SparkJiao

dify

Dify is an open-source LLM app development platform. Dify's intuitive interface combines AI workflow, RAG pipeline, agent capabilities, model management, observability features and more, letting you quickly go from prototype to production.

llama-recipes

Scripts for fine-tuning Meta Llama3 with composable FSDP & PEFT methods to cover single/multi-node GPUs. Supports default & custom datasets for applications such as summarization and Q&A. Supporting a number of candid inference solutions such as HF TGI, VLLM for local or cloud deployment. Demo apps to showcase Meta Llama3 for WhatsApp & Messenger.

FlexGen

Running large language models on a single GPU for throughput-oriented scenarios.

llm-foundry

LLM training code for Databricks foundation models

Megatron-DeepSpeed

Ongoing research training transformer language models at scale, including: BERT & GPT-2

Emu

Emu Series: Generative Multimodal Models from BAAI

Laws

MOSS-RLHF

MOSS-RLHF

MeZO

[NeurIPS 2023] MeZO: Fine-Tuning Language Models with Just Forward Passes. https://arxiv.org/abs/2305.17333

Megatron-LLaMA

Best practice for training LLaMA models in Megatron-LM

LongBench

LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding

VQ-VAE

Minimalist implementation of VQ-VAE in Pytorch

BLoRA

batched loras

MathGLM

Official Pytorch Implementation for MathGLM

tree-of-thought-puzzle-solver

The Tree of Thoughts (ToT) framework for solving complex reasoning tasks using LLMs

SearchAnything

A semantic local search engine powered by AI models.

Program-of-Thoughts

Data and Code for Program of Thoughts (TMLR 2023)

world-models

RAP

Reasoning with Language Model is Planning with World Model

Awesome-LegalAI-Resources

This repository aims to collect all LegalAI data to facilitate the development of intelligent justice systems

llama-ssp

Experiments on speculative sampling with Llama models

Glot500

Glot500: Scaling Multilingual Corpora and Language Models to 500 Languages -- ACL 2023

reStructured-Pretraining

reStructured Pre-training

gdGPT

Train llm (bloom, llama, baichuan2-7b, chatglm3-6b) with deepspeed pipeline mode. Faster than zero/zero++/fsdp.

M3Exam

Data and code for paper "M3Exam: A Multilingual, Multimodal, Multilevel Benchmark for Examining Large Language Models"

Quark

Hrrformer

Hrrformer: A Neuro-symbolic Self-attention Model (ICML23)

CoTConsistency

The released data for paper "Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models".

IDOL

Repo for paper "IDOL: Indicator-oriented Logic Pre-training for Logical Reasoning" accepted to the Findings of ACL 2023

IDOL

Repo for paper "IDOL: Indicator-oriented Logic Pre-training for Logical Reasoning" accepted to the Findings of ACL 2023