daniperfer opened a new issue #7971:
URL: https://github.com/apache/tvm/issues/7971
Hi:
I am trying to follow the tutorial in
`tutorials/frontend/deploy_object_detection_pytorch.py`, but got the following
error:
```
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torch/tensor.py:593:
RuntimeWarning: Iterating over a tensor might cause the trace to be incorrect.
Passing a tensor of different shape won't change the number of iterations
executed (and might lead to errors or silently give incorrect results).
'incorrect results).', category=RuntimeWarning)
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torch/nn/functional.py:3123:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
dtype=torch.float32)).float())) for i in range(dim)]
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torchvision/models/detection/anchor_utils.py:147:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
torch.tensor(image_size[1] // g[1], dtype=torch.int64, device=device)] for
g in grid_sizes]
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torchvision/ops/boxes.py:128:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
boxes_x = torch.min(boxes_x, torch.tensor(width, dtype=boxes.dtype,
device=boxes.device))
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torchvision/ops/boxes.py:130:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
boxes_y = torch.min(boxes_y, torch.tensor(height, dtype=boxes.dtype,
device=boxes.device))
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torchvision/models/detection/transform.py:271:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
for s, s_orig in zip(new_size, original_size)
/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torchvision/models/detection/roi_heads.py:372:
UserWarning: To copy construct from a tensor, it is recommended to use
sourceTensor.clone().detach() or
sourceTensor.clone().detach().requires_grad_(True), rather than
torch.tensor(sourceTensor).
return torch.tensor(M + 2 * padding).to(torch.float32) /
torch.tensor(M).to(torch.float32)
Traceback (most recent call last):
File "tutorials/frontend/deploy_object_detection_pytorch.py", line 95, in
<module>
script_module = do_trace(model, inp)
File "tutorials/frontend/deploy_object_detection_pytorch.py", line 65, in
do_trace
model_trace = torch.jit.trace(model, inp)
File
"/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torch/jit/_trace.py",
line 742, in trace
_module_class,
File
"/home/vitis-ai-user/.conda/envs/my-vitis-ai-pytorch/lib/python3.6/site-packages/torch/jit/_trace.py",
line 940, in trace_module
_force_outplace,
RuntimeError: Only tensors, lists, tuples of tensors, or dictionary of
tensors can be output from traced functions
```
First of all, I have built and installed TVM according to the steps
described in the [Host setup and docker
build](https://tvm.apache.org/docs/deploy/vitis_ai.html#host-setup-and-docker-build)
section of Vitis-AI integration tutorial.
Then, I slightly modified the `deploy_object_detection_pytorch.py` script
(see code below), and launched it under the pytorch conda environment, inside
docker:
> `python tutorials/frontend/deploy_object_detection_pytorch.py`
And I got this error: `RuntimeError: Only tensors, lists, tuples of tensors,
or dictionary of tensors can be output from traced functions`
Code of the slightly modified script `deploy_object_detection_pytorch.py`:
```
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
"""
Compile PyTorch Object Detection Models
=======================================
This article is an introductory tutorial to deploy PyTorch object
detection models with Relay VM.
For us to begin with, PyTorch should be installed.
TorchVision is also required since we will be using it as our model zoo.
A quick solution is to install via pip
.. code-block:: bash
pip install torch==1.7.0
pip install torchvision==0.8.1
or please refer to official site
https://pytorch.org/get-started/locally/
PyTorch versions should be backwards compatible but should be used
with the proper TorchVision version.
Currently, TVM supports PyTorch 1.7 and 1.4. Other versions may
be unstable.
"""
import tvm
from tvm import relay
from tvm import relay
from tvm.runtime.vm import VirtualMachine
from tvm.contrib.download import download
import numpy as np
import cv2
# PyTorch imports
import torch
import torchvision
######################################################################
# Load pre-trained maskrcnn from torchvision and do tracing
# ---------------------------------------------------------
in_size = 300
input_shape = (1, 3, in_size, in_size)
def do_trace(model, inp):
model_trace = torch.jit.trace(model, inp)
model_trace.eval()
return model_trace
def dict_to_tuple(out_dict):
if "masks" in out_dict.keys():
return out_dict["boxes"], out_dict["scores"], out_dict["labels"],
out_dict["masks"]
return out_dict["boxes"], out_dict["scores"], out_dict["labels"]
class TraceWrapper(torch.nn.Module):
def __init__(self, model):
super().__init__()
self.model = model
def forward(self, inp):
out = self.model(inp)
return dict_to_tuple(out[0])
# model_func = torchvision.models.detection.maskrcnn_resnet50_fpn
# model = TraceWrapper(model_func(pretrained=True))
####################################################################
# THIS IS THE ONLY MODIFICATION I MADE TO THE ORIGINAL TUTORIAL CODE
model = torchvision.models.detection.maskrcnn_resnet50_fpn(pretrained=True)
model.eval()
inp = torch.Tensor(np.random.uniform(0.0, 250.0, size=(1, 3, in_size,
in_size)))
with torch.no_grad():
out = model(inp)
script_module = do_trace(model, inp)
######################################################################
# Download a test image and pre-process
# -------------------------------------
img_path = "test_street_small.jpg"
img_url = (
"https://raw.githubusercontent.com/dmlc/web-data/";
"master/gluoncv/detection/street_small.jpg"
)
download(img_url, img_path)
img = cv2.imread(img_path).astype("float32")
img = cv2.resize(img, (in_size, in_size))
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
img = np.transpose(img / 255.0, [2, 0, 1])
img = np.expand_dims(img, axis=0)
######################################################################
# Import the graph to Relay
# -------------------------
input_name = "input0"
shape_list = [(input_name, input_shape)]
mod, params = relay.frontend.from_pytorch(script_module, shape_list)
######################################################################
# Compile with Relay VM
# ---------------------
# Note: Currently only CPU target is supported. For x86 target, it is
# highly recommended to build TVM with Intel MKL and Intel OpenMP to get
# best performance, due to the existence of large dense operator in
# torchvision rcnn models.
# Add "-libs=mkl" to get best performance on x86 target.
# For x86 machine supports AVX512, the complete target is
# "llvm -mcpu=skylake-avx512 -libs=mkl"
target = "llvm"
with tvm.transform.PassContext(opt_level=3, disabled_pass=["FoldScaleAxis"]):
vm_exec = relay.vm.compile(mod, target=target, params=params)
######################################################################
# Inference with Relay VM
# -----------------------
dev = tvm.cpu()
vm = VirtualMachine(vm_exec, dev)
vm.set_input("main", **{input_name: img})
tvm_res = vm.run()
######################################################################
# Get boxes with score larger than 0.9
# ------------------------------------
score_threshold = 0.9
boxes = tvm_res[0].asnumpy().tolist()
valid_boxes = []
for i, score in enumerate(tvm_res[1].asnumpy().tolist()):
if score > score_threshold:
valid_boxes.append(boxes[i])
else:
break
print("Get {} valid boxes".format(len(valid_boxes)))
```
I would like to know if there is anything wrong, and what I can do to
successfully run the PyTorch Object Detection Tutorial.
Thanks.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
