Tune in to the first Women in ML Symposium this Tuesday, October 19 at 9am PST Register now

BlockLSTMV2

public final class BlockLSTMV2

Computes the LSTM cell forward propagation for all the time steps.

This is equivalent to applying LSTMBlockCell in a loop, like so:

for x1 in unpack(x):
   i1, cs1, f1, o1, ci1, co1, h1 = LSTMBlock(
     x1, cs_prev, h_prev, w, wci, wcf, wco, b)
   cs_prev = cs1
   h_prev = h1
   i.append(i1)
   cs.append(cs1)
   f.append(f1)
   o.append(o1)
   ci.append(ci1)
   co.append(co1)
   h.append(h1)
 return pack(i), pack(cs), pack(f), pack(o), pack(ci), pack(ch), pack(h)
 
 Note that unlike LSTMBlockCell (and BlockLSTM) which uses ICFO gate layout,
 this op uses IFCO. So in order for the following snippet to be equivalent
 all gate-related outputs should be reordered.
 

Nested Classes

class BlockLSTMV2.Options Optional attributes for BlockLSTMV2  

Public Methods

static BlockLSTMV2.Options
cellClip(Float cellClip)
Output<T>
ci()
The cell input over the whole time sequence.
Output<T>
co()
The cell after the tanh over the whole time sequence.
static <T extends Number> BlockLSTMV2<T>
create(Scope scope, Operand<Long> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Options... options)
Factory method to create a class wrapping a new BlockLSTMV2 operation.
Output<T>
cs()
The cell state before the tanh over the whole time sequence.
Output<T>
f()
The forget gate over the whole time sequence.
Output<T>
h()
The output h vector over the whole time sequence.
Output<T>
i()
The input gate over the whole time sequence.
Output<T>
o()
The output gate over the whole time sequence.
static BlockLSTMV2.Options
usePeephole(Boolean usePeephole)

Inherited Methods

Public Methods

public static BlockLSTMV2.Options cellClip (Float cellClip)

Parameters
cellClip Value to clip the 'cs' value to.

public Output<T> ci ()

The cell input over the whole time sequence.

public Output<T> co ()

The cell after the tanh over the whole time sequence.

public static BlockLSTMV2<T> create (Scope scope, Operand<Long> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Options... options)

Factory method to create a class wrapping a new BlockLSTMV2 operation.

Parameters
scope current scope
seqLenMax Maximum time length actually used by this input. Outputs are padded with zeros beyond this length.
x The sequence input to the LSTM, shape (timelen, batch_size, num_inputs).
csPrev Value of the initial cell state.
hPrev Initial output of cell (to be used for peephole).
w The weight matrix.
wci The weight matrix for input gate peephole connection.
wcf The weight matrix for forget gate peephole connection.
wco The weight matrix for output gate peephole connection.
b The bias vector.
options carries optional attributes values
Returns
  • a new instance of BlockLSTMV2

public Output<T> cs ()

The cell state before the tanh over the whole time sequence.

public Output<T> f ()

The forget gate over the whole time sequence.

public Output<T> h ()

The output h vector over the whole time sequence.

public Output<T> i ()

The input gate over the whole time sequence.

public Output<T> o ()

The output gate over the whole time sequence.

public static BlockLSTMV2.Options usePeephole (Boolean usePeephole)

Parameters
usePeephole Whether to use peephole weights.