runanywhere
/

whisper_tiny_4bit

Model card Files Files and versions

whisper_tiny_4bit / Readme.md

AlgoX's picture

Flatten structure: remove subdirectory

8eb3dc8 2 days ago

|

history blame contribute delete

185 Bytes

OpenAI's Whisper large-v3 converted to MLX format and quantized up to 4 bits, with group size 64.