(wip)(do not merge) Run all tests on relay

myia/compile/backends/pytorch.py

@@ -381,9 +381,7 @@ def _impl(input, weight, stride, padding, dilation, groups):
 def pytorch_conv_transpose2d(op):
     """Implementation of conv_transpose2d."""
 
-    def _impl(
-        input, weight, bias, stride, padding, output_padding, groups, dilation
-    ):
+    def _impl(input, weight, stride, padding, output_padding, groups, dilation):
         stride = tuple(_x.item() for _x in stride)
         padding = tuple(_x.item() for _x in padding)
         output_padding = tuple(_x.item() for _x in output_padding)
@@ -392,7 +390,7 @@ def _impl(
         return torch.conv_transpose2d(
             input,
             weight,
-            bias,
+            None,
             stride,
             padding,
             output_padding,

myia/compile/backends/relay.py

@@ -374,26 +374,29 @@ def relay_conv2d_weight_grad(c, data, wsize, dout, stride, pad, dil, groups):
 
 
 def relay_conv_transpose2d(
-    c, input, weight, bias, stride, padding, output_padding, groups, dilation
+    c, input, weight, stride, padding, output_padding, groups, dilation
 ):
-
-    if not bias.is_constant(type(None)):
-        raise NotImplementedError(
-            "conv_transpose2d: bias not yet supported " "in relay backend."
-        )
-
     assert stride.is_constant(tuple)
     assert padding.is_constant(tuple)
     assert output_padding.is_constant(tuple)
     assert dilation.is_constant(tuple)
     assert groups.is_constant(int)
 
+    # Only groups==1 and dilation==(1, 1) is supported, on both CPU and GPU.
+    if groups.value != 1:
+        raise RuntimeError(f"Only support groups=1, got {groups.value}")
+    if dilation.value != (1, 1):
+        raise RuntimeError(
+            f"Only support dilation=(1, 1), got {dilation.value}"
+        )
+
     data_shape = input.abstract.xshape()
     weight_shape = weight.abstract.xshape()
     _, in_channels, _, _ = data_shape
     _, w_c, filter_h, filter_w = weight_shape
     _i = c.ref(input)
     _w = c.ref(weight)
+    output_channels = w_c * groups.value
     return relay.nn.conv2d_transpose(
         _i,
         _w,
@@ -403,7 +406,7 @@ def relay_conv_transpose2d(
         groups=groups.value,
         output_padding=output_padding.value,
         kernel_size=(filter_h, filter_w),
-        channels=in_channels,
+        channels=output_channels,
     )
 
 

myia/operations/macro_conv2d_grad_input.py

@@ -67,7 +67,6 @@ async def conv2d_grad_input(
         P.conv_transpose2d,
         r_grad_output.node,
         r_weight.node,
-        Constant(None),
         r_stride.node,
         r_padding.node,
         Constant(grad_input_padding),

myia/operations/prim_conv_transpose2d.py

@@ -18,7 +18,6 @@ async def infer_conv_transpose2d(
     engine,
     input: AbstractArray,
     weight: AbstractArray,
-    bias,  # un-typed, because it may be either None or an abstract array.
     stride: AbstractTuple,
     padding: AbstractTuple,
     output_padding: AbstractTuple,

myia/public_api.py

@@ -313,7 +313,7 @@ def conv_transpose2d(
 ):
     """Map of Pytorch method torch.nn.functional.conv_transpose2d."""
     return P.conv_transpose2d(
-        input, weight, bias, stride, padding, output_padding, groups, dilation
+        input, weight, stride, padding, output_padding, groups, dilation
     )