JetBrains/Mellum2-12B-A2.5B-Thinking — 12B / 2.5B active · MOE · 128K ctx
JetBrains’ reasoning-augmented code MoE (12B total / 2.5B active) that emits explicit <think> chains for debugging, planning, and agentic coding
AI Model Documentation
Summary
This Hugging Face page documents the Qwopus3.6-27B-Coder-MTP-GGUF model, a 27B parameter multimodal coding model. It provides detailed instructions for local and server deployment using Transformers, vLLM, and SGLang, including API examples.
Highlights
- Multimodal coding model supporting text and image inputs.
- Includes code snippets for Hugging Face Transformers pipelines.
- Guides for serving via vLLM and SGLang with OpenAI-compatible APIs.
- Deployment options via Docker, Colab, and Kaggle.
- Available in GGUF format for efficient local inference.
auto-generated
Context
Audience
Machine Learning Engineers and AI Developers
DomainMachine Learning
Formattechnical documentation
Accessopen source
Source Overview
Jackrong/Qwopus3.6-27B-Coder-MTP-GGUF is a 27B-parameter coding-focused multimodal model hosted on Hugging Face, provided in GGUF format for local and server deployments.
The page documents how to run the model via Transformers pipelines, vLLM, SGLang, Unsloth Studio, Docker, Google Colab, Kaggle, and Hugging Face local apps. It includes concrete command-line and Python examples for serving OpenAI-compatible chat/image APIs and for loading the model programmatically.
Topics
Related
Hugging FacevLLMSGLangTransformersGGUFLarge Language Models
