Orri
Orri is a fine-tuned language model based on Qwen2.5-3B, designed to provide a more focused, natural, and instruction-following conversational experience while retaining the capabilities of the original Qwen2.5 architecture.
The goal of Orri is to create a compact model that is useful for everyday conversations, creative tasks, reasoning, writing, and general-purpose assistance without requiring a large amount of hardware.
Model Details
- Model Name: Orri
- Base Model: Qwen2.5-3B
- Model Type: Causal Language Model
- Parameters: ~3B
- Architecture: Qwen2.5
- Language: English
- Primary Use: Text generation and conversational AI
- Fine-tuning: Supervised fine-tuning
- Quantized Versions: Depending on the release
What is Orri?
Orri is intended to be a lightweight alternative to larger language models.
The model was fine-tuned with an emphasis on producing responses that feel natural and useful while maintaining the relatively small footprint of a 3B parameter model.
It can be used for:
- 💬 General conversation
- ✍️ Creative writing
- 📝 Rewriting and editing
- 💡 Brainstorming
- 👨💻 Programming assistance
- 📚 Summarization
- ❓ Question answering
- 🧠 General reasoning
- 🎭 Roleplay and character-based conversations
Usage
Orri can be loaded using the Hugging Face Transformers library.
For more deterministic responses:
Temperature: 0.2–0.4 Top P: 0.8–0.9
For more creative responses:
Temperature: 0.8–1.0 Top P: 0.9–0.95
- Downloads last month
- 45