Liu Jun's repositories
delve
PyTorch model training and layer saturation monitor
xtreme1
Xtreme1 - The Next GEN Platform for Multimodal Training Data. #3D annotation, 3D segmentation, lidar-camera fusion annotation, image annotation and rlhf tools are supported!
segment-anything
The repository provides code for running inference with the SegmentAnything Model (SAM), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.
pdfplumber
Plumb a PDF for detailed information about each char, rectangle, line, et cetera — and easily extract text and tables.
llama-recipes
Examples and recipes for Llama model
xformers
Hackable and optimized Transformers building blocks, supporting a composable construction.
flash-attention
Fast and memory-efficient exact attention
vector-quantize-pytorch
Vector Quantization, in Pytorch
muse-maskgit-pytorch
Implementation of Muse: Text-to-Image Generation via Masked Generative Transformers, in Pytorch
StableSR
Exploiting Diffusion Prior for Real-World Image Super-Resolution
img2dataset
Easily turn large sets of image urls to an image dataset. Can download, resize and package 100M urls in 20h on one machine.
mteb
MTEB: Massive Text Embedding Benchmark
stablediffusion
High-Resolution Image Synthesis with Latent Diffusion Models
Splice
Official Pytorch Implementation for "Splicing ViT Features for Semantic Appearance Transfer" presenting "Splice" (CVPR 2022 Oral)
spg
[ICML 2023] Parameter-Level Soft-Masking for Continual Learning
can-ai-code
Self-evaluating interview for AI coders
causal_reasoning_of_entities_and_events
Data and code for the paper Causal Reasoning of Entities and Events in Procedural Texts.
FastChat
An open platform for training, serving, and evaluating large languages. Release repo for Vicuna and FastChat-T5.
tree-of-thought-llm
Official Implementation of "Tree of Thoughts: Deliberate Problem Solving with Large Language Models"
train_custom_LLM
Train your custom LLMs like Llama, baichuan-7b, GPT
ToolQA
ToolQA, a new dataset to evaluate the capabilities of LLMs in answering challenging questions with external tools. It offers two levels (easy/hard) across eight real-life scenarios.
sgpt
SGPT: GPT Sentence Embeddings for Semantic Search
ALCE
Enabling Large Language Models to Generate Text with Citations. Paper: https://arxiv.org/abs/2305.14627
ReWOO
Decoupling Reasoning from Observations for Efficient Augmented Language Models
x-stable-diffusion
Real-time inference for Stable Diffusion - 0.88s latency. Covers AITemplate, nvFuser, TensorRT, FlashAttention.
manifest
Prompt programming with FMs.
so-vits-svc
SoftVC VITS Singing Voice Conversion