Qwen2.5-Omni is an end-to-end multimodal model by Qwen team at Alibaba Cloud, capable of understanding text, audio, vision, video, and performing real-time speech generation.

MLX-VLM is a package for inference and fine-tuning of Vision Language Models (VLMs) on your Mac using MLX.

Artificial Neural Engine Machine Learning Library

An OpenAI API compatible FastAPI server that sits on top of the Anemll repo. Tested with Open WebUI.

Qualitative data analysis for text, images, audio, video. Cross platform. Python 3.10 or newer and PyQt6.

Unified Efficient Fine-Tuning of 100+ LLMs (ACL 2024)

The simplest, fastest repository for training/finetuning small-sized VLMs.

LLaVE: Large Language and Vision Embedding Models with Hardness-Weighted Contrastive Learning

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

Official inference framework for 1-bit LLMs

A high-throughput and memory-efficient inference and serving engine for LLMs

Finetune Llama 3.3, DeepSeek-R1, Gemma 3 & Reasoning LLMs 2x faster with 70% less memory! 🦥

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

FastMLX is a high performance production ready API to host MLX models.

Blazingly fast LLM inference.

Official inference repo for FLUX.1 models

Represent, send, store and search multimodal data

The simplest, fastest repository for training/finetuning medium-sized GPTs.

Train transformer language models with reinforcement learning.

Use PEFT or Full-parameter to finetune 400+ LLMs or 100+ MLLMs. (LLM: Qwen2.5, Llama3.2, GLM4, Internlm2.5, Yi1.5, Mistral, Baichuan2, DeepSeek, Gemma2, ...; MLLM: Qwen2-VL, Qwen2-Audio, Llama3.2-V…

SiLLM simplifies the process of training and running Large Language Models (LLMs) on Apple Silicon by leveraging the MLX framework.

A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training

Visual Instruction Tuning: Large Language-and-Vision Assistant built towards multimodal GPT-4 level capabilities.

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.

A Framework of Small-scale Large Multimodal Models

Port of Facebook's LLaMA model in C/C++

Python bindings for llama.cpp