Latent On-Policy Self-Distillation
Paper • 2608.13040 • Published
Qwen3-8B-LOPD
This model is fine-tuned from Qwen3-8B using Latent On-Policy Self-Distillation (LOPD) on the agentic tool-use domain.
For more details, see our GitHub repository and paper.
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("liunanfu1992/Qwen3-8B-LOPD", torch_dtype="bfloat16", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("liunanfu1992/Qwen3-8B-LOPD")
- Downloads last month
- 35
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support