👋 Hello @mazatov, thank you for your interest in Ultralytics 🚀! We recommend taking a look at the Docs for detailed guidance and examples. You may find the sections on Python and CLI usage particularly useful as they explain training configurations in detail.

If this is a 🐛 Bug Report, please provide a minimum reproducible example that we can use to debug the issue.

From your description, it seems like you’re attempting to adjust the number of epochs while resuming training. Please ensure:

1. You use the correct syntax for overriding the parameters when running the training command.
2. Your modifications persist in the correct locations—e.g., args.yaml or explicitly passed via the command line.

If this is a ❓ Question, please share the following to help us assist you better:

• The exact command you're using
• Any relevant training logs or outputs
• The content of your args.yaml file

To explore similar topics or get direct help from the community, consider joining the Ultralytics community:

• Discord 🎧 for real-time conversation
• Discourse for in-depth discussions
• Subreddit for topic-based sharing

Upgrade

First, ensure you’re using the latest version of the ultralytics package. Issues are often resolved in newer updates, so upgrading might solve this problem:

pip install -U ultralytics

Make sure you're running this in a Python>=3.8 environment with PyTorch>=1.8 installed.

Environments

For running YOLO, verified environments include:

• Notebooks with free GPU:
• Cloud environments: Options like GCP or AWS
• Local setups: Including Docker. See our Docker Quickstart Guide

Status

When this badge is green, all Ultralytics CI tests are passing. CI runs validate all YOLO Modes and Tasks across macOS, Windows, and Ubuntu systems on a 24-hour cycle and after every commit.

🤖 This is an automated response, but an Ultralytics engineer will follow up with you shortly to provide further assistance!