whisper-large-stt4sg

This model is a fine-tuned version of openai/whisper-large-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2355
• Wer: 15.5402

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-06
• 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: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2994	0.0801	1000	0.3001	19.6658
0.2953	0.1602	2000	0.2741	18.0230
0.2638	0.2403	3000	0.2575	17.0951
0.2456	0.3205	4000	0.2421	15.9823
0.2442	0.4006	5000	0.2355	15.5402

Framework versions

• Transformers 4.44.1
• Pytorch 2.3.1+cu118
• Datasets 2.20.0
• Tokenizers 0.19.1