Add comprehensive model card with training details and usage examples

Add model card with training details

Processor for refined Whisper model

Refined Whisper model - continued fine-tuning on 49 samples, loss reduced from 2.14 to 0.12

initial commit