Great question. I, (and other people from the Clojure community) think that the Clojure language has a lot to offer to the Deep Learning/ Machine Learning space.
Step 1 - Build bridges to other Deep Learning communities using Jupyter notebooks, etc.
Step 2 - Bring innovation. In this area I see Clojure particularly useful in joining the symbolic logic and the deep learning model realms.
Step 3 - TBD (Depending on how the various time lines shake out to the singularity ;) )
After a quick look at the library, there does seem to be some Java libraries being used - like java.io.BufferedReader. But having a ClojureScript port is a great idea!
Step 1 - Build bridges to other Deep Learning communities using Jupyter notebooks, etc.
Step 2 - Bring innovation. In this area I see Clojure particularly useful in joining the symbolic logic and the deep learning model realms.
Step 3 - TBD (Depending on how the various time lines shake out to the singularity ;) )