pytorch-patterns
PyTorch deep learning patterns and best practices for building robust
by affaan-mFree
Description
pytorch-patterns is an agent skill from the ECC collection by affaan-m, listed on skills.sh. It sets out PyTorch practice for building deep learning training pipelines, model architectures and data loading that are robust, efficient and reproducible. Its first principle is device-agnostic code, which runs on either a CPU or a GPU without hard-coding a device. The skill applies when writing new PyTorch models or training scripts, reviewing deep learning code, debugging training loops or data pipelines, optimising GPU memory use or training speed, and setting up reproducible experiments. Because it is a SKILL.md file of instructions, an AI coding agent loads it for those tasks and applies the same conventions each time. It is free and open source, and the skills.sh listing shows all three security audits passing.…
Highlights
- Device-agnostic CPU and GPU code
- Training pipeline and model architecture guidance
- Data loading practice
- Reproducible experiments and GPU memory tuning
Ratings & Reviews
No ratings yet
Be the first to rate pytorch-patterns.
People also visit
pytorch-patterns FAQ
What does the pytorch-patterns skill do?
It guides an agent in writing and reviewing PyTorch models and training scripts, with reproducible pipelines and efficient data loading.
Does the code need a GPU?
No. The first principle is device-agnostic code that works on both CPU and GPU.
What can it help debug?
Training loops, data pipelines, GPU memory use and training speed.
How do I install it?
Run: npx skills add https://github.com/affaan-m/ecc --skill pytorch-patterns
Is it free?
Yes. It is free and open source.
Are there security findings?
The skills.sh listing shows Gen Agent Trust Hub, Socket and Snyk all passing.
What are the best pytorch-patterns alternatives?
Popular alternatives to pytorch-patterns include pytorch-lightning, torch-geometric and data-scraper-agent. See all pytorch-patterns alternatives