whisper-medium-stt4sg

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

• Loss: 0.2665
• Wer: 17.3141

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: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3463	0.0801	1000	0.3711	24.6919
0.3607	0.1602	2000	0.3283	21.6146
0.2997	0.2403	3000	0.2997	19.4824
0.2831	0.3205	4000	0.2768	17.9290
0.2636	0.4006	5000	0.2665	17.3141

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu118
• Datasets 2.20.0
• Tokenizers 0.19.1