Thinkless is a learnable framework that empowers a Language and Reasoning Model (LLM) to adaptively select between short-form and long-form reasoning based on task complexity and model's ability. It is trained under a reinforcement learning paradigm and employs control tokens for concise responses and detailed reasoning. The core method uses a Decoupled Group Relative Policy Optimization (DeGRPO) algorithm to govern reasoning mode selection and improve answer accuracy, reducing long-chain thinking by 50% - 90% on benchmarks like Minerva Algebra and MATH-500. Thinkless enhances computational efficiency of Reasoning Language Models.

This repository contains an unofficial implementation of the xLSTM model introduced in Beck et al. (2024). It serves as a didactic tool to explain the details of a modern Long-Short Term Memory model with competitive performance against Transformers or State-Space models. The repository also includes a Lightning-based implementation of a basic LLM for multi-GPU training. It provides modules for scalar-LSTM and matrix-LSTM, as well as an xLSTM LLM built using Pytorch Lightning for easy training on multi-GPUs.

simple_GRPO is a very simple implementation of the GRPO algorithm for reproducing r1-like LLM thinking. It provides a codebase that supports saving GPU memory, understanding RL processes, trying various improvements like multi-answer generation, regrouping, penalty on KL, and parameter tuning. The project focuses on simplicity, performance, and core loss calculation based on Hugging Face's trl. It offers a straightforward setup with minimal dependencies and efficient training on multiple GPUs.

LLM Reasoners is a library that enables LLMs to conduct complex reasoning, with advanced reasoning algorithms. It approaches multi-step reasoning as planning and searches for the optimal reasoning chain, which achieves the best balance of exploration vs exploitation with the idea of "World Model" and "Reward". Given any reasoning problem, simply define the reward function and an optional world model (explained below), and let LLM reasoners take care of the rest, including Reasoning Algorithms, Visualization, LLM calling, and more!

Minic is a chess engine developed for learning about chess programming and modern C++. It is compatible with CECP and UCI protocols, making it usable in various software. Minic has evolved from a one-file code to a more classic C++ style, incorporating features like evaluation tuning, perft, tests, and more. It has integrated NNUE frameworks from Stockfish and Seer implementations to enhance its strength. Minic is currently ranked among the top engines with an Elo rating around 3400 at CCRL scale.

LangBridge is a tool that bridges mT5 encoder and the target LM together using only English data. It enables models to effectively solve multilingual reasoning tasks without the need for multilingual supervision. The tool provides pretrained models like Orca 2, MetaMath, Code Llama, Llemma, and Llama 2 for various instruction-tuned and not instruction-tuned scenarios. Users can install the tool to replicate evaluations from the paper and utilize the models for multilingual reasoning tasks. LangBridge is particularly useful for low-resource languages and may lower performance in languages where the language model is already proficient.

This repository contains code for training reward models for Deep Reinforcement Learning-based Reward-modulated Hierarchical Fine-tuning (DRL-based RLHF), Iterative Selection Fine-tuning (Rejection sampling fine-tuning), and iterative Decision Policy Optimization (DPO). The reward models are trained using a Bradley-Terry model based on the Gemma and Mistral language models. The resulting reward models achieve state-of-the-art performance on the RewardBench leaderboard for reward models with base models of up to 13B parameters.

BAML is a config file format for declaring LLM functions that you can then use in TypeScript or Python. With BAML you can Classify or Extract any structured data using Anthropic, OpenAI or local models (using Ollama) ## Resources  [Discord Community](https://discord.gg/boundaryml)  [Follow us on Twitter](https://twitter.com/boundaryml) * Discord Office Hours - Come ask us anything! We hold office hours most days (9am - 12pm PST). * Documentation - Learn BAML * Documentation - BAML Syntax Reference * Documentation - Prompt engineering tips * Boundary Studio - Observability and more #### Starter projects * BAML + NextJS 14 * BAML + FastAPI + Streaming ## Motivation Calling LLMs in your code is frustrating: * your code uses types everywhere: classes, enums, and arrays * but LLMs speak English, not types BAML makes calling LLMs easy by taking a type-first approach that lives fully in your codebase: 1. Define what your LLM output type is in a .baml file, with rich syntax to describe any field (even enum values) 2. Declare your prompt in the .baml config using those types 3. Add additional LLM config like retries or redundancy 4. Transpile the .baml files to a callable Python or TS function with a type-safe interface. (VSCode extension does this for you automatically). We were inspired by similar patterns for type safety: protobuf and OpenAPI for RPCs, Prisma and SQLAlchemy for databases. BAML guarantees type safety for LLMs and comes with tools to give you a great developer experience:  Jump to BAML code or how Flexible Parsing works without additional LLM calls. | BAML Tooling | Capabilities | | ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | BAML Compiler install | Transpiles BAML code to a native Python / Typescript library (you only need it for development, never for releases) Works on Mac, Windows, Linux  | | VSCode Extension install | Syntax highlighting for BAML files Real-time prompt preview Testing UI | | Boundary Studio open (not open source) | Type-safe observability Labeling |

Quiz Generator is a plugin for Obsidian that uses AI models to create interactive exam-style questions from notes. It supports various question types and provides real-time feedback. Users can save questions, generate in multiple languages, and use math support. The tool is suitable for students preparing for exams and educators designing assessments.

LLM-Pruner is a tool for structural pruning of large language models, allowing task-agnostic compression while retaining multi-task solving ability. It supports automatic structural pruning of various LLMs with minimal human effort. The tool is efficient, requiring only 3 minutes for pruning and 3 hours for post-training. Supported LLMs include Llama-3.1, Llama-3, Llama-2, LLaMA, BLOOM, Vicuna, and Baichuan. Updates include support for new LLMs like GQA and BLOOM, as well as fine-tuning results achieving high accuracy. The tool provides step-by-step instructions for pruning, post-training, and evaluation, along with a Gradio interface for text generation. Limitations include issues with generating repetitive or nonsensical tokens in compressed models and manual operations for certain models.

PaperQA is a minimal package for question and answering from PDFs or text files, providing very good answers with in-text citations. It uses OpenAI Embeddings to embed and search documents, and follows a process of embedding docs and queries, searching for top passages, creating summaries, scoring and selecting relevant summaries, putting summaries into prompt, and generating answers. Users can customize prompts and use various models for embeddings and LLMs. The tool can be used asynchronously and supports adding documents from paths, files, or URLs.

LangEvals is an all-in-one Python library for testing and evaluating LLM models. It can be used in notebooks for exploration, in pytest for writing unit tests, or as a server API for live evaluations and guardrails. The library is modular, with 20+ evaluators including Ragas for RAG quality, OpenAI Moderation, and Azure Jailbreak detection. LangEvals powers LangWatch evaluations and provides tools for batch evaluations on notebooks and unit test evaluations with PyTest. It also offers LangEvals evaluators for LLM-as-a-Judge scenarios and out-of-the-box evaluators for language detection and answer relevancy checks.

LlamaBot is a Pythonic bot interface to Large Language Models (LLMs), providing an easy way to experiment with LLMs in Jupyter notebooks and build Python apps utilizing LLMs. It supports all models available in LiteLLM. Users can access LLMs either through local models with Ollama or by using API providers like OpenAI and Mistral. LlamaBot offers different bot interfaces like SimpleBot, ChatBot, QueryBot, and ImageBot for various tasks such as rephrasing text, maintaining chat history, querying documents, and generating images. The tool also includes CLI demos showcasing its capabilities and supports contributions for new features and bug reports from the community.

aiorun is a Python package that provides a `run()` function as the starting point of your `asyncio`-based application. The `run()` function handles everything needed during the shutdown sequence of the application, such as creating a `Task` for the given coroutine, running the event loop, adding signal handlers for `SIGINT` and `SIGTERM`, cancelling tasks, waiting for the executor to complete shutdown, and closing the loop. It automates standard actions for asyncio apps, eliminating the need to write boilerplate code. The package also offers error handling options and tools for specific scenarios like TCP server startup and smart shield for shutdown.

Trace is a new AutoDiff-like tool for training AI systems end-to-end with general feedback. It generalizes the back-propagation algorithm by capturing and propagating an AI system's execution trace. Implemented as a PyTorch-like Python library, users can write Python code directly and use Trace primitives to optimize certain parts, similar to training neural networks.

Thinkless is a learnable framework that empowers a Language and Reasoning Model (LLM) to adaptively select between short-form and long-form reasoning based on task complexity and model's ability. It is trained under a reinforcement learning paradigm and employs control tokens for concise responses and detailed reasoning. The core method uses a Decoupled Group Relative Policy Optimization (DeGRPO) algorithm to govern reasoning mode selection and improve answer accuracy, reducing long-chain thinking by 50% - 90% on benchmarks like Minerva Algebra and MATH-500. Thinkless enhances computational efficiency of Reasoning Language Models.

LightEval is a lightweight LLM evaluation suite that Hugging Face has been using internally with the recently released LLM data processing library datatrove and LLM training library nanotron. We're releasing it with the community in the spirit of building in the open. Note that it is still very much early so don't expect 100% stability ^^' In case of problems or question, feel free to open an issue!

Firefly is an open-source large model training project that supports pre-training, fine-tuning, and DPO of mainstream large models. It includes models like Llama3, Gemma, Qwen1.5, MiniCPM, Llama, InternLM, Baichuan, ChatGLM, Yi, Deepseek, Qwen, Orion, Ziya, Xverse, Mistral, Mixtral-8x7B, Zephyr, Vicuna, Bloom, etc. The project supports full-parameter training, LoRA, QLoRA efficient training, and various tasks such as pre-training, SFT, and DPO. Suitable for users with limited training resources, QLoRA is recommended for fine-tuning instructions. The project has achieved good results on the Open LLM Leaderboard with QLoRA training process validation. The latest version has significant updates and adaptations for different chat model templates.

Awesome Text2SQL is a curated repository containing tutorials and resources for Large Language Models, Text2SQL, Text2DSL, Text2API, Text2Vis, and more. It provides guidelines on converting natural language questions into structured SQL queries, with a focus on NL2SQL. The repository includes information on various models, datasets, evaluation metrics, fine-tuning methods, libraries, and practice projects related to Text2SQL. It serves as a comprehensive resource for individuals interested in working with Text2SQL and related technologies.

The 'create-million-parameter-llm-from-scratch' repository provides a detailed guide on creating a Large Language Model (LLM) with 2.3 million parameters from scratch. The blog replicates the LLaMA approach, incorporating concepts like RMSNorm for pre-normalization, SwiGLU activation function, and Rotary Embeddings. The model is trained on a basic dataset to demonstrate the ease of creating a million-parameter LLM without the need for a high-end GPU.

StableToolBench is a new benchmark developed to address the instability of Tool Learning benchmarks. It aims to balance stability and reality by introducing features such as a Virtual API System with caching and API simulators, a new set of solvable queries determined by LLMs, and a Stable Evaluation System using GPT-4. The Virtual API Server can be set up either by building from source or using a prebuilt Docker image. Users can test the server using provided scripts and evaluate models with Solvable Pass Rate and Solvable Win Rate metrics. The tool also includes model experiments results comparing different models' performance.

The Beta Machine Learning Toolkit is a package containing various algorithms and utilities for implementing machine learning workflows in multiple languages, including Julia, Python, and R. It offers a range of supervised and unsupervised models, data transformers, and assessment tools. The models are implemented entirely in Julia and are not wrappers for third-party models. Users can easily contribute new models or request implementations. The focus is on user-friendliness rather than computational efficiency, making it suitable for educational and research purposes.

AI-TOD is a dataset for tiny object detection in aerial images, containing 700,621 object instances across 28,036 images. Objects in AI-TOD are smaller with a mean size of 12.8 pixels compared to other aerial image datasets. To use AI-TOD, download xView training set and AI-TOD_wo_xview, then generate the complete dataset using the provided synthesis tool. The dataset is publicly available for academic and research purposes under CC BY-NC-SA 4.0 license.

Weave is a toolkit for developing Generative AI applications, built by Weights & Biases. With Weave, you can log and debug language model inputs, outputs, and traces; build rigorous, apples-to-apples evaluations for language model use cases; and organize all the information generated across the LLM workflow, from experimentation to evaluations to production. Weave aims to bring rigor, best-practices, and composability to the inherently experimental process of developing Generative AI software, without introducing cognitive overhead.

LLMStack is a no-code platform for building generative AI agents, workflows, and chatbots. It allows users to connect their own data, internal tools, and GPT-powered models without any coding experience. LLMStack can be deployed to the cloud or on-premise and can be accessed via HTTP API or triggered from Slack or Discord.

The VisionCraft API is a free API for using over 100 different AI models. From images to sound.

Kaito is an operator that automates the AI/ML inference model deployment in a Kubernetes cluster. It manages large model files using container images, avoids tuning deployment parameters to fit GPU hardware by providing preset configurations, auto-provisions GPU nodes based on model requirements, and hosts large model images in the public Microsoft Container Registry (MCR) if the license allows. Using Kaito, the workflow of onboarding large AI inference models in Kubernetes is largely simplified.

PyRIT is an open access automation framework designed to empower security professionals and ML engineers to red team foundation models and their applications. It automates AI Red Teaming tasks to allow operators to focus on more complicated and time-consuming tasks and can also identify security harms such as misuse (e.g., malware generation, jailbreaking), and privacy harms (e.g., identity theft). The goal is to allow researchers to have a baseline of how well their model and entire inference pipeline is doing against different harm categories and to be able to compare that baseline to future iterations of their model. This allows them to have empirical data on how well their model is doing today, and detect any degradation of performance based on future improvements.

Tabby is a self-hosted AI coding assistant, offering an open-source and on-premises alternative to GitHub Copilot. It boasts several key features: * Self-contained, with no need for a DBMS or cloud service. * OpenAPI interface, easy to integrate with existing infrastructure (e.g Cloud IDE). * Supports consumer-grade GPUs.

SPEAR (Simulator for Photorealistic Embodied AI Research) is a powerful tool for training embodied agents. It features 300 unique virtual indoor environments with 2,566 unique rooms and 17,234 unique objects that can be manipulated individually. Each environment is designed by a professional artist and features detailed geometry, photorealistic materials, and a unique floor plan and object layout. SPEAR is implemented as Unreal Engine assets and provides an OpenAI Gym interface for interacting with the environments via Python.

Magick is a groundbreaking visual AIDE (Artificial Intelligence Development Environment) for no-code data pipelines and multimodal agents. Magick can connect to other services and comes with nodes and templates well-suited for intelligent agents, chatbots, complex reasoning systems and realistic characters.