whisper-large-v2-3swissdatasets

This model is a fine-tuned version of openai/whisper-large-v2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2418
• Wer: 16.0707

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: 5e-06
• 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: 7000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2962	0.0727	1000	0.3163	20.0659
0.2756	0.1454	2000	0.2962	19.2670
0.2405	0.2181	3000	0.2771	18.1353
0.2917	0.2908	4000	0.2644	17.5769
0.2117	0.3635	5000	0.2536	16.7275
0.2334	0.4362	6000	0.2455	16.3825
0.2408	0.5089	7000	0.2418	16.0707

Framework versions

• Transformers 4.44.1
• Pytorch 2.3.1+cu118
• Datasets 3.2.0
• Tokenizers 0.19.1