tensorflow-0.1.0.0: TensorFlow bindings.

Safe HaskellNone
LanguageHaskell2010

TensorFlow.Session

Contents

Synopsis

Documentation

data Session a

Instances

Opaque value created via sessionConfig and sessionTarget.

data SessionOption

Setting of an option for the session (see runSessionWithOptions).

sessionConfig :: ConfigProto -> SessionOption

Uses the specified config for the created session.

sessionTarget :: ByteString -> SessionOption

Target can be: "local", ip:port, host:port. The set of supported factories depends on the linked in libraries. REQUIRES "/learningbrain/public:tensorflow_remote" dependency for the binary.

runSession :: Session a -> IO a

Run Session actions in a new TensorFlow session.

runSessionWithOptions :: [SessionOption] -> Session a -> IO a

Run Session actions in a new TensorFlow session created with the given option setter actions (sessionTarget, sessionConfig).

build :: Build a -> Session a

Lift a Build action into a Session, including any explicit op renderings.

buildAnd :: (a -> Session b) -> Build a -> Session b

Helper combinator for doing something with the result of a Build action. Example usage:

buildAnd run :: Fetchable t a => Build t -> Session a

buildWithSummary :: forall a. Build a -> Session (a, [SummaryTensor])

Lift a Build action into a Session, including any explicit op renderings. Returns the merged summary ops which can be used for logging, see build for a convenient wrapper.

extend :: Session ()

Add all pending rendered nodes to the TensorFlow graph and runs any pending initializers.

Note that run, runWithFeeds, etc. will all call this function implicitly.

addGraphDef :: GraphDef -> Build ()

run :: Fetchable t a => t -> Session a

Run a subgraph t, rendering any dependent nodes that aren't already rendered, and fetch the corresponding values for a.

runWithFeeds :: Fetchable t a => [Feed] -> t -> Session a

Run a subgraph t, rendering any dependent nodes that aren't already rendered, feed the given input values, and fetch the corresponding result values for a.

run_ :: Nodes t => t -> Session ()

Run a subgraph t, rendering and extending any dependent nodes that aren't already rendered. This behaves like run except that it doesn't do any fetches.

runWithFeeds_ :: Nodes t => [Feed] -> t -> Session ()

Run a subgraph t, rendering any dependent nodes that aren't already rendered, feed the given input values, and fetch the corresponding result values for a. This behaves like runWithFeeds except that it doesn't do any fetches.

asyncProdNodes

Arguments

:: Nodes t 
=> t

Node to evaluate concurrently.

-> Session () 

Starts a concurrent thread which evaluates the given Nodes forever until runSession exits or an exception occurs. Graph extension happens synchronously, but the resultant run proceeds as a separate thread.