VingeNie's picture
Upload tokenizer
226837e verified
metadata
language:
  - zh
license: apache-2.0
tags:
  - generated_from_trainer
base_model: openai/whisper-tiny
datasets:
  - mozilla-foundation/common_voice_16_1
model-index:
  - name: Whisper Tiny chinese - VingeNie
    results: []

Whisper Tiny chinese - VingeNie

This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 16.1 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7223
  • Cer Ortho: 40.6344
  • Cer: 29.5803

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: 0.0001
  • train_batch_size: 64
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant_with_warmup
  • lr_scheduler_warmup_steps: 25
  • training_steps: 650
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Cer Ortho Cer
0.7743 0.3594 225 0.7942 40.4431 31.4481
0.7344 0.7188 450 0.7223 40.6344 29.5803

Framework versions

  • Transformers 4.40.1
  • Pytorch 2.2.1+cu121
  • Datasets 2.19.0
  • Tokenizers 0.19.1