Hugging Face's logo

rdomanski
/
training

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
Metrics Community
rdomanski commited on
Commit
63557d3
·
verified ·
1 Parent(s): c77c8f6

Model save

Browse files
Files changed (2) hide show
  1. README.md +11 -11
  2. model.safetensors +1 -1
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Wer
24
  type: wer
25
- value: 37.745098039215684
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the audiofolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.5488
36
- - Wer: 37.7451
37
 
38
  ## Model description
39
 
@@ -58,18 +58,18 @@ The following hyperparameters were used during training:
58
  - seed: 42
59
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
  - lr_scheduler_type: linear
61
- - lr_scheduler_warmup_steps: 100
62
- - training_steps: 800
63
  - mixed_precision_training: Native AMP
64
 
65
  ### Training results
66
 
67
- | Training Loss | Epoch | Step | Validation Loss | Wer |
68
- |:-------------:|:-----:|:----:|:---------------:|:-------:|
69
- | 0.2488 | 3.08 | 200 | 0.4975 | 38.7255 |
70
- | 0.0138 | 6.15 | 400 | 0.5237 | 39.7059 |
71
- | 0.0024 | 9.23 | 600 | 0.5270 | 38.2353 |
72
- | 0.0004 | 12.31 | 800 | 0.5488 | 37.7451 |
73
 
74
 
75
  ### Framework versions
 
22
  metrics:
23
  - name: Wer
24
  type: wer
25
+ value: 54.10855824219926
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the audiofolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.3691
36
+ - Wer: 54.1086
37
 
38
  ## Model description
39
 
 
58
  - seed: 42
59
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
  - lr_scheduler_type: linear
61
+ - lr_scheduler_warmup_steps: 500
62
+ - training_steps: 4000
63
  - mixed_precision_training: Native AMP
64
 
65
  ### Training results
66
 
67
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
68
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
69
+ | 0.142 | 1.43 | 1000 | 0.3087 | 114.8992 |
70
+ | 0.0678 | 2.87 | 2000 | 0.3034 | 81.2481 |
71
+ | 0.0121 | 4.3 | 3000 | 0.3543 | 51.2291 |
72
+ | 0.0038 | 5.74 | 4000 | 0.3691 | 54.1086 |
73
 
74
 
75
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6151327b2181dd0c2d2dbfe030ac44fbf794a83675cdc7a512ce8fa586c9f4b
3
  size 3055544304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e81c583cfc0552ca1bb01e59faf72017dea2de60d35a00ebbfee8ba271a15d69
3
  size 3055544304