I use mobilenet_v1_1.0_224.tflite and try to partition it like test_dnn.py case: (1) relay.frontend.from_tflite (2) partition_for_dnnl(processed_mod, params, alter_layout)
but some error happens: in mod = seq(mod) Check failed: oshape_sum == data_shape_sum (9216 vs. 288) : Input tensor shape(32,3,3,32,1) and reshaped shape(3,3,32,1) are not compatible!
How to fix it? Thanks