End of training
Browse files- README.md +6 -3
- adapter.toi.safetensors +3 -0
- all_results.json +15 -0
- eval_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +483 -0
README.md
CHANGED
|
@@ -3,6 +3,9 @@ library_name: transformers
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
|
|
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- wer
|
|
@@ -16,10 +19,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# mms-1b-toigen-balanced-model
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Wer: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
| 6 |
+
- automatic-speech-recognition
|
| 7 |
+
- toigen
|
| 8 |
+
- mms
|
| 9 |
- generated_from_trainer
|
| 10 |
metrics:
|
| 11 |
- wer
|
|
|
|
| 19 |
|
| 20 |
# mms-1b-toigen-balanced-model
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the TOIGEN - TOI dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.3740
|
| 25 |
+
- Wer: 0.3990
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
adapter.toi.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92c459e7b229b8b12b75b104db63cdcae0a13dfde2053a2dbaaf47ec6f45d4a9
|
| 3 |
+
size 8793408
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.053571428571429,
|
| 3 |
+
"eval_loss": 0.37403130531311035,
|
| 4 |
+
"eval_runtime": 17.8093,
|
| 5 |
+
"eval_samples": 204,
|
| 6 |
+
"eval_samples_per_second": 11.455,
|
| 7 |
+
"eval_steps_per_second": 2.864,
|
| 8 |
+
"eval_wer": 0.39895882631329865,
|
| 9 |
+
"total_flos": 9.163482510982138e+18,
|
| 10 |
+
"train_loss": 0.8344545293737341,
|
| 11 |
+
"train_runtime": 2390.0152,
|
| 12 |
+
"train_samples": 894,
|
| 13 |
+
"train_samples_per_second": 11.222,
|
| 14 |
+
"train_steps_per_second": 2.812
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.053571428571429,
|
| 3 |
+
"eval_loss": 0.37403130531311035,
|
| 4 |
+
"eval_runtime": 17.8093,
|
| 5 |
+
"eval_samples": 204,
|
| 6 |
+
"eval_samples_per_second": 11.455,
|
| 7 |
+
"eval_steps_per_second": 2.864,
|
| 8 |
+
"eval_wer": 0.39895882631329865
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.053571428571429,
|
| 3 |
+
"total_flos": 9.163482510982138e+18,
|
| 4 |
+
"train_loss": 0.8344545293737341,
|
| 5 |
+
"train_runtime": 2390.0152,
|
| 6 |
+
"train_samples": 894,
|
| 7 |
+
"train_samples_per_second": 11.222,
|
| 8 |
+
"train_steps_per_second": 2.812
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,483 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3741886615753174,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-toigen-balanced-model/checkpoint-2400",
|
| 4 |
+
"epoch": 12.053571428571429,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 2700,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.44642857142857145,
|
| 13 |
+
"grad_norm": 3.282467842102051,
|
| 14 |
+
"learning_rate": 0.00028199999999999997,
|
| 15 |
+
"loss": 7.7726,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.44642857142857145,
|
| 20 |
+
"eval_loss": 3.8109493255615234,
|
| 21 |
+
"eval_runtime": 17.8235,
|
| 22 |
+
"eval_samples_per_second": 11.446,
|
| 23 |
+
"eval_steps_per_second": 2.861,
|
| 24 |
+
"eval_wer": 0.9938476100331283,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.8928571428571429,
|
| 29 |
+
"grad_norm": 2.164923667907715,
|
| 30 |
+
"learning_rate": 0.00029574018126888213,
|
| 31 |
+
"loss": 2.5726,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.8928571428571429,
|
| 36 |
+
"eval_loss": 0.8106288313865662,
|
| 37 |
+
"eval_runtime": 17.6933,
|
| 38 |
+
"eval_samples_per_second": 11.53,
|
| 39 |
+
"eval_steps_per_second": 2.882,
|
| 40 |
+
"eval_wer": 0.616658778987222,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 1.3392857142857144,
|
| 45 |
+
"grad_norm": 1.3534202575683594,
|
| 46 |
+
"learning_rate": 0.0002912084592145015,
|
| 47 |
+
"loss": 0.7986,
|
| 48 |
+
"step": 300
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.3392857142857144,
|
| 52 |
+
"eval_loss": 0.5409455299377441,
|
| 53 |
+
"eval_runtime": 17.7815,
|
| 54 |
+
"eval_samples_per_second": 11.473,
|
| 55 |
+
"eval_steps_per_second": 2.868,
|
| 56 |
+
"eval_wer": 0.5257927117841931,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.7857142857142856,
|
| 61 |
+
"grad_norm": 0.9854668378829956,
|
| 62 |
+
"learning_rate": 0.00028667673716012085,
|
| 63 |
+
"loss": 0.6324,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.7857142857142856,
|
| 68 |
+
"eval_loss": 0.5256258845329285,
|
| 69 |
+
"eval_runtime": 17.7084,
|
| 70 |
+
"eval_samples_per_second": 11.52,
|
| 71 |
+
"eval_steps_per_second": 2.88,
|
| 72 |
+
"eval_wer": 0.5054424988168481,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 2.232142857142857,
|
| 77 |
+
"grad_norm": 18.835981369018555,
|
| 78 |
+
"learning_rate": 0.00028214501510574015,
|
| 79 |
+
"loss": 0.603,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.232142857142857,
|
| 84 |
+
"eval_loss": 0.4854464828968048,
|
| 85 |
+
"eval_runtime": 17.7046,
|
| 86 |
+
"eval_samples_per_second": 11.522,
|
| 87 |
+
"eval_steps_per_second": 2.881,
|
| 88 |
+
"eval_wer": 0.4831992427827733,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.678571428571429,
|
| 93 |
+
"grad_norm": 32.54256820678711,
|
| 94 |
+
"learning_rate": 0.0002776132930513595,
|
| 95 |
+
"loss": 0.59,
|
| 96 |
+
"step": 600
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.678571428571429,
|
| 100 |
+
"eval_loss": 0.47332894802093506,
|
| 101 |
+
"eval_runtime": 17.6431,
|
| 102 |
+
"eval_samples_per_second": 11.563,
|
| 103 |
+
"eval_steps_per_second": 2.891,
|
| 104 |
+
"eval_wer": 0.4846190250828206,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 3.125,
|
| 109 |
+
"grad_norm": 1.3252086639404297,
|
| 110 |
+
"learning_rate": 0.0002730815709969788,
|
| 111 |
+
"loss": 0.5489,
|
| 112 |
+
"step": 700
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 3.125,
|
| 116 |
+
"eval_loss": 0.4439888894557953,
|
| 117 |
+
"eval_runtime": 17.8297,
|
| 118 |
+
"eval_samples_per_second": 11.442,
|
| 119 |
+
"eval_steps_per_second": 2.86,
|
| 120 |
+
"eval_wer": 0.46568859441552296,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.571428571428571,
|
| 125 |
+
"grad_norm": 0.5452375411987305,
|
| 126 |
+
"learning_rate": 0.00026854984894259817,
|
| 127 |
+
"loss": 0.5173,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.571428571428571,
|
| 132 |
+
"eval_loss": 0.43219566345214844,
|
| 133 |
+
"eval_runtime": 17.6382,
|
| 134 |
+
"eval_samples_per_second": 11.566,
|
| 135 |
+
"eval_steps_per_second": 2.891,
|
| 136 |
+
"eval_wer": 0.45764316138192146,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 4.017857142857143,
|
| 141 |
+
"grad_norm": 0.7151035070419312,
|
| 142 |
+
"learning_rate": 0.0002640181268882175,
|
| 143 |
+
"loss": 0.5315,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 4.017857142857143,
|
| 148 |
+
"eval_loss": 0.4285721480846405,
|
| 149 |
+
"eval_runtime": 17.7542,
|
| 150 |
+
"eval_samples_per_second": 11.49,
|
| 151 |
+
"eval_steps_per_second": 2.873,
|
| 152 |
+
"eval_wer": 0.44533838144817794,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 4.464285714285714,
|
| 157 |
+
"grad_norm": 1.8268319368362427,
|
| 158 |
+
"learning_rate": 0.0002594864048338368,
|
| 159 |
+
"loss": 0.4912,
|
| 160 |
+
"step": 1000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 4.464285714285714,
|
| 164 |
+
"eval_loss": 0.42536306381225586,
|
| 165 |
+
"eval_runtime": 17.765,
|
| 166 |
+
"eval_samples_per_second": 11.483,
|
| 167 |
+
"eval_steps_per_second": 2.871,
|
| 168 |
+
"eval_wer": 0.4458116422148604,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.910714285714286,
|
| 173 |
+
"grad_norm": 0.850709080696106,
|
| 174 |
+
"learning_rate": 0.0002549546827794562,
|
| 175 |
+
"loss": 0.4728,
|
| 176 |
+
"step": 1100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.910714285714286,
|
| 180 |
+
"eval_loss": 0.43455594778060913,
|
| 181 |
+
"eval_runtime": 17.7563,
|
| 182 |
+
"eval_samples_per_second": 11.489,
|
| 183 |
+
"eval_steps_per_second": 2.872,
|
| 184 |
+
"eval_wer": 0.44297207761476576,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.357142857142857,
|
| 189 |
+
"grad_norm": 0.7361202836036682,
|
| 190 |
+
"learning_rate": 0.00025042296072507554,
|
| 191 |
+
"loss": 0.4989,
|
| 192 |
+
"step": 1200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 5.357142857142857,
|
| 196 |
+
"eval_loss": 0.40502411127090454,
|
| 197 |
+
"eval_runtime": 17.6139,
|
| 198 |
+
"eval_samples_per_second": 11.582,
|
| 199 |
+
"eval_steps_per_second": 2.895,
|
| 200 |
+
"eval_wer": 0.42924751538097494,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 5.803571428571429,
|
| 205 |
+
"grad_norm": 1.305498719215393,
|
| 206 |
+
"learning_rate": 0.00024589123867069484,
|
| 207 |
+
"loss": 0.4661,
|
| 208 |
+
"step": 1300
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.803571428571429,
|
| 212 |
+
"eval_loss": 0.4019148647785187,
|
| 213 |
+
"eval_runtime": 17.792,
|
| 214 |
+
"eval_samples_per_second": 11.466,
|
| 215 |
+
"eval_steps_per_second": 2.866,
|
| 216 |
+
"eval_wer": 0.4254614292475154,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 6.25,
|
| 221 |
+
"grad_norm": 1.875386357307434,
|
| 222 |
+
"learning_rate": 0.00024135951661631417,
|
| 223 |
+
"loss": 0.4755,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 6.25,
|
| 228 |
+
"eval_loss": 0.4128676652908325,
|
| 229 |
+
"eval_runtime": 17.7386,
|
| 230 |
+
"eval_samples_per_second": 11.5,
|
| 231 |
+
"eval_steps_per_second": 2.875,
|
| 232 |
+
"eval_wer": 0.44486512068149553,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 6.696428571428571,
|
| 237 |
+
"grad_norm": 1.3318761587142944,
|
| 238 |
+
"learning_rate": 0.0002368277945619335,
|
| 239 |
+
"loss": 0.4603,
|
| 240 |
+
"step": 1500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 6.696428571428571,
|
| 244 |
+
"eval_loss": 0.40455254912376404,
|
| 245 |
+
"eval_runtime": 17.8115,
|
| 246 |
+
"eval_samples_per_second": 11.453,
|
| 247 |
+
"eval_steps_per_second": 2.863,
|
| 248 |
+
"eval_wer": 0.4254614292475154,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 7.142857142857143,
|
| 253 |
+
"grad_norm": 1.7303593158721924,
|
| 254 |
+
"learning_rate": 0.00023229607250755283,
|
| 255 |
+
"loss": 0.4229,
|
| 256 |
+
"step": 1600
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 7.142857142857143,
|
| 260 |
+
"eval_loss": 0.3939039707183838,
|
| 261 |
+
"eval_runtime": 17.7505,
|
| 262 |
+
"eval_samples_per_second": 11.493,
|
| 263 |
+
"eval_steps_per_second": 2.873,
|
| 264 |
+
"eval_wer": 0.41504969238050166,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 7.589285714285714,
|
| 269 |
+
"grad_norm": 0.9812105894088745,
|
| 270 |
+
"learning_rate": 0.0002277643504531722,
|
| 271 |
+
"loss": 0.455,
|
| 272 |
+
"step": 1700
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.589285714285714,
|
| 276 |
+
"eval_loss": 0.41328728199005127,
|
| 277 |
+
"eval_runtime": 17.7719,
|
| 278 |
+
"eval_samples_per_second": 11.479,
|
| 279 |
+
"eval_steps_per_second": 2.87,
|
| 280 |
+
"eval_wer": 0.41552295314718407,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 8.035714285714286,
|
| 285 |
+
"grad_norm": 0.6829022765159607,
|
| 286 |
+
"learning_rate": 0.00022323262839879152,
|
| 287 |
+
"loss": 0.4501,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 8.035714285714286,
|
| 292 |
+
"eval_loss": 0.3978167176246643,
|
| 293 |
+
"eval_runtime": 17.743,
|
| 294 |
+
"eval_samples_per_second": 11.497,
|
| 295 |
+
"eval_steps_per_second": 2.874,
|
| 296 |
+
"eval_wer": 0.4065309985802177,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 8.482142857142858,
|
| 301 |
+
"grad_norm": 1.3150678873062134,
|
| 302 |
+
"learning_rate": 0.00021870090634441088,
|
| 303 |
+
"loss": 0.45,
|
| 304 |
+
"step": 1900
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 8.482142857142858,
|
| 308 |
+
"eval_loss": 0.3925248682498932,
|
| 309 |
+
"eval_runtime": 17.6795,
|
| 310 |
+
"eval_samples_per_second": 11.539,
|
| 311 |
+
"eval_steps_per_second": 2.885,
|
| 312 |
+
"eval_wer": 0.42309512541410316,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 8.928571428571429,
|
| 317 |
+
"grad_norm": 1.0976217985153198,
|
| 318 |
+
"learning_rate": 0.0002141691842900302,
|
| 319 |
+
"loss": 0.4226,
|
| 320 |
+
"step": 2000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.928571428571429,
|
| 324 |
+
"eval_loss": 0.3901020586490631,
|
| 325 |
+
"eval_runtime": 17.7714,
|
| 326 |
+
"eval_samples_per_second": 11.479,
|
| 327 |
+
"eval_steps_per_second": 2.87,
|
| 328 |
+
"eval_wer": 0.4098438239469948,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 9.375,
|
| 333 |
+
"grad_norm": 0.727407693862915,
|
| 334 |
+
"learning_rate": 0.00020963746223564954,
|
| 335 |
+
"loss": 0.3973,
|
| 336 |
+
"step": 2100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 9.375,
|
| 340 |
+
"eval_loss": 0.38098010420799255,
|
| 341 |
+
"eval_runtime": 17.825,
|
| 342 |
+
"eval_samples_per_second": 11.445,
|
| 343 |
+
"eval_steps_per_second": 2.861,
|
| 344 |
+
"eval_wer": 0.4055844770468528,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 9.821428571428571,
|
| 349 |
+
"grad_norm": 2.031233072280884,
|
| 350 |
+
"learning_rate": 0.00020510574018126884,
|
| 351 |
+
"loss": 0.4038,
|
| 352 |
+
"step": 2200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 9.821428571428571,
|
| 356 |
+
"eval_loss": 0.41775575280189514,
|
| 357 |
+
"eval_runtime": 17.8559,
|
| 358 |
+
"eval_samples_per_second": 11.425,
|
| 359 |
+
"eval_steps_per_second": 2.856,
|
| 360 |
+
"eval_wer": 0.4117368670137246,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 10.267857142857142,
|
| 365 |
+
"grad_norm": 1.3557627201080322,
|
| 366 |
+
"learning_rate": 0.0002005740181268882,
|
| 367 |
+
"loss": 0.4559,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 10.267857142857142,
|
| 372 |
+
"eval_loss": 0.38752201199531555,
|
| 373 |
+
"eval_runtime": 17.7454,
|
| 374 |
+
"eval_samples_per_second": 11.496,
|
| 375 |
+
"eval_steps_per_second": 2.874,
|
| 376 |
+
"eval_wer": 0.40747752011358257,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 10.714285714285714,
|
| 381 |
+
"grad_norm": 0.7480702996253967,
|
| 382 |
+
"learning_rate": 0.00019604229607250753,
|
| 383 |
+
"loss": 0.4399,
|
| 384 |
+
"step": 2400
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 10.714285714285714,
|
| 388 |
+
"eval_loss": 0.3741886615753174,
|
| 389 |
+
"eval_runtime": 17.7283,
|
| 390 |
+
"eval_samples_per_second": 11.507,
|
| 391 |
+
"eval_steps_per_second": 2.877,
|
| 392 |
+
"eval_wer": 0.39895882631329865,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 11.160714285714286,
|
| 397 |
+
"grad_norm": 1.065514087677002,
|
| 398 |
+
"learning_rate": 0.00019151057401812688,
|
| 399 |
+
"loss": 0.3545,
|
| 400 |
+
"step": 2500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 11.160714285714286,
|
| 404 |
+
"eval_loss": 0.38181087374687195,
|
| 405 |
+
"eval_runtime": 17.8745,
|
| 406 |
+
"eval_samples_per_second": 11.413,
|
| 407 |
+
"eval_steps_per_second": 2.853,
|
| 408 |
+
"eval_wer": 0.40132513014671084,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 11.607142857142858,
|
| 413 |
+
"grad_norm": 1.19502854347229,
|
| 414 |
+
"learning_rate": 0.0001869788519637462,
|
| 415 |
+
"loss": 0.4452,
|
| 416 |
+
"step": 2600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 11.607142857142858,
|
| 420 |
+
"eval_loss": 0.3905617594718933,
|
| 421 |
+
"eval_runtime": 17.8748,
|
| 422 |
+
"eval_samples_per_second": 11.413,
|
| 423 |
+
"eval_steps_per_second": 2.853,
|
| 424 |
+
"eval_wer": 0.39801230477993377,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 12.053571428571429,
|
| 429 |
+
"grad_norm": 0.8653120994567871,
|
| 430 |
+
"learning_rate": 0.00018244712990936554,
|
| 431 |
+
"loss": 0.4014,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 12.053571428571429,
|
| 436 |
+
"eval_loss": 0.3751629889011383,
|
| 437 |
+
"eval_runtime": 17.7431,
|
| 438 |
+
"eval_samples_per_second": 11.497,
|
| 439 |
+
"eval_steps_per_second": 2.874,
|
| 440 |
+
"eval_wer": 0.39990534784666354,
|
| 441 |
+
"step": 2700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 12.053571428571429,
|
| 445 |
+
"step": 2700,
|
| 446 |
+
"total_flos": 9.163482510982138e+18,
|
| 447 |
+
"train_loss": 0.8344545293737341,
|
| 448 |
+
"train_runtime": 2390.0152,
|
| 449 |
+
"train_samples_per_second": 11.222,
|
| 450 |
+
"train_steps_per_second": 2.812
|
| 451 |
+
}
|
| 452 |
+
],
|
| 453 |
+
"logging_steps": 100,
|
| 454 |
+
"max_steps": 6720,
|
| 455 |
+
"num_input_tokens_seen": 0,
|
| 456 |
+
"num_train_epochs": 30,
|
| 457 |
+
"save_steps": 400,
|
| 458 |
+
"stateful_callbacks": {
|
| 459 |
+
"EarlyStoppingCallback": {
|
| 460 |
+
"args": {
|
| 461 |
+
"early_stopping_patience": 3,
|
| 462 |
+
"early_stopping_threshold": 0.0
|
| 463 |
+
},
|
| 464 |
+
"attributes": {
|
| 465 |
+
"early_stopping_patience_counter": 0
|
| 466 |
+
}
|
| 467 |
+
},
|
| 468 |
+
"TrainerControl": {
|
| 469 |
+
"args": {
|
| 470 |
+
"should_epoch_stop": false,
|
| 471 |
+
"should_evaluate": false,
|
| 472 |
+
"should_log": false,
|
| 473 |
+
"should_save": true,
|
| 474 |
+
"should_training_stop": false
|
| 475 |
+
},
|
| 476 |
+
"attributes": {}
|
| 477 |
+
}
|
| 478 |
+
},
|
| 479 |
+
"total_flos": 9.163482510982138e+18,
|
| 480 |
+
"train_batch_size": 4,
|
| 481 |
+
"trial_name": null,
|
| 482 |
+
"trial_params": null
|
| 483 |
+
}
|