[ONNX] Fix issues for Clip and RoiAlign (apache#7237)

python/tvm/relay/frontend/onnx.py

@@ -2046,7 +2046,7 @@ def _impl_v1(cls, inputs, attr, params):
         x = inputs[0]
         rois = inputs[1]
         batch_indices = inputs[2]
-        mode = attr.get("mode", "avg")
+        mode = attr.get("mode", b"avg")
         if mode != b"avg":
             raise ValueError("RoiAlign in Relay only uses avg mode")
         output_height = attr.get("output_height", 1)
@@ -2056,7 +2056,7 @@ def _impl_v1(cls, inputs, attr, params):
         spatial_scale = attr.get("spatial_scale", 1.0)
 
         batch_indices = _op.expand_dims(batch_indices, axis=1, num_newaxis=1)
-        batch_indices = _op.cast(batch_indices, infer_type(rois).type_annotation.dtype)
+        batch_indices = _op.cast(batch_indices, infer_type(rois).checked_type.dtype)
         rois = _op.concatenate([batch_indices, rois], 1)
 
         return _vision.roi_align(
@@ -2074,6 +2074,10 @@ def convert_attributes(inputs, attr, params):
 
     @classmethod
     def _impl_v1(cls, inputs, attr, params):
+        if "min" not in attr:
+            attr["min"] = -np.inf
+        if "max" not in attr:
+            attr["max"] = np.inf
         return Clip.convert_attributes(inputs, attr, params)
 
     @classmethod

tests/python/frontend/onnx/test_forward.py

@@ -840,7 +840,7 @@ def test_slice():
     )
 
 
-def _test_onnx_op_elementwise(inshape, outfunc, npargs, dtype, opname, kwargs):
+def _test_onnx_op_elementwise(inshape, outfunc, npargs, dtype, opname, kwargs, opset=None):
     indata = np.random.uniform(-1, 1, size=inshape).astype(dtype)
     outdata = outfunc(indata, **npargs)
 
@@ -856,7 +856,7 @@ def _test_onnx_op_elementwise(inshape, outfunc, npargs, dtype, opname, kwargs):
     model = helper.make_model(graph, producer_name=opname + "_test")
 
     for target, ctx in tvm.testing.enabled_targets():
-        tvm_out = get_tvm_output(model, indata, target, ctx, outdata.shape, dtype)
+        tvm_out = get_tvm_output(model, indata, target, ctx, outdata.shape, dtype, opset=opset)
         tvm.testing.assert_allclose(outdata, tvm_out)
 
 
@@ -881,6 +881,26 @@ def test_clip():
         {"min": -1.0, "max": 1.0},
     )
 
+    _test_onnx_op_elementwise(
+        (2, 4, 5, 6),
+        np.clip,
+        {"a_min": -np.inf, "a_max": 1.0},
+        "float32",
+        "Clip",
+        {"max": 1.0},
+        opset=1,
+    )
+
+    _test_onnx_op_elementwise(
+        (2, 4, 5, 6),
+        np.clip,
+        {"a_min": -1.0, "a_max": np.inf},
+        "float32",
+        "Clip",
+        {"min": -1.0},
+        opset=1,
+    )
+
 
 @tvm.testing.uses_gpu
 def test_clip_min_max_as_inputs():