large_config

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

• Loss: 0.2441
• Wer: 17.8570
• Cer: 8.0073

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.5155	0.1	1000	0.4045	27.7845	12.1308
0.5436	1.02	2000	0.3784	27.0983	11.9181
0.265	1.12	3000	0.3227	22.8815	10.0236
0.4765	2.04	4000	0.3021	21.9744	9.6005
0.1818	2.14	5000	0.3002	21.0336	9.1830
0.4031	3.05	6000	0.2496	18.0914	7.9181
0.1991	3.15	7000	0.2971	21.7029	9.9984
0.3023	4.07	8000	0.2445	17.7946	7.9248
0.2185	4.17	9000	0.2842	20.3760	9.2891
0.2114	5.09	10000	0.2441	17.8570	8.0073

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu117
• Datasets 2.9.0
• Tokenizers 0.13.2