Bespoke Automata is a GUI and deployment pipline for making complex AI agents locally and offline

Finetune Llama 3, Mistral, Phi & Gemma LLMs 2-5x faster with 80% less memory

A minimalistic R wrapper for the Reddit API

The game of Balderdash: weird words only

Soul Engine SDK

A User Interface for XTTS-2 Text-Based Voice Cloning using only 10 seconds of speech

Emote Portrait Alive: Generating Expressive Portrait Videos with Audio2Video Diffusion Model under Weak Conditions

A fast and lightweight framework for creating decentralized agents with ease.

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

An Open Source text-to-speech system built by inverting Whisper.

TinyGPT-V: Efficient Multimodal Large Language Model via Small Backbones

Unified framework for building enterprise RAG pipelines with small, specialized models

Flexible, scalable and customizable agents to do your bidding.

Go ahead and axolotl questions

A collection of standardized JSON descriptors for Large Language Model (LLM) files.

TruthfulQA: Measuring How Models Imitate Human Falsehoods

Code release for Dataless Knowledge Fusion by Merging Weights of Language Models (https://openreview.net/forum?id=FCnohuR6AnM)

🎨 Free custom elements for the WordPress Theme Bricks Builder.

Open-source framework to make AI agents' team collaboration as effective as human collaboration.

some common Huggingface transformers in maximal update parametrization (µP)

maximal update parametrization (µP)

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

Prefix-Tuning: Optimizing Continuous Prompts for Generation

Code for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"

Drag & drop UI to build your customized LLM flow

Longformer: The Long-Document Transformer

Notion-style WYSIWYG editor with AI-powered autocompletion.

A tool used to create 3D content for rooms by calling OpenAI's API

A project demonstrating Lidar related AI solutions, including three GPU accelerated Lidar/camera DL networks (PointPillars, CenterPoint, BEVFusion) and the related libs (cuPCL, 3D SparseConvolution…