whisper-tiny-ashok

This model is a fine-tuned version of openai/whisper-tiny on the PolyAI/minds14 dataset. It achieves the following results on the evaluation set:

• Loss: 0.8440
• Wer Ortho: 34.6847
• Wer: 0.3530

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-05
• train_batch_size: 32
• 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: 50
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.0234	6.67	100	0.6639	34.2986	0.3383
0.003	13.33	200	0.7587	33.9768	0.3401
0.0005	20.0	300	0.7870	34.2342	0.3475
0.0003	26.67	400	0.8045	35.1351	0.3567
0.0002	33.33	500	0.8144	35.5856	0.3610
0.0001	40.0	600	0.8262	35.5212	0.3604
0.0001	46.67	700	0.8341	35.3282	0.3592
0.0001	53.33	800	0.8397	35.1995	0.3579
0.0001	60.0	900	0.8426	34.7490	0.3536
0.0001	66.67	1000	0.8440	34.6847	0.3530

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3