[TOPI] Improve conv2d for resnet18 workload (#427)

* relu activation migrated to topi

* reviews addressed

* relu compute deleted

* conv2d_nchw updated

* resnet18 hand tuned schedule added

* pylint error fixed

* one more workload test for conv2d_nchw

* conv2d schedule subfunctions added for different patterns

* reviews addressed

topi/python/topi/cuda/elemwise.py

@@ -22,7 +22,6 @@ def schedule_elemwise(outs):
     tvm.schedule.AutoInlineInjective(s)
 
     x = outs[0]
-    num_dim = len(x.shape)
     fused = s[x].fuse(*x.op.axis)
     num_thread = 64
     bx, tx = s[x].split(fused, factor=num_thread)

topi/tests/python/test_topi_conv2d_nchw.py

@@ -55,6 +55,7 @@ def verify_conv2d_nchw(batch, in_channel, in_size, num_filter, kernel, stride, p
 
 
 def test_conv2d_nchw():
+    verify_conv2d_nchw(1, 3, 224, 64, 7, 3, 2)
     verify_conv2d_nchw(1, 64, 56, 64, 3, 1, 1)
     verify_conv2d_nchw(1, 64, 56, 64, 1, 1, 0)
     verify_conv2d_nchw(1, 64, 56, 128, 3, 2, 1)