": 66
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8774193548387097,
|
| 244 |
+
"grad_norm": 1.0730479955673218,
|
| 245 |
+
"learning_rate": 0.00015584415584415587,
|
| 246 |
+
"loss": 5.4338,
|
| 247 |
+
"step": 68
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9032258064516129,
|
| 251 |
+
"grad_norm": 1.4834214448928833,
|
| 252 |
+
"learning_rate": 0.00015454545454545454,
|
| 253 |
+
"loss": 5.1802,
|
| 254 |
+
"step": 70
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9290322580645162,
|
| 258 |
+
"grad_norm": 2.2870066165924072,
|
| 259 |
+
"learning_rate": 0.00015324675324675325,
|
| 260 |
+
"loss": 6.3573,
|
| 261 |
+
"step": 72
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9548387096774194,
|
| 265 |
+
"grad_norm": 0.935196578502655,
|
| 266 |
+
"learning_rate": 0.00015194805194805196,
|
| 267 |
+
"loss": 5.0633,
|
| 268 |
+
"step": 74
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9806451612903225,
|
| 272 |
+
"grad_norm": 0.994687020778656,
|
| 273 |
+
"learning_rate": 0.00015064935064935066,
|
| 274 |
+
"loss": 5.8048,
|
| 275 |
+
"step": 76
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.0064516129032257,
|
| 279 |
+
"grad_norm": 2.274411201477051,
|
| 280 |
+
"learning_rate": 0.00014935064935064934,
|
| 281 |
+
"loss": 4.6835,
|
| 282 |
+
"step": 78
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.032258064516129,
|
| 286 |
+
"grad_norm": 1.2067323923110962,
|
| 287 |
+
"learning_rate": 0.00014805194805194807,
|
| 288 |
+
"loss": 5.4713,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.0580645161290323,
|
| 293 |
+
"grad_norm": 0.9473636150360107,
|
| 294 |
+
"learning_rate": 0.00014675324675324675,
|
| 295 |
+
"loss": 5.8508,
|
| 296 |
+
"step": 82
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.0838709677419356,
|
| 300 |
+
"grad_norm": 5.7191996574401855,
|
| 301 |
+
"learning_rate": 0.00014545454545454546,
|
| 302 |
+
"loss": 4.9011,
|
| 303 |
+
"step": 84
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.1096774193548387,
|
| 307 |
+
"grad_norm": 2.201740026473999,
|
| 308 |
+
"learning_rate": 0.00014415584415584416,
|
| 309 |
+
"loss": 5.4717,
|
| 310 |
+
"step": 86
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.135483870967742,
|
| 314 |
+
"grad_norm": 1.0441229343414307,
|
| 315 |
+
"learning_rate": 0.00014285714285714287,
|
| 316 |
+
"loss": 5.5206,
|
| 317 |
+
"step": 88
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.1612903225806452,
|
| 321 |
+
"grad_norm": 1.0501593351364136,
|
| 322 |
+
"learning_rate": 0.00014155844155844155,
|
| 323 |
+
"loss": 5.2199,
|
| 324 |
+
"step": 90
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.1870967741935483,
|
| 328 |
+
"grad_norm": 1.1907446384429932,
|
| 329 |
+
"learning_rate": 0.00014025974025974028,
|
| 330 |
+
"loss": 4.8319,
|
| 331 |
+
"step": 92
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.2129032258064516,
|
| 335 |
+
"grad_norm": 4.133155345916748,
|
| 336 |
+
"learning_rate": 0.00013896103896103896,
|
| 337 |
+
"loss": 5.1049,
|
| 338 |
+
"step": 94
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.238709677419355,
|
| 342 |
+
"grad_norm": 1.399916172027588,
|
| 343 |
+
"learning_rate": 0.00013766233766233766,
|
| 344 |
+
"loss": 5.5704,
|
| 345 |
+
"step": 96
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.2645161290322582,
|
| 349 |
+
"grad_norm": 1.0567469596862793,
|
| 350 |
+
"learning_rate": 0.00013636363636363637,
|
| 351 |
+
"loss": 5.3263,
|
| 352 |
+
"step": 98
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.2903225806451613,
|
| 356 |
+
"grad_norm": 1.2196253538131714,
|
| 357 |
+
"learning_rate": 0.00013506493506493507,
|
| 358 |
+
"loss": 5.2236,
|
| 359 |
+
"step": 100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3161290322580645,
|
| 363 |
+
"grad_norm": 2.465505838394165,
|
| 364 |
+
"learning_rate": 0.00013376623376623375,
|
| 365 |
+
"loss": 5.0547,
|
| 366 |
+
"step": 102
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.3419354838709676,
|
| 370 |
+
"grad_norm": 0.9022129774093628,
|
| 371 |
+
"learning_rate": 0.00013246753246753249,
|
| 372 |
+
"loss": 5.8387,
|
| 373 |
+
"step": 104
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.367741935483871,
|
| 377 |
+
"grad_norm": 1.6524704694747925,
|
| 378 |
+
"learning_rate": 0.0001311688311688312,
|
| 379 |
+
"loss": 5.4351,
|
| 380 |
+
"step": 106
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.3935483870967742,
|
| 384 |
+
"grad_norm": 1.2133516073226929,
|
| 385 |
+
"learning_rate": 0.00012987012987012987,
|
| 386 |
+
"loss": 5.4708,
|
| 387 |
+
"step": 108
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.4193548387096775,
|
| 391 |
+
"grad_norm": 1.019903302192688,
|
| 392 |
+
"learning_rate": 0.00012857142857142858,
|
| 393 |
+
"loss": 5.8655,
|
| 394 |
+
"step": 110
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.4451612903225808,
|
| 398 |
+
"grad_norm": 1.1822901964187622,
|
| 399 |
+
"learning_rate": 0.00012727272727272728,
|
| 400 |
+
"loss": 5.3089,
|
| 401 |
+
"step": 112
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.4709677419354839,
|
| 405 |
+
"grad_norm": 0.8840041160583496,
|
| 406 |
+
"learning_rate": 0.000125974025974026,
|
| 407 |
+
"loss": 5.0657,
|
| 408 |
+
"step": 114
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.4967741935483871,
|
| 412 |
+
"grad_norm": 0.6745492219924927,
|
| 413 |
+
"learning_rate": 0.00012467532467532467,
|
| 414 |
+
"loss": 4.8004,
|
| 415 |
+
"step": 116
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.5225806451612902,
|
| 419 |
+
"grad_norm": 1.4586549997329712,
|
| 420 |
+
"learning_rate": 0.0001233766233766234,
|
| 421 |
+
"loss": 5.0572,
|
| 422 |
+
"step": 118
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.5483870967741935,
|
| 426 |
+
"grad_norm": 1.5153924226760864,
|
| 427 |
+
"learning_rate": 0.00012207792207792208,
|
| 428 |
+
"loss": 4.4382,
|
| 429 |
+
"step": 120
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.5741935483870968,
|
| 433 |
+
"grad_norm": 0.9981886148452759,
|
| 434 |
+
"learning_rate": 0.0001207792207792208,
|
| 435 |
+
"loss": 4.8087,
|
| 436 |
+
"step": 122
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.6,
|
| 440 |
+
"grad_norm": 0.7403278350830078,
|
| 441 |
+
"learning_rate": 0.00011948051948051949,
|
| 442 |
+
"loss": 5.2025,
|
| 443 |
+
"step": 124
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.6258064516129034,
|
| 447 |
+
"grad_norm": 1.1075459718704224,
|
| 448 |
+
"learning_rate": 0.0001181818181818182,
|
| 449 |
+
"loss": 4.946,
|
| 450 |
+
"step": 126
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.6516129032258065,
|
| 454 |
+
"grad_norm": 0.6300385594367981,
|
| 455 |
+
"learning_rate": 0.00011688311688311689,
|
| 456 |
+
"loss": 5.6817,
|
| 457 |
+
"step": 128
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.6774193548387095,
|
| 461 |
+
"grad_norm": 0.8701953291893005,
|
| 462 |
+
"learning_rate": 0.00011558441558441559,
|
| 463 |
+
"loss": 5.6561,
|
| 464 |
+
"step": 130
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.7032258064516128,
|
| 468 |
+
"grad_norm": 1.133817434310913,
|
| 469 |
+
"learning_rate": 0.00011428571428571428,
|
| 470 |
+
"loss": 5.6528,
|
| 471 |
+
"step": 132
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.729032258064516,
|
| 475 |
+
"grad_norm": 1.6048352718353271,
|
| 476 |
+
"learning_rate": 0.000112987012987013,
|
| 477 |
+
"loss": 4.8602,
|
| 478 |
+
"step": 134
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.7548387096774194,
|
| 482 |
+
"grad_norm": 1.2783055305480957,
|
| 483 |
+
"learning_rate": 0.00011168831168831168,
|
| 484 |
+
"loss": 4.1662,
|
| 485 |
+
"step": 136
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.7806451612903227,
|
| 489 |
+
"grad_norm": 1.6137133836746216,
|
| 490 |
+
"learning_rate": 0.0001103896103896104,
|
| 491 |
+
"loss": 5.2295,
|
| 492 |
+
"step": 138
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.8064516129032258,
|
| 496 |
+
"grad_norm": 0.6374461054801941,
|
| 497 |
+
"learning_rate": 0.00010909090909090909,
|
| 498 |
+
"loss": 5.3992,
|
| 499 |
+
"step": 140
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.832258064516129,
|
| 503 |
+
"grad_norm": 0.8323061466217041,
|
| 504 |
+
"learning_rate": 0.0001077922077922078,
|
| 505 |
+
"loss": 5.1712,
|
| 506 |
+
"step": 142
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.8580645161290321,
|
| 510 |
+
"grad_norm": 0.6804484724998474,
|
| 511 |
+
"learning_rate": 0.00010649350649350649,
|
| 512 |
+
"loss": 5.3369,
|
| 513 |
+
"step": 144
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.8838709677419354,
|
| 517 |
+
"grad_norm": 0.9051455855369568,
|
| 518 |
+
"learning_rate": 0.0001051948051948052,
|
| 519 |
+
"loss": 5.3726,
|
| 520 |
+
"step": 146
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.9096774193548387,
|
| 524 |
+
"grad_norm": 0.5890415906906128,
|
| 525 |
+
"learning_rate": 0.00010389610389610389,
|
| 526 |
+
"loss": 5.4903,
|
| 527 |
+
"step": 148
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.935483870967742,
|
| 531 |
+
"grad_norm": 0.5433252453804016,
|
| 532 |
+
"learning_rate": 0.00010259740259740261,
|
| 533 |
+
"loss": 5.4437,
|
| 534 |
+
"step": 150
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.9612903225806453,
|
| 538 |
+
"grad_norm": 1.353121042251587,
|
| 539 |
+
"learning_rate": 0.0001012987012987013,
|
| 540 |
+
"loss": 4.1137,
|
| 541 |
+
"step": 152
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.9870967741935484,
|
| 545 |
+
"grad_norm": 1.0199609994888306,
|
| 546 |
+
"learning_rate": 0.0001,
|
| 547 |
+
"loss": 5.2558,
|
| 548 |
+
"step": 154
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.0129032258064514,
|
| 552 |
+
"grad_norm": 0.6021209359169006,
|
| 553 |
+
"learning_rate": 9.870129870129871e-05,
|
| 554 |
+
"loss": 5.5338,
|
| 555 |
+
"step": 156
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.0387096774193547,
|
| 559 |
+
"grad_norm": 0.9102515578269958,
|
| 560 |
+
"learning_rate": 9.74025974025974e-05,
|
| 561 |
+
"loss": 4.5988,
|
| 562 |
+
"step": 158
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.064516129032258,
|
| 566 |
+
"grad_norm": 0.7101506590843201,
|
| 567 |
+
"learning_rate": 9.610389610389611e-05,
|
| 568 |
+
"loss": 5.2622,
|
| 569 |
+
"step": 160
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.0903225806451613,
|
| 573 |
+
"grad_norm": 0.631308913230896,
|
| 574 |
+
"learning_rate": 9.480519480519481e-05,
|
| 575 |
+
"loss": 5.4718,
|
| 576 |
+
"step": 162
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.1161290322580646,
|
| 580 |
+
"grad_norm": 0.7462102770805359,
|
| 581 |
+
"learning_rate": 9.35064935064935e-05,
|
| 582 |
+
"loss": 5.0117,
|
| 583 |
+
"step": 164
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.141935483870968,
|
| 587 |
+
"grad_norm": 0.6367154717445374,
|
| 588 |
+
"learning_rate": 9.220779220779221e-05,
|
| 589 |
+
"loss": 4.9968,
|
| 590 |
+
"step": 166
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.167741935483871,
|
| 594 |
+
"grad_norm": 0.6959227323532104,
|
| 595 |
+
"learning_rate": 9.090909090909092e-05,
|
| 596 |
+
"loss": 5.6022,
|
| 597 |
+
"step": 168
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.193548387096774,
|
| 601 |
+
"grad_norm": 0.6119377613067627,
|
| 602 |
+
"learning_rate": 8.961038961038961e-05,
|
| 603 |
+
"loss": 4.9405,
|
| 604 |
+
"step": 170
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.2193548387096773,
|
| 608 |
+
"grad_norm": 0.6411863565444946,
|
| 609 |
+
"learning_rate": 8.831168831168831e-05,
|
| 610 |
+
"loss": 5.1413,
|
| 611 |
+
"step": 172
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.2451612903225806,
|
| 615 |
+
"grad_norm": 1.615324854850769,
|
| 616 |
+
"learning_rate": 8.701298701298701e-05,
|
| 617 |
+
"loss": 4.7308,
|
| 618 |
+
"step": 174
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.270967741935484,
|
| 622 |
+
"grad_norm": 0.5708146691322327,
|
| 623 |
+
"learning_rate": 8.571428571428571e-05,
|
| 624 |
+
"loss": 5.4497,
|
| 625 |
+
"step": 176
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.296774193548387,
|
| 629 |
+
"grad_norm": 0.8313891291618347,
|
| 630 |
+
"learning_rate": 8.441558441558442e-05,
|
| 631 |
+
"loss": 5.3687,
|
| 632 |
+
"step": 178
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.3225806451612905,
|
| 636 |
+
"grad_norm": 1.1731419563293457,
|
| 637 |
+
"learning_rate": 8.311688311688312e-05,
|
| 638 |
+
"loss": 5.3314,
|
| 639 |
+
"step": 180
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.3483870967741938,
|
| 643 |
+
"grad_norm": 0.7381497025489807,
|
| 644 |
+
"learning_rate": 8.181818181818183e-05,
|
| 645 |
+
"loss": 5.1178,
|
| 646 |
+
"step": 182
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.3741935483870966,
|
| 650 |
+
"grad_norm": 0.6883618831634521,
|
| 651 |
+
"learning_rate": 8.051948051948052e-05,
|
| 652 |
+
"loss": 5.9266,
|
| 653 |
+
"step": 184
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.4,
|
| 657 |
+
"grad_norm": 0.735080361366272,
|
| 658 |
+
"learning_rate": 7.922077922077923e-05,
|
| 659 |
+
"loss": 4.7453,
|
| 660 |
+
"step": 186
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.425806451612903,
|
| 664 |
+
"grad_norm": 0.6184589862823486,
|
| 665 |
+
"learning_rate": 7.792207792207793e-05,
|
| 666 |
+
"loss": 5.0924,
|
| 667 |
+
"step": 188
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.4516129032258065,
|
| 671 |
+
"grad_norm": 0.6305899024009705,
|
| 672 |
+
"learning_rate": 7.662337662337662e-05,
|
| 673 |
+
"loss": 5.1167,
|
| 674 |
+
"step": 190
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.47741935483871,
|
| 678 |
+
"grad_norm": 0.815334677696228,
|
| 679 |
+
"learning_rate": 7.532467532467533e-05,
|
| 680 |
+
"loss": 5.4719,
|
| 681 |
+
"step": 192
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.5032258064516126,
|
| 685 |
+
"grad_norm": 0.6805923581123352,
|
| 686 |
+
"learning_rate": 7.402597402597404e-05,
|
| 687 |
+
"loss": 4.4392,
|
| 688 |
+
"step": 194
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.5290322580645164,
|
| 692 |
+
"grad_norm": 0.6633741855621338,
|
| 693 |
+
"learning_rate": 7.272727272727273e-05,
|
| 694 |
+
"loss": 5.1926,
|
| 695 |
+
"step": 196
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.554838709677419,
|
| 699 |
+
"grad_norm": 0.6042450666427612,
|
| 700 |
+
"learning_rate": 7.142857142857143e-05,
|
| 701 |
+
"loss": 4.7079,
|
| 702 |
+
"step": 198
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.5806451612903225,
|
| 706 |
+
"grad_norm": 0.5892207622528076,
|
| 707 |
+
"learning_rate": 7.012987012987014e-05,
|
| 708 |
+
"loss": 4.9938,
|
| 709 |
+
"step": 200
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.606451612903226,
|
| 713 |
+
"grad_norm": 0.6753908395767212,
|
| 714 |
+
"learning_rate": 6.883116883116883e-05,
|
| 715 |
+
"loss": 5.0097,
|
| 716 |
+
"step": 202
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.632258064516129,
|
| 720 |
+
"grad_norm": 0.5345973968505859,
|
| 721 |
+
"learning_rate": 6.753246753246754e-05,
|
| 722 |
+
"loss": 5.1074,
|
| 723 |
+
"step": 204
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.6580645161290324,
|
| 727 |
+
"grad_norm": 0.4924313724040985,
|
| 728 |
+
"learning_rate": 6.623376623376624e-05,
|
| 729 |
+
"loss": 5.0817,
|
| 730 |
+
"step": 206
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.6838709677419352,
|
| 734 |
+
"grad_norm": 0.5989976525306702,
|
| 735 |
+
"learning_rate": 6.493506493506494e-05,
|
| 736 |
+
"loss": 4.2039,
|
| 737 |
+
"step": 208
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.709677419354839,
|
| 741 |
+
"grad_norm": 1.0730255842208862,
|
| 742 |
+
"learning_rate": 6.363636363636364e-05,
|
| 743 |
+
"loss": 4.2224,
|
| 744 |
+
"step": 210
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.735483870967742,
|
| 748 |
+
"grad_norm": 0.991038978099823,
|
| 749 |
+
"learning_rate": 6.233766233766233e-05,
|
| 750 |
+
"loss": 4.8869,
|
| 751 |
+
"step": 212
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.761290322580645,
|
| 755 |
+
"grad_norm": 0.5952357053756714,
|
| 756 |
+
"learning_rate": 6.103896103896104e-05,
|
| 757 |
+
"loss": 5.1753,
|
| 758 |
+
"step": 214
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.7870967741935484,
|
| 762 |
+
"grad_norm": 0.6798732876777649,
|
| 763 |
+
"learning_rate": 5.9740259740259744e-05,
|
| 764 |
+
"loss": 4.3396,
|
| 765 |
+
"step": 216
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.8129032258064517,
|
| 769 |
+
"grad_norm": 0.5361295342445374,
|
| 770 |
+
"learning_rate": 5.844155844155844e-05,
|
| 771 |
+
"loss": 5.2857,
|
| 772 |
+
"step": 218
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.838709677419355,
|
| 776 |
+
"grad_norm": 0.6326772570610046,
|
| 777 |
+
"learning_rate": 5.714285714285714e-05,
|
| 778 |
+
"loss": 4.7385,
|
| 779 |
+
"step": 220
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.864516129032258,
|
| 783 |
+
"grad_norm": 0.639589786529541,
|
| 784 |
+
"learning_rate": 5.584415584415584e-05,
|
| 785 |
+
"loss": 4.3803,
|
| 786 |
+
"step": 222
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.8903225806451616,
|
| 790 |
+
"grad_norm": 0.7248474955558777,
|
| 791 |
+
"learning_rate": 5.4545454545454546e-05,
|
| 792 |
+
"loss": 4.6668,
|
| 793 |
+
"step": 224
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.9161290322580644,
|
| 797 |
+
"grad_norm": 0.7551538348197937,
|
| 798 |
+
"learning_rate": 5.3246753246753245e-05,
|
| 799 |
+
"loss": 5.016,
|
| 800 |
+
"step": 226
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.9419354838709677,
|
| 804 |
+
"grad_norm": 0.4990728497505188,
|
| 805 |
+
"learning_rate": 5.1948051948051944e-05,
|
| 806 |
+
"loss": 5.0843,
|
| 807 |
+
"step": 228
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.967741935483871,
|
| 811 |
+
"grad_norm": 0.5729503035545349,
|
| 812 |
+
"learning_rate": 5.064935064935065e-05,
|
| 813 |
+
"loss": 4.573,
|
| 814 |
+
"step": 230
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.9935483870967743,
|
| 818 |
+
"grad_norm": 1.4705737829208374,
|
| 819 |
+
"learning_rate": 4.9350649350649355e-05,
|
| 820 |
+
"loss": 4.9812,
|
| 821 |
+
"step": 232
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.0193548387096776,
|
| 825 |
+
"grad_norm": 0.7053755521774292,
|
| 826 |
+
"learning_rate": 4.8051948051948054e-05,
|
| 827 |
+
"loss": 4.7651,
|
| 828 |
+
"step": 234
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.0451612903225804,
|
| 832 |
+
"grad_norm": 0.6120907664299011,
|
| 833 |
+
"learning_rate": 4.675324675324675e-05,
|
| 834 |
+
"loss": 5.2057,
|
| 835 |
+
"step": 236
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.0709677419354837,
|
| 839 |
+
"grad_norm": 0.6173492074012756,
|
| 840 |
+
"learning_rate": 4.545454545454546e-05,
|
| 841 |
+
"loss": 5.0342,
|
| 842 |
+
"step": 238
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.096774193548387,
|
| 846 |
+
"grad_norm": 0.5435605049133301,
|
| 847 |
+
"learning_rate": 4.415584415584416e-05,
|
| 848 |
+
"loss": 4.9538,
|
| 849 |
+
"step": 240
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.1225806451612903,
|
| 853 |
+
"grad_norm": 1.4004778861999512,
|
| 854 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 855 |
+
"loss": 4.3856,
|
| 856 |
+
"step": 242
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.1483870967741936,
|
| 860 |
+
"grad_norm": 1.1300957202911377,
|
| 861 |
+
"learning_rate": 4.155844155844156e-05,
|
| 862 |
+
"loss": 5.0085,
|
| 863 |
+
"step": 244
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.174193548387097,
|
| 867 |
+
"grad_norm": 0.6826758980751038,
|
| 868 |
+
"learning_rate": 4.025974025974026e-05,
|
| 869 |
+
"loss": 5.3525,
|
| 870 |
+
"step": 246
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.2,
|
| 874 |
+
"grad_norm": 0.6162336468696594,
|
| 875 |
+
"learning_rate": 3.8961038961038966e-05,
|
| 876 |
+
"loss": 5.1639,
|
| 877 |
+
"step": 248
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.225806451612903,
|
| 881 |
+
"grad_norm": 0.7551366686820984,
|
| 882 |
+
"learning_rate": 3.7662337662337665e-05,
|
| 883 |
+
"loss": 5.1913,
|
| 884 |
+
"step": 250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.2516129032258063,
|
| 888 |
+
"grad_norm": 0.5899360775947571,
|
| 889 |
+
"learning_rate": 3.6363636363636364e-05,
|
| 890 |
+
"loss": 5.022,
|
| 891 |
+
"step": 252
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.2774193548387096,
|
| 895 |
+
"grad_norm": 0.6666110157966614,
|
| 896 |
+
"learning_rate": 3.506493506493507e-05,
|
| 897 |
+
"loss": 4.697,
|
| 898 |
+
"step": 254
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.303225806451613,
|
| 902 |
+
"grad_norm": 0.5686184167861938,
|
| 903 |
+
"learning_rate": 3.376623376623377e-05,
|
| 904 |
+
"loss": 4.8758,
|
| 905 |
+
"step": 256
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.329032258064516,
|
| 909 |
+
"grad_norm": 0.7564727663993835,
|
| 910 |
+
"learning_rate": 3.246753246753247e-05,
|
| 911 |
+
"loss": 4.9468,
|
| 912 |
+
"step": 258
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.3548387096774195,
|
| 916 |
+
"grad_norm": 0.5773691534996033,
|
| 917 |
+
"learning_rate": 3.1168831168831166e-05,
|
| 918 |
+
"loss": 5.1122,
|
| 919 |
+
"step": 260
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.3806451612903228,
|
| 923 |
+
"grad_norm": 0.642393171787262,
|
| 924 |
+
"learning_rate": 2.9870129870129872e-05,
|
| 925 |
+
"loss": 5.233,
|
| 926 |
+
"step": 262
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.4064516129032256,
|
| 930 |
+
"grad_norm": 0.6513245701789856,
|
| 931 |
+
"learning_rate": 2.857142857142857e-05,
|
| 932 |
+
"loss": 4.2779,
|
| 933 |
+
"step": 264
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.432258064516129,
|
| 937 |
+
"grad_norm": 0.8267136812210083,
|
| 938 |
+
"learning_rate": 2.7272727272727273e-05,
|
| 939 |
+
"loss": 5.2998,
|
| 940 |
+
"step": 266
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.458064516129032,
|
| 944 |
+
"grad_norm": 0.5494163632392883,
|
| 945 |
+
"learning_rate": 2.5974025974025972e-05,
|
| 946 |
+
"loss": 5.4667,
|
| 947 |
+
"step": 268
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.4838709677419355,
|
| 951 |
+
"grad_norm": 0.7057967782020569,
|
| 952 |
+
"learning_rate": 2.4675324675324678e-05,
|
| 953 |
+
"loss": 5.1893,
|
| 954 |
+
"step": 270
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.509677419354839,
|
| 958 |
+
"grad_norm": 0.5400364398956299,
|
| 959 |
+
"learning_rate": 2.3376623376623376e-05,
|
| 960 |
+
"loss": 4.9972,
|
| 961 |
+
"step": 272
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.535483870967742,
|
| 965 |
+
"grad_norm": 0.4732670485973358,
|
| 966 |
+
"learning_rate": 2.207792207792208e-05,
|
| 967 |
+
"loss": 4.896,
|
| 968 |
+
"step": 274
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.5612903225806454,
|
| 972 |
+
"grad_norm": 0.5432953834533691,
|
| 973 |
+
"learning_rate": 2.077922077922078e-05,
|
| 974 |
+
"loss": 5.0452,
|
| 975 |
+
"step": 276
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.587096774193548,
|
| 979 |
+
"grad_norm": 0.5461270213127136,
|
| 980 |
+
"learning_rate": 1.9480519480519483e-05,
|
| 981 |
+
"loss": 4.7124,
|
| 982 |
+
"step": 278
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.6129032258064515,
|
| 986 |
+
"grad_norm": 0.6231604218482971,
|
| 987 |
+
"learning_rate": 1.8181818181818182e-05,
|
| 988 |
+
"loss": 4.3152,
|
| 989 |
+
"step": 280
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.638709677419355,
|
| 993 |
+
"grad_norm": 0.42820078134536743,
|
| 994 |
+
"learning_rate": 1.6883116883116884e-05,
|
| 995 |
+
"loss": 4.9029,
|
| 996 |
+
"step": 282
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.664516129032258,
|
| 1000 |
+
"grad_norm": 0.7605751752853394,
|
| 1001 |
+
"learning_rate": 1.5584415584415583e-05,
|
| 1002 |
+
"loss": 4.8563,
|
| 1003 |
+
"step": 284
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.6903225806451614,
|
| 1007 |
+
"grad_norm": 0.8210684061050415,
|
| 1008 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 1009 |
+
"loss": 3.7461,
|
| 1010 |
+
"step": 286
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.7161290322580647,
|
| 1014 |
+
"grad_norm": 0.6016200184822083,
|
| 1015 |
+
"learning_rate": 1.2987012987012986e-05,
|
| 1016 |
+
"loss": 4.7194,
|
| 1017 |
+
"step": 288
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.741935483870968,
|
| 1021 |
+
"grad_norm": 0.7171183228492737,
|
| 1022 |
+
"learning_rate": 1.1688311688311688e-05,
|
| 1023 |
+
"loss": 4.5957,
|
| 1024 |
+
"step": 290
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.767741935483871,
|
| 1028 |
+
"grad_norm": 0.6558433175086975,
|
| 1029 |
+
"learning_rate": 1.038961038961039e-05,
|
| 1030 |
+
"loss": 4.5994,
|
| 1031 |
+
"step": 292
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.793548387096774,
|
| 1035 |
+
"grad_norm": 0.5295835733413696,
|
| 1036 |
+
"learning_rate": 9.090909090909091e-06,
|
| 1037 |
+
"loss": 5.5671,
|
| 1038 |
+
"step": 294
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.8193548387096774,
|
| 1042 |
+
"grad_norm": 0.6055201292037964,
|
| 1043 |
+
"learning_rate": 7.792207792207792e-06,
|
| 1044 |
+
"loss": 5.0694,
|
| 1045 |
+
"step": 296
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.8451612903225807,
|
| 1049 |
+
"grad_norm": 0.6041186451911926,
|
| 1050 |
+
"learning_rate": 6.493506493506493e-06,
|
| 1051 |
+
"loss": 5.0007,
|
| 1052 |
+
"step": 298
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.870967741935484,
|
| 1056 |
+
"grad_norm": 0.7281818985939026,
|
| 1057 |
+
"learning_rate": 5.194805194805195e-06,
|
| 1058 |
+
"loss": 4.3153,
|
| 1059 |
+
"step": 300
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.896774193548387,
|
| 1063 |
+
"grad_norm": 0.4925851821899414,
|
| 1064 |
+
"learning_rate": 3.896103896103896e-06,
|
| 1065 |
+
"loss": 5.5631,
|
| 1066 |
+
"step": 302
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.9225806451612906,
|
| 1070 |
+
"grad_norm": 0.6916934847831726,
|
| 1071 |
+
"learning_rate": 2.5974025974025976e-06,
|
| 1072 |
+
"loss": 5.3932,
|
| 1073 |
+
"step": 304
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.9483870967741934,
|
| 1077 |
+
"grad_norm": 0.6561426520347595,
|
| 1078 |
+
"learning_rate": 1.2987012987012988e-06,
|
| 1079 |
+
"loss": 5.3373,
|
| 1080 |
+
"step": 306
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.9741935483870967,
|
| 1084 |
+
"grad_norm": 0.6102042198181152,
|
| 1085 |
+
"learning_rate": 0.0,
|
| 1086 |
+
"loss": 5.1064,
|
| 1087 |
+
"step": 308
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.9741935483870967,
|
| 1091 |
+
"step": 308,
|
| 1092 |
+
"total_flos": 723765039226044.0,
|
| 1093 |
+
"train_loss": 5.270187915145577,
|
| 1094 |
+
"train_runtime": 746.2778,
|
| 1095 |
+
"train_samples_per_second": 3.318,
|
| 1096 |
+
"train_steps_per_second": 0.413
|
| 1097 |
}
|
| 1098 |
],
|
| 1099 |
"logging_steps": 2,
|
| 1100 |
+
"max_steps": 308,
|
| 1101 |
"num_input_tokens_seen": 0,
|
| 1102 |
"num_train_epochs": 4,
|
| 1103 |
"save_steps": 500,
|
|
|
|
| 1113 |
"attributes": {}
|
| 1114 |
}
|
| 1115 |
},
|
| 1116 |
+
"total_flos": 723765039226044.0,
|
| 1117 |
"train_batch_size": 2,
|
| 1118 |
"trial_name": null,
|
| 1119 |
"trial_params": null
|