whisper-turbo-tr_All_datasets_finetune_5

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2090

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use 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_ratio: 0.1
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.0769	0.2923	1500	0.2720
0.0768	0.5846	3000	0.3229
0.0626	0.8769	4500	0.3273
0.042	1.1691	6000	0.3147
0.0439	1.4614	7500	0.3314
0.0358	1.7537	9000	0.2714
0.0169	2.0460	10500	0.2531
0.016	2.3383	12000	0.2445
0.0155	2.6306	13500	0.2286
0.0119	2.9228	15000	0.2090

Framework versions

• PEFT 0.15.1
• Transformers 4.51.3
• Pytorch 2.5.1
• Datasets 3.0.0
• Tokenizers 0.21.1