whisper-large-v3-turbo-seg

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

• Loss: 0.9616
• Wer: 210.0224

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
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	1.0	23	4.6351	146.6717
No log	2.0	46	2.1577	130.8153
No log	3.0	69	1.2595	200.9723
No log	4.0	92	1.0324	132.4607
No log	5.0	115	0.9518	111.4435
No log	6.0	138	0.9300	111.9671
No log	7.0	161	0.9466	101.1219
No log	8.0	184	0.9552	133.7322
No log	9.0	207	0.9677	147.5692
No log	10.0	230	0.9616	210.0224

Framework versions

• Transformers 4.46.2
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.0