varghese

I would like to run a fully convolution neural network with SpatialCrossEntropy as criteria. Does a cpu version of the same exists? require 'cudnn' require 'cunn' require 'nn' model=nn.Sequential() model:add(nn.SpatialConvolutionMM(3,2,3,3,1,1,1,1))...

Possibility to share the code for retraining

2

Hi; Thank you for sharing your reimplementation. Is it possible to share the code which you made use for training the network.

Hello, I was wondering if it is necessary to have dataset on device. For example to train a model for a resource constrained device (Say Tx2 or mobile phone), does...

Hello Robert; I would like to change the number of the blocks from [3,4,8,6] to [a,b,c,d]. Then I would like to use this model for training peele. Could let me...

Hi; Is it required that the targets and labels need to be one hot encoded?

{ "error": "Malformed request: GET /v1/models/mask:predict" }

1

when I use curl -X POST http://localhost:8501/v1/models/123456:predict { "error": "JSON Parse error: The document is empty" } any directions or pointers please why this is coming

I have about 10,00,000 t7 files as training data and 3,00,000 t7 files as validation. I am pretty impressed with the training speed however the testing seems extremely slow. Could...

Inflating pre-trained models

1

Hello All, Upon reading the paper, I see that the authors have repeated the same image numerous time to create a "boring" video. What I fail to understand is the...

Installation error

1

Hello, I am facing the following error when I run sh build.sh [ 96%] Building CUDA object CMakeFiles/dark.dir/src/im2col_kernels.cu.o /mnt/darknet.CG/src/im2col_kernels.cu:125:18: warning: "/*" within comment [-Wcomment] //*data_col_ptr = (h >= 0 &&...

Object detection

1

Hello; Were you able to detect bounding boxes ?