whisper-tiny-minds14-en_US-test-finetuned

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: 1.0871
• Wer Ortho: 26.8342
• Wer: 0.2638

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 50
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.0021	17.86	500	0.7863	26.6304	0.2534
0.0002	35.71	1000	0.8689	26.7663	0.2612
0.0001	53.57	1500	0.9230	27.2418	0.2664
0.0001	71.43	2000	0.9637	27.1739	0.2664
0.0	89.29	2500	0.9977	26.9022	0.2638
0.0	107.14	3000	1.0277	27.1739	0.2664
0.0	125.0	3500	1.0571	27.1739	0.2671
0.0	142.86	4000	1.0871	26.8342	0.2638

Framework versions

• Transformers 4.38.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.15.2