kgourgou

kgourgou

setfit-integrated-gradients

Hacking SetFit so that it works with integrated gradients.

cookiecutter-python-project

mlx-examples

Examples in the MLX framework

mygithubpage

My website, served with jekyll.

nanoGPT

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

raggler

Simple, local-first, retrieval-augmented generation without tears (or API keys). Powered by MLX and AlphaMonarch.

alignment-handbook

Robust recipes to align language models with human and AI preferences

chess_llm_interpretability

Evaluating an LLM trained on chess PGN strings using techniques from the Othello World Models paper.

diff_history

[arXiv preprint 2024] Official code release accompanying the paper "diff History for Neural Language Agents" (Piterbarg, Pinto, Fergus)

diffuse-distributions

Forcing Diffuse Distributions out of Language Models

folio-jekyll

guidance

A guidance language for controlling large language models.

hades

Fast singularity detection with kernel

llama2.c

Inference Llama 2 in one file of pure C

LLM-LieDetector

Code for the paper "How to catch an AI liar: Lie detection in black-box LLMs by asking unrelated questions"

LookaheadDecoding

mamba.py

An efficient Mamba implementation in PyTorch and MLX.

maxtext

A simple, performant and scalable Jax LLM!

mistral-src

mlx

MLX: An array framework for Apple silicon

pdf-renamer-server

A python tool to automatically rename the pdf files of scientific publications by looking up the publication metadata on the web.

pigeonXT

🐦 Quickly annotate data from the comfort of your Jupyter notebook

posteriors

Uncertainty quantification with PyTorch

starting-kit

Starting kit for the NeurIPS 2023 unlearning challenge

streamlit-chess

Bi-directional component to play chess in streamlit apps

TART

TART: A plug-and-play Transformer module for task-agnostic reasoning

time_interpret

Unified Model Interpretability Library for Time Series

topical

yet-another-applied-llm-benchmark

A benchmark to evaluate language models on questions I've previously asked them to solve.