Hugging Face's logo

thiagobarbosa
/
whisper-base-google-fleurs-pt-br

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Safetensors
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
thiagobarbosa commited on
Commit
aecd128
•
1 Parent(s): 609840c

Model save

Browse files
Files changed (2) hide show
  1. README.md +15 -15
  2. pytorch_model.bin +1 -1
README.md CHANGED
@@ -17,8 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 0.6283
21
- - Wer: 25.9071
 
22
 
23
  ## Model description
24
 
@@ -37,26 +38,25 @@ More information needed
37
  ### Training hyperparameters
38
 
39
  The following hyperparameters were used during training:
40
- - learning_rate: 2.5e-05
41
- - train_batch_size: 12
42
- - eval_batch_size: 12
43
  - seed: 42
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
- - lr_scheduler_warmup_steps: 120
47
- - training_steps: 2400
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
- | Training Loss | Epoch | Step | Validation Loss | Wer |
53
- |:-------------:|:-----:|:----:|:---------------:|:-------:|
54
- | 0.0871 | 2.72 | 400 | 0.4838 | 24.4078 |
55
- | 0.0066 | 5.44 | 800 | 0.5647 | 25.5452 |
56
- | 0.0013 | 8.16 | 1200 | 0.5981 | 25.6110 |
57
- | 0.0008 | 10.88 | 1600 | 0.6143 | 25.6533 |
58
- | 0.0006 | 13.61 | 2000 | 0.6245 | 25.7661 |
59
- | 0.0006 | 16.33 | 2400 | 0.6283 | 25.9071 |
60
 
61
 
62
  ### Framework versions
 
17
 
18
  This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 0.4063
21
+ - Wer: 21.6112
22
+ - Wer Normalized: 18.0010
23
 
24
  ## Model description
25
 
 
38
  ### Training hyperparameters
39
 
40
  The following hyperparameters were used during training:
41
+ - learning_rate: 2.05e-05
42
+ - train_batch_size: 32
43
+ - eval_batch_size: 16
44
  - seed: 42
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 80
48
+ - training_steps: 800
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
 
53
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Wer Normalized |
54
+ |:-------------:|:-----:|:----:|:---------------:|:-------:|:--------------:|
55
+ | 0.6738 | 0.5 | 100 | 0.3943 | 21.7334 | 17.9487 |
56
+ | 0.4816 | 1.01 | 200 | 0.3762 | 20.9203 | 17.1352 |
57
+ | 0.2652 | 1.51 | 300 | 0.3872 | 21.1882 | 17.2827 |
58
+ | 0.2901 | 2.01 | 400 | 0.3912 | 21.4608 | 17.7061 |
59
+ | 0.1408 | 2.51 | 500 | 0.4063 | 21.6112 | 18.0010 |
 
60
 
61
 
62
  ### Framework versions
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3df3ff8fad7a8d821c278f2c0c942794505e774f1afdba57bc9f1aadacc75a97
3
  size 290459230
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b15b29d63443d314967b1209a3221fdd9cd6307c18bc7468c9307f187985ceee
3
  size 290459230