AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs

媒体报道 Hugging Face Papers 原文语言:中文

论文摘要(作者报告,未经本站复测):Kernel optimization is essential for efficient large language model (LLM) training and inference on Ascend NPUs.

全文

AscendGYM provides a benchmark for assessing kernel optimization capabilities on Ascend NPUs using real-workload test cases, two evaluation modes, and a Speedup-Token Score that balances speedup against token cost.

正文以来源语言显示。

参考来源

← 返回每日AI报告