miosipof/asr2_aug_IT_v5_full

Whisper Medium IT

This model is a fine-tuned version of miosipof/asr2_aug_IT_v4_merged on the b-brave-balanced-augmented dataset. It achieves the following results on the evaluation set:

• Loss: 0.1611
• Wer: 19.0255
• Cer: 12.3586
• Lr: 0.0000

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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: 0.3
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer	Lr
No log	0.9829	43	0.2443	25.7541	18.4332	0.0000
0.7954	1.9829	86	0.2087	24.3619	16.3385	0.0000
0.4587	2.9829	129	0.1720	22.9698	15.7101	0.0000
0.197	3.9829	172	0.1611	19.0255	12.3586	0.0000

Framework versions

• Transformers 4.47.1
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1