SPEAK-ASR/whisper-si-exp-8

This model is a fine-tuned version of openai/whisper-small on the SPEAK-ASR/openslr-sinhala-asr-preprocessed dataset. It achieves the following results on the evaluation set:

• Loss: 0.1671
• Wer: 19.7550

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 256
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 128
• total_eval_batch_size: 1024
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.8759	1.2942	1500	0.2156	24.5941
0.7228	2.5884	3000	0.1802	21.0896
0.6525	3.8827	4500	0.1671	19.7550

Framework versions

• PEFT 0.18.1
• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2