whisperFinetuneTakeTwo

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

• Loss: 0.5685
• Wer: 26.8493

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: 0.001
• train_batch_size: 128
• eval_batch_size: 8
• 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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.8768	0.2778	10	1.8749	35.4033
0.8301	0.5556	20	0.6687	24.9619
0.5543	0.8333	30	0.5484	22.9224
0.3902	1.1111	40	0.5108	20.3044
0.3395	1.3889	50	0.4900	20.6088
0.3255	1.6667	60	0.4830	20.7915
0.362	1.9444	70	0.4867	20.6393
0.1228	2.2222	80	0.5114	20.8524
0.1288	2.5	90	0.5299	21.3090
0.1513	2.7778	100	0.5685	26.8493

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.1.dev0
• Tokenizers 0.19.1