Merge pull request #6 from Passiolife/main

rename new function

use torch.matmul over torch.mm for coreml compatibility. Add apply_lora_by_name

Update README, add nanoGPT example

Refactor, Add advanced exapmles and weight tying

fix embedding fan_in/fan_out

quality improvement

Init