whisper-base-mzn

This model is a fine-tuned version of openai/whisper-base on the razhan/DOLMA-speech mazanderani dataset. It achieves the following results on the evaluation set:

• Loss: 1.3368
• Wer: 0.8150
• Cer: 0.3173

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: 192
• eval_batch_size: 128
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• total_train_batch_size: 384
• total_eval_batch_size: 256
• 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_steps: 1
• num_epochs: 5.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
No log	1.0	3	2.1575	1.3606	0.7428
No log	2.0	6	2.1575	1.3606	0.7428
No log	3.0	9	1.6346	0.9451	0.6618
2.0051	4.0	12	1.4320	0.8414	0.3594
2.0051	5.0	15	1.3368	0.8150	0.3173

Framework versions

• Transformers 4.49.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0