Hugging Face's logo

ylacombe
/
wav2vec2-bert-CV16-en-libri

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2-bert
librispeech_asr
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
ylacombe HF staff commited on
Commit
02daf7d
•
1 Parent(s): 284f523

Model save

Browse files
Files changed (4) hide show
  1. README.md +12 -10
  2. emissions.csv +1 -0
  3. model.safetensors +1 -1
  4. runs/Jan16_14-53-45_vorace/events.out.tfevents.1705417021.vorace.510473.0 +2 -2
README.md CHANGED
@@ -2,8 +2,6 @@
2
  license: mit
3
  base_model: facebook/w2v-bert-2.0
4
  tags:
5
- - automatic-speech-recognition
6
- - librispeech_asr
7
  - generated_from_trainer
8
  metrics:
9
  - wer
@@ -17,11 +15,11 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  # wav2vec2-bert-CV16-en-libri
19
 
20
- This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on the LIBRISPEECH_ASR - CLEAN dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 0.1331
23
- - Wer: 0.0997
24
- - Cer: 0.0264
25
 
26
  ## Model description
27
 
@@ -52,7 +50,7 @@ The following hyperparameters were used during training:
52
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
53
  - lr_scheduler_type: linear
54
  - lr_scheduler_warmup_steps: 10000
55
- - num_epochs: 5.0
56
  - mixed_precision_training: Native AMP
57
 
58
  ### Training results
@@ -63,9 +61,13 @@ The following hyperparameters were used during training:
63
  | 1.2899 | 1.26 | 500 | 0.2563 | 1.1471 | 0.7030 |
64
  | 0.5276 | 1.89 | 750 | 0.1127 | 0.4687 | 0.4114 |
65
  | 0.3313 | 2.52 | 1000 | 0.0659 | 0.2870 | 0.2577 |
66
- | 0.2089 | 3.16 | 1250 | 0.2079 | 0.1766 | 0.0445 |
67
- | 0.1634 | 3.79 | 1500 | 0.1687 | 0.1411 | 0.0366 |
68
- | 0.163 | 4.42 | 1750 | 0.1490 | 0.1163 | 0.0298 |
 
 
 
 
69
 
70
 
71
  ### Framework versions
 
2
  license: mit
3
  base_model: facebook/w2v-bert-2.0
4
  tags:
 
 
5
  - generated_from_trainer
6
  metrics:
7
  - wer
 
15
 
16
  # wav2vec2-bert-CV16-en-libri
17
 
18
+ This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 0.1077
21
+ - Wer: 0.0717
22
+ - Cer: 0.0197
23
 
24
  ## Model description
25
 
 
50
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
51
  - lr_scheduler_type: linear
52
  - lr_scheduler_warmup_steps: 10000
53
+ - num_epochs: 7.0
54
  - mixed_precision_training: Native AMP
55
 
56
  ### Training results
 
61
  | 1.2899 | 1.26 | 500 | 0.2563 | 1.1471 | 0.7030 |
62
  | 0.5276 | 1.89 | 750 | 0.1127 | 0.4687 | 0.4114 |
63
  | 0.3313 | 2.52 | 1000 | 0.0659 | 0.2870 | 0.2577 |
64
+ | 0.2089 | 3.16 | 1250 | 0.0445 | 0.2079 | 0.1766 |
65
+ | 0.1634 | 3.79 | 1500 | 0.0366 | 0.1687 | 0.1411 |
66
+ | 0.1546 | 4.42 | 1750 | 0.1452 | 0.1138 | 0.0294 |
67
+ | 0.1245 | 5.05 | 2000 | 0.1316 | 0.0973 | 0.0260 |
68
+ | 0.1341 | 5.68 | 2250 | 0.1196 | 0.0867 | 0.0234 |
69
+ | 0.0942 | 6.31 | 2500 | 0.1128 | 0.0794 | 0.0213 |
70
+ | 0.0848 | 6.94 | 2750 | 0.1077 | 0.0717 | 0.0197 |
71
 
72
 
73
  ### Framework versions
emissions.csv CHANGED
@@ -1,3 +1,4 @@
1
  timestamp,experiment_id,project_name,duration,emissions,energy_consumed,country_name,country_iso_code,region,on_cloud,cloud_provider,cloud_region
2
  2024-01-16T12:18:10,100ce0c1-abd3-4e89-a129-ac49a10532a2,codecarbon,5727.051881551743,0.48059596920304154,1.131922185780931,France,FRA,île-de-france,N,,
3
  2024-01-16T14:32:52,fc5dbc38-c7d4-40f7-94ab-dd0de4d1b430,codecarbon,4902.239959001541,0.33650212333732155,0.7925455962511117,France,FRA,île-de-france,N,,
 
 
1
  timestamp,experiment_id,project_name,duration,emissions,energy_consumed,country_name,country_iso_code,region,on_cloud,cloud_provider,cloud_region
2
  2024-01-16T12:18:10,100ce0c1-abd3-4e89-a129-ac49a10532a2,codecarbon,5727.051881551743,0.48059596920304154,1.131922185780931,France,FRA,île-de-france,N,,
3
  2024-01-16T14:32:52,fc5dbc38-c7d4-40f7-94ab-dd0de4d1b430,codecarbon,4902.239959001541,0.33650212333732155,0.7925455962511117,France,FRA,île-de-france,N,,
4
+ 2024-01-16T16:44:29,be3ee4e5-8b23-482a-802c-91b438696e7d,codecarbon,6446.471297740936,0.453908136580529,1.069065749664906,France,FRA,île-de-france,N,,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f58404b4a65eae1b7b7d0f6d2696b2782e06518c8795c2d048abe09a1b4e480b
3
  size 2422949860
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e8d6b0f45bf6460e6e9d355b6abdd3f9ad9a5490c7e339c91d1e969743f9ba6d
3
  size 2422949860
runs/Jan16_14-53-45_vorace/events.out.tfevents.1705417021.vorace.510473.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:39e8d07cf77f2e4e4500db6b23c66dd843827752e8d3d4bf05ac83d7be2ae800
3
- size 38500
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:906547ea0256268e6ac0fb1f84ade72ba013870df7040bfda83ed31fcb1cd365
3
+ size 47697