facebookresearch / facebookresearch/sonata

what is the best-practice of deploy?

Open
#40 1 comment 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
795
Forks
56
PR merge metrics
No merged PRs in 30d

Description

Hi thanks for this greak work.

If I want to deploy sonota/ point transformer to my NVIDIA CUDA device, what is the best practice and minimum requirements?

I see
- torch-scatter --find-links https://data.pyg.org/whl/torch-2.5.0+cu124.html
- git+https://github.com/Dao-AILab/flash-attention.git
- spconv-cu124

these packages used, so is it impossible to convert point transformer to something like JIT/onnx/tvm

Thank you

Contributor guide

Open the contributing guide

Research direction

No file, test, or entry point is named. First clarify the supported NVIDIA CUDA requirements, deployment target, and whether JIT, ONNX, or TVM conversion is expected; done would be a documented, reproducible deployment path with minimum requirements and known limitations.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
computer-vision, machine-learning
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.