Modelpublic

connorolson/qwen35-9b-lean4-geometry-lora

sourceHugging Faceupdated 4mo agoView on Hugging Face

0likes3downloads

Model Card

connorolson/qwen35-9b-lean4-geometry-lora

LoRA adapter for Lean 4 subgoal completion, tuned for geometry.

python

from transformers import AutoModelForCausalLM
from peft import PeftModel

base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-9B", trust_remote_code=True)
model = PeftModel.from_pretrained(base, "connorolson/qwen35-9b-lean4-geometry-lora")

Adapter only. Load on top of the base model.