Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Beacon: Post-Training Quantization with Integrated Grid Selection

About

Quantization is a widely used compression technique for reducing the memory and computation costs of large pre-trained models. A key challenge in per-channel post-training quantization (PTQ) is selecting appropriate scaling factors to replace weight values with values from a scaled integer grid. Existing methods typically fix the scale at the outset via heuristic tuning or grid search. We propose Beacon, a simple and effective algorithm that eliminates the need for such manual tuning. Beacon performs per-channel PTQ directly using an unscaled grid and automatically determines the optimal scaling factors by exploiting the geometry of scalar quantization. It does not rely on back-propagation or large calibration sets. Despite its simplicity and tuning-free nature, Beacon achieves competitive performance compared to state-of-the-art methods, making it a practical solution for efficient model deployment.

Shihao Zhang, Rayan Saab• 2025

Related benchmarks

TaskDatasetResultRank
Language ModelingWikiText-2
Perplexity (PPL)7.53
2862
Image ClassificationImageNet (val)--
1206
Language ModelingWikiText-2 (val)
Perplexity (BVS)21.9
179
Commonsense ReasoningARC-Easy, ARC-Challenge, and HellaSwag Average
Zero-Shot Accuracy36.8
60
Zero-shot EvaluationZero-shot Benchmarks Average
Average Accuracy59.7
60
Image ClassificationImageNet (val)
Accuracy Drop0.34
12
Showing 6 of 6 rows

Other info

Follow for update