LightEval is a lightweight LLM evaluation suite that Hugging Face has been using internally with the recently released LLM data processing library datatrove and LLM training library nanotron.

QJL: 1-Bit Quantized JL transform for KV Cache Quantization with Zero Overhead

Interpretability for sequence generation models 🐛 🔍

This repository provides the dataset used in "ExPUNations: Augmenting puns with keywords and explanations" by Jiao Sun, Anjali Narayan-Chen, Shereen Oraby, Alessandra Cervone, Tagyoung Chung, Jing …

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

Aligning pretrained language models with instruction data generated by themselves.

DecompX: Explaining Transformers Decisions by Propagating Token Decomposition

OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.

Measuring the Mixing of Contextual Information in the Transformer

This repo contains Data Science course material, taught at Amirkabir U of Tech, winter 2020 and 2021.

Collections of CS PhD Application Fee Waivers of schools in North America

[NAACL 2022] GlobEnc: Quantifying Global Token Attribution by Incorporating the Whole Encoder Layer in Transformers

Tools and Modeling Code for the MASSIVE dataset

Official style files for papers submitted to venues of the Association for Computational Linguistics

Probing and Generalization of Metaphorical Knowledge in Pre-Trained Language Modelss[ACL 2022]

Explain, analyze, and visualize NLP language models. Ecco creates interactive visualizations directly in Jupyter notebooks explaining the behavior of Transformer-based language models (like GPT2, B…

📺 Discover the latest machine learning / AI courses on YouTube.

Flax is a neural network library for JAX that is designed for flexibility.

Minimal keyword extraction with BERT

🎓 无需编写任何代码即可轻松创建漂亮的学术网站 Easily create a beautiful academic résumé or educational website using Hugo and GitHub. No code.

A beautiful, simple, clean, and responsive Jekyll theme for academics

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

This is a repository with the code for the EMNLP 2020 paper "Information-Theoretic Probing with Minimum Description Length"

LaTeX template for BSc/MSc/PhD theses of University of Tehran - قالب لاتک پایان‌نامه دانشگاه تهران