Hugging Face's logo

legendary2910
/
Mnong-ASR

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Vietnamese
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Mnong-ASR / README.md
legendary2910's picture
legendary2910
Upload tokenizer
2d7bf65 verified
preview code
|
raw
history blame
1.66 kB
metadata
language:
  - vi
license: apache-2.0
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
metrics:
  - wer
base_model: openai/whisper-small
model-index:
  - name: Whisper Small Mnong
    results: []

Whisper Small Mnong

This model is a fine-tuned version of openai/whisper-small on the MnongAudio dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0749
  • Wer: 59.5216

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: 16
  • 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: 500
  • training_steps: 4000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.293 5.92 1000 1.0033 73.6891
0.0091 11.83 2000 1.0491 66.0534
0.0005 17.75 3000 1.0669 59.4296
0.0004 23.67 4000 1.0749 59.5216

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2