Skip to content

bump PyTorch version to 1.11 #10794

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Mar 30, 2022
Merged

bump PyTorch version to 1.11 #10794

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f333889
bump PyTorch version to 1.11
t-vi Mar 26, 2022
9511ff1
disable some caffe2 ci
t-vi Mar 28, 2022
d253476
Fix sub conversion in PyTorch frontend
t-vi Mar 28, 2022
b3e0fc0
Merge branch 'main' into bump_pytorch
t-vi Mar 30, 2022
e728c3d
use fuse_modules_qat if available, fallback to fuse_modules for older…
t-vi Mar 30, 2022
fec6897
Re-Run CI
t-vi Mar 30, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 0 additions & 3 deletions docker/Dockerfile.ci_gpu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,9 +77,6 @@ RUN bash /install/ubuntu_install_onnx.sh
COPY install/ubuntu_install_tflite.sh /install/ubuntu_install_tflite.sh
RUN bash /install/ubuntu_install_tflite.sh

COPY install/ubuntu_install_caffe2.sh /install/ubuntu_install_caffe2.sh
RUN bash /install/ubuntu_install_caffe2.sh

COPY install/ubuntu_install_dgl.sh /install/ubuntu_install_dgl.sh
RUN bash /install/ubuntu_install_dgl.sh

Expand Down
4 changes: 2 additions & 2 deletions docker/install/ubuntu_install_onnx.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,5 +36,5 @@ pip3 install \
pip3 install future

pip3 install \
torch==1.10.1 \
torchvision==0.11.2
torch==1.11.0 \
torchvision==0.12.0
145 changes: 0 additions & 145 deletions gallery/how_to/compile_models/from_caffe2.py
View file
Open in desktop

This file was deleted.

15 changes: 13 additions & 2 deletions python/tvm/relay/frontend/pytorch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2039,6 +2039,14 @@ def stack(self, inputs, input_types):
assert isinstance(ty, tvm.ir.TypeCall) and ty.func == list_ty, msg
return self.tensor_array_stack(inputs, input_types)

def sub(self, inputs, input_types):
if len(inputs) == 3:
data0, data1, alpha = self.pytorch_promote_types(inputs, input_types)
return get_relay_op("subtract")(data0, alpha * data1)
else:
data0, data1= self.pytorch_promote_types(inputs, input_types)
return get_relay_op("subtract")(data0, data1)

def rsub(self, inputs, input_types):
data0, data1, alpha = self.pytorch_promote_types(inputs, input_types)

Expand Down Expand Up @@ -2859,7 +2867,10 @@ def all_any_common(self, op, inputs, input_types):
inp = inputs[0]
return op(inp, axis=dim, keepdims=keepdim)

def searchsorted_common(self, sorted_sequence, values, out_int32, right):
def searchsorted_common(
self, sorted_sequence, values, out_int32, right, side=None, out=None, sorter=None
):
assert side is None and out is None and sorter is None, "unsupported parameters"
dtype = "int32" if out_int32 else "int64"
values_shape = _infer_shape(values)

Expand Down Expand Up @@ -2959,7 +2970,7 @@ def create_convert_map(self):
"aten::pixel_shuffle": self.pixel_shuffle,
"aten::device": self.none,
"prim::device": self.none,
"aten::sub": self.make_elemwise("subtract"),
"aten::sub": self.sub,
Copy link
Member

@masahi masahi Mar 26, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It seems this change breaks test_lstm.py:

FAILED test_lstm.py::test_custom_lstm - AttributeError: 'function' object has no attribute 'dtype'

Copy link
Contributor Author

@t-vi t-vi Mar 26, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'll see to fixing it. Thank you.

"aten::max": self.max,
"aten::min": self.min,
"aten::mul": self.make_elemwise("multiply"),
Expand Down
3 changes: 2 additions & 1 deletion tests/python/frontend/pytorch/qnn_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -743,9 +743,10 @@ def __init__(self, inputsize=(128, 128)):
self.backbone = Backbone()

def fuse_model(self):
fuse_modules_qat = getattr(torch.ao.quantization, "fuse_modules_qat", fuse_modules)
for idx, m in enumerate(self.modules()):
if type(m) == ConvBnRelu:
torch.quantization.fuse_modules(m, ["conv", "bn", "relu"], inplace=True)
fuse_modules_qat(m, ["conv", "bn", "relu"], inplace=True)

def forward(self, input):
input = self.quant(input)
Expand Down
3 changes: 0 additions & 3 deletions tests/scripts/task_python_frontend.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,9 +56,6 @@ function shard2 {
i=$((i+1))
done

echo "Running relay caffe2 frontend test..."
run_pytest cython python-frontend-caffe2 tests/python/frontend/caffe2

echo "Running relay DarkNet frontend test..."
run_pytest cython python-frontend-darknet tests/python/frontend/darknet

Expand Down