# whwangovo/pyre-code

A self-hosted ML coding practice platform. 68 problems from ReLU to flow matching — attention, training, RLHF, diffusion, and more. Instant feedback in the browser.

Repository: https://github.com/whwangovo/pyre-code
Canonical: https://ross.abutalabs.com/products/pyre-code
Language: Python
License: NOASSERTION
License Family: other
Last push: 2026-05-12T14:55:30+00:00

## Health v2 (maintenance only)
Score: 51/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 82, release rhythm 35, longevity 10
- inputs: {"age_days": 146, "days_push": 113, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1253, forks 113 (observed 2026-08-28T04:04:08.443378+00:00)

## What it is
A self-hosted coding practice platform with ~76 ML implementation problems covering attention, training, RLHF, diffusion, and GNNs, graded instantly by a local FastAPI judge service. It runs entirely in the browser with a Monaco editor and requires no GPU.

## Use cases
- practice implementing transformers attention from scratch
- prepare for machine learning interview coding questions
- learn how vLLM and TRL internals work by coding them
- implement diffusion and flow matching models step by step
- practice writing graph neural networks like GCN and GAT
- get instant test feedback on ML exercise submissions
- self-hosted LeetCode-style platform for ML code

## When to choose
- you learn best by writing code rather than reading papers or watching lectures
- you are preparing for ML engineering interviews and need hands-on implementation practice
- you want a fully local, GPU-free environment to practice deep learning internals
- you want reference solutions and per-test-case feedback for ML exercises

## When to avoid
- you need a production ML training or inference framework rather than practice problems
- you want guided lectures or courses instead of self-directed exercises
- you need GPU-accelerated workloads or large-scale experiments
- you require a permissively licensed project, since the license is not a standard OSI license

## Facets
- artifact type: learning-resource
- maturity: active
- function: machine-learning, deep-learning, llm-training, llm-inference, rag, developer-tools, self-hosted
- domain: machine-learning, deep-learning, large-language-models, education, tutorials, artificial-intelligence
- platform: self-hosted, python, cross-platform
- tags: coding-practice, ml-interview-prep, exercises, grading-service, transformers, diffusion-models, reinforcement-learning, graph-neural-networks, browser-editor, monaco-editor, fastapi, sqlite, web-server, docker

## Member repositories
- whwangovo/pyre-code (main) score 51

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:08.443378+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T05:07:31.022944+00:00, confidence not recorded.
  - readme: https://github.com/whwangovo/pyre-code (fetched 2026-08-28T04:04:08.443378+00:00, sha 57a9c36e9432)
- Data as of 2026-08-30T08:39:29.467469+00:00.