ayeganov

DeepSeek-VL

DeepSeek-VL: Towards Real-World Vision-Language Understanding

pyo3

Rust bindings for the Python interpreter

pepin

A probabilistic approximate DNF counter

Bend

A massively parallel, high-level programming language

LWM

NeMo-Guardrails

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.

jepsen

A framework for distributed systems verification, with fault injection

Fooocus

Focus on prompting and generating

swan

This project aims to enable language model inference on FPGAs, supporting AI applications in edge devices and environments with limited resources.

Consistency_LLM

[ICML 2024] CLLMs: Consistency Large Language Models

guardrails

Adding guardrails to large language models.

drl-zh

Deep Reinforcement Learning: Zero to Hero!

shuttle

Build & ship backends without writing any infrastructure files.

VoiceCraft

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

version2

Vector class library, latest version

pykan

Kolmogorov Arnold Networks

HarmBench

HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal

deepeval

The LLM Evaluation Framework

SALAD-BENCH

【ACL 2024】 SALAD benchmark & MD-Judge

OpenVoice

Instant voice cloning by MyShell.

unsloth

Finetune Llama 3.1, Mistral, Phi & Gemma LLMs 2-5x faster with 80% less memory

penzai

A JAX research toolkit for building, editing, and visualizing neural networks.

lunatic-rs

This library contains higher level Rust wrappers for low level Lunatic syscalls.

PurpleLlama

Set of tools to assess and improve LLM security.

moondream

tiny vision language model

Ethos-Hate-Speech-Dataset

This repository contains a dataset for hate speech detection on social media platforms.

TyDiP

TyDiP Multilingual Politeness dataset and code

capsule_net_pytorch

Readable implementation of a Capsule Network as described in "Dynamic Routing Between Capsules" [Hinton et. al.]

free-exercise-db

Open Public Domain Exercise Dataset in JSON format, over 800 exercises with a browsable public searchable frontend

NDP