---
language:
- sa
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- addy88/sanskrit-asr-84
metrics:
- wer
model-index:
- name: whisper-small-sanskasr-37000-V3
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: addy88/sanskrit-asr-84
      type: addy88/sanskrit-asr-84
      config: null
      split: None
      args: 'split: test'
    metrics:
    - name: Wer
      type: wer
      value: 87.80468001493968
---

<!-- 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-small-sanskasr-37000-V3

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the addy88/sanskrit-asr-84 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7560
- Wer: 87.8047

## 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: 10
- eval_batch_size: 10
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 5
- training_steps: 100

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 1.5226        | 0.0   | 10   | 1.1482          | 96.8367 |
| 1.0625        | 0.01  | 20   | 0.9524          | 96.0036 |
| 0.8662        | 0.01  | 30   | 0.8896          | 94.1134 |
| 0.8763        | 0.01  | 40   | 0.8467          | 90.5783 |
| 0.8338        | 0.02  | 50   | 0.8134          | 90.5734 |
| 0.802         | 0.02  | 60   | 0.7916          | 88.7546 |
| 0.7823        | 0.02  | 70   | 0.7781          | 89.0323 |
| 0.7241        | 0.03  | 80   | 0.7681          | 89.0486 |
| 0.7404        | 0.03  | 90   | 0.7585          | 88.1489 |
| 0.7086        | 0.03  | 100  | 0.7560          | 87.8047 |


### Framework versions

- Transformers 4.30.2
- Pytorch 2.0.0
- Datasets 2.13.0
- Tokenizers 0.13.3