mikr/whisper-small-ro-cv11

Whisper Small Romanian CV11

This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_11_0 ro dataset. It achieves the following results on the evaluation set:

• Loss: 0.2808
• Wer: 17.5713

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: 64
• eval_batch_size: 32
• seed: 42
• distributed_type: multi-GPU
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0046	8.01	1000	0.2808	17.5713
0.0006	17.01	2000	0.3091	17.6916
0.0003	25.02	3000	0.3267	17.6881
0.0002	34.01	4000	0.3381	17.7977
0.0002	43.0	5000	0.3428	17.8048

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.1+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2