Zixuan Huang

Results 5 issues of Zixuan Huang

想请问论文中的意思是说,对于每种类型的依赖,比如intra+coref,inter+relate,文中的意思是说每种类型共享一个神经网络?这边好像分了6种类型,就是这一块总共有6个独立的神经网络?然后根据sij的值来选择,具体使用哪个网络计算?

文档中有提到说在很罕见的情况下会发现在数据前递时虚地址和实地址会匹配到不同的项,此时认为存在前递失败,因此需要进行重定向。 但是到底是怎样的一种罕见情况下会出现这种前递失败呢,文档里好像没有说明?当虚实地址的映射关系改变的时候难道不会清空store buffer/store Queue吗?能否具体说明一下? [TRANSLATION] The documentation mentions that in rare cases, the virtual address and the real address match different items during data forwarding, in which case a forwarding...

question

I am trying to move ara on VCS, but met too many errors, and they are hard to fix. Do you have the correct Compilation Options on VCS or irun?...

**Describe the bug** I am trying to use run onnx models on gemmini by spike. I have successfully run the inference on googlenet_quantized.onnx and mobilenet_quantized.onnx, but failed in [resnet50_opt_quant.onnx] and...