Update README

Remove int4 variant which had too poor quality

Tidy up repo and conversion scripts

Remove unnecessary retry for prefill

Use named args in CLI

Tidy up example CLI

Fix incorrect comment

Workaround for bad ML state on MacOS

WIP mac compatibility

Performance improvements on iOS

Add swift test CLI

Drop prefill model, add tokenizer config

Tidy up the repo

Fix audio decoder, fix broken KV cache

WIP audio decoder

Fix a NaN attention issue in converted model

Add CoreML prefill and decode models

Inherit Plapre license

Create README.md

initial commit