dmrgo is a Go library for writing map/reduce jobs.
It can be used with Hadoop's streaming protocol, but also includes a standalone map/reduce implementation (including partitioner) for 'small' jobs (~5G-10G).
It is partially based on ideas from Yelp's MrJob package for Python, but since the Go is statically typed I've tried to make the API match more closely with Hadoop's Java API.
The traditional "word count" example is in the examples directory.
This code is licensed under the GPLv3, or at your option any later version.
Hadoop map/reduce tutorial: http://hadoop.apache.org/common/docs/current/mapred_tutorial.html
Hadoop streaming protocol: http://hadoop.apache.org/common/docs/current/streaming.html