The authors train a diffusion-based solver on small TSP and CVRP instances using supervised learning on optimal solutions.
Reinforcement learning adapts diffusion models to large routing problems
ResizeCO fine-tunes on large instances without optimal solution labels and uses GPU beam search to improve inference.
Big Tech
Barak Itzchak Davidovitch · Eli Meirom · Ethan Fetaya · Gal Chechik
Bar-Ilan University · NVIDIA
Research Digest··2 min read
The authors present ResizeCO, a method that uses reinforcement learning to adapt a diffusion model trained on small routing instances to solve large ones.
Why this paper
From NVIDIA and Bar-Ilan University
In one line
ResizeCO adapts a diffusion model trained on small routing instances to large TSP and CVRP instances via reinforcement learning, setting a new state of the art for short run times.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§