My function works fine by itself but not when I run it as a worker

I am trying to run multiple Yolos in parallel in Python.

I have a function that I am trying to use as the worker. something like this:

import darknet as dn

def yolo(input_queue, net, meta):

    while True:

        task = input_queue.get_nowait()

        c = task.shape[0]

        h = task.shape[1]

        w = task.shape[2]

        task = (task/255.0)

        task = task.flatten()

        new_arr = np.copy(task).astype(np.float32) 

        new_arr = np.ctypeslib.as_ctypes(new_arr)

        im = dn.IMAGE(w,h,c,new_arr)

        dn.rgbgr_image(im)



        print(dn.detect(net, meta, im))

You can find darknet file here.

Let's don't talk about the way I convert the array to a ctype array. it has been discussed here.

If I run this function as a normal function works like a charm. But if I run it as a worker like this:

p = Process(target=yolo, args=(input_queue, net, meta))

p.start()

it goes all the way down to the last line then it gives me this error:

CUDA Error: initialization error

python: ./src/cuda.c:36: check_error: Assertion `0' failed.

I couldn't figure out why exactly this error happens.
can anyone help? what's the difference between these two way of using this function that cause this problem?

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

1

Check this out.

– GeeTransit
Jan 3 at 1:18

You may may need to put the "entry point" of the main script within an if __name__ == '__main__': conditional as described in the "Safe importing of main module" section of the multiprocessing Programming guidelines documentation.

– martineau
Jan 3 at 1:19

@GeeTransit it is different but answer by sagarwal gave me the idea to solve the problem

– Ehsan Fathi
Jan 3 at 22:25

add a comment |

I am trying to run multiple Yolos in parallel in Python.

I have a function that I am trying to use as the worker. something like this:

import darknet as dn

def yolo(input_queue, net, meta):

    while True:

        task = input_queue.get_nowait()

        c = task.shape[0]

        h = task.shape[1]

        w = task.shape[2]

        task = (task/255.0)

        task = task.flatten()

        new_arr = np.copy(task).astype(np.float32) 

        new_arr = np.ctypeslib.as_ctypes(new_arr)

        im = dn.IMAGE(w,h,c,new_arr)

        dn.rgbgr_image(im)



        print(dn.detect(net, meta, im))

You can find darknet file here.

Let's don't talk about the way I convert the array to a ctype array. it has been discussed here.

If I run this function as a normal function works like a charm. But if I run it as a worker like this:

p = Process(target=yolo, args=(input_queue, net, meta))

p.start()

it goes all the way down to the last line then it gives me this error:

CUDA Error: initialization error

python: ./src/cuda.c:36: check_error: Assertion `0' failed.

I couldn't figure out why exactly this error happens.
can anyone help? what's the difference between these two way of using this function that cause this problem?

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

1

Check this out.

– GeeTransit
Jan 3 at 1:18

You may may need to put the "entry point" of the main script within an if __name__ == '__main__': conditional as described in the "Safe importing of main module" section of the multiprocessing Programming guidelines documentation.

– martineau
Jan 3 at 1:19

@GeeTransit it is different but answer by sagarwal gave me the idea to solve the problem

– Ehsan Fathi
Jan 3 at 22:25

add a comment |

I am trying to run multiple Yolos in parallel in Python.

I have a function that I am trying to use as the worker. something like this:

import darknet as dn

def yolo(input_queue, net, meta):

    while True:

        task = input_queue.get_nowait()

        c = task.shape[0]

        h = task.shape[1]

        w = task.shape[2]

        task = (task/255.0)

        task = task.flatten()

        new_arr = np.copy(task).astype(np.float32) 

        new_arr = np.ctypeslib.as_ctypes(new_arr)

        im = dn.IMAGE(w,h,c,new_arr)

        dn.rgbgr_image(im)



        print(dn.detect(net, meta, im))

You can find darknet file here.

Let's don't talk about the way I convert the array to a ctype array. it has been discussed here.

If I run this function as a normal function works like a charm. But if I run it as a worker like this:

p = Process(target=yolo, args=(input_queue, net, meta))

p.start()

it goes all the way down to the last line then it gives me this error:

CUDA Error: initialization error

python: ./src/cuda.c:36: check_error: Assertion `0' failed.

I couldn't figure out why exactly this error happens.
can anyone help? what's the difference between these two way of using this function that cause this problem?

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

I am trying to run multiple Yolos in parallel in Python.

I have a function that I am trying to use as the worker. something like this:

import darknet as dn

def yolo(input_queue, net, meta):

    while True:

        task = input_queue.get_nowait()

        c = task.shape[0]

        h = task.shape[1]

        w = task.shape[2]

        task = (task/255.0)

        task = task.flatten()

        new_arr = np.copy(task).astype(np.float32) 

        new_arr = np.ctypeslib.as_ctypes(new_arr)

        im = dn.IMAGE(w,h,c,new_arr)

        dn.rgbgr_image(im)



        print(dn.detect(net, meta, im))

You can find darknet file here.

Let's don't talk about the way I convert the array to a ctype array. it has been discussed here.

If I run this function as a normal function works like a charm. But if I run it as a worker like this:

p = Process(target=yolo, args=(input_queue, net, meta))

p.start()

it goes all the way down to the last line then it gives me this error:

CUDA Error: initialization error

python: ./src/cuda.c:36: check_error: Assertion `0' failed.

I couldn't figure out why exactly this error happens.
can anyone help? what's the difference between these two way of using this function that cause this problem?

python multiprocessing yolo

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

edited Jan 3 at 2:15

talonmies

59.8k17132199

edited Jan 3 at 2:15

talonmies

59.8k17132199

edited Jan 3 at 2:15

talonmies

59.8k17132199

asked Jan 3 at 1:05

Ehsan Fathi

378

asked Jan 3 at 1:05

Ehsan Fathi

378

asked Jan 3 at 1:05

Ehsan Fathi

378

1

Check this out.

– GeeTransit
Jan 3 at 1:18

You may may need to put the "entry point" of the main script within an if __name__ == '__main__': conditional as described in the "Safe importing of main module" section of the multiprocessing Programming guidelines documentation.

– martineau
Jan 3 at 1:19

@GeeTransit it is different but answer by sagarwal gave me the idea to solve the problem

– Ehsan Fathi
Jan 3 at 22:25

add a comment |

1

Check this out.

– GeeTransit
Jan 3 at 1:18

You may may need to put the "entry point" of the main script within an if __name__ == '__main__': conditional as described in the "Safe importing of main module" section of the multiprocessing Programming guidelines documentation.

– martineau
Jan 3 at 1:19

@GeeTransit it is different but answer by sagarwal gave me the idea to solve the problem

– Ehsan Fathi
Jan 3 at 22:25

Check this out.

– GeeTransit
Jan 3 at 1:18

You may may need to put the "entry point" of the main script within an if __name__ == '__main__': conditional as described in the "Safe importing of main module" section of the multiprocessing Programming guidelines documentation.

– martineau
Jan 3 at 1:19

@GeeTransit it is different but answer by sagarwal gave me the idea to solve the problem

– Ehsan Fathi
Jan 3 at 22:25

add a comment |

1 Answer
1

active

oldest

votes

Thanks to this I found the problem is that you can't share the context in CUDA between processes with different PIDs which makes sense. So I found that if I run related things (in my case, net) in the same process then it works okay. before this I was loading weights before I start the worker.

answered Jan 3 at 22:31

Ehsan Fathi

378

add a comment |

Your Answer

StackExchange.ifUsing("editor", function () {
StackExchange.using("externalEditor", function () {
StackExchange.using("snippets", function () {
StackExchange.snippets.init();
});
});
}, "code-snippets");

StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "1"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});

function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});

}
});

draft saved

draft discarded

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f54015121%2fmy-function-works-fine-by-itself-but-not-when-i-run-it-as-a-worker%23new-answer', 'question_page');
}
);

Post as a guest

Name

Required, but never shown

1 Answer
1

active

oldest

votes

1 Answer
1

active

oldest

votes

answered Jan 3 at 22:31

Ehsan Fathi

378

add a comment |

answered Jan 3 at 22:31

Ehsan Fathi

378

add a comment |

answered Jan 3 at 22:31

Ehsan Fathi

378

answered Jan 3 at 22:31

Ehsan Fathi

378

answered Jan 3 at 22:31

Ehsan Fathi

378

answered Jan 3 at 22:31

Ehsan Fathi

378

answered Jan 3 at 22:31

Ehsan Fathi

378

add a comment |

draft saved

draft discarded

Thanks for contributing an answer to Stack Overflow!

Please be sure to answer the question. Provide details and share your research!

But avoid …

Asking for help, clarification, or responding to other answers.

Making statements based on opinion; back them up with references or personal experience.

To learn more, see our tips on writing great answers.

draft saved

draft discarded

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Sign up or log in

StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});

Post as a guest

Name

Required, but never shown

Name

Required, but never shown

Name

Required, but never shown

This page is only for reference, If you need detailed information, please check here

Search This Blog

Ufyukyu