Bugs with caffe-0.14-cnmem branch about caffe HOT 2 CLOSED

Also, when I try to integrate this with DIGITS I get a crash during inference (training is fine).

F1009 18:25:11.085157 18400 CuMem.cpp:30] Check failed: initialized_

This is when I'm using pycaffe like this [basically]:

import caffe
net = caffe.Net(deploy_file, caffemodel, caffe.TEST)
net.blobs['data'].data[0] = image_data
net.forward()

This type of invocation works for Caffe v0.11, v0.12, and v0.13.

from caffe.

Hi Soumith,

Pleas hold on testing that branch - there are some errors due to test inconsistencies that we are working on.
I will let you know once it's ready.

Best,
-Boris.

From: Soumith Chintala <[email protected]mailto:[email protected]>
Reply-To: drnikolaev/caffe <[email protected]mailto:[email protected]>
Date: Friday, October 9, 2015 at 6:33 PM
To: NVIDIA/caffe <[email protected]mailto:[email protected]>
Subject: Re: [caffe] Bugs with caffe-0.14-cnmem branch (#4)

Also, when I try to integrate this with DIGITS I get a crash during inference (training is fine).

F1009 18:25:11.085157 18400 CuMem.cpp:30] Check failed: initialized_

This is when I'm using pycaffe like this [basically]:

import caffe
net = caffe.Net(deploy_file, caffemodel, caffe.TEST)
net.blobs['data'].data[0] = image_data
net.forward()

This type of invocation works for Caffe v0.11, v0.12, and v0.13.

Reply to this email directly or view it on GitHubhttps://github.com//issues/4#issuecomment-147020858.

This email message is for the sole use of the intended recipient(s) and may contain
confidential information. Any unauthorized review, use, disclosure or distribution
is prohibited. If you are not the intended recipient, please contact the sender by

reply email and destroy all copies of the original message.

from caffe.