Issues with deployment on google app engine example

Part 1 (2019)
23

First of all thank you for replying!
So are you saying that I should go to source code and change this line:
“_solve_func = getattr(torch, ‘solve’, torch.gesv)”

Manually? I think I would face the same issues andres is facing.
See his answer below mine.

I tried / trying another path, but while, faced another error, maybe you can help.
So I went to my VM JupyterHub.
Checked ! pip install list.
And transformed my requirments.txt file to mimic what ! pip install returned.
So I know I have a perfect match of what I am deploying vs what I trained on.

So I ended up with the following requirements doc:"

‘’’
numpy==1.16.2
torchvision==0.3.0
https://download.pytorch.org/whl/cpu/torch-1.1.0-cp37-cp37m-linux_x86_64.whl
fastai==1.0.55
starlette==0.11.4
uvicorn==0.3.32
python-multipart
aiofiles==0.4.0
aiohttp==3.5.4

‘’’

And here is my current docker file

‘’’
FROM python:3.6-slim-stretch

RUN apt update

RUN apt install -y python3-dev gcc

ADD requirements.txt requirements.txt

RUN pip install -r requirements.txt

COPY app app/

RUN python app/server.py

EXPOSE 8080

CMD [“python”, “app/server.py”, “serve”]

#CMD [“gunicorn”, “-b”, “:8080”, “–chdir”, “app/”, “main:app”]

#RUN python app/server.py
#gcloud app update --no-split-health-check

‘’’

However, when I tried to launch I got the following error:

‘’’
Step 5/9 : RUN pip install -r requirements.txt
—> Running in 0132b685c40c
ERROR: torch-1.1.0-cp37-cp37m-linux_x86_64.whl is not a supported wheel on this platform.
WARNING: You are using pip version 19.2.1, however version 19.2.2 is available.
You should consider upgrading via the ‘pip install --upgrade pip’ command.
The command ‘/bin/sh -c pip install -r requirements.txt’ returned a non-zero code: 1
ERROR
ERROR: build step 0 “gcr.io/cloud-builders/docker” failed: exit status 1

‘’’

That was due to =
https://download.pytorch.org/whl/cpu/torch-1.1.0-cp37-cp37m-linux_x86_64.whl

trying to install version for Python 3.7 (CP37) vs Docker file specifying Python 3.6

So I changed it to:
https://download.pytorch.org/whl/cpu/torch-1.1.0-cp36-cp36m-linux_x86_64.whl

And now facing this error.

from torchvision import _C
ImportError: libcudart.so.9.0: cannot open shared object file: No such file or directory
The command ‘/bin/sh -c python app/server.py’ returned a non-zero code: 1
ERROR
ERROR: build step 0 “gcr.io/cloud-builders/docker” failed: exit status 1

Looked into it, and it has to do with CUDA versioning…
Looks like this will never end.
I am giving up for now.

PS.
The app did deploy on render…
So an easier way out.

Here is my GIT if it helps anyone here:

24

The timeout error for me is solved by increasing the requested disk space in the app.yaml file, like this:

resources:
    disk_size_gb: 12

12 GB seems to be enough for me.

25

could you solve the problem

26

@joaquinmaya could you solve the problem

27

Hi All ,

I’ve down graded pillow to <7.0.0 and moved to docker file install step.

here is docker file

FROM python:3.6-slim-stretch
MAINTAINER Emre Cavunt <emre.cavunt@gmail.com>
RUN apt-get update
RUN apt-get install --yes software-properties-common
RUN apt-add-repository contrib
RUN apt-get update

RUN apt-get install --yes \
    python \
    python-dev \
    python-pip \
    build-essential \
    git \
    bash \
    strace \
  && pip install virtualenv \
  && pip install 'pillow<7.0.0' \
  && rm -rf /var/cache/apk/* \
  && apt-get clean

ADD requirements.txt requirements.txt
RUN pip --no-cache-dir install -r requirements.txt

COPY app app/

RUN python app/server.py

EXPOSE 5000

CMD ["python", "app/server.py", "serve"]

You can check the our github repo

https://github.com/emrecavunt/detect-dental-problem

28

Hello, i’m a bit of a beginner and don’t understand this part of the tutorial. He mentions to open up server.py on instructions and change url and classes. I am using Windows and using Ubuntu app as terminal. I follow all the previous instruction, but then i am having trouble opening and thus modifying the file from the terminal. I’m not sure if there is also a GUI to do it more easily. But, i’m stuck here. Some help with beginner details exactly on that part would be appreciated

image
image1814×725 87.3 KB

29

Yes, you can open server.py script in notepad or any other text editor on windows. Then you can edit whatever you need to edit.

30

I finally got it deployed but it never passes the readiness checks. Only main changes are on the app.yaml which can be seen below. Any suggestions on how to fix?

runtime: custom
env: flex
service: first-class
readiness_check:
  app_start_timeout_sec: 1800
resources:
 disk_size_gb: 15
31

The service works if I use the suggested download URL in the tutorial, but not with my own parameters.

I’m using Google Drive and I’m creating a request from here:

Because my file is over 100MB, I’ve attached an API key to my request. I am not receiving an ‘Unpickling error’ about my response returning with ‘<’.

Here’s what my Google Drive link looks like:

All results occurred in AppEngine. I can’t run the service locally to test because I receive an error that Starlette cannot be found, which I attribute to a lack of a download even though I’ve downloaded the dependencies multiple times.

In order to test this locally, I’ve been starting a docker process and running docker build .

Is there a better way to run the service locally given that the suggested command doesn’t work for me? python app/server.py serve

I’m on a Mac

32

Hi nswekosk hope your having a jolly day!

See this link Deployment Platform: Render ✅ There are many posts of the issues people have deploying their models. I use exactly the same code to deploy locally on my mac, via docker on my mac in a virtual directory on my mac or on an online host.

Pay particular notice to posts about pip list in relation to requirements.txt, your torch version and server.py config.

I nearly always start with this repository starter code https://course.fast.ai/deployment_render.html and build on it when I have it working.

Lastly this Running, Deploying Fastai successes: Links, Tutorials, Stories is a thread started recently which shows various ways people have deployed their models.

Hope this helps

mrfabulous1 :smiley: :smiley:

1 Like
33

Thanks!

The link to deploying on ‘Render’ gave a helpful example of the right command to run the Dockerfile locally. I’ll give that a shot