I am new to docker and learning about it recently. When I was trying to use it for one of my assignments. I got the following error of which I did not find any good solution anywhere. I have properly installed nvidia-docker and build the image before trying out the following command as given in the link. The nvidia-docker version is:
ii nvidia-docker 1.0.1-1 amd64 NVIDIA Docker container tools
This is the error I got after running the above command.
Unable to find image 'retinanet:latest' locallydocker: Error response from
daemon: pull access denied for retinanet, repository does not exist or may
require 'docker login': denied: requested access to the resource is denied.
The following is what I got after building the image.
$ sudo docker build -t retinanet:latest retinanet/
Sending build context to Docker daemon 208.9kB
Step 1/3 : FROM nvcr.io/nvidia/pytorch:19.05-py3
---> 7e98758d4777
Step 2/3 : COPY . retinanet/
---> Using cache
---> a32277843b1f
Step 3/3 : RUN pip install --no-cache-dir -e retinanet/
---> Using cache
---> a0195cf77814
Successfully built a0195cf77814
Successfully tagged retinanet:latest
The following is the result I got after docker images.
$ sudo docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
<none> <none> 7cd7d8e7cedc 2 days ago
7.59GB
nvcr.io/nvidia/pytorch 19.09-py3 9d6f9ccfbe31 6 weeks ago 9.15GB
nvidia/cuda 9.0-base 1443caa429f9 7 weeks ago 137MB
retinanet latest a0195cf77814 7 weeks ago 7.59GB
nvcr.io/nvidia/pytorch 19.07-py3 71df86c191f8 3 months ago 8.32GB
nvcr.io/nvidia/pytorch 19.05-py3 7e98758d4777 5 months ago 7.55GB
When I use the command
docker run -ti retinanet:latest
The workspace is created but I get the following warnings which indicates that I am not utilizing the GPU facility.
WARNING: The NVIDIA Driver was not detected. GPU functionality will not be available.
Use 'nvidia-docker run' to start this container; see https://github.com/NVIDIA/nvidia-docker/wiki/nvidia-docker .
NOTE: MOFED driver for multi-node communication was not detected.
Multi-node communication performance may be reduced.
NOTE: The SHMEM allocation limit is set to the default of 64MB. This may be
insufficient for PyTorch. NVIDIA recommends the use of the following flags:
nvidia-docker run --ipc=host ...
Note: I followed the solution to this link but it is not helpful in my case. Any ideas and suggestions would be very helpful. Thanks in advance.