KevinPanJun
Follow
Stars
Renderer for the harmony response format to be used with gpt-oss
gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI
Everything about the SmolLM and SmolVLM family of models
Vision Document Retrieval (ViDoRe): Benchmark. Evaluation code for the ColPali paper.
Lightweight coding agent that runs in your terminal
ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learning & ReCall: Learning to Reason with Tool Call for LLMs via Reinforcement Learning
Official Repo for Open-Reasoner-Zero
Fully open reproduction of DeepSeek-R1
The code used to train and run inference with the ColVision models, e.g. ColPali, ColQwen2, and ColSmol.
Efficient Triton Kernels for LLM Training
LAVIS - A One-stop Library for Language-Vision Intelligence
A high-throughput and memory-efficient inference and serving engine for LLMs
The repository provides code for running inference with the Meta Segment Anything Model 2 (SAM 2), links for downloading the trained model checkpoints, and example notebooks that show how to use th…
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
PyTorch implementation of InstructDiffusion, a unifying and generic framework for aligning computer vision tasks with human instructions.
PyTorch implementation of MAE https//arxiv.org/abs/2111.06377
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.
An open-source framework for training large multimodal models.
Port of OpenAI's Whisper model in C/C++
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Tools to download and cleanup Common Crawl data
Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.
Fault-tolerant, highly scalable GPU orchestration, and a machine learning framework designed for training models with billions to trillions of parameters