Stack Overflow
Stack Overflow

Questions tagged [caffe2]

Caffe2 is an open-source deep learning framework, which is developed and maintained by Facebook, and is based on the Caffe framework.

Caffe2 is a deep learning framework, which is based on the framework. It is developed and maintained by Facebook, and released under the Apache 2.0 license, on GitHub.

Caffe2 is an improved / redesigned version of Caffe. A presentation by Caffe2's lead developer Yangqing Jia, held at CVPR 2015, shows some of the consideration of the redesign.

Useful links

116 questions
0
votes
1 answer

How to build caffe2 with specified cuda and cudnn?

In the server,there are cudnn 7.0 and cuda 9.1 at /usr/local/cuda. I need to use detectron but it do not support cuda 9.1, so I need to install cuda 8.0. I am using the server so I do not have root privilege.I tried to install cuda 8.0 and cudnn as…
qimeng wang
  • 83
  • 2
  • 6
0
votes
2 answers

build caffe2 with conda failed

I build the caffe2 with anaconda following the page. In the server with a single titanx, has cudnn7 and cuda9 but do not have nccl, so I download the nccl2 from nvidia and extract it to path/to/local/nccl2, and then edit the…
qimeng wang
  • 83
  • 2
  • 6
0
votes
0 answers

How to change the padding for semantic segmentation?

I am trying to run UNet on my data, which is grayscale images with 256x256 resolution. UNet is downsampling the image to 1-by-5-by-84-by-84 (5 is number of classes). and I am getting the following error: 0501 02:16:17.345309 2433 net.cpp:400] loss…
user630317
  • 31
  • 8
0
votes
1 answer

Caffe to Caffe2 Conversion

I'm trying to run a prediction on a model of caffe2, convert from caffe using the script included with caffe2. The script terminates correctly, giving me "init_net.pb" and "predict_net.pb". When I try to call the Predictor method run() I get the…
Gionata Benelli
  • 357
  • 1
  • 3
  • 20
0
votes
1 answer

Detectron: no module named c2

I am using Detectron from caffe2, but after running python2 tests/test_spatial_narrow_as_op.py I get the following error: no module named c2 I have checked the lib/utils and it has c2.py. I have also added the Detectron/lib folder to…
ycAlex110
  • 11
  • 7
0
votes
1 answer

Hdf5 error while installing the caffe on Ubuntu 16.04?

I am trying to install the caffe in Ubuntu 16.04. I followed the steps as mentioned in the docs. Everything is fine, but my Machine is installed with CUDA toolkit 9, but doc says ver 8 - Hopefully this is not a problem(I guess). But when I try to…
ajayramesh
  • 3,576
  • 8
  • 50
  • 75
0
votes
1 answer

Trying to use Caffe2 to add two blobs together that contain matrices

I'm trying to add the values of two blobs together. These blobs contain a matrix that is 2,2. workspace.FeedBlob("X", np.random.randn(2, 2).astype(np.float32)) workspace.FeedBlob("Y", np.random.randn(2, 2).astype(np.float32)) net =…
user2743
  • 1,423
  • 3
  • 22
  • 34
0
votes
1 answer

Using to Caffe2 to create a model that uses dropout but getting an error related to dropout code

I'm trying to create a model in Caffe2 that uses dropout. But I'm getting a model an error that refers to my code dropout code. def someModel(model, data): conv1 = brew.conv(model, data, 'conv1', dim_in=1, dim_out=20, kernel=5) conv_relu_1…
user2743
  • 1,423
  • 3
  • 22
  • 34
0
votes
1 answer

Got confused after I extracted weights from Trained caffenet

So basically this are the dimensions of the weights from trained caffenet: conv1: (96,3,11,11) conv2: (256,48,5,5) conv3:(384,256,3,3) conv4: (384,192,3,3) conv5:(256, 192, 3 , 3) I am confused that although conv1 gives 96 channels as output why…
sunjeet95
  • 13
  • 6
0
votes
1 answer

Onnx 'helper' and 'checker' attribute are not defined

I have onnx 0.2.1 installed on my conda virtual environment conda list | grep onnx packages in environment at /Users/aanirud/anaconda2/envs/onnx: onnx 0.2.1 py27_1 ezyang onnx-caffe2 0.2.1 py27hbe716ef_1 ezyang onnx-mxnet 0.4.1 But I am not able…
Anirudh
  • 52
  • 1
  • 7
0
votes
2 answers

Detectron Error during saving of training model file

I'm training Detectron with the following command python2 tools/train_net.py \ --cfg configs/getting_started/tutorial_1gpu_e2e_faster_rcnn_R-50-FPN.yaml \ OUTPUT_DIR /tmp/detectron-output and I'm getting the error below. Everything goes well…
anggagoko
  • 1
0
votes
1 answer

caffe2 - loss not working and accuracy crashes

I am doing ramp up to caffe2 and I try a toy problem: # Read the data from the database def read_input(model, batch_size, db, db_type): # load the data data, binary, dist = model.TensorProtosDBInput( [], ['data', 'binary'],…
Itamar
  • 21
  • 3
0
votes
0 answers

In a Anaconda virtual environment，what‘s the difference between "conda install package_name", "pip install package_name"?

The default PyThon environment in my linux is: Python 3.6.2 :: Anaconda custom (64-bit), I created the virtual environment: Python 2.7.14 :: Anaconda custom (64-bit) First, I enter the Aanconda virtual envirenment: source activate python27 Then, I…
S.Liu
  • 9
  • 1
0
votes
1 answer

How to change blob value in Caffe2 while using GPU (C++)

I am trying to manually access blobs value and change it. I have a blob called "1conv1_w" and I access it by: auto 1conv1_w = caffe2::TensorCPU((*workspace.GetBlob("1conv1_w")).Get()).data(); this will return a pointer…
吳順成
  • 13
  • 1
  • 4
0
votes
1 answer

Build CUDA library using Xcode and CMake fail

I am trying to use CMake to generate a Xcode project and build a CUDA library with it. The code I used for collecting & building CUDA library "caffe2_cpp_gpu" is as follows: list(APPEND CUDA_NVCC_FLAGS "-std=c++11" "-Wno-deprecated-gpu-targets") …
吳順成
  • 13
  • 1
  • 4
1 2 3 4 5 6
7
8