CaffeSpark configuration
the layer index in the network protocol file
stop mark to indicate source is exhausted
make a dummy data blob to be used by Solver threads
make a dummy data blob to be used by Solver threads
a dummy data blob
make a dummy data blob to be used by Solver threads
make a dummy data blob to be used by Solver threads
a dummy data blob
construct a sample RDD
construct a sample RDD
spark context
RDD created from this source
create a batch of samples extracted from source queue
create a batch of samples extracted from source queue
This method is Invoked by Transformer thread. You should extract samples from source queue, parse it and produce a batch.
holder for sample Ids
holder for data blob
holder for label blob
true if successful
stop mark to indicate source is exhausted
batch size
CaffeSpark configuration
initialization of a Source within a process
initialization of a Source within a process
true if successfully initialized
the layer index in the network protocol file
adjust batch size
adjust batch size
the new batch size
Base class for various data sources.
Each subclass must have a constructor with the following signature: (conf: Config, layerId: Int, isTrain: Boolean). This is required by CaffeOnSpark at startup.
class of entries extracted from RDD
class of data blob in batch