Hugging Face's logo

juanjucm
/
whisper-small-FLEURS-GL-EN

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Galician
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
whisper-small-FLEURS-GL-EN / README.md
juanjucm's picture
juanjucm
Update README.md
20700af verified
preview code
|
raw
history blame contribute delete
2.04 kB
metadata
library_name: transformers
license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
metrics:
  - wer
  - bleu
model-index:
  - name: whisper-small-FLEURS-GL-EN
    results: []
datasets:
  - juanjucm/FLEURS-SpeechT-GL-EN
language:
  - gl

whisper-small-FLEURS-GL-EN

This model is a fine-tuned version of openai/whisper-small on juanjucm/FLEURS-SpeechT-GL-EN. It achieves the following results on the evaluation set:

  • Loss: 1.6607
  • Wer: 67.1683
  • Bleu: 22.6201

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1.25e-05
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 2
  • total_train_batch_size: 32
  • total_eval_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 8
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer Bleu
1.3189 1.0 86 1.6608 67.1683 22.6201
0.6613 2.0 172 1.6643 68.5990 21.1576
0.3492 3.0 258 1.7873 69.7046 20.7371
0.1416 4.0 344 1.9098 69.9090 20.5952
0.0974 5.0 430 2.0487 70.0948 20.6740
0.061 6.0 516 2.1565 73.4578 19.2411
0.0384 7.0 602 2.2107 73.6622 19.5413
0.0203 8.0 688 2.2476 73.9874 19.4512

Framework versions

  • Transformers 4.45.1
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.20.0