Luandrie/whisper-afrispeech-callcenter

Distill Whisper Call Center Afri

This model is a fine-tuned version of distil-whisper/distil-large-v3 on the www_call_center_english_merged_w_afrispeech dataset. It achieves the following results on the evaluation set:

• Loss: 0.7519
• Wer: 16.8840

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
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.171	3.3613	1000	0.4087	15.9136
0.0691	6.7227	2000	0.4971	16.6982
0.0121	10.0840	3000	0.6424	16.6087
0.0025	13.4454	4000	0.7519	16.8840

Framework versions

• Transformers 4.45.2
• Pytorch 2.6.0+cu124
• Datasets 3.4.0
• Tokenizers 0.20.3