Single File, Single GPU, From Scratch, Efficient, Full Parameter Tuning library for "RL for LLMs"
☆644Oct 7, 2025Updated last year
Alternatives and similar repositories for nano-aha-moment
Users that are interested in nano-aha-moment are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Implementing DeepSeek R1's GRPO algorithm from scratch☆1,920Apr 18, 2025Updated last year
- Minimalistic 4D-parallelism distributed training framework for education purpose☆2,346Aug 26, 2025Updated last year
- Simple repository for training small reasoning models☆53Aug 4, 2026Updated 3 months ago
- Our library for RL environments + evals☆4,793Updated this week
- A basic pure pytorch implementation of flash attention☆17Oct 28, 2024Updated 2 years ago
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- Kinetics: Rethinking Test-Time Scaling Laws☆87Jul 11, 2025Updated last year
- Focused on fast experimentation and simplicity☆77Dec 24, 2024Updated last year
- A bibliography and survey of the papers surrounding o1☆1,213Jul 7, 2026Updated 4 months ago
- Understanding R1-Zero-Like Training: A Critical Perspective☆1,288Aug 27, 2025Updated last year
- Build your own visual reasoning model☆423Jan 13, 2026Updated 10 months ago
- ☆17Apr 9, 2025Updated last year
- NanoGPT (124M) in 90 seconds☆6,051Sep 28, 2026Updated last month
- Minimal reproduction of DeepSeek R1-Zero☆13,258Feb 27, 2026Updated 8 months ago
- FlexAttention based, minimal vllm-style inference engine for fast Gemma 2 inference.☆367Nov 2, 2025Updated last year
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- Official Repo for Open-Reasoner-Zero☆2,102Jun 2, 2025Updated last year
- AllenAI's post-training codebase☆3,939Updated this week
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆132Apr 17, 2024Updated 2 years ago
- Codebase for the paper "A Percolation Model of Emergence: Analyzing Transformers Trained on a Formal Language"☆11Oct 11, 2024Updated 2 years ago
- Minimalistic large language model 3D-parallelism training☆2,864Oct 8, 2026Updated last month
- Democratizing Reinforcement Learning for LLMs☆5,925Oct 16, 2026Updated last month
- [NeurIPS 2025 Spotlight] Reasoning Environments for Reinforcement Learning with Verifiable Rewards☆1,555Nov 9, 2026Updated last week
- Simple RL training for reasoning☆3,880Dec 23, 2025Updated 10 months ago
- Muon is Scalable for LLM Training☆1,562Aug 3, 2025Updated last year
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- SkyRL: A Modular Full-stack RL Library for LLMs☆2,539Updated this week
- rl from zero pretrain, can it be done? yes.☆297Sep 28, 2025Updated last year
- simple grpo☆13May 28, 2025Updated last year
- Agentic RL Training at Scale☆2,257Updated this week
- Meta Lingua: a lean, efficient, and easy-to-hack codebase to research LLMs.☆4,773Jul 18, 2025Updated last year
- ☆33Jan 6, 2025Updated last year
- What would you do with 1000 H100s...☆1,205Jan 10, 2024Updated 2 years ago
- ☆267Nov 24, 2025Updated 11 months ago
- Exploring Applications of GRPO☆252Aug 25, 2025Updated last year
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Minimal and annotated implementations of key ideas from modern deep learning research.☆1,361Jan 29, 2026Updated 9 months ago
- ☆1,323May 20, 2026Updated 5 months ago
- 🌾 OAT: A research-friendly framework for LLM online alignment, including reinforcement learning, preference learning, etc.☆678Jan 29, 2026Updated 9 months ago
- j1-micro (1.7B) & j1-nano (600M) are absurdly tiny but mighty reward models.☆109Jul 19, 2025Updated last year
- An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Asy…☆10,167Nov 3, 2026Updated 2 weeks ago
- Train transformer language models with reinforcement learning.☆19,745Updated this week
- verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework☆24,266Updated this week
Are these results useful?