sirus20x6

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

autogen

A programming framework for agentic AI. Discord: https://aka.ms/autogen-dc. Roadmap: https://aka.ms/autogen-roadmap

unilm

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

dspy

DSPy: The framework for programming—not prompting—foundation models

haystack

:mag: LLM orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your data. With advanced retrieval methods, it's best suited for building RAG, question answering, semantic search or conversational agent chatbots.

instructor

structured outputs for llms

introtodeeplearning

Lab Materials for MIT 6.S191: Introduction to Deep Learning

skywater-pdk

Open source process design kit for usage with SkyWater Technology Foundry's 130nm node.

Devon

Devon: An open-source pair programmer

mistral-finetune

RAGatouille

Easily use and train state of the art late-interaction retrieval methods (ColBERT) in any RAG pipeline. Designed for modularity and ease-of-use, backed by research.

langroid

Harness LLMs with Multi-Agent Programming

jailbreak_llms

[CCS'24] A dataset consists of 15,140 ChatGPT prompts from Reddit, Discord, websites, and open-source datasets (including 1,405 jailbreak prompts).

HippoRAG

HippoRAG is a novel RAG framework inspired by human long-term memory that enables LLMs to continuously integrate knowledge across external documents. RAG + Knowledge Graphs + Personalized PageRank.

burr

Build applications that make decisions (chatbots, agents, simulations, etc...). Monitor, trace, persist, and execute on your own infrastructure.

llama3.np

llama3.np is a pure NumPy implementation for Llama 3 model.

YaFSDP

YaFSDP: Yet another Fully Sharded Data Parallel

augmentoolkit

Convert Compute And Books Into Instruct-Tuning Datasets (or classifiers)!

text-clustering

Easily embed, cluster and semantically label text datasets

7900xtx

omnichain

Efficient visual programming for AI language models

llama-zip

LLM-powered lossless compression tool

autogenstudio-skills

Repo of skills for autogenstudio

Yuan2.0-M32

Mixture-of-Experts (MoE) Language Model

CodeR

Mitten

Infinite canvas drawing application.

Note

Machine learning library, Distributed training, Deep learning, Reinforcement learning, Models, TensorFlow, PyTorch

autogen_skills

python skills for autogen

caravel_fulgor_opamp

Test Chip General Purpose OpAmp using Skywater SKY130 PDK

blaz

Blaz: a library for frugal matrix computations. Blaz provides compression/uncompression functions for matrices of floating-point numbers and makes it possible to perform basic linear algebra on the compressed matrices, without uncompressing them.