sirus20x6

data-is-better-together

Let's build better datasets, together!

Vodalus-Expert-LLM-Forge

Dataset Crafting w/ RAG/Wikipedia ground truth and Efficient Fine-Tuning Using MLX and Unsloth. Includes configurable dataset annotation editor Gradio UI.

optuna-dashboard

Real-time Web Dashboard for Optuna.

graph-studio-next

GraphStudioNext is a tool for developers to build and test DirectShow Graphs

nanoXLSTM

The simplest, fastest repository for training/finetuning medium-sized xLSTMs.

llm

My attempt at an LLM from complete scratch!

nox

Efficient fine-tuning for ko-llm models

DeepSeek-V2

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

marker

Convert PDF to markdown quickly with high accuracy

simdjson

Parsing gigabytes of JSON per second : used by Facebook/Meta Velox, the Node.js runtime, ClickHouse, WatermelonDB, Apache Doris, Milvus, StarRocks

dynet

DyNet: The Dynamic Neural Network Toolkit

vkvg

Vulkan 2D graphics library

awesome-electron-alternatives

A curated list of awesome Electron alternatives.

nnstreamer

:twisted_rightwards_arrows: Neural Network (NN) Streamer, Stream Processing Paradigm for Neural Network Apps/Devices.

phidata

Build AI Assistants with memory, knowledge and tools.

ragflow

RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding.

Cyberpunk-Neon

Cyberpunk Neon Themes for KDE Plasma, GTK, Telegram, Tilix, Vim, Zim and more.

dotfiles

llm-chain

`llm-chain` is a powerful rust crate for building chains in large language models allowing you to summarise text and complete complex tasks

library-of-phi

devika

Devika is an Agentic AI Software Engineer that can understand high-level human instructions, break them down into steps, research relevant information, and write code to achieve the given objective. Devika aims to be a competitive open-source alternative to Devin by Cognition AI.

SWE-agent

SWE-agent takes a GitHub issue and tries to automatically fix it, using GPT-4, or your LM of choice. It solves 12.47% of bugs in the SWE-bench evaluation set and takes just 1 minute to run.

How-To-Secure-A-Linux-Server

An evolving how-to guide for securing a Linux server.

mergekit

Tools for merging pretrained large language models.

fbpca

Fast Randomized PCA/SVD

whisperX

WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)

neurallambda

Reasoning Computers. Lambda Calculus, Fully Differentiable. Also Neural Stacks, Queues, Arrays, Lists, Trees, and Latches.

VoiceCraft

Zero-Shot Speech Editing and Text-to-Speech in the Wild

deepsparse

Sparsity-aware deep learning inference runtime for CPUs

Cadetwriter

Developed as a by-product of the IBM 1620 Jr. project, the Cadetwriter is a general-purpose ASCII terminal. It can be connected via RS-232 or USB to any mini, micro, mainframe, or replica computer as a terminal device. A commercial quality IBM/Lexmark Wheelwriter 1000 was adapted by interposing a circuit board containing a Teensy 3.5 microcontroller between the typewriter's keyboard and motherboard. Custom firmware controls the typewriter and communicates with the host computer. To support the full ASCII character set, characters not on the printwheel are synthesized using overprinted characters and "period graphics". The Cadetwriter can print up to 16cps and is a reliable, low-maintenance, low-cost substitute for Teletype, DECwriter, Diablo, Spinwriter, Imagewriter, etc. teleprinters.