Hugging Face's logo

Redve
/
BengaliModel

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Bengali
whisper
hf-asr-leaderboard
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
 Metrics Community
Redve commited on
Commit
f3e7bd4
·
1 Parent(s): 7bf282b

End of training

Browse files
Files changed (1) hide show
  1. README.md +15 -12
README.md CHANGED
@@ -25,7 +25,7 @@ model-index:
25
  metrics:
26
  - name: Wer
27
  type: wer
28
- value: 43.03652465083491
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -35,8 +35,8 @@ should probably proofread and complete it, then remove this comment. -->
35
 
36
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
37
  It achieves the following results on the evaluation set:
38
- - Loss: 0.2531
39
- - Wer: 43.0365
40
 
41
  ## Model description
42
 
@@ -61,21 +61,24 @@ The following hyperparameters were used during training:
61
  - seed: 42
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
- - training_steps: 4000
 
65
  - mixed_precision_training: Native AMP
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Wer |
70
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
71
- | 0.1467 | 1.06 | 500 | 0.1944 | 52.8704 |
72
- | 0.0769 | 2.12 | 1000 | 0.1651 | 47.0106 |
73
- | 0.044 | 3.18 | 1500 | 0.1693 | 45.1381 |
74
- | 0.0203 | 4.24 | 2000 | 0.1860 | 43.4903 |
75
- | 0.0097 | 5.3 | 2500 | 0.2096 | 43.9618 |
76
- | 0.0039 | 6.36 | 3000 | 0.2311 | 43.2877 |
77
- | 0.0019 | 7.42 | 3500 | 0.2454 | 43.2612 |
78
- | 0.0009 | 8.47 | 4000 | 0.2531 | 43.0365 |
 
 
79
 
80
 
81
  ### Framework versions
 
25
  metrics:
26
  - name: Wer
27
  type: wer
28
+ value: 42.212627219456316
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
35
 
36
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
37
  It achieves the following results on the evaluation set:
38
+ - Loss: 0.2751
39
+ - Wer: 42.2126
40
 
41
  ## Model description
42
 
 
61
  - seed: 42
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
+ - lr_scheduler_warmup_steps: 200
65
+ - training_steps: 5000
66
  - mixed_precision_training: Native AMP
67
 
68
  ### Training results
69
 
70
  | Training Loss | Epoch | Step | Validation Loss | Wer |
71
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
72
+ | 0.163 | 1.06 | 500 | 0.2064 | 55.6065 |
73
+ | 0.0862 | 2.12 | 1000 | 0.1675 | 47.1869 |
74
+ | 0.0475 | 3.18 | 1500 | 0.1696 | 44.8561 |
75
+ | 0.0239 | 4.24 | 2000 | 0.1848 | 43.2436 |
76
+ | 0.0119 | 5.3 | 2500 | 0.2081 | 43.5608 |
77
+ | 0.0058 | 6.36 | 3000 | 0.2262 | 43.0718 |
78
+ | 0.0024 | 7.42 | 3500 | 0.2427 | 42.4726 |
79
+ | 0.0009 | 8.47 | 4000 | 0.2611 | 42.6356 |
80
+ | 0.0005 | 9.53 | 4500 | 0.2709 | 42.3492 |
81
+ | 0.0004 | 10.59 | 5000 | 0.2751 | 42.2126 |
82
 
83
 
84
  ### Framework versions