Hugging Face's logo

Baselhany
/
Graduation_Project_Whisper_tiny3

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Arabic
whisper
Generated from Trainer
Model card Files
xet
Metrics Community
Baselhany commited on
Commit
db10da9
·
verified ·
1 Parent(s): 542cb29

Training finished

Browse files
Files changed (3) hide show
  1. README.md +21 -26
  2. model.safetensors +1 -1
  3. runs/Feb10_16-54-26_bf70c8764c91/events.out.tfevents.1739221847.bf70c8764c91.18.1 +3 -0
README.md CHANGED
@@ -20,9 +20,9 @@ should probably proofread and complete it, then remove this comment. -->
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the quran-ayat-speech-to-text dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.0067
24
- - Wer: 0.0764
25
- - Cer: 0.0301
26
 
27
  ## Model description
28
 
@@ -50,33 +50,28 @@ The following hyperparameters were used during training:
50
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
51
  - lr_scheduler_type: linear
52
  - lr_scheduler_warmup_steps: 500
53
- - num_epochs: 20
54
  - mixed_precision_training: Native AMP
55
 
56
  ### Training results
57
 
58
- | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
59
- |:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
60
- | 0.0073 | 1.0 | 157 | 0.0068 | 0.0708 | 0.0279 |
61
- | 0.0068 | 2.0 | 314 | 0.0066 | 0.0684 | 0.0266 |
62
- | 0.0057 | 3.0 | 471 | 0.0066 | 0.0702 | 0.0292 |
63
- | 0.0042 | 4.0 | 628 | 0.0068 | 0.0715 | 0.0291 |
64
- | 0.0036 | 5.0 | 785 | 0.0072 | 0.0731 | 0.0285 |
65
- | 0.0021 | 6.0 | 942 | 0.0082 | 0.0731 | 0.0297 |
66
- | 0.0003 | 7.0 | 1099 | 0.0091 | 0.0715 | 0.0282 |
67
- | 0.0005 | 8.0 | 1256 | 0.0096 | 0.0737 | 0.0299 |
68
- | 0.0001 | 9.0 | 1413 | 0.0102 | 0.0755 | 0.0309 |
69
- | 0.0001 | 10.0 | 1570 | 0.0104 | 0.0729 | 0.0286 |
70
- | 0.0 | 11.0 | 1727 | 0.0107 | 0.0746 | 0.0298 |
71
- | 0.0 | 12.0 | 1884 | 0.0110 | 0.0715 | 0.0289 |
72
- | 0.0 | 13.0 | 2041 | 0.0113 | 0.0764 | 0.0308 |
73
- | 0.0 | 14.0 | 2198 | 0.0116 | 0.0753 | 0.0298 |
74
- | 0.0 | 15.0 | 2355 | 0.0118 | 0.0733 | 0.0290 |
75
- | 0.0 | 16.0 | 2512 | 0.0121 | 0.0748 | 0.0298 |
76
- | 0.0 | 17.0 | 2669 | 0.0123 | 0.0773 | 0.0318 |
77
- | 0.0 | 18.0 | 2826 | 0.0124 | 0.0751 | 0.0307 |
78
- | 0.0 | 19.0 | 2983 | 0.0132 | 0.0847 | 0.0329 |
79
- | 0.0 | 19.8768 | 3120 | 0.0126 | 0.0757 | 0.0307 |
80
 
81
 
82
  ### Framework versions
 
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the quran-ayat-speech-to-text dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.0065
24
+ - Wer: 0.0736
25
+ - Cer: 0.0296
26
 
27
  ## Model description
28
 
 
50
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
51
  - lr_scheduler_type: linear
52
  - lr_scheduler_warmup_steps: 500
53
+ - num_epochs: 15
54
  - mixed_precision_training: Native AMP
55
 
56
  ### Training results
57
 
58
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
59
+ |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
60
+ | 0.0066 | 1.0 | 125 | 0.0065 | 0.0700 | 0.0291 |
61
+ | 0.0051 | 2.0 | 250 | 0.0064 | 0.0681 | 0.0281 |
62
+ | 0.0043 | 3.0 | 375 | 0.0064 | 0.0655 | 0.0270 |
63
+ | 0.0045 | 4.0 | 500 | 0.0067 | 0.0664 | 0.0270 |
64
+ | 0.0027 | 5.0 | 625 | 0.0071 | 0.0679 | 0.0277 |
65
+ | 0.0017 | 6.0 | 750 | 0.0081 | 0.0691 | 0.0282 |
66
+ | 0.0005 | 7.0 | 875 | 0.0089 | 0.0697 | 0.0286 |
67
+ | 0.0004 | 8.0 | 1000 | 0.0097 | 0.0715 | 0.0282 |
68
+ | 0.0001 | 9.0 | 1125 | 0.0102 | 0.0690 | 0.0277 |
69
+ | 0.0001 | 10.0 | 1250 | 0.0105 | 0.0693 | 0.0282 |
70
+ | 0.0001 | 11.0 | 1375 | 0.0109 | 0.0679 | 0.0270 |
71
+ | 0.0 | 12.0 | 1500 | 0.0110 | 0.0686 | 0.0274 |
72
+ | 0.0 | 13.0 | 1625 | 0.0112 | 0.0670 | 0.0273 |
73
+ | 0.0 | 14.0 | 1750 | 0.0115 | 0.0752 | 0.0333 |
74
+ | 0.0 | 15.0 | 1875 | 0.0114 | 0.0679 | 0.0271 |
 
 
 
 
 
75
 
76
 
77
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d969193ec74a646fea4074b1ef0202395b099d625e1093b989c56de4954f9c34
3
  size 151061672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f6746418ab24630d3f943ad91bb0c217b4e6147f208bc621b2465319d5aed52
3
  size 151061672
runs/Feb10_16-54-26_bf70c8764c91/events.out.tfevents.1739221847.bf70c8764c91.18.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea7e09151e6f0b7107ed8d63e58b12432572c3f5492359f9ed2faf1a73de6e8b
3
+ size 453