The hub for EleutherAI's work on interpretability and learning dynamics

Pandas DataFrames as Interactive DataTables

ReFT: Representation Finetuning for Language Models

MetaQA: Combining Expert Agents for Multi-Skill Question Answering

[CCS'24] A dataset consists of 15,140 ChatGPT prompts from Reddit, Discord, websites, and open-source datasets (including 1,405 jailbreak prompts).

A high-throughput and memory-efficient inference and serving engine for LLMs

A package to evaluate factuality of long-form generation. Original implementation of our EMNLP 2023 paper "FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation"

Fuzzy String Matching in Python

The simplest way to run LLaMA on your local machine

Implementation of the LLaMA language model based on nanoGPT. Supports flash attention, Int8 and GPTQ 4bit quantization, LoRA and LLaMA-Adapter fine-tuning, pre-training. Apache 2.0-licensed.

Instruct-tune LLaMA on consumer hardware

Code to extract KB triples from given contexts using Wikidata API

SQuARE: Software for question answering research.

Solve puzzles. Improve your pytorch.

Persistent dict, backed by sqlite3 and pickle, multithread-safe.

Acceptance rates for the major AI conferences

Must-read Papers on pre-trained language models.

ACL2020 Tutorial: Open-Domain Question Answering

Parsers for scientific papers (PDF2JSON, TEX2JSON, JATS2JSON)

Rich Context knowledge graph management

LUKE -- Language Understanding with Knowledge-based Embeddings

A BERT model for scientific text.

A tool for extracting plain text from Wikipedia dumps

🖥 Delightful Hyper plugins, themes, and resources

A terminal built on web technologies

Personal Webpage