Skip to content
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.

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

Add PAD operator to relay tflite frontend #3310

Merged
merged 1 commit into from
Jun 10, 2019

Conversation

apivovarov
Copy link
Contributor

PAD operator is used in TFLite SSD Resnet 50 model (ssd_resnet_50_fpn_coco).
Model graph: ssd_resnet_50_fpn_coco.tflite.pdf

### ssd_resnet_50_fpn_coco summary:
# op_id: op_name - count
 0: ADD - 58
 2: CONCATENATION - 2
 3: CONV_2D - 110
14: LOGISTIC - 1
17: MAX_POOL_2D - 4
18: MUL - 42
22: RESHAPE - 14
32: CUSTOM - 1
34: PAD - 1

This PR adds PAD operator support to relay tflite frontend.

@kevinthesun kevinthesun requested a review from srkreddy1238 June 7, 2019 07:00
in_expr = self.get_expr(input_tensor.tensor_idx)

# paddings
padlist = self.get_tensor_value(input_tensors[1])
Copy link
Member

Choose a reason for hiding this comment

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

pad_list

Copy link
Contributor Author

Choose a reason for hiding this comment

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

fixed

Copy link
Member

@yongwww yongwww left a comment

Choose a reason for hiding this comment

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

lgtm

@apivovarov apivovarov force-pushed the tflite_pad branch 2 times, most recently from c4ad7a8 to 573729e Compare June 8, 2019 05:31
@kevinthesun
Copy link
Contributor

@apivovarov Can you fix ci failure?

@apivovarov
Copy link
Contributor Author

apivovarov commented Jun 10, 2019

The test which I added to tests/python/frontend/tflite/test_forward.py run in frontend: GPU
tests/scripts/task_python_frontend.sh
It run successfully
http://ci.tvm.ai:8080/blue/organizations/jenkins/tvm/detail/PR-3310/4/pipeline/239#step-293-log-969

Shell Script
Running relay TFLite frontend test... ok

Failed test is topi: GPU which runs tests/scripts/task_python_topi.sh
python3 -m nose -v topi/tests/python
I did not even touch topi code
the last line in the failed test is
missing workspace /home/ubuntu/jenkins/workspace/tvm/topi-python-gpu on g2.n0
It tried to run test_topi_conv2d_nchw.test_conv2d_nchw

@apivovarov
Copy link
Contributor Author

apivovarov commented Jun 10, 2019

There is "Missing workspace - workspace deleted during concurrent build" issue
https://issues.jenkins-ci.org/browse/JENKINS-44659

workspace path which tvm ci uses to test topi is not unique - /home/ubuntu/jenkins/workspace/tvm/topi-python-gpu
this is why Jenkins can delete this folder in the middle of the execution of particular build if another build which run in parallel is already done.

@kevinthesun kevinthesun merged commit 8f219b9 into apache:master Jun 10, 2019
wweic pushed a commit to wweic/tvm that referenced this pull request Jun 26, 2019
wweic pushed a commit to neo-ai/tvm that referenced this pull request Jun 27, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants