---
language:
- dj
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- Fahd1199/MergeV2
model-index:
- name: Whisper medium Darija
  results: []
---

<!-- 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 medium Darija

This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the MergeV2 dataset.
It achieves the following results on the evaluation set:
- Loss: 5.4036

## 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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.4992        | 0.2415 | 100  | 1.2243          |
| 0.8767        | 0.4831 | 200  | 1.6230          |
| 0.7651        | 0.7246 | 300  | 1.5771          |
| 0.8255        | 0.9662 | 400  | 2.2942          |
| 1.0079        | 1.2077 | 500  | 1.5600          |
| 1.6485        | 1.4493 | 600  | 3.9792          |
| 1.9125        | 1.6908 | 700  | 4.1289          |
| 1.2813        | 1.9324 | 800  | 4.8182          |
| 0.8415        | 2.1739 | 900  | 4.9184          |
| 0.6234        | 2.4155 | 1000 | 5.0654          |
| 0.5348        | 2.6570 | 1100 | 5.2614          |
| 0.4577        | 2.8986 | 1200 | 5.4036          |


### Framework versions

- Transformers 4.43.4
- Pytorch 2.2.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1