korbih/whisper-small-hi

Whisper Small Hi - Sanchit Gandhi

This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4377
• Wer: 32.9848

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: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0.0	1	2.2652	86.7857
0.1858	1.22	500	0.3301	39.7317
0.0881	2.44	1000	0.2966	34.9065
0.0457	3.67	1500	0.3160	33.8695
0.0195	4.89	2000	0.3571	33.9287
0.0047	6.11	2500	0.3913	33.4843
0.0014	7.33	3000	0.4186	32.9637
0.0005	8.56	3500	0.4286	33.0737
0.0005	9.78	4000	0.4377	32.9848

Framework versions

• Transformers 4.40.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2