Hugging Face's logo

5roop
/
output

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Croatian
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
output / README.md
5roop's picture
5roop
End of training
b39289f verified
preview code
|
raw
history blame
1.6 kB
metadata
language:
  - hr
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: whisper-large-v3-mici-princ
    results: []

whisper-large-v3-mici-princ

This model is a fine-tuned version of openai/whisper-large-v3 on the Mići Princ dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1816
  • Wer: 36.2647

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: 7e-06
  • train_batch_size: 4
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • training_steps: 618
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.0014 17.66 309 1.1196 29.9832
0.0007 35.31 618 1.1816 36.2647

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.0.0+cu117
  • Datasets 2.18.0
  • Tokenizers 0.15.2