Hugging Face's logo

dmatekenya
/
wav2vec2-large-mms-1b-chich-colab

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
dmatekenya commited on
Commit
f90fa8e
•
1 Parent(s): f30355b

End of training

Browse files
Files changed (3) hide show
  1. README.md +21 -48
  2. adapter.nya.safetensors +2 -2
  3. pytorch_model.bin +1 -1
README.md CHANGED
@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
18
  This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: inf
21
- - Wer: 0.7342
22
 
23
  ## Model description
24
 
@@ -44,63 +44,36 @@ The following hyperparameters were used during training:
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
  - lr_scheduler_warmup_steps: 100
47
- - num_epochs: 20
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:-----:|:----:|:---------------:|:------:|
54
- | 3.4078 | 0.44 | 100 | inf | 1.0010 |
55
- | 3.3151 | 0.88 | 200 | inf | 0.9997 |
56
- | 1.758 | 1.32 | 300 | inf | 0.7672 |
57
- | 1.3904 | 1.75 | 400 | inf | 0.7329 |
58
- | 1.2602 | 2.19 | 500 | inf | 0.7159 |
59
- | 1.1833 | 2.63 | 600 | inf | 0.7149 |
60
- | 1.263 | 3.07 | 700 | inf | 0.7166 |
61
- | 1.1387 | 3.51 | 800 | inf | 0.6955 |
62
- | 1.0807 | 3.95 | 900 | inf | 0.7006 |
63
- | 1.2017 | 4.39 | 1000 | inf | 0.6910 |
64
- | 1.0561 | 4.82 | 1100 | inf | 0.6858 |
65
- | 1.099 | 5.26 | 1200 | inf | 0.6961 |
66
- | 1.0189 | 5.7 | 1300 | inf | 0.6726 |
67
- | 1.026 | 6.14 | 1400 | inf | 0.7083 |
68
- | 1.0125 | 6.58 | 1500 | inf | 0.6934 |
69
- | 0.977 | 7.02 | 1600 | inf | 0.6781 |
70
- | 0.988 | 7.46 | 1700 | inf | 0.6783 |
71
- | 0.999 | 7.89 | 1800 | inf | 0.6880 |
72
- | 0.9678 | 8.33 | 1900 | inf | 0.6807 |
73
- | 0.9463 | 8.77 | 2000 | inf | 0.6744 |
74
- | 0.8928 | 9.21 | 2100 | inf | 0.6701 |
75
- | 0.9399 | 9.65 | 2200 | inf | 0.6813 |
76
- | 0.9535 | 10.09 | 2300 | inf | 0.6862 |
77
- | 0.8886 | 10.53 | 2400 | inf | 0.6694 |
78
- | 0.9367 | 10.96 | 2500 | inf | 0.6825 |
79
- | 0.9035 | 11.4 | 2600 | inf | 0.6724 |
80
- | 0.8658 | 11.84 | 2700 | inf | 0.6554 |
81
- | 0.9042 | 12.28 | 2800 | inf | 0.6644 |
82
- | 0.8668 | 12.72 | 2900 | inf | 0.6618 |
83
- | 0.8843 | 13.16 | 3000 | inf | 0.6689 |
84
- | 0.8504 | 13.6 | 3100 | inf | 0.6589 |
85
- | 0.8462 | 14.04 | 3200 | inf | 0.6619 |
86
- | 0.8472 | 14.47 | 3300 | inf | 0.6729 |
87
- | 0.8375 | 14.91 | 3400 | inf | 0.6575 |
88
- | 0.7992 | 15.35 | 3500 | inf | 0.6850 |
89
- | 0.8593 | 15.79 | 3600 | inf | 0.6905 |
90
- | 0.8069 | 16.23 | 3700 | inf | 0.7018 |
91
- | 0.8884 | 16.67 | 3800 | inf | 0.6963 |
92
- | 0.8578 | 17.11 | 3900 | inf | 0.7070 |
93
- | 0.8411 | 17.54 | 4000 | inf | 0.7214 |
94
- | 0.8276 | 17.98 | 4100 | inf | 0.6968 |
95
- | 0.8659 | 18.42 | 4200 | inf | 0.6958 |
96
- | 0.9367 | 18.86 | 4300 | inf | 0.7568 |
97
- | 0.9912 | 19.3 | 4400 | inf | 0.7080 |
98
- | 0.9682 | 19.74 | 4500 | inf | 0.7342 |
99
 
100
 
101
  ### Framework versions
102
 
103
  - Transformers 4.35.0.dev0
104
  - Pytorch 2.1.0+cu118
105
- - Datasets 2.14.5
106
  - Tokenizers 0.14.1
 
18
  This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: inf
21
+ - Wer: 0.6583
22
 
23
  ## Model description
24
 
 
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
  - lr_scheduler_warmup_steps: 100
47
+ - num_epochs: 8
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:-----:|:----:|:---------------:|:------:|
54
+ | 6.9371 | 0.44 | 100 | inf | 0.7521 |
55
+ | 1.4647 | 0.88 | 200 | inf | 0.7314 |
56
+ | 1.2928 | 1.32 | 300 | inf | 0.7388 |
57
+ | 1.3529 | 1.75 | 400 | inf | 0.7320 |
58
+ | 1.1927 | 2.19 | 500 | inf | 0.6984 |
59
+ | 1.34 | 2.63 | 600 | inf | 0.6992 |
60
+ | 1.2707 | 3.07 | 700 | inf | 0.6922 |
61
+ | 1.105 | 3.51 | 800 | inf | 0.7185 |
62
+ | 1.1334 | 3.95 | 900 | inf | 0.6838 |
63
+ | 1.0915 | 4.39 | 1000 | inf | 0.6751 |
64
+ | 1.133 | 4.82 | 1100 | inf | 0.6934 |
65
+ | 1.1295 | 5.26 | 1200 | inf | 0.6752 |
66
+ | 1.0214 | 5.7 | 1300 | inf | 0.6700 |
67
+ | 0.9687 | 6.14 | 1400 | inf | 0.6692 |
68
+ | 1.0553 | 6.58 | 1500 | inf | 0.6664 |
69
+ | 0.9686 | 7.02 | 1600 | inf | 0.6618 |
70
+ | 0.9565 | 7.46 | 1700 | inf | 0.6605 |
71
+ | 1.001 | 7.89 | 1800 | inf | 0.6583 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
72
 
73
 
74
  ### Framework versions
75
 
76
  - Transformers 4.35.0.dev0
77
  - Pytorch 2.1.0+cu118
78
+ - Datasets 2.14.6
79
  - Tokenizers 0.14.1
adapter.nya.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e4d36476a41b42098f848a309acd154119babd5b655e6643b455a2080c6b28ba
3
- size 8890772
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23f4fdd721e59656a4c6f613f844b6de303eca0b278783a34c649638b63c6313
3
+ size 8870276
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:28ca815f4de4a2ba22b5c3577e4a6209c3eb7a1110fdab127127330f34ea9b32
3
  size 3859206794
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00dad28bd2dab37cc0c26f1bf0f4a99b591e1bd4918686f0f91e746f79fb4147
3
  size 3859206794