5 Commits

Author	SHA1	Message	Date
gdh1995	1eaa7a632b	fix Reduce may crash if it follows Reshape ... When converting a caffe model, `ReductionTransform` thinks there're always 4 dims in the input of a `Reduction` layer. However, when forwarding `OpCommonUtils::computeReduceDims` had an assumption that dim numbers wouldn't exceed `inputs[0]->dimensions()`. So, MNN would write unexpected addresses and then crash, if a Reduction op follows a Reshape / InnerProduct op, since both ops may output a tensor of 2 dimensions. _Comment_: I know `TransformInnerProduct` often converts caffe's InnerProduct into 1x1 Convolution, but Reshape can also cause a tensor has only 2 dimensions, so this still deserves a fix.	2020-12-19 02:42:07 +08:00
Hui Shu	d6795ad031	Github release 1.1.0	2020-11-05 16:49:17 +08:00
玄裳	0df31a8667	MNN 1.0.0 release sync. ... - Added Python Express API implemented with pbind11 - Added demos for Python Express API - Performance improvements for ARM64, ARMv8.2, x86. - README update.	2020-05-07 18:22:11 +08:00
xiaying	4dec63d499	[PATCH 103/160] [Vulkan:Feature, MNN:Refract] Support Reduce for vulkan	2020-03-22 19:02:10 +08:00
xiaying	956d528026	[PATCH 094/160] [MNN:Refractor] Sepearte compute dim from CPUBinary	2020-03-22 19:02:09 +08:00