Any start Guide for those who do not know PY at all ?

[yarekc]

Hi,

Your project looks great. However I have not Python knowledge at all (nodejs/js only) and was unable to test it.

I think it would be great if you can provide the simplest possible way to deploy that on a dedicated server and a small snippet on how to use that in html/js page (a kind of rest API ajax call I guess)

A kind of "demo" page.

Thanks a lot.

[bedapudi6788]

@yarekc the simplest way to deploy this on a server is to

1. https://docs.docker.com/engine/install/ Install docker
2. Run docker run -p8080:8080 notaitech/nudenet:detector or :classifier if you want the classifier.
3. This will start a api server with endpoint at http://localhost:8080/sync to which you can send any image as a base64 encoded file and get the prediction back.
4. https://fastdeploy.notai.tech/api#request-type-file this is the API documentation.

[yarekc]

Sounds great This will start a api server with endpoint at http://localhost:8080/sync to which you can send any image as a base64 encoded file and get the prediction back.

Send = POST ? Maybe a sample about what to Send ? Just base64 data ? Other params ?

[bedapudi6788]

Send = POST ?

yes

https://fastdeploy.notai.tech/api#request-type-file this details the request formats.

{
  "data": {
    "1.png": "base64 encoded 1.png",
    "2.png": "base64 encoded 2.png"
  }
}

this is the format of the json.

[yarekc]

Hi,

docker run -p8080:8080 notaitech/nudenet:detector

2021-02-18:13:01:40,720 INFO     [_utils.py:129] AVAILABLE FREE MEMORY: 3384207.09375; CACHE: 169210.35468750002 MB
2021-02-18:13:01:40,830 INFO     [_utils.py:129] AVAILABLE FREE MEMORY: 3384207.08984375; CACHE: 169210.3544921875 MB
2021-02-18:13:01:41,533 INFO     [_utils.py:186] Warming up ..
Waiting for prediction loop to begin.
2021-02-18:13:01:47,507 INFO     [_utils.py:226] Time per sample for batch_size: 1 is 0.9713496367136637
Waiting for prediction loop to begin.
2021-02-18:13:01:53,307 INFO     [_utils.py:226] Time per sample for batch_size: 2 is 0.9665474096934
2021-02-18:13:01:53,307 INFO     [_utils.py:241] optimum batch size is 1
2021-02-18:13:01:53,339 INFO     [_loop.py:35] Starting prediction loop
Waiting for prediction loop to begin.
2021-02-18:13:01:55,817 INFO     [_utils.py:129] AVAILABLE FREE MEMORY: 3384207.0390625; CACHE: 169210.351953125 MB
2021-02-18:13:01:55,823 INFO     [_generate_run_sh.py:11] WORKERS=2; batch_size=1; cpu_count=12
2021-02-18:13:01:56,234 INFO     [_utils.py:129] AVAILABLE FREE MEMORY: 3384207.015625; CACHE: 169210.35078125002 MB
[2021-02-18 13:01:56 +0000] [65] [INFO] Starting gunicorn 20.0.4
[2021-02-18 13:01:56 +0000] [65] [INFO] Listening at: http://0.0.0.0:8080 (65)
[2021-02-18 13:01:56 +0000] [65] [INFO] Using worker: gevent
[2021-02-18 13:01:56 +0000] [69] [INFO] Booting worker with pid: 69
[2021-02-18 13:01:56 +0000] [70] [INFO] Booting worker with pid: 7

Server runs on : http://46.105.117.18:8080/ (which gives me a 404 error with direct access, which I believe is normal)

Then I try to POST data (2 php methods)

<?php
ini_set('display_errors', 1);error_reporting(E_ALL);
$url = "http://46.105.117.18:8080/";
$image1 = base64_encode(file_get_contents('https://www.rezocoquin.com/uploads/big/801590326639.jpg'));
$data = array(
    '1.png'=>$image1
);
$payload = json_encode($data);

$ch = curl_init($url);
curl_setopt($ch, CURLOPT_POSTFIELDS, $payload);
curl_setopt($ch, CURLOPT_HTTPHEADER, array('Content-Type:application/json'));
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
$result1 = curl_exec($ch);
print_r($result1);

$options = array(
    'http' => array(
        'method'  => 'POST',
        'content' => $payload,
        'header'=>  "Content-Type: application/json\r\n" .
            "Accept: application/json\r\n"
    )
);
$context  = stream_context_create( $options );
$result = file_get_contents( $url, false, $context );
$result2 = json_decode( $result );
print_r($result2);

method1: blank result method2: 404 error

Any clue/tip/help is welcome.

Regards

[bedapudi6788]

are you posting to /sync end point? Also, the json your are sending should have "data" key.

something like

$data = array('data' => array( '1.png'=>$image1 ););

[yarekc]

I just changed $url and set it to $url = "http://46.105.117.18:8080/sync/";

and got the same result

[bedapudi6788]

remove / at the end

[yarekc]

You are a GENIOUS ! May I suggest you to publish this php same ? (I am sure there will be lot of people interested in !) Regards

{"prediction": {"1.png": [{"box": [194, 335, 380, 495], "score": 0.8863019943237305, "label": "EXPOSED_BREAST_F"}, {"box": [64, 444, 253, 571], "score": 0.7764702439308167, "label": "EXPOSED_BELLY"}, {"box": [42, 220, 246, 426], "score": 0.6950249075889587, "label": "EXPOSED_BREAST_F"}]}, "success": true}stdClass Object ( [prediction] => stdClass Object ( [1.png] => Array ( [0] => stdClass Object ( [box] => Array ( [0] => 194 [1] => 335 [2] => 380 [3] => 495 ) [score] => 0.88630199432373 [label] => EXPOSED_BREAST_F ) [1] => stdClass Object ( [box] => Array ( [0] => 64 [1] => 444 [2] => 253 [3] => 571 ) [score] => 0.77647024393082 [label] => EXPOSED_BELLY ) [2] => stdClass Object ( [box] => Array ( [0] => 42 [1] => 220 [2] => 246 [3] => 426 ) [score] => 0.69502490758896 [label] => EXPOSED_BREAST_F ) ) ) [success] => 1 )

[bedapudi6788]

@yarekc can you post the full php snippet for this. I will link it in the readme.

[AlexUrrutia]

I'm using the following PHP script:

<?php
header("Content-Type: application/json");
$url = "http://localhost:8080/sync";
$image1 = base64_encode(file_get_contents('https://domain.com/16955673673220129769.jpg'));

    $data = array('data' => array('image'=>$image1));
    $payload = json_encode($data);
    $ch = curl_init($url);
    curl_setopt($ch, CURLOPT_POSTFIELDS, $payload);
    curl_setopt($ch, CURLOPT_HTTPHEADER, array('Content-Type:application/json'));
    curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
    $result1 = curl_exec($ch);
    print_r($result1);

?>

and all I get is this:

{
    "prediction": {
        "image": {
            "safe": 0.07022832334041595,
            "unsafe": 0.9297716617584229
        }
    },
    "success": true
}

Am I missing something?