h3nock/scriptify-api

remove unused vars

try the full model (no quantization)

remove priming

support only single batch for now

add updated model checkpoints

refactor: change sampling to torch.inference_mode

refactor: shorten style1 for better inference time on cpu

fix: add style files and fix their path

refactor: update inference to use priming by default

feature(utils): add utilities to support priming logic

feature: add style1 files (.txt and .npy)

refactor: add latest models with priming logic

reafactor: increase valid bias range

fix: update dockerfile to copy quantized model

fix: remove invalid import

feature: add quantized model for cpu inference

fix pytorch version in req.txt

add docker config

fix: fix REAME.md format

configure git lfs to track model files

initial commit

feat: init commit for hugging face space

initial commit