These are chat archives for beniz/deepdetect

25th
Jun 2017
roysG
@roysG
Jun 25 2017 07:02 UTC

I am trying to use the googlenet model but i got error

The Request:

curl -X PUT "http://192.168.1.25:8080/services/images1" -d '{
"mllib":"caffe",
"description":"image classification service",
"type":"supervised",
"parameters":{
"input":{
"connector":"image","width":"224","height":"224"
},
"mllib":{
"template":"googlenet",
"nclasses":1000
}
},
"model":{
"templates":"/home/roy/deepdetect/templates/caffe",
"repository":"/home/roy/deepdetect/templates/caffe/googlenet/"
}
}'

The response in client:
{"status":{"code":500,"msg":"InternalError","dd_code":1007,"dd_msg":"in get()"}}

The response in server:
ERROR - 02:59:03 - service creation call failed

Emmanuel Benazera
@beniz
Jun 25 2017 07:15 UTC
Height and width should be int
roysG
@roysG
Jun 25 2017 07:22 UTC
Ok i replaced it to integer
now i get the error:
{"status":{"code":500,"msg":"InternalError","dd_code":1007,"dd_msg":"CHECK failed: (index) < (size()): "}}
roysG
@roysG
Jun 25 2017 07:40 UTC
Upadte:

Now i success to create service with this configuration:

curl -X PUT "http://192.168.1.25:8080/services/images3" -d '{
"mllib":"caffe",
"description":"image classification service",
"type":"supervised",
"parameters":{
"input":{
"connector":"image","width":224,"height":224
},
"mllib":{
"nclasses":1000
}
},
"model":{
"repository":"/home/roy/models/xx"
}
}'

But when i do predict this is the result i get:

{"status":{"code":200,"msg":"OK"},"head":{"method":"/predict","service":"images3","time":2635.0},"body":{"predictions":[{"uri":"https://deepdetect.com/img/ambulance.jpg","classes":[{"prob":0.9928520321846008,"cat":"407"},{"prob":0.006932162679731846,"cat":"734"},{"prob":0.000069531706685666,"last":true,"cat":"654"}]}]}}

what is "407" and "734" ? why i do no get "ambulance" ?
Ok i figure it out
roysG
@roysG
Jun 25 2017 08:02 UTC
I want to improve the result i get, is there any way to do that?
roysG
@roysG
Jun 25 2017 10:16 UTC
I want to use in face detect according to this link:
I have problem to understand which and where are the changes that i need to do in deploy file
Emmanuel Benazera
@beniz
Jun 25 2017 16:34 UTC
It's in the FAQ