jethrowang/whisper-tiny_tat-esc_exp_una-gan

Whisper Tiny Taiwanese (exp_una-gan)

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

• Loss: 1.3251
• Cer: 29.5038

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 32
• 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_steps: 681
• training_steps: 6810
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.3994	0.9985	681	0.9444	60.1427
0.2352	1.9971	1362	0.8823	31.7033
0.1455	2.9956	2043	0.9547	32.2316
0.0828	3.9941	2724	1.0160	32.6732
0.0457	4.9927	3405	1.1148	32.5119
0.0247	5.9912	4086	1.1846	31.0990
0.013	6.9897	4767	1.2245	31.3227
0.0047	7.9883	5448	1.2999	30.9868
0.0014	8.9868	6129	1.3053	29.6578
0.0008	9.9853	6810	1.3251	29.5038

Framework versions

• Transformers 4.49.0
• Pytorch 2.0.0.post304
• Datasets 3.3.2
• Tokenizers 0.21.0