whisper-md-catala / README.md
JulioCastro's picture
Update README.md
3064c67
metadata
language:
  - ca
license: apache-2.0
tags:
  - whisper-event
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: Whisper Medium Catala 1k steps
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0, Fleurs, SLR69, tb3_parla, parlament_parla
          type: >-
            mozilla-foundation/common_voice_11_0, google/fleurs, openslr,
            collectivat/tv3_parla, projecte-aina/parlament_parla
          config: ca
          split: test
          args: ca
        metrics:
          - name: Wer
            type: wer
            value: 10.9688

Whisper Md Ca - 1k This model is a fine-tuned version of openai/whisper-medium on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

Loss: 0.2554 Wer: 10.9688 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: 32 eval_batch_size: 8 seed: 42 gradient_accumulation_steps: 2 total_train_batch_size: 64 optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 lr_scheduler_type: linear lr_scheduler_warmup_steps: 100 training_steps: 1000 mixed_precision_training: Native AMP

Training results Training Loss Epoch Step Validation Loss Wer 0.2554 1.0 1000 0.2554 10.9688 Framework versions Transformers 4.26.0.dev0 Pytorch 1.13.1+cu117 Datasets 2.7.1.dev0 Tokenizers 0.13.2