Wingmark’s first-generation chat models, achieving performance comparable to Google Gemini or GPT 4o-Mini across CRP(Character Roleplay), special-tasks.

小显存友好设计，部分版本模型已经预先量化。

Design for personal app, some version of model has been pre-quantified.

Models

OpenCyberFurry-lite huge

ollama run wingmark/opencyberfurry-lite:72b

OpenCyberFurry-lite large

ollama run wingmark/opencyberfurry-lite:32b

OpenCyberFurry-lite standard

ollama run wingmark/opencyberfurry-lite:14b

OpenCyberFurry-lite minimal

ollama run wingmark/opencyberfurry-lite:7b

OpenCyberFurry-lite micro

ollama run wingmark/opencyberfurry-lite:1.5b

OpenCyberFurry-lite nano

ollama run wingmark/opencyberfurry-lite:0.5b

Features

• Native Chinese Support: Pre-trained for Chinese scene.
• Furry knowledge-base: Fine-tuned with furry base knowledge.
• System prompt: role playing, language style transfer, task setting, and behavior-setting by using a system prompt.

License

重要提醒：

如果您基于OpenCyberFurry进行二次开发，请在显著位置使用CyberFurry商标。

在使用时，请遵循如下开源许可证：

This model is a fine-tuned version of Qwen (千问), originally developed by Alibaba Cloud under the Apache 2.0 License. Therefore, this model is also released under the Apache 2.0 License, allowing free use, modification, and distribution under the terms of Apache 2.0.

To comply with the Apache 2.0 license, please retain the original license and copyright notice in any derivative works based on this model.

Original Qwen Repository: https://github.com/QwenLM

For the full Apache 2.0 License, please refer to the LICENSE file.