Struct tensorflow::Session
source · pub struct Session { /* private fields */ }
Expand description
Manages a single graph and execution.
Implementations§
source§impl Session
impl Session
sourcepub fn new(options: &SessionOptions, graph: &Graph) -> Result<Self>
pub fn new(options: &SessionOptions, graph: &Graph) -> Result<Self>
Creates a session.
graph
will be be kept alive for the lifetime of the returned session.
New nodes can still be added to graph
after this call.
sourcepub fn from_saved_model<P: AsRef<Path>, Tag: AsRef<str>, Tags: IntoIterator<Item = Tag>>(
options: &SessionOptions,
tags: Tags,
graph: &mut Graph,
export_dir: P
) -> Result<Self>
👎Deprecated since 0.17.0: Please use SavedModelBundle::load() instead
pub fn from_saved_model<P: AsRef<Path>, Tag: AsRef<str>, Tags: IntoIterator<Item = Tag>>( options: &SessionOptions, tags: Tags, graph: &mut Graph, export_dir: P ) -> Result<Self>
Loads a session from an exported model.
sourcepub fn run(&self, step: &mut SessionRunArgs<'_>) -> Result<()>
pub fn run(&self, step: &mut SessionRunArgs<'_>) -> Result<()>
Runs the graph, feeding the inputs and then fetching the outputs requested in the step. Note that the session has interior mutability; this may mutate variables in the graph, and the caller is responsible for handling race conditions.
sourcepub fn device_list(&self) -> Result<Vec<Device>>
pub fn device_list(&self) -> Result<Vec<Device>>
Lists all devices in a session.
Trait Implementations§
impl Send for Session
impl Sync for Session
Auto Trait Implementations§
Blanket Implementations§
source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere T: ?Sized,
source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more