CheeLi03/whisper-5b

Whisper Small Five - Chee Li

This model is a fine-tuned version of openai/whisper-base on the Google Fleurs dataset. It achieves the following results on the evaluation set:

• Loss: 0.4348
• Wer: 21.6569

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 850
• training_steps: 8000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3047	1.0560	1000	0.4050	22.7955
0.1771	2.1119	2000	0.3811	21.0509
0.1176	3.1679	3000	0.3864	21.3429
0.0852	4.2239	4000	0.3962	20.7720
0.0443	5.2798	5000	0.4100	21.4523
0.0265	6.3358	6000	0.4234	21.3823
0.0254	7.3918	7000	0.4310	21.7783
0.0188	8.4477	8000	0.4348	21.6569

Framework versions

• Transformers 4.43.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1