Hugging Face's logo

ctaguchi
/
ssc-cgg-mms-model-mix-adapt-max3-devtrain

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Model card Files
xet
Community
ctaguchi commited on
Commit
c7a5efe
·
verified ·
1 Parent(s): 660bc24

End of training

Browse files
Files changed (4) hide show
  1. README.md +31 -75
  2. adapter.cgg.safetensors +1 -1
  3. model.safetensors +1 -1
  4. training_args.bin +1 -1
README.md CHANGED
@@ -1,7 +1,5 @@
1
  ---
2
  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:
@@ -16,11 +14,11 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  # ssc-cgg-mms-model-mix-adapt-max3-devtrain
18
 
19
- This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.3144
22
- - Cer: 0.1074
23
- - Wer: 0.5
24
 
25
  ## Model description
26
 
@@ -40,85 +38,43 @@ More information needed
40
 
41
  The following hyperparameters were used during training:
42
  - learning_rate: 0.0005
43
- - train_batch_size: 8
44
  - eval_batch_size: 6
45
  - seed: 42
46
  - gradient_accumulation_steps: 2
47
- - total_train_batch_size: 16
48
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
49
  - lr_scheduler_type: linear
50
  - lr_scheduler_warmup_steps: 100
51
- - num_epochs: 20
52
  - mixed_precision_training: Native AMP
53
 
54
  ### Training results
55
 
56
- | Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
57
- |:-------------:|:-------:|:-----:|:---------------:|:------:|:------:|
58
- | 2.8396 | 0.3118 | 200 | 0.7349 | 0.1882 | 0.7851 |
59
- | 0.808 | 0.6235 | 400 | 0.5227 | 0.1524 | 0.6640 |
60
- | 0.7968 | 0.9353 | 600 | 0.4932 | 0.1473 | 0.6426 |
61
- | 0.7337 | 1.2463 | 800 | 0.4622 | 0.1393 | 0.6229 |
62
- | 0.6984 | 1.5581 | 1000 | 0.4511 | 0.1387 | 0.6231 |
63
- | 0.7177 | 1.8698 | 1200 | 0.4409 | 0.1348 | 0.6050 |
64
- | 0.7086 | 2.1808 | 1400 | 0.4373 | 0.1351 | 0.6060 |
65
- | 0.6591 | 2.4926 | 1600 | 0.4194 | 0.1309 | 0.5879 |
66
- | 0.6929 | 2.8044 | 1800 | 0.4356 | 0.1325 | 0.5909 |
67
- | 0.686 | 3.1154 | 2000 | 0.4189 | 0.1296 | 0.5772 |
68
- | 0.6588 | 3.4271 | 2200 | 0.4104 | 0.1266 | 0.5709 |
69
- | 0.6267 | 3.7389 | 2400 | 0.4076 | 0.1280 | 0.5772 |
70
- | 0.6633 | 4.0499 | 2600 | 0.4059 | 0.1254 | 0.5677 |
71
- | 0.6479 | 4.3617 | 2800 | 0.3915 | 0.1246 | 0.5658 |
72
- | 0.6445 | 4.6734 | 3000 | 0.3884 | 0.1254 | 0.5673 |
73
- | 0.6554 | 4.9852 | 3200 | 0.3809 | 0.1223 | 0.5540 |
74
- | 0.6029 | 5.2962 | 3400 | 0.3913 | 0.1246 | 0.5662 |
75
- | 0.6445 | 5.6080 | 3600 | 0.3869 | 0.1224 | 0.5565 |
76
- | 0.5975 | 5.9197 | 3800 | 0.3775 | 0.1208 | 0.5505 |
77
- | 0.5668 | 6.2307 | 4000 | 0.3732 | 0.1201 | 0.5485 |
78
- | 0.6055 | 6.5425 | 4200 | 0.3715 | 0.1195 | 0.5464 |
79
- | 0.593 | 6.8542 | 4400 | 0.3723 | 0.1188 | 0.5450 |
80
- | 0.614 | 7.1652 | 4600 | 0.3653 | 0.1183 | 0.5446 |
81
- | 0.5791 | 7.4770 | 4800 | 0.3627 | 0.1198 | 0.5508 |
82
- | 0.5815 | 7.7888 | 5000 | 0.3634 | 0.1175 | 0.5386 |
83
- | 0.6083 | 8.0998 | 5200 | 0.3582 | 0.1167 | 0.5371 |
84
- | 0.5951 | 8.4115 | 5400 | 0.3568 | 0.1160 | 0.5325 |
85
- | 0.5479 | 8.7233 | 5600 | 0.3563 | 0.1160 | 0.5354 |
86
- | 0.5761 | 9.0343 | 5800 | 0.3494 | 0.1154 | 0.5329 |
87
- | 0.5584 | 9.3461 | 6000 | 0.3519 | 0.1150 | 0.5312 |
88
- | 0.5523 | 9.6578 | 6200 | 0.3508 | 0.1157 | 0.5361 |
89
- | 0.5625 | 9.9696 | 6400 | 0.3428 | 0.1139 | 0.5272 |
90
- | 0.5237 | 10.2806 | 6600 | 0.3448 | 0.1140 | 0.5268 |
91
- | 0.5102 | 10.5924 | 6800 | 0.3417 | 0.1137 | 0.5249 |
92
- | 0.5779 | 10.9041 | 7000 | 0.3414 | 0.1135 | 0.5232 |
93
- | 0.525 | 11.2151 | 7200 | 0.3385 | 0.1128 | 0.5232 |
94
- | 0.5485 | 11.5269 | 7400 | 0.3372 | 0.1120 | 0.5185 |
95
- | 0.5494 | 11.8387 | 7600 | 0.3333 | 0.1114 | 0.5177 |
96
- | 0.5044 | 12.1496 | 7800 | 0.3308 | 0.1113 | 0.5140 |
97
- | 0.5421 | 12.4614 | 8000 | 0.3342 | 0.1106 | 0.5121 |
98
- | 0.5278 | 12.7732 | 8200 | 0.3269 | 0.1113 | 0.5135 |
99
- | 0.5121 | 13.0842 | 8400 | 0.3393 | 0.1120 | 0.5178 |
100
- | 0.5305 | 13.3959 | 8600 | 0.3332 | 0.1109 | 0.5155 |
101
- | 0.5061 | 13.7077 | 8800 | 0.3306 | 0.1102 | 0.5122 |
102
- | 0.5042 | 14.0187 | 9000 | 0.3264 | 0.1098 | 0.5127 |
103
- | 0.4954 | 14.3305 | 9200 | 0.3256 | 0.1099 | 0.5128 |
104
- | 0.504 | 14.6422 | 9400 | 0.3261 | 0.1097 | 0.5104 |
105
- | 0.5442 | 14.9540 | 9600 | 0.3199 | 0.1087 | 0.5070 |
106
- | 0.4786 | 15.2650 | 9800 | 0.3255 | 0.1091 | 0.5072 |
107
- | 0.5219 | 15.5768 | 10000 | 0.3225 | 0.1084 | 0.5053 |
108
- | 0.4802 | 15.8885 | 10200 | 0.3227 | 0.1087 | 0.5067 |
109
- | 0.5 | 16.1995 | 10400 | 0.3230 | 0.1083 | 0.5062 |
110
- | 0.4789 | 16.5113 | 10600 | 0.3210 | 0.1085 | 0.5061 |
111
- | 0.4613 | 16.8231 | 10800 | 0.3224 | 0.1081 | 0.5047 |
112
- | 0.4906 | 17.1341 | 11000 | 0.3196 | 0.1085 | 0.5061 |
113
- | 0.495 | 17.4458 | 11200 | 0.3190 | 0.1082 | 0.5048 |
114
- | 0.4796 | 17.7576 | 11400 | 0.3174 | 0.1084 | 0.5041 |
115
- | 0.4687 | 18.0686 | 11600 | 0.3178 | 0.1082 | 0.5043 |
116
- | 0.4707 | 18.3804 | 11800 | 0.3167 | 0.1081 | 0.5028 |
117
- | 0.5062 | 18.6921 | 12000 | 0.3145 | 0.1077 | 0.5034 |
118
- | 0.4891 | 19.0031 | 12200 | 0.3149 | 0.1079 | 0.5044 |
119
- | 0.4901 | 19.3149 | 12400 | 0.3150 | 0.1072 | 0.5018 |
120
- | 0.4705 | 19.6267 | 12600 | 0.3147 | 0.1075 | 0.5004 |
121
- | 0.4639 | 19.9384 | 12800 | 0.3144 | 0.1074 | 0.5 |
122
 
123
 
124
  ### Framework versions
 
1
  ---
2
  library_name: transformers
 
 
3
  tags:
4
  - generated_from_trainer
5
  metrics:
 
14
 
15
  # ssc-cgg-mms-model-mix-adapt-max3-devtrain
16
 
17
+ This model was trained from scratch on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.3101
20
+ - Cer: 0.1185
21
+ - Wer: 0.5337
22
 
23
  ## Model description
24
 
 
38
 
39
  The following hyperparameters were used during training:
40
  - learning_rate: 0.0005
41
+ - train_batch_size: 1
42
  - eval_batch_size: 6
43
  - seed: 42
44
  - gradient_accumulation_steps: 2
45
+ - total_train_batch_size: 2
46
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
47
  - lr_scheduler_type: linear
48
  - lr_scheduler_warmup_steps: 100
49
+ - num_epochs: 5
50
  - mixed_precision_training: Native AMP
51
 
52
  ### Training results
53
 
54
+ | Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
55
+ |:-------------:|:------:|:----:|:---------------:|:------:|:------:|
56
+ | 0.4048 | 0.2261 | 200 | 0.3375 | 0.1202 | 0.5490 |
57
+ | 0.4134 | 0.4522 | 400 | 0.3257 | 0.1247 | 0.5674 |
58
+ | 0.3972 | 0.6783 | 600 | 0.3291 | 0.1219 | 0.5564 |
59
+ | 0.3561 | 0.9045 | 800 | 0.3348 | 0.1231 | 0.5552 |
60
+ | 0.3581 | 1.1300 | 1000 | 0.3231 | 0.1232 | 0.5608 |
61
+ | 0.3902 | 1.3561 | 1200 | 0.3248 | 0.1187 | 0.5403 |
62
+ | 0.355 | 1.5822 | 1400 | 0.3365 | 0.1140 | 0.5262 |
63
+ | 0.3532 | 1.8084 | 1600 | 0.3273 | 0.1217 | 0.5584 |
64
+ | 0.3802 | 2.0339 | 1800 | 0.3210 | 0.1243 | 0.5551 |
65
+ | 0.3517 | 2.2600 | 2000 | 0.3221 | 0.1219 | 0.5513 |
66
+ | 0.3492 | 2.4862 | 2200 | 0.3227 | 0.1212 | 0.5466 |
67
+ | 0.3459 | 2.7123 | 2400 | 0.3187 | 0.1200 | 0.5417 |
68
+ | 0.3261 | 2.9384 | 2600 | 0.3166 | 0.1198 | 0.5382 |
69
+ | 0.33 | 3.1639 | 2800 | 0.3149 | 0.1194 | 0.5424 |
70
+ | 0.3256 | 3.3901 | 3000 | 0.3172 | 0.1186 | 0.5389 |
71
+ | 0.3335 | 3.6162 | 3200 | 0.3105 | 0.1196 | 0.5386 |
72
+ | 0.3246 | 3.8423 | 3400 | 0.3101 | 0.1188 | 0.5361 |
73
+ | 0.3195 | 4.0678 | 3600 | 0.3119 | 0.1198 | 0.5395 |
74
+ | 0.3034 | 4.2940 | 3800 | 0.3109 | 0.1186 | 0.5321 |
75
+ | 0.2654 | 4.5201 | 4000 | 0.3125 | 0.1187 | 0.5370 |
76
+ | 0.2981 | 4.7462 | 4200 | 0.3102 | 0.1188 | 0.5363 |
77
+ | 0.3048 | 4.9723 | 4400 | 0.3101 | 0.1185 | 0.5337 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
78
 
79
 
80
  ### Framework versions
adapter.cgg.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3dee7750002ac3a0e8000a2962a820e181ecde8242ed4c2d6bd129510339b82a
3
  size 8947144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c48d731944cc4a0ada8b676b31e86b06b4529e4882b49956af3d7f019699be9
3
  size 8947144
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:50909cae219c5a1562f0f53a7c30ebed7cd80a2065b9b887248aeb9b40282ff7
3
  size 3859039520
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba1a3bd8804b78b587664cf574bca91798891a7c98de09c80df3b265d6d371ec
3
  size 3859039520
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bf620c349d8b5c19466ee976d2835ddf13c816b15af03c9e86f1245f441d4062
3
  size 5969
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc503e90e752b235933d374365671f12f17a06344132ecd6d3913308d1a65aa9
3
  size 5969