Have you tried Sacred[1]? It definitely doesn't answer the "infrastructure for deep learning" challenge but it is helpful for understanding what experiments have been run/where did this model come from (including what version of the code/parameters produced it)
It's great to see people talking about the infrastructure they use to manage their deep learning workloads.
One area where we've had trouble with other orchestration tools (e.g. Docker Swarm) was in managing resources at anything beyond whole boxes. They are all good at managing CPU/RAM/Disk but we've had trouble with give this task GPU2. We had planned to try Mesos (given that we already run it for other things) but it sounds like maybe we should take a harder look at Kubernetes first.
It's not quite the same (since it doesn't become a Map-Reduce job) but if you're mostly interested in the programming paradigm/scalability the Python API for Apache Spark might be a good alternative
There are actually dockerized versions of many (maybe all) of the deep learning libraries. The docker containers can take advantage of the GPU for training. You still have to install CUDA on the box (outside the docker container) but then you can try out different deep learning libraries.