rishabhjain16/whisper_small_en_to_pf10h

openai/whisper-small.en

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

• Loss: 0.1764
• Wer: 2.9777

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: 64
• 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
0.0179	10.0	500	0.1422	3.4691
0.0006	20.0	1000	0.1530	3.0001
0.0004	30.01	1500	0.1631	3.0150
0.0002	40.01	2000	0.1672	2.9777
0.0001	51.0	2500	0.1717	2.9703
0.0001	61.0	3000	0.1742	2.9926
0.0001	71.01	3500	0.1759	2.9852
0.0001	81.01	4000	0.1764	2.9777

Framework versions

• Transformers 4.27.0.dev0
• Pytorch 1.13.1+cu117
• Datasets 2.9.1.dev0
• Tokenizers 0.13.2