Ganaa0614/whisper-tiny-mongolian-ver_0.3

whisper-tiny-mongolian-ver_0.2

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

• Loss: 1.5086
• Wer: 5.9374
• Cer: 4.5374

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: 5e-06
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
5.5777	0.7828	200	2.6264	3.8311	2.8421
3.8879	1.5636	400	1.9959	6.9300	5.2355
3.2323	2.3444	600	1.7458	7.3196	5.7269
2.9454	3.1252	800	1.6397	5.8093	4.4308
2.7639	3.9080	1000	1.5831	7.1481	5.4729
2.6266	4.6888	1200	1.5506	6.6786	5.1936
2.5721	5.4697	1400	1.5296	5.7130	4.3788
2.4195	6.2505	1600	1.5180	5.6403	4.3207
2.3462	7.0313	1800	1.5105	5.7683	4.4404
2.3161	7.8141	2000	1.5086	5.9374	4.5374

Framework versions

• Transformers 5.3.0.dev0
• Pytorch 2.7.1+cu128
• Datasets 3.6.0
• Tokenizers 0.22.2