OPAL: a reactive optimizer. A coding agent rewrites the optimizer while your model trains.
- deep-learning
- llm-agents
- optimizer
- pytorch
98,836 visible repositories in the index
241–264 of 284 · Most stars
OPAL: a reactive optimizer. A coding agent rewrites the optimizer while your model trains.
TranslatedOne zero-dependency MCP gateway CLI for every server and agent skill. Token optimization, tool discovery and context compression: 71,929 -> 581 tokens (-99.2%, measured), schemas stay out of context. One config for Claude Code, Codex, Cursor, every agent. 561KB, pure Python. | Zero-dependency CLI: manages all MCP tools and skills, and tool discovery saves 99.2% of tokens.
A model generator for energy system modelling and optimisation (LP/MILP).
Iterative Linear Quadratic Regulator with auto-differentiatiable dynamics models
Hyperparameter optimization and feature selection for scikit-learn using evolutionary algorithms. A modern alternative to GridSearchCV and RandomizedSearchCV.
Free global SEO/GEO skill for Claude Code, Codex & 10 agent hosts — 18 market portals, 56 stdlib scripts, 672 golden tests, self-update intel loop, monitoring daemon, CI gates. Multilingual GEO, pSEO, CWV, penalty recovery.
Self-improving code for any git repo: give it a measurable goal, a coding agent (Claude Code or Aider) proposes changes, the kernel scores them, keeps winners and rolls back failures. Open-source AlphaEvolve / FunSearch-style evolution loop, sandboxed, with an auditable ledger.
An autonomous agent that profiles, optimizes, and validates AI training code.
Official repository for the paper “RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution”
A high-performance, general-purpose toolkit for versatile robotic tasks.
Maze Applied Reinforcement Learning Framework
OpenAeroStruct is a lightweight tool that performs aerostructural optimization using OpenMDAO.
Bayesian Optimization using GPflow
TOMC Memory: CPU context compression and task memory for Codex, Claude Code, Cursor, TRAE, WorkBuddy and other MCP clients.
zoofs is a python library for performing feature selection using a variety of nature-inspired wrapper algorithms. The algorithms range from swarm-intelligence to physics-based to Evolutionary. It's easy to use , flexible and powerful tool to reduce your feature size.
Trading Environment(OpenAI Gym) + PPO(TensorForce)
Ankh: Optimized Protein Language Model
Open source compiler for your whole ML repo (support python, CUDA C/C++) into sass and cuda binary (which are closed sourced by Nvidia); 3-40% faster than Openai Triton for Neural Network (primarily LLM) based training/inferencing
SpecPilot: Autonomous Algorithm Discovery and Systems Optimization for Speculative Decoding
PrismQuant: Optimal Null-Space Rotations for Grouped Quantizers
Safe Bayesian Optimization
Translated🍟 McD Saver (麦麦省钱助手, "McDonald's money-saving assistant") — a money-saving ordering Agent Skill built on McDonald's China's official MCP: real-time price verification via the official calculate-price, coupon combination optimization, sunk-cost check-up and multi-order coupon scheduling, and final prices calculated on the same basis across four channels: official coupons, set meals, group-buy vouchers, and points. Online price estimation without login.
Official code for "Reinforcement Fine-Tuning of Flow-Matching Policies for Vision-Language-Action Models" (ICRA 2026): Flow Policy Optimization for pi0
Universal LLM router for AI coding tools. Works with Claude Code, Cursor, Codex, Gemini CLI, Copilot and more. Free-first fallback chain cuts costs 35-80%.