This curated guide brings together 16 vetted Reinforcement Learning resources, including 13 with structured technical analysis and 0 with practitioner field notes. Primary focus areas include Python Library, AI Agent Framework, AI Agent Infrastructure, AI Development Tool, AI Development Toolkit. Primary sources include github.com, docs.unsloth.ai, osmosis.ai, rikiyat.github.io, rohanbansal.com.
| Resource | Category & Format |
|---|
| Key Takeaway & Notes |
|---|
| 0xSero/home-rl-wiki: Wikipedia-style wiki on RL post-training with LoRA on home hardwaregithub.com | Technical Wiki / self-hosted wiki with public static site / open source (GitHub), free to run locally or view online | Covers RL post-training concepts, algorithms, hardware specs, models, frameworks, and run-time estimates for home setups |
|---|---|---|
| Sparks of Weight-Space Programming: LLMs can natively read & write neural network weightsrikiyat.github.io | Research Blog Post / Research blog post / Free online | Frontier LLMs can directly author neural network weights without conventional training or gradient-based optimization |
| GitHub - togethercomputer/HybridAgentgithub.com | AI Agent Framework / Open-source code repository with documentation / Open source on GitHub, models on Hugging Face Hub | Dynamically alternates between latent reasoning and text-based interaction within a single generation process |
| Training a 4B model to produce 81% faster query plans than Postgresrohanbansal.com | Machine Learning Research / long-form technical blog post / free online | 44.7% latency reduction across 113 join-heavy queries from a 4B model that initially failed on 99 of them |
| GitHub - Human-Agent-Society/reef: Continual learning infra for self-improving agentsgithub.com | AI Agent Infrastructure / open-source software project / open source, free on PyPI | Four-step learning cycle: Serve, Observe, Grow, Commit — connecting inference, feedback, learning, and versioned delivery |
| microsoft/agent-lightning: The absolute trainer to light up AI agents.github.com | Python Library / open-source software library with documentation and research papers / free, open source on GitHub; installable via pip | Zero code change integration with any agent framework (LangChain, AutoGen, CrewAI, OpenAI Agent SDK, or raw OpenAI) via a lightweight tracer |
| thinking-machines-lab/tinker-cookbook: Post-training with Tinkergithub.com | Python Library / open-source library with documentation and interactive notebooks / open source on GitHub; requires Tinker API key for training access | Offloads distributed training to Thinking Machines Lab infrastructure via a simple API, abstracting away cluster management |
| fw-ai/cookbook: Recipes and resources for building, deploying, and fine-tuning generative AI with Fireworks.github.com | AI Development Toolkit / open source repository / open source | Provides ready-to-run training recipes for RL (GRPO, DAPO), preference optimization (DPO, ORPO), and supervised fine-tuning (SFT) |
| GitHub - Danau5tin/ai-trains-ai: RL-training an AI agent to RL-train AI agents.github.com | Machine Learning Project / open source repository / open source | Implements a dual-loop RL architecture (Outer loop for the trainer, Inner loop for the target model) |
| GitHub - g023/g023_agentic_chat: g023’s Agentic Chatgithub.com | AI Development Tool / open source software / open source | Single-Tool Architecture using python_exec for all operations |
| Tongyi DeepResearch: A New Era of Open-Source AI Researcherstongyi-agent.github.io | Open-Source AI Agent / Technical blog post with model release / Open source on GitHub, HuggingFace, and ModelScope | First fully open-source web agent to match OpenAI's DeepResearch across comprehensive benchmarks |
| Applying RL: Fixing Structured Outputsosmosis.ai | AI/ML Research Blog / Technical blog post with benchmark comparisons / Free online article; model available on Ollama and Hugging Face | Structured output mode enforces a schema that prevents models from thinking freely, causing accuracy drops on complex reasoning tasks |
| GitHub - FoundationAgents/OpenManus: No fortress, purely open ground. OpenManus is Coming.github.com | Open Source AI Framework / Open source repository / Open source | Provides an open-source alternative to the proprietary Manus.im system |
| Introducing System One Models & Jev - TypeSafe AI Blogtypesafe.ai | Reference | TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in… |
| GLM-5: From Vibe Coding to Agentic Engineeringz.ai | Reference | GLM-5 is a 744B-parameter MoE model (40B active) from Zhipu AI, scaled up from GLM-4.5's 355B with 28.5T pre-training tokens and DeepSeek Sparse Attention for long-context… |
A collection of bookmarks filtered by the tag "Reinforcement Learning".
Showing 1-16 of 16 bookmarks tagged with "Reinforcement Learning"
All Stories
Community recipes for serving LLMs on RTX 3090. Multi-engine (vLLM, llama.cpp, SGLang) and model-agnostic. Currently shipping Qwen3.6-27B configs for ...
