---
library_name: peft
language:
- tr
license: mit
base_model: openai/whisper-large-v3-turbo
tags:
- generated_from_trainer
datasets:
- dcl-ai-team/Cagri-kayitlar-relabeled-06s-200ms-padded
metrics:
- wer
model-index:
- name: 'Large v3 turbo 0.6s 200mspad  augmentedv1 - Alperitoo '
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: Cagrilar-full-relabeled
      type: dcl-ai-team/Cagri-kayitlar-relabeled-06s-200ms-padded
      args: 'config: turkish, split: test'
    metrics:
    - type: wer
      value: 29.49686892679767
      name: Wer
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# Large v3 turbo 0.6s 200mspad  augmentedv1 - Alperitoo 

This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the Cagrilar-full-relabeled dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4556
- Wer: 29.4969

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 4
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 10
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.3879        | 1.0   | 477  | 0.3409          | 28.6979 |
| 0.2692        | 2.0   | 954  | 0.3272          | 27.9853 |
| 0.1854        | 3.0   | 1431 | 0.3403          | 28.0285 |
| 0.1384        | 4.0   | 1908 | 0.3524          | 28.2444 |
| 0.0976        | 5.0   | 2385 | 0.3778          | 28.3740 |
| 0.0707        | 6.0   | 2862 | 0.3889          | 28.3092 |
| 0.0523        | 7.0   | 3339 | 0.4196          | 28.9786 |
| 0.0385        | 8.0   | 3816 | 0.4418          | 29.3241 |
| 0.0324        | 9.0   | 4293 | 0.4532          | 29.8424 |
| 0.0268        | 10.0  | 4770 | 0.4556          | 29.4969 |


### Framework versions

- PEFT 0.14.0
- Transformers 4.48.3
- Pytorch 2.4.1+cu121
- Datasets 3.1.0
- Tokenizers 0.21.0