Asgat
Pytorch benchmarking source code for decimation model
Install / Use
/learn @seanli3/AsgatREADME
ASGAT
Installation
You can install the dependencies by one of the following two methods.
Conda
The conda env file assumes you would use CUDA11.
conda env create -f environment.yml
pip
- Install PyTorch 1.9+
- Install PyTorch Geometrics
pip install -r requirements.txt
Run
./run.sh
Note
The accuracy performance might differ a little depending on the GPU vs. CPU, CUDA driver, Pytorch version, etc. We try to make the results as deterministic as possible using torch.use_deterministic but this feature is experimental and not stable.
For details, see
https://github.com/rusty1s/pytorch_geometric/issues/2788
https://pytorch.org/docs/stable/notes/randomness.html
Attribution
Some of the benchmark implimentations are forked from https://github.com/rusty1s/pytorch_geometric/tree/master/benchmark
Related Skills
node-connect
348.2kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
108.9kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
348.2kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
348.2kQQBot 富媒体收发能力。使用 <qqmedia> 标签,系统根据文件扩展名自动识别类型(图片/语音/视频/文件)。
