-
So I've been playing around with moving preprocessing to CPU, in order to save GPU RAM on big models. I notice different behaviour of the CropMirrorNormalize operator between CPU & GPU modes. I've m…
-
I install DALI 0.2.0 from pip. I write some code similar to https://github.com/NVIDIA/DALI/blob/master/docs/examples/external_input.ipynb .
However, I get very strange behavior. The first batch wi…
-
I'm testing the `mxnet-resnet50.ipynb` example by converting it into a python script.
I use a record io file with no image resize, to preserve the image in original quality.
Then I got the follo…
-
Hi,
I used DALI in the purpose of improving the performance of my deep learning application .But it seems that I can't take the fully advantage of CPU core when I use DALI , which is feasible while I…
-
There are 2 reasons.
1) Our training environments are still CUDA 8.0 and it may take too much effort to upgrade all training tools to CUDA 9.0 until new machines are deployed.
2) Since image dec…
-
1、I have compiled libjpeg-turbo-1.5.9 and installed it at /home/guangyuan/install/libjpeg_turbo_1.5.9, but the import was not successful.
2、If I debug libjpeg-turbo when compiling dali from source co…
-
Hello!
I've noticed the comment in the headers of ```Workspace``` ( ```SampleWorkspace```, ```DeviceWorkspace``` etc), and corresponding ```OpType``` (CPU, GPU, etc).
My question is:
a) …
-
I'm trying to run libfreenect2 on an Nvidia Jetson TX1 Developer Kit and can't seem to get it working. Here's the output I'm getting:
```
$ LIBUSB_DEBUG=3 ./bin/Protonect
Version: 0.2.0
Environment …