Dan Biderman comments

Results 16 comments of


                                            Dan Biderman

Question on video loading

@JanuszL @szalpal was traveling, will get you a repro as soon as I can. a question in the meanwhile: say I want to sequentially read 16 frames from a video,...

Trying to add custom-trained YOLOv8n-pose model

Possible it's a CUDA/DALI issue... I also see that you use `python=3.11.6`. First, let's try to create a new conda environment with `python=3.8` ```conda create --name python=3.8``` Then re-install lightning-pose...

Feature/peft compatible models

refactored the hf convertor to a single function as suggested by @dakinggg. tested it on my end and ran pre-commit successfully. I want to move forward and push the code...

Adding filtering for optimizer parameters to only trainable parameters

looks good! will make fsdp lighter for models with some frozen weights

over confident pseudo tracking when animal exits field of view

@DavidGill159 -- the way we calculate the confidence score is different from DeepLabCut, and our scores will by definition be higher. Using our scaling, it is advised to use `0.95`...

ImportError: cannot import name 'Array' from 'jaxtyping'

@MilesCranmer thanks. @patrick-kidger the `jax` requirement was relaxed? I don't see it anymore in `pyproject.toml`. If so i'll modify my code according to the syntax suggested by Miles