rasbt/reasoning-from-scratch
Implement a reasoning LLM in PyTorch from scratch, step by step
This project provides practical code examples to understand and implement reasoning capabilities in large language models (LLMs). Starting with a base LLM, you'll add features like inference-time scaling and reinforcement learning to improve its ability to reason. It's intended for AI practitioners and researchers who want to deeply understand how advanced LLMs are built to perform complex reasoning tasks.
3,452 stars. Actively maintained with 16 commits in the last 30 days.
Use this if you are an AI practitioner or researcher who wants to learn the step-by-step process of adding reasoning capabilities to an existing large language model using hands-on code.
Not ideal if you are a non-technical user looking for a pre-built tool to apply reasoning to your data without coding or understanding the underlying mechanics.
Stars
3,452
Forks
495
Language
Jupyter Notebook
License
Apache-2.0
Category
Last pushed
Mar 12, 2026
Commits (30d)
16
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/rasbt/reasoning-from-scratch"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related models
AI-Hypercomputer/maxtext
A simple, performant and scalable Jax LLM!
mindspore-lab/mindnlp
MindSpore + 🤗Huggingface: Run any Transformers/Diffusers model on MindSpore with seamless...
mosaicml/llm-foundry
LLM training code for Databricks foundation models
rickiepark/llm-from-scratch
<밑바닥부터 만들면서 공부하는 LLM>(길벗, 2025)의 코드 저장소
CASE-Lab-UMD/LLM-Drop
The official implementation of the paper "Uncovering the Redundancy in Transformers via a...