AlpinDale

ssh-dashboard. Monitor GPU/CPU Usage on SSH servers, with NVIDIA and AMD support.

gptslop. A community list of common phrases generated by GPT and Claude models

sparsegpt-for-LLaMA. Code for the paper "SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot" with LLaMA implementation.

pygmalion.cpp. C/C++ implementation of PygmalionAI/pygmalion-6b

QuIP-for-Llama. Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees" adapted for Llama models

LLM-Shearing. Preprint: Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning

RPTQ-for-LLaMA. Efficient 3bit/4bit quantization of LLaMA models

gptq-gptj. Code for the paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers" with GPT-J implementation.

kizuna. Fast TTS Library for Kokoro

mergekit-LGPL. Tools for merging pretrained large language models.

MixtureOfExperts. Python

turbokit. High-performance C++ utilities

waifu. REST API server for serving images of your waifu

VNParser. Project aimed at make a universal VN parser

tplib. A torch-based, universal tensor-parallel library.

pygbooru. Booru style imageboard for Pygmalion Character Cards

flash-attention. Fast and memory-efficient exact attention

GPT-J_INT4_LoRA. Code for fine-tuning GPT-J LoRAs in 4bit

dry_sampling_kernel. CUDA kernels for Don't Repeat Yourself (DRY) Sampling

AutoQuIP. Easy to use LLM quantization via the QuIP# technique

AlpinDale. Config files for my GitHub profile.

lora-merge. Script for merging LoRA adapters with base models.

misc-scripts. Miscellaneous scripts for various stuff.

toes. Various experiments with building an AI-based Tic Tac Toe.

litellm. Call all LLM APIs using the OpenAI format. Use Bedrock, Azure, OpenAI, Cohere, Anthropic, Ollama, Sagemaker, HuggingFace, Replicate (100+ LLMs)

mersenne. C

vastai-kobold. Script to automate KoboldAI on a Vast.ai VM instance

huggingface_hub. The official Python client for the Huggingface Hub.

diplomacy. A modern web app for the Diplomacy Board Game

RapidLLaMA. Runs LLaMA with Extremely HIGH speed

prompt-tuner. Soft Prompt training library for Causal Transformer Language Models

llama.cpp. Port of Facebook's LLaMA model in C/C++

causal_llm_tuning. Jupyter notebook for when I want to quickly tune a causal LLM

alpinator. Ongoing research training transformer models at scale

causal-lm-lora. Training code for training LoRAs

TalOS-Reborn. LLM Powered discord bot, Character Card enabled Chat page, Stable Diffusion discord bot, and overall AI tool. All from one app, TalOS: Reborn.

exllamav2. A fast inference library for running LLMs locally on modern consumer-class GPUs

rentry_actions. Automatically create rentries from markdown files in your GitHub repository.