Hugging Face's logo

mmtg
/
train-test-valid

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
train-test-valid / README.md
mmtg's picture
mmtg
Upload tokenizer
7a768e7 verified
preview code
|
raw
history blame contribute delete
1.69 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
base_model: openai/whisper-base
metrics:
  - wer
model-index:
  - name: train-test-valid
    results: []

train-test-valid

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

  • Loss: 1.2197
  • Wer: 73.5930

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: 3e-05
  • train_batch_size: 12
  • eval_batch_size: 4
  • 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: 5000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.0002 23.8095 1000 1.1097 77.0127
0.0001 47.6190 2000 1.1609 74.0991
0.0001 71.4286 3000 1.1919 73.0716
0.0 95.2381 4000 1.2115 73.3783
0.0 119.0476 5000 1.2197 73.5930

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.1.2+cu121
  • Datasets 2.19.2
  • Tokenizers 0.19.1