Edit model card

Whisper Medium Medical

This model is a fine-tuned version of openai/whisper-medium on the Medical ASR dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0995
  • Wer: 12.4065

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 50
  • training_steps: 500
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.7532 0.1357 100 0.2695 13.0099
0.2155 0.2714 200 0.2053 9.8238
0.2392 0.4071 300 0.1567 8.9549
0.151 0.5427 400 0.1159 6.9273
0.1439 0.6784 500 0.0995 12.4065

Framework versions

  • Transformers 4.45.1
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.20.0
Downloads last month
8
Safetensors
Model size
764M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for jilangdi/whisper-medium-medical

Finetuned
(455)
this model

Dataset used to train jilangdi/whisper-medium-medical

Evaluation results