def forward(self, inputs):
x, lstm_state = inputs
# LSTM state consists of c and h.
c, h = jnp.split(lstm_state, 2, axis=-1)
# Dense layer on the concatenation of x and h.
w, b = self.weights
y = jnp.dot(jnp.concatenate([x, h], axis=-1), w) + b
# i = input_gate, j = new_input, f = forget_gate, o = output_gate
i, j, f, o = jnp.split(y, 4, axis=-1)
new_c = c * fastmath.sigmoid(f) + fastmath.sigmoid(i) * jnp.tanh(j)
new_h = jnp.tanh(new_c) * fastmath.sigmoid(o)
return new_h, jnp.concatenate([new_c, new_h], axis=-1)
def _shard_fn(x):
axis = _axis_to_shard_heuristic(x.shape)
if int(x.shape[axis]) % n_shards != 0:
raise ValueError(
f'Cannot split x with shape {x.shape} into {n_shards}.')
split_x = jnp.split(x, n_shards, axis=axis)
split_x = [split_x[i % n_shards] for i in indices]
return np.stack(split_x, axis=0)
def forward(self, inputs):
x, gru_state = inputs
# Dense layer on the concatenation of x and h.
w1, b1, w2, b2 = self.weights
y = jnp.dot(jnp.concatenate([x, gru_state], axis=-1), w1) + b1
# Update and reset gates.
u, r = jnp.split(fastmath.sigmoid(y), 2, axis=-1)
# Candidate.
c = jnp.dot(jnp.concatenate([x, r * gru_state], axis=-1), w2) + b2
new_gru_state = u * gru_state + (1 - u) * jnp.tanh(c)
return new_gru_state, new_gru_state
def forward(self, x):
"""Executes this layer as part of a forward pass through the model.
Args:
x: Tensor of same shape and dtype as the input signature used to
initialize this layer.
Returns:
Tensor of same shape and dtype as the input, except the final dimension
is the layer's `filters` value, and the second to last dimension is
shrinked if 'VALID' padding is used with kernel_size bigger than one.
"""
if self._use_bias:
if not isinstance(self.weights, (tuple, list)):
raise ValueError(f'Weights should be a (w, b) tuple or list; '
f'instead got: {self.weights}')
w, b = self.weights
else:
w = self.weights
linear_results_before_shifting = jnp.einsum('...lp,lkpd->...lkd', x, w)
# TODO(jaszczur): this could be run after padding for better efficiency
if self._kernel_size == 1:
# With kernel size 1 we don't have to split or shift anything.
linear_result = jnp.squeeze(linear_results_before_shifting,
axis=-2)
else:
# We computed a result for every "pixel", but each direction from the
# receptive field (there are 'self._kernel_size' such directions) must be
# shifted by a different amount. The easiest way to do it is to split
# the tensor to 'self._kernel_size' smaller tensors, shift each one
# appropriately, and then sum them together.
split_shifting_linear_results = jnp.split(
linear_results_before_shifting, self._kernel_size, axis=-2)
for i in range(self._kernel_size):
# Each tensor has to be shifted a different amount.
if self._padding == 'WRAP':
# We can shift by padding and cutting. With 'wrap' padding we
# essentially have a torus.
padding = [(0, 0)
for i in split_shifting_linear_results[i].shape]
padding[-3] = ((self._kernel_size - 1) - i, i)
split_shifting_linear_results[i] = jnp.pad(
split_shifting_linear_results[i], padding, mode='wrap')
split_shifting_linear_results[
i] = split_shifting_linear_results[i][
..., (self._kernel_size - 1) //
2:-(self._kernel_size - 1) // 2, :, :]
elif self._padding == 'SAME':
# We can shift by padding and cutting.
padding = [(0, 0)
for i in split_shifting_linear_results[i].shape]
padding[-3] = ((self._kernel_size - 1) - i, i)
split_shifting_linear_results[i] = jnp.pad(
split_shifting_linear_results[i], padding)
split_shifting_linear_results[
i] = split_shifting_linear_results[i][
..., (self._kernel_size - 1) //
2:-(self._kernel_size - 1) // 2, :, :]
# TODO(jaszczur): improve efficiency by not padding things to cut
elif self._padding == 'VALID':
# We don't need to shift - just cut the leftmost and rightmost values.
cut_left = (self._kernel_size - 1) - i
cut_right = split_shifting_linear_results[i].shape[-3] - i
split_shifting_linear_results[
i] = split_shifting_linear_results[i][
..., cut_left:cut_right, :, :]
else:
raise ValueError(f'Invalid padding {self._padding}')
# After shifting.
shifted_linear_results = jnp.concatenate(
split_shifting_linear_results, axis=-2)
linear_result = jnp.sum(shifted_linear_results, axis=-2)
if self._use_bias:
return linear_result + b
else:
return linear_result
def forward(self, inputs):
return tuple(jnp.split(inputs, self._n_items, self._axis))
def forward(self, inputs):
"""Executes this layer as part of a forward pass through the model."""
return tuple(jnp.split(inputs, self._n_items, self._axis))
def _unshard_fn(x):
y = jax.lax.all_gather(x, 'batch', axis_index_groups=groups)
split_y = jnp.split(y, n_shards, axis=0)
split_y = [jnp.squeeze(sy, axis=0) for sy in split_y]
axis = _axis_to_shard_heuristic(split_y[0].shape)
return jnp.concatenate(split_y, axis=axis)
def _f(x, axis=-1): # pylint: disable=invalid-name
size = x.shape[axis]
assert size % 2 == 0, f'axis {axis} of size {size} is not be divisible by 2'
a, b = jnp.split(x, 2, axis)
return a * fastmath.expit(b)
