Pragmatic framework to build LLM Copilots.
PokerStars bot that plays real-time poker using GPT-4 for decision-making in cash games.
A fast communication-overlapping library for tensor/expert parallelism on GPUs.
Autonomously train research-agent LLMs on custom data using reinforcement learning and self-verification.
An open-source solution for full parameter fine-tuning of DeepSeek-V3/R1 671B, including complete code and scripts.
A collaborative repository for sharing examples showcasing Mistral models and tools developed by the community.
An experimental tool for controlling reasoning depth in language models using explicit thinking tokens.
A reflection agent that uses a two-agent system to validate and improve outputs in AI applications.
Agent S is an open-source framework that enables autonomous computer interactions using intelligent agents.
A browser extension that helps users publish content to multiple social media platforms with one click.
A powerful Obsidian plugin for automatic note synchronization across various cloud services with end-to-end encryption.
A Model Context Protocol server for automating browsers and APIs in various IDEs.
