Afficher la pageAnciennes révisionsHaut de page Cette page est en lecture seule. Vous pouvez afficher le texte source, mais ne pourrez pas le modifier. Contactez votre administrateur si vous pensez qu'il s'agit d'une erreur. ====== Caffe ====== Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by Berkeley AI Research (BAIR) and by community contributors. ====== How to use Caffe ====== Caffe is available within [[singularity| Singularity container]] Please check how to [[caffe_singularity|build caffe (CPU only) container]] with singularity ===== GPU version ===== We will use **caffe-gpu.img** container image. Here is an example of SGE script that use caffe GPU version withing singularity: <code bash> #!/bin/bash #$ -o $JOB_NAME.o$JOB_ID #$ -N job_caffe i#$ -q tesla.q ## we request tesla.q #$ -l h_vmem=8G ## we request 8G ## load singularity module load tools/singularity ## run TF with singularity singularity exec --nv $SING_SPOOL/caffe-gpu.img caffe train -solver /opt/caffe/examples/mnist/lenet_solver.prototxt </code> ===== How to use caffe in interactive mode with qlogin ===== Open a terminal and connect to GPU node with qlogin: <code bash> $ qlogin -q tesla.q -l h_vmem=8G node1-50:~$ module load tools/singularity node1-50:~$ singularity shell --nv $SING_SPOOL/caffe-gpu.img singularity:~> caffe train -solver /opt/caffe/examples/mnist/lenet_solver.prototxt </code> caffe.txt Dernière modification : 2017/12/18 10:07de admin