Orenis 3B (Demo)
Orenis is an AI assistant developed by OrenCraft Labs, founded by Lee Zinu. This is the 3B demo tier — a LoRA fine-tune on top of Qwen2.5-3B-Instruct. A larger 7-9B flagship model is in development.
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-3B-Instruct")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-3B-Instruct")
model = PeftModel.from_pretrained(base, "your-username/orenis-3b-demo")
Known limitations
- Occasional inaccuracies on edge-case Python semantics (e.g. for-else loops)
- May sometimes over-explain rather than decline sensitive meta-questions
- 3B scale — general knowledge and reasoning are more limited than larger models
Training
Fine-tuned with LoRA (r=16, alpha=32) on a mixed dataset of general instruction data (alpaca-cleaned) plus custom identity, reasoning, and safety-calibration examples.
- Downloads last month
- 23
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support