Multi node training of YOLOv8 Stuck in the following steps about ultralytics HOT 2 OPEN

👋 Hello @Python123wbl456, thank you for your interest in Ultralytics YOLOv8 🚀! We recommend a visit to the Docs for new users where you can find many Python and CLI usage examples and where many of the most common questions may already be answered.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Join the vibrant Ultralytics Discord 🎧 community for real-time conversations and collaborations. This platform offers a perfect space to inquire, showcase your work, and connect with fellow Ultralytics users.

Install

Pip install the ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8.

pip install ultralytics

Environments

YOLOv8 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all Ultralytics CI tests are currently passing. CI tests verify correct operation of all YOLOv8 Modes and Tasks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

from ultralytics.

@Python123wbl456 it looks like your multi-node training setup is mostly correct, but the logs indicate a potential issue with the NCCL communication library, specifically with the failure to open libnccl-net.so and libibverbs.so. These libraries are essential for efficient GPU communication in multi-node setups.

Here are a couple of steps you can take to resolve this:

1. Ensure NCCL is installed correctly: Verify that NCCL is installed and correctly configured on both nodes. You might need to reinstall or update NCCL.

2. Check Network Configuration: Ensure that the network interface specified (NCCL_SOCKET_IFNAME=eno4) is correct and consistent across all nodes. You can check available interfaces using ifconfig.

3. Permissions and Paths: Check that the paths to NCCL libraries are correctly set and that your user has the necessary permissions to access them.

If these steps don't resolve the issue, you might want to consider posting a more detailed question with specific error logs on platforms like Stack Overflow or NVIDIA's developer forums, where a community of developers can help troubleshoot NCCL-specific issues.

from ultralytics.