Xudong Wang, Quanlu Zhang, Li, Yujing Wang, Mao Yang, Yuqing Yang, Jiahang Xu, Ting Cao, Ningxin Zheng, Lyna Zhang
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
The combination of Neural Architecture Search (NAS) and quantization has proven successful in automatically designing low-FLOPs INT8 quantized neural networks (QNN). However, directly applying NAS to design accurate QNN models that achieve low latency on real-world devices leads to inferior performance. In this work, we identify that the poor INT8 latency is due to the quantization-unfriendly issue: the operator and configuration (e.g., channel width) choices in prior art search spaces lead to diverse quantization efficiency and can slow down the INT8 inference speed. To address this challenge, we propose SpaceEvo, an automatic method for designing a dedicated, quantizationfriendly search space for each target hardware. The key idea of SpaceEvo is to automatically search hardwarepreferred operators and configurations to construct the search space, guided by a metric called Q-T score to quantify how quantization-friendly a candidate search space is. We further train a quantized-for-all supernet over our discovered search space, enabling the searched models to be directly deployed without extra retraining or quantization. Our discovered models, SEQnet, establish new SOTA INT8 quantized accuracy under various latency constraints, achieving up to 10.1% accuracy improvement on ImageNet than prior art CNNs under the same latency. Extensive experiments on real devices show that SpaceEvo consistently outperforms manually-designed search spaces with up to 2.5× faster speed while achieving the same accuracy.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2303.08308")
get_code_for_paper("2303.08308")
have("2303.08308")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2303.08308.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2303.08308)
Connect an agent — have() is free.