distil-whisper-large-fo-100h-5k-steps

This model is a fine-tuned version of distil-whisper/distil-large-v3 on the ravnursson_asr dataset. It achieves the following results on the evaluation set:

• Loss: 0.2075
• Wer: 13.5545

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
0.4469	0.2320	1000	0.5128	30.3201
0.2995	0.4640	2000	0.3383	21.1295
0.2338	0.6961	3000	0.2666	17.4351
0.2009	0.9281	4000	0.2270	14.9940
0.0963	1.1601	5000	0.2075	13.5545

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1