[Relay][UnitTest] Removed redundant unit test.

tests/python/relay/test_any.py

@@ -419,30 +419,6 @@ def test_any_reduce(
         check_result([data_np], mod, ref_out_shape, assert_shape=True, targets=[(target, dev)])
 
 
-def verify_any_reduce(
-    reduce_op, data_shape, axis, exclude, keepdims, static_data_shape, ref_out_shape
-):
-    mod = tvm.IRModule()
-    dtype = "bool" if reduce_op == relay.all else "float32"
-    data = relay.var("data", shape=data_shape, dtype=dtype)
-    y = reduce_op(data, axis, keepdims, exclude)
-    mod["main"] = relay.Function([data], y)
-    data_np = np.random.uniform(size=static_data_shape).astype(dtype)
-    check_result([data_np], mod, ref_out_shape, assert_shape=True)
-
-
-@tvm.testing.uses_gpu
-def test_any_reduce():
-    verify_any_reduce(relay.argmax, any_dims(3), None, False, False, (3, 4, 5), ())
-    verify_any_reduce(relay.argmin, any_dims(4), 1, False, True, (3, 4, 5, 6), (3, 1, 5, 6))
-    verify_any_reduce(relay.all, any_dims(3), (1, 2), True, False, (3, 4, 5), (4, 5))
-    verify_any_reduce(relay.max, any_dims(4), -1, True, True, (3, 4, 5, 6), (1, 1, 1, 6))
-    verify_any_reduce(relay.min, any_dims(3), (0, 1), False, False, (4, 5, 6), (6,))
-    verify_any_reduce(relay.prod, any_dims(4), 2, True, True, (3, 4, 5, 6), (1, 1, 5, 1))
-    verify_any_reduce(relay.mean, any_dims(2), 0, False, False, (1, 2), (2,))
-    verify_any_reduce(relay.variance, any_dims(5), (2, 4), False, False, (3, 4, 5, 6, 7), (3, 4, 6))
-
-
 def verify_any_layout_transform(
     data_shape, src_layout, dst_layout, static_data_shape, ref_out_shape
 ):

tests/python/relay/test_op_level2.py

@@ -357,121 +357,6 @@ def test_run(
         tvm.testing.assert_allclose(op_res1.numpy(), ref_res, rtol=1e-4, atol=1e-4)
 
 
-def test_conv2d_run(target, dev):
-    def run_test_conv2d(
-        dtype,
-        out_dtype,
-        scale,
-        dshape,
-        kshape,
-        padding=(1, 1),
-        fref=None,
-        groups=1,
-        dilation=(1, 1),
-        channels=32,
-        kernel_size=(3, 3),
-    ):
-        x = relay.var("x", shape=dshape, dtype=dtype)
-        w = relay.var("w", shape=kshape, dtype=dtype)
-        y = relay.nn.conv2d(
-            x,
-            w,
-            padding=padding,
-            dilation=dilation,
-            groups=groups,
-            channels=channels,
-            kernel_size=kernel_size,
-        )
-        func = relay.Function([x, w], y)
-        data = np.random.uniform(-scale, scale, size=dshape).astype(dtype)
-        kernel = np.random.uniform(-scale, scale, size=kshape).astype(dtype)
-        dkernel = tvm.topi.testing.dilate_python(kernel, (1, 1) + dilation)
-        ref_res = tvm.topi.testing.conv2d_nchw_python(
-            data.astype(out_dtype), dkernel.astype(out_dtype), 1, padding, groups=groups
-        )
-
-        op_res1 = relay.create_executor("graph", device=dev, target=target).evaluate(func)(
-            data, kernel
-        )
-        tvm.testing.assert_allclose(op_res1.numpy(), ref_res, rtol=1e-4, atol=1e-4)
-
-    # group conv2d
-    run_test_conv2d(
-        dtype="float32",
-        out_dtype="float32",
-        scale=1,
-        dshape=(1, 32, 18, 18),
-        kshape=(32, 4, 3, 3),
-        padding=(1, 1),
-        channels=32,
-        groups=8,
-        kernel_size=(3, 3),
-        dilation=(1, 1),
-    )
-    # also group conv2d
-    run_test_conv2d(
-        dtype="float32",
-        out_dtype="float32",
-        scale=1,
-        dshape=(1, 32, 18, 18),
-        kshape=(64, 1, 3, 3),
-        padding=(1, 1),
-        channels=64,
-        groups=32,
-        kernel_size=(3, 3),
-        dilation=(1, 1),
-    )
-
-    # normal conv2d
-    run_test_conv2d(
-        dtype="float32",
-        out_dtype="float32",
-        scale=1,
-        dshape=(1, 3, 224, 224),
-        kshape=(10, 3, 3, 3),
-        padding=(1, 1),
-        channels=10,
-        kernel_size=(3, 3),
-        dilation=(1, 1),
-    )
-    # mixed precision
-    run_test_conv2d(
-        dtype="int8",
-        out_dtype="int32",
-        scale=1,
-        dshape=(1, 3, 224, 224),
-        kshape=(10, 3, 3, 3),
-        padding=(1, 1),
-        channels=10,
-        kernel_size=(3, 3),
-        dilation=(1, 1),
-    )
-    # mixed precision.
-    run_test_conv2d(
-        dtype="int8",
-        out_dtype="int32",
-        scale=1,
-        dshape=(1, 3, 224, 224),
-        kshape=(10, 3, 1, 3),
-        padding=(0, 1),
-        channels=10,
-        kernel_size=(1, 3),
-        dilation=(1, 1),
-    )
-    # dilated conv2d
-    run_test_conv2d(
-        dtype="float32",
-        out_dtype="float32",
-        scale=1,
-        dshape=(1, 3, 18, 18),
-        kshape=(10, 3, 3, 3),
-        padding=(1, 1),
-        channels=10,
-        kernel_size=(3, 3),
-        dilation=(3, 3),
-    )
-
-
 def test_compile_depthwise_conv2d_arm_cpu():
     dtype = "float32"
     out_dtype = "float32"