Hugging Face's logo

mahwizzzz
/
medwhishper

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
English
whisper
hf-asr-leaderboard
Generated from Trainer
Model card Files
xet
Metrics Community
mahwizzzz commited on
Commit
89edf85
·
verified ·
1 Parent(s): 6aac906

End of training

Browse files
Files changed (2) hide show
  1. README.md +20 -8
  2. model.safetensors +1 -1
README.md CHANGED
@@ -21,10 +21,10 @@ should probably proofread and complete it, then remove this comment. -->
21
 
22
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 0.0358
25
- - Cer: 2.7788
26
- - Wer: 4.0416
27
- - Ser: 18.2836
28
 
29
  ## Model description
30
 
@@ -52,14 +52,26 @@ The following hyperparameters were used during training:
52
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
53
  - lr_scheduler_type: linear
54
  - lr_scheduler_warmup_steps: 500
55
- - num_epochs: 5
56
  - mixed_precision_training: Native AMP
57
 
58
  ### Training results
59
 
60
- | Training Loss | Epoch | Step | Validation Loss | Cer | Wer | Ser |
61
- |:-------------:|:------:|:----:|:---------------:|:------:|:------:|:-------:|
62
- | 0.0114 | 3.6384 | 1000 | 0.0358 | 2.7788 | 4.0416 | 18.2836 |
 
 
 
 
 
 
 
 
 
 
 
 
63
 
64
 
65
  ### Framework versions
 
21
 
22
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 0.0303
25
+ - Cer: 1.9023
26
+ - Wer: 2.6537
27
+ - Ser: 10.6810
28
 
29
  ## Model description
30
 
 
52
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
53
  - lr_scheduler_type: linear
54
  - lr_scheduler_warmup_steps: 500
55
+ - num_epochs: 50
56
  - mixed_precision_training: Native AMP
57
 
58
  ### Training results
59
 
60
+ | Training Loss | Epoch | Step | Validation Loss | Cer | Wer | Ser |
61
+ |:-------------:|:-------:|:-----:|:---------------:|:------:|:------:|:-------:|
62
+ | 0.0105 | 3.6384 | 1000 | 0.0343 | 3.2308 | 4.5581 | 17.5840 |
63
+ | 0.0022 | 7.2736 | 2000 | 0.0307 | 1.7735 | 2.5874 | 12.7332 |
64
+ | 0.0009 | 10.9120 | 3000 | 0.0308 | 1.9396 | 2.7782 | 12.5933 |
65
+ | 0.0006 | 14.5472 | 4000 | 0.0301 | 1.8891 | 2.7200 | 11.9869 |
66
+ | 0.0008 | 18.1824 | 5000 | 0.0299 | 1.8632 | 2.6676 | 12.2201 |
67
+ | 0.0008 | 21.8208 | 6000 | 0.0300 | 1.9788 | 2.7945 | 11.7537 |
68
+ | 0.0003 | 25.4560 | 7000 | 0.0286 | 1.8727 | 2.6211 | 11.1940 |
69
+ | 0.0002 | 29.0912 | 8000 | 0.0289 | 1.7497 | 2.4722 | 10.9142 |
70
+ | 0.0002 | 32.7296 | 9000 | 0.0291 | 1.5872 | 2.2861 | 10.6343 |
71
+ | 0.0001 | 36.3648 | 10000 | 0.0298 | 1.7848 | 2.5199 | 10.8209 |
72
+ | 0.0001 | 40.0 | 11000 | 0.0299 | 1.8900 | 2.6420 | 10.8209 |
73
+ | 0.0001 | 43.6384 | 12000 | 0.0301 | 1.9319 | 2.6897 | 10.6343 |
74
+ | 0.0001 | 47.2736 | 13000 | 0.0303 | 1.9023 | 2.6537 | 10.6810 |
75
 
76
 
77
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:67933124fe7b24e7c30c6a70dff1d441d26f22905c3745a6f2ce084158adf87a
3
  size 966995080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8759b6c48511f04c676fbd72656e22b34e427d025c5423ff8d9c6963ed430ebb
3
  size 966995080