This is an automated email from the ASF dual-hosted git repository.
mbrookhart pushed a commit to branch ci-docker-staging
in repository https://gitbox.apache.org/repos/asf/tvm.git
commit f4445dff9e9410c5bcbfc58d2d5ccf9383a94460
Author: Matthew <mbrookh...@octoml.ai>
AuthorDate: Wed Jun 16 16:03:40 2021 -0600
support convtranspose opset 11 autopadding
---
python/tvm/relay/frontend/onnx.py | 66 +++++++++++++++++++++++++++++++++++++++
1 file changed, 66 insertions(+)
diff --git a/python/tvm/relay/frontend/onnx.py
b/python/tvm/relay/frontend/onnx.py
index f876b1d..49d58fa 100644
--- a/python/tvm/relay/frontend/onnx.py
+++ b/python/tvm/relay/frontend/onnx.py
@@ -581,6 +581,72 @@ class ConvTranspose(OnnxOpConverter):
out = _op.nn.bias_add(out, inputs[2])
return out
+ @classmethod
+ def _impl_v11(cls, inputs, attr, params):
+ # get number of channels
+ out_type = infer_type(inputs[1])
+ out_shapes = [get_const_tuple(out_type.checked_type.shape)]
+ channels = out_shapes[0][1]
+ attr["channels"] = channels
+ groups = attr.get("group", 1)
+
+ if "kernel_shape" not in attr:
+ attr["kernel_shape"] = out_shapes[0][2:]
+
+ attr["groups"] = groups
+ # infer pads for auto_pad
+ data = inputs[0]
+ input_shape = infer_shape(data)
+ ndim = len(input_shape)
+ if "auto_pad" in attr:
+ attr["auto_pad"] = attr["auto_pad"].decode("utf-8")
+ if attr["auto_pad"] in ("SAME_UPPER", "SAME_LOWER"):
+ # Warning: Convolution does not yet support dynamic shapes,
+ # one will need to run dynamic_to_static on this model after
import
+ kernel_shape = attr["kernel_shape"]
+ kndim = len(kernel_shape)
+ dilations = attr.get("dilations", [1] * kndim)
+ output_padding = attr.get("output_padding", [0] * kndim)
+ strides = attr["strides"]
+ total_pad = [0] * kndim
+ print(kernel_shape, dilations, output_padding, strides)
+ for i in range(kndim):
+ total_pad[i] = (
+ output_padding[i] + ((kernel_shape[i] - 1) *
dilations[i] + 1) - strides[i]
+ )
+ left = [p // 2 for p in total_pad]
+ right = [total_pad[i] - left[i] for i in range(kndim)]
+ print(left, right)
+ if "LOWER" in attr["auto_pad"]:
+ pad = left + right
+ else:
+ pad = right + left
+ attr["pads"] = pad
+ elif attr["auto_pad"] == "VALID":
+ attr["pads"] = tuple([0 for i in range(ndim - 2)])
+ elif attr["auto_pad"] == "NOTSET":
+ pass
+ else:
+ msg = 'Value {} in attribute "auto_pad" of operator Conv is
invalid.'
+ raise
tvm.error.OpAttributeInvalid(msg.format(attr["auto_pad"]))
+ attr.pop("auto_pad")
+
+ out = AttrCvt(
+ op_name=dimension_picker("conv", "_transpose"),
+ transforms={
+ "kernel_shape": "kernel_size",
+ "dilations": ("dilation", 1),
+ "pads": ("padding", 0),
+ "group": ("groups", 1),
+ },
+ disables=["output_shape"],
+ custom_check=dimension_constraint(),
+ )([data, inputs[1]], attr, params)
+ use_bias = len(inputs) == 3
+ if use_bias:
+ out = _op.nn.bias_add(out, inputs[2])
+ return out
+
class GlobalAveragePool(OnnxOpConverter):
"""Operator converter for GlobalAveragePool"""
