Although we'd like to have a bulk loader in the future, we currently don't support bulk loading in NoSQL DB. There are a couple of other options that might work for you.
In NoSQL DB, you could load data from a text file by scripting the data into put commands, and running the script with the Admin CLI load -file option. That's not a bulk load, but is one way of loading the text file in. Another option might be to use Hadoop in this path. In NoSQL DB R3, if you have used the table api, you would be able to insert data in JSON format.
We're always interested in getting input about requirements for features. For example, in your case, would a basic bulk load tool be sufficient, or do you need to do significant ETL/field mapping? Are simple formats like CSV sufficient? Are your source file sorted on the primary key? And are you familiar with Oracle Data Pump, and do you use any of the features provided by that tool?