Whisper large cs - jan_zizka
This model is a fine-tuned version of openai/whisper-large-v3 on the combined dataset. It achieves the following results on the evaluation set:
- eval_loss: 0.1242
- eval_wer: 11.3608
- eval_runtime: 14167.2564
- eval_samples_per_second: 0.624
- eval_steps_per_second: 0.078
- epoch: 0.3795
- step: 1000
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: 16
- 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: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1
- Downloads last month
- 22
Model tree for Cem13/whisper-large-v3-czech
Base model
openai/whisper-large-v3