Hugging Face's logo

Baselhany
/
fine_tune_Whisper_tiny2

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Arabic
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Baselhany commited on
Commit
374efe4
·
verified ·
1 Parent(s): 7a42caf

Training finished

Browse files
Files changed (3) hide show
  1. README.md +9 -8
  2. model.safetensors +1 -1
  3. runs/Jan21_16-24-41_aac7088bc6e6/events.out.tfevents.1737486257.aac7088bc6e6.18.1 +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: 0.0065
24
- - Wer: 0.0838
25
- - Cer: 0.0327
26
 
27
  ## Model description
28
 
@@ -48,7 +48,7 @@ The following hyperparameters were used during training:
48
  - gradient_accumulation_steps: 4
49
  - total_train_batch_size: 64
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
@@ -57,10 +57,11 @@ The following hyperparameters were used during training:
57
 
58
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
59
  |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
60
- | 0.0065 | 1.0 | 235 | 0.0060 | 0.0766 | 0.0292 |
61
- | 0.0071 | 2.0 | 470 | 0.0060 | 0.0802 | 0.0301 |
62
- | 0.0067 | 3.0 | 705 | 0.0060 | 0.0767 | 0.0298 |
63
- | 0.0041 | 4.0 | 940 | 0.0061 | 0.0814 | 0.0341 |
 
64
 
65
 
66
  ### 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: 0.0063
24
+ - Wer: 0.0826
25
+ - Cer: 0.0324
26
 
27
  ## Model description
28
 
 
48
  - gradient_accumulation_steps: 4
49
  - total_train_batch_size: 64
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: cosine
52
  - lr_scheduler_warmup_steps: 500
53
  - num_epochs: 20
54
  - mixed_precision_training: Native AMP
 
57
 
58
  | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
59
  |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
60
+ | 0.0085 | 1.0 | 274 | 0.0059 | 0.0755 | 0.0293 |
61
+ | 0.0067 | 2.0 | 548 | 0.0058 | 0.0789 | 0.0305 |
62
+ | 0.0062 | 3.0 | 822 | 0.0059 | 0.0760 | 0.0281 |
63
+ | 0.0033 | 4.0 | 1096 | 0.0061 | 0.0769 | 0.0293 |
64
+ | 0.0024 | 5.0 | 1370 | 0.0063 | 0.0746 | 0.0285 |
65
 
66
 
67
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8454128905d6f8b99f4138375a77b9356bd23e933c3448565dd6572105b53c98
3
  size 151061672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:744c4ec0a6b1a8b3bcafa8d6399fd9f7e76905b71468d1617261a498320e7141
3
  size 151061672
runs/Jan21_16-24-41_aac7088bc6e6/events.out.tfevents.1737486257.aac7088bc6e6.18.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:024340aafefd876ff4260186cdd8f0cbf170296bbe9c97c0f4e8e3c6b5e8fa04
3
+ size 453