bjelkenhed/whisper-medium-sv

openai/whisper-medium

This model is a fine-tuned version of openai/whisper-medium trained on NST Swedish ASR and evaluated on Common Voice 11 testset. It achieves the following results on the evaluation set:

• Loss: 0.2636
• Wer: 10.7122

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: 16
• 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: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0746	0.2	1000	0.2904	13.4695
0.0564	0.4	2000	0.3121	13.2384
0.0532	0.6	3000	0.2862	11.9726
0.0387	0.8	4000	0.2629	11.6931
0.0279	1.14	5000	0.2636	10.7122

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.0+cu117
• Datasets 2.7.1.dev0
• Tokenizers 0.13.2