qwen2-5

Local CLI Copilot, powered by Ollama. 💻🦙

An open-source implementaion for fine-tuning Qwen2-VL and Qwen2.5-VL series by Alibaba Cloud.

OmniThink: Expanding Knowledge Boundaries in Machine Writing through Thinking

Experimental tools to backdoor large language models by re-writing their system prompts at a raw parameter level. This allows you to potentially execute offline remote code execution without running any actual code on the victim's machine or thwart LLM-based fraud/moderation systems.

GPU-accelerated Llama3.java inference in pure Java using TornadoVM.

A light llama-like llm inference framework based on the triton kernel.

Hand-derived memory-efficient super lazy PyTorch VJPs for training LLMs on laptop, all using one op (bundled scaled matmuls).

Deploy open-source LLMs on AWS in minutes — with OpenAI-compatible APIs and a powerful CLI/SDK toolkit.

1st Place Solution for Eedi - Mining Misconceptions in Mathematics Kaggle Competition

Java 23, SpringBoot 3.4.1 Examples using Deep Learning 4 Java & LangChain4J for Generative AI using ChatGPT LLM, RAG and other open source LLMs. Sentiment Analysis, Application Context based ChatBots. Custom Data Handling. LLMs - GPT 3.5 / 4o, Gemini Pro 1.5, Claude 3, Llama 3.1, Phi-3, Gemma 2, Falcon 3, Qwen 2.5, Mistral Nemo, Wizard Math

Exploring Agno framework for building AI agents.

Community-built Qwen AI Provider for Vercel AI SDK - Integrate Alibaba Cloud's Qwen models with Vercel's AI application framework

Introducing Project Zephyrine: Elevating Your Interaction Plug and Play, and Employing GPU Acceleration within a Modernized Automata Local Graphical User Interface.

Simple RAG system.

Silver Medal Solution for the Kaggle Competition: Eedi - Mining Misconceptions in Mathematics

Get Clothes from image

Models: Deepseek R1 models, Llama3.2, Qwen2.5. Integrations: Ollama, Gradio. Supports Local LLM. Test and deploy the latest LLM models in the fastest and most efficient way

grpo to train long form QA and instructions with long-form reward model

FastLongSpeech is a novel framework designed to extend the capabilities of Large Speech-Language Models for efficient long-speech processing without necessitating dedicated long-speech training data.

A framework for using local LLMs (Qwen2.5-coder 7B) that are fine-tuned using RL to generate, debug, and optimize code solutions through iterative refinement.