pradachan/whisper-large-v3-turbo-disfluency-lora

meta: add pipeline_tag, wer metric, model-index eval result (3.4% WER), link bespoke dataset

docs: credit LibriSpeech as the synthetic-data source, clarify it is not trained on DisfluencySpeech, add dataset links

fix: drop null task_type key (Whisper LoRA loads via generic PeftModel; clears Hub config-parsing warning)

Upload README.md with huggingface_hub

Upload folder using huggingface_hub

initial commit