Hugging Face's logo

gweltou
/
wav2vec2-xls-r-300m-MCV15

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
gweltou commited on
Commit
e2238ed
•
1 Parent(s): 23885e1

Model save

Browse files
Files changed (3) hide show
  1. README.md +19 -17
  2. model.safetensors +1 -1
  3. runs/Dec23_08-33-38_630beb74b32f/events.out.tfevents.1703320471.630beb74b32f.162.0 +2 -2
README.md CHANGED
@@ -17,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 1.0523
21
- - Wer: 0.7437
22
- - Cer: 0.2712
23
 
24
  ## Model description
25
 
@@ -38,7 +38,7 @@ More information needed
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
- - learning_rate: 3e-05
42
  - train_batch_size: 24
43
  - eval_batch_size: 8
44
  - seed: 42
@@ -47,29 +47,31 @@ The following hyperparameters were used during training:
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
  - lr_scheduler_warmup_ratio: 0.1
50
- - num_epochs: 50
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
56
  |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
57
- | 12.5819 | 4.5 | 250 | 4.5566 | 1.0 | 1.0000 |
58
- | 3.4789 | 9.01 | 500 | 3.1356 | 1.0 | 1.0000 |
59
- | 3.0456 | 13.51 | 750 | 2.9915 | 1.0 | 1.0000 |
60
- | 2.7595 | 18.02 | 1000 | 2.2266 | 0.9648 | 0.6214 |
61
- | 1.9067 | 22.52 | 1250 | 1.4719 | 0.8477 | 0.3298 |
62
- | 1.3924 | 27.03 | 1500 | 1.2345 | 0.8086 | 0.2978 |
63
- | 1.168 | 31.53 | 1750 | 1.1614 | 0.7843 | 0.2893 |
64
- | 1.0616 | 36.04 | 2000 | 1.1134 | 0.7690 | 0.2824 |
65
- | 0.9895 | 40.54 | 2250 | 1.0685 | 0.7489 | 0.2749 |
66
- | 0.9316 | 45.05 | 2500 | 1.0537 | 0.7451 | 0.2715 |
67
- | 0.9192 | 49.55 | 2750 | 1.0523 | 0.7437 | 0.2712 |
 
 
68
 
69
 
70
  ### Framework versions
71
 
72
  - Transformers 4.35.2
73
  - Pytorch 2.1.0+cu121
74
- - Datasets 2.15.0
75
  - Tokenizers 0.15.0
 
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 1.1155
21
+ - Wer: 0.6060
22
+ - Cer: 0.2242
23
 
24
  ## Model description
25
 
 
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
+ - learning_rate: 8e-05
42
  - train_batch_size: 24
43
  - eval_batch_size: 8
44
  - seed: 42
 
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
  - lr_scheduler_warmup_ratio: 0.1
50
+ - num_epochs: 60
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
56
  |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
57
+ | 9.8843 | 4.5 | 250 | 3.3243 | 1.0 | 1.0000 |
58
+ | 3.066 | 9.01 | 500 | 2.8141 | 1.0 | 1.0000 |
59
+ | 1.8006 | 13.51 | 750 | 1.1647 | 0.7936 | 0.3028 |
60
+ | 0.8925 | 18.02 | 1000 | 1.0696 | 0.7174 | 0.2679 |
61
+ | 0.6306 | 22.52 | 1250 | 1.0330 | 0.6792 | 0.2509 |
62
+ | 0.512 | 27.03 | 1500 | 0.9988 | 0.6560 | 0.2405 |
63
+ | 0.4275 | 31.53 | 1750 | 1.0428 | 0.6345 | 0.2332 |
64
+ | 0.3837 | 36.04 | 2000 | 1.0555 | 0.6267 | 0.2315 |
65
+ | 0.3481 | 40.54 | 2250 | 1.1165 | 0.6231 | 0.2312 |
66
+ | 0.3081 | 45.05 | 2500 | 1.0772 | 0.6142 | 0.2255 |
67
+ | 0.2906 | 49.55 | 2750 | 1.1146 | 0.6085 | 0.2267 |
68
+ | 0.2923 | 54.05 | 3000 | 1.1030 | 0.6058 | 0.2228 |
69
+ | 0.259 | 58.56 | 3250 | 1.1155 | 0.6060 | 0.2242 |
70
 
71
 
72
  ### Framework versions
73
 
74
  - Transformers 4.35.2
75
  - Pytorch 2.1.0+cu121
76
+ - Datasets 2.16.0
77
  - Tokenizers 0.15.0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ff2580c85ab3e4197c4b7b72f1270cb9f1fe76389c112dc4c37cdbdbf05a6f60
3
  size 1261975580
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c64e772b8de9304f94556b6dfdd5a8a4a338819f3a6b19a88115f8d954c48679
3
  size 1261975580
runs/Dec23_08-33-38_630beb74b32f/events.out.tfevents.1703320471.630beb74b32f.162.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dcb89b5a272e8698d98902d0793a34a6514b805246f89af391281aa2899ac523
3
- size 11980
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c35667fb425dec077ef2627a55e8cf80ba95b919bd188edf21df8f495c5eb82d
3
+ size 12856