How to run within NVIDIA Container

[TommasoBendinelli]

Hello, I have a Grace Hopper Chip, and I would like to install and run LitGPT. I was planning to use this dockerfile to build the container:

FROM nvcr.io/nvidia/pytorch:24.10-py3
ENV DEBIAN_FRONTEND=noninteractive
RUN apt-get update && apt-get install -y python3.10-venv && apt-get clean && rm -rf /var/lib/apt/lists/*
RUN pip install 'litgpt[all]'

The base image nvcr.io/nvidia/pytorch:24.10-py3 comes with torch 2.5.0, so it should meet the requirement. However, it looks like LitGPT is uninstalling and then reinstalling torch. Do you have any idea why this is happening and how it can be avoided?