The absolute trainer to light up AI agents.

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

A framework for efficient model inference with omni-modality models

Official inference framework for 1-bit LLMs

MoVa - Generalizable Classification of Human Morals and Values - EMNLP 2025

Cleaning up the code

OmniGen2: Exploration to Advanced Multimodal Generation. https://arxiv.org/abs/2506.18871

The source codes of GRU model for Chinese poetry generation (CCL 2017).

BERT-CCPoem is an BERT-based pre-trained model particularly for Chinese classical poetry

Awesome papers involving LLMs in Social Science.

Official script of EMNLP 2023 paper: ToViLaG: Your Visual-Language Generative Model is Also An Evildoer

Code for the ICLR 2023 paper: Unified Detoxifying and Debiasing in Language Generation via Inference-time Adaptive Optimization

This repository contains the code for "Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLP".

[ACL 2020] Towards Debiasing Sentence Representations

List of Dirty, Naughty, Obscene, and Otherwise Bad Words

Plug and Play Language Model implementation. Allows to steer topic and attributes of GPT-2 models.

[ICML 2021] Towards Understanding and Mitigating Social Biases in Language Models

Source codes of Neural Quality Estimation with Multiple Hypotheses for Grammatical Error Correction

Scripts to evaluate various bias metrics for different NLG models + decoding algorithms

Dataset + classifier tools to study social perception biases in natural language generation

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

Tensors and Dynamic neural networks in Python with strong GPU acceleration

Code for Concrete Dropout as presented in https://arxiv.org/abs/1705.07832

Google Research

BERT score for text generation