Supporting for Deconvolution Layer?

wyxsky commented 6 years ago

Try to compile YOLO v3 network, get error information in Deconvolution layer, this layer is used as Upsample layer in Caffe.

Compiler log:

creating new wisdom context...
opening wisdom context...
parsing caffe network...
libnvdla<1> error: parsing layer type Deconvolution index 261
(DLA) Error 0x00000004: Unable to parse caffemodel: "yolo-voc.prototxt" (in ParseTest.cpp, function parseCaffeNetwork(), line 91)
(DLA) Error 0x00000004: (propagating from ParseTest.cpp, function parseAndCompile(), line 164)
(DLA) Error 0x00000004: (propagating from main.cpp, function launchTest(), line 87)

prototxt content of the Deconvolution layer

layer {
    bottom: "layer85-conv"
    type: "Deconvolution"
    top: "Upsample_layer86"
    name: "Upsample_layer86"
    convolution_param {
        num_output: 256
        group: 256
        stride: 2
        kernel_size: 4
        pad: 1
        weight_filler {
            type: "bilinear"
        }
        bias_term: false
    }
    param {
        lr_mult: 0
        decay_mult: 0
    }
}

prasshantg commented 6 years ago

Compiler does not support de-convolution layer yet. Keeping it open to notify when this feature is available.

MrHaiCoder commented 4 years ago

功

Is this feature supported now? Yolov3 is so important!

nvdla / sw

Supporting for Deconvolution Layer? #39