---
language:
- hi
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- konnakol
metrics:
- wer
model-index:
- name: Whisper Hi - Gopika
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Konnakol
      type: konnakol
      args: 'config: hi, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 0.026845637583892617
---

<!-- 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. -->

# Whisper Hi - Gopika

This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the Konnakol dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2424
- Wer: 0.0268
- Cer: 0.0281

## 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: 5e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 1000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch    | Step | Validation Loss | Wer    | Cer    |
|:-------------:|:--------:|:----:|:---------------:|:------:|:------:|
| 2.6928        | 8.3333   | 50   | 0.7108          | 0.0593 | 0.0502 |
| 0.1456        | 16.6667  | 100  | 0.1978          | 0.0503 | 0.0448 |
| 0.0104        | 25.0     | 150  | 0.1970          | 0.0291 | 0.0293 |
| 0.0145        | 33.3333  | 200  | 0.2114          | 0.0503 | 0.0387 |
| 0.0123        | 41.6667  | 250  | 0.2174          | 0.0515 | 0.0511 |
| 0.0086        | 50.0     | 300  | 0.2339          | 0.0246 | 0.0263 |
| 0.0077        | 58.3333  | 350  | 0.2737          | 0.0336 | 0.0303 |
| 0.0132        | 66.6667  | 400  | 0.1764          | 0.0268 | 0.0269 |
| 0.005         | 75.0     | 450  | 0.2107          | 0.0336 | 0.0339 |
| 0.0207        | 83.3333  | 500  | 0.2167          | 0.4955 | 0.4702 |
| 0.0153        | 91.6667  | 550  | 0.1948          | 0.0358 | 0.0342 |
| 0.013         | 100.0    | 600  | 0.1882          | 0.0257 | 0.0230 |
| 0.001         | 108.3333 | 650  | 0.2405          | 0.0302 | 0.0324 |
| 0.0018        | 116.6667 | 700  | 0.2377          | 0.0302 | 0.0290 |
| 0.0           | 125.0    | 750  | 0.2385          | 0.0268 | 0.0281 |
| 0.0           | 133.3333 | 800  | 0.2398          | 0.0268 | 0.0281 |
| 0.0           | 141.6667 | 850  | 0.2409          | 0.0268 | 0.0281 |
| 0.0           | 150.0    | 900  | 0.2418          | 0.0268 | 0.0281 |
| 0.0           | 158.3333 | 950  | 0.2422          | 0.0268 | 0.0281 |
| 0.0           | 166.6667 | 1000 | 0.2424          | 0.0268 | 0.0281 |


### Framework versions

- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1