Showing 101-120 of 154 projects
Open academic research on improving the LLaMA language model to state-of-the-art performance.
A research paper on representation alignment for training diffusion models for AI-assisted code generation.
High-quality implementations of standard and state-of-the-art methods for Bayesian and probabilistic machine learning.
A high-performance text-to-3D generation model for building immersive 3D experiences with AI tools.
Measuring Massive Multitask Language Understanding with GPT-3 and few-shot learning techniques.
Official repository for the 'Big Transfer (BiT): General Visual Representation Learning' paper, focused on transfer learning for computer vision.
A repository for training large language models (LLMs) using QLoRA and FSDP techniques.
This repository provides fine-tuning code for the ChatGLM 6B language model and the Alpaca model, enabling developers to customize these AI models for their own use cases.
A Vision-and-Language Transformer model for multimodal tasks without the need for convolution or region supervision.
CLUENER2020 is a Chinese fine-grained named entity recognition dataset and benchmark for AI-powered NLP development.
Adapts Meta AI's Segment Anything model to downstream tasks using adapters and prompts.
AgentTuning: A Python library for enabling generalized agent abilities in large language models (LLMs).
An implementation for detailed localized image and video captioning using large multimodal models.
A Python library for exploring secrets of RLHF (Reward-Weighted Maximum Likelihood Estimation) in large language models
A simple image captioning model built using the CLIP neural network for generating captions for images.
A Python package for chatting with an AI model and executing the InstructLab workflow to train a model using custom taxonomy data.
WebThinker is a powerful framework for building large language models with deep research capabilities.
An open-source framework for building knowledgeable large language models with fine-tuning capabilities.
A PyTorch implementation of a BERT-style pretraining method for convolutional networks, enabling more efficient self-supervised learning.
A library that helps developers train BERT-type language models with limited compute resources.
Get weekly updates on trending AI coding tools and projects.