PaddlePaddle / PARL

A high-performance distributed training framework for Reinforcement Learning
https://parl.readthedocs.io/
Apache License 2.0
3.22k stars 817 forks source link

Use the nvidia-supplied nvidia-ml-py instead of pynvml #1101

Open mattip opened 1 year ago

mattip commented 1 year ago

This project depends on pynvml, where the nvidia-blessed bindings to nvml are in the nvidia-ml-py package instead. This is causing problems for the gpustat project, which needs some internals from nvidia-ml-py to properly support older nvml version. Would it be possible to switch?