whisper-large-v3-genbed-m

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

• Loss: 0.6146
• Wer: 33.0189

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: 1.75e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 2500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.7275	0.6596	250	0.7019	56.8503
0.469	1.3193	500	0.6319	47.4164
0.4453	1.9789	750	0.5507	42.0133
0.2294	2.6385	1000	0.5573	38.9473
0.1087	3.2982	1250	0.5727	38.6364
0.1139	3.9578	1500	0.5532	36.3422
0.0421	4.6174	1750	0.5786	35.5274
0.0173	5.2770	2000	0.5795	34.0159
0.0108	5.9367	2250	0.5977	33.5549
0.0023	6.5963	2500	0.6146	33.0189

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1