returnn.tf.frontend_layers._backend

High-level backend for RETURNN layers

class returnn.tf.frontend_layers._backend.ReturnnLayersBackend

RETURNN layers backend (using TF), where raw_tensor represents a RETURNN layer

name: str | None = 'returnn_layers_tf'

RawTensorType

alias of Layer

is_tensorflow: bool = True

is_backend_raw_tensor_dim_tag_independent: bool = False

static executing_eagerly() → bool

executing eagerly

static get_tensor_dependencies(x: Tensor[Layer]) → Sequence[Tensor]

get tensor inputs

static get_tensor_consumers(x: Tensor[Layer]) → Sequence[Tensor]

get tensor consumers

static cond(pred: Tensor, true_fn: Callable, false_fn: Callable)

static while_loop(cond: Callable[[S], bool | Tensor], body: Callable[[S], S], initial: S) → S

while loop

static set_random_seed(seed: int)

Parameters:

seed

static get_random_state() → Dict[str, bytes]

Returns:

random state

static set_random_state(state: Dict[str, bytes])

Parameters:

state – as returned by get_random_state(). This might not always be successful (e.g. different hardware, different backend version), so the calling code should always have called set_random_seed before to have the random generators in a reasonable fallback state.

static get_dtype_name_raw(raw_tensor: Layer) → str

dtype

static as_dtype_raw(dtype_name: str) → str

dtype

static get_ndim_raw(raw_tensor: Layer) → int

ndim

static get_shape_raw(raw_tensor: Layer) → Layer

shape

static get_shape_tuple_raw(raw_tensor: Layer) → Tuple[int | Layer, ...]

shape

static get_known_shape_raw(raw_tensor: Layer) → Tuple[int | None, ...]

known shape

static fill_raw(shape: Sequence[int | Layer] | Layer, value: Any | Layer) → Layer

fill raw

static compare_raw(a: Layer, kind: str, b: Layer) → Layer

compare

static combine_raw(a: Layer, kind: str, b: Layer) → Layer

combine

static reshape_raw(raw_tensor: Layer, shape: Sequence[int | Layer] | Layer) → Layer

static make_output_tensor(tensor: Tensor, dims: Sequence[Dim], *, name: str) → Tensor

only func where we have explicitly defined dim order in the output

static copy(tensor: Tensor) → Tensor

static cast(tensor: Tensor, dtype: str) → Tensor

static set_requires_gradient(tensor: Tensor)

set requires gradient; not needed for TensorFlow, will always calculate whatever is needed

static gradient(y: Tensor, x: Tensor) → Tensor

static stop_gradient(tensor: Tensor) → Tensor

stop grad

static scaled_gradient(tensor: Tensor, scale: float | Tensor) → Tensor

scaled gradient

static scaled_gradient_ext(x: Tensor, *, scale: float | Tensor = 1.0, shift: float | Tensor | None = None, scale_shift_by_sum_over_axis: Dim | None = None)

scaled gradient ext

static merge_dims(source: Tensor, *, dims: Sequence[Dim], out_dim: Dim) → Tensor

Merges a list of axes into a single one. (Flatten the dims.) E.g. input is (batch, width, height, dim) and dims=(width,height), then we get (batch, width*height, dim). Or input is (batch, time, height, dim) and axes=(height,dim), then we get (batch, time, height*dim).

Parameters:

• source

• dims

• out_dim

Returns:

tensor

static split_dims(source: Tensor, *, axis: Dim, dims: Sequence[Dim], pad_to_multiples: bool | None = None, pad_value: None | int | float = None) → Tensor

split dims

static reshape(source: Tensor, in_dims: Sequence[Dim], out_dims: Sequence[Dim]) → Tensor

static split(source: Tensor, *, axis: Dim, out_dims: Sequence[Dim]) → Tuple[Tensor, ...]

static expand_dim(source: Tensor, dim: Dim) → Tensor

expand dim

static squeeze(source: Tensor, axis: Dim) → Tensor

static concat(*sources: Tuple[Tensor, Dim], allow_broadcast: bool = False, out_dim: Dim) → Tensor

static pad(source: Tensor, *, axes: Sequence[Dim], padding: Sequence[Tuple[Dim | int | Tensor, Dim | int | Tensor]], out_dims: Sequence[Dim], handle_dynamic_dims: bool, mode: str = 'constant', value: int | float | complex | number | ndarray | bool | str | Tensor | None = None) → Tensor

static activation(tensor: Tensor, func: str) → Tensor

static activation_raw(raw_tensor: Layer, func: str) → Layer

activation

static safe_log(tensor: Tensor, *, eps: float) → Tensor

safe log

static softmax(tensor: Tensor, *, axis: Dim, use_mask: bool = True) → Tensor

static log_softmax(tensor: Tensor, *, axis: Dim, use_mask: bool = True) → Tensor

log softmax

static softmax_cross_entropy_with_logits(*, logits: Tensor, targets: Tensor, axis: Dim)

Efficient cross entropy.

Parameters:

• logits – target estimates given as inputs to softmax (i.e. unnormalized)

• targets – probabilities, i.e. normalized, can also be sparse

• axis – class labels dim over which softmax is computed

Returns:

cross entropy (same Dims as ‘logits’ but without ‘axis’)

static ctc_loss(*, logits: Tensor, logits_normalized: bool = False, targets: Tensor, input_spatial_dim: Dim, targets_spatial_dim: Dim, blank_index: int, max_approx: bool = False) → Tensor

CTC

static create_parameter_raw(tensor: Parameter, *, device: str | None = None) → Layer

create parameter

static set_parameter_initial_value(param: Parameter[Layer], value: None | Tensor | int | float | complex | number | ndarray | bool | str) → None

set parameter initial value

static set_parameter_trainable(param: Parameter, trainable: bool) → None

set parameter trainable

static parameter_assign(param: Parameter, value: Tensor, *, op: str = 'assign') → None

param assign

static convert_to_tensor(value: Tensor | Layer | int | float | complex | number | ndarray | bool | str, *, dims: Sequence[Dim], dtype: str, sparse_dim: Dim | None = None, feature_dim: Dim | None = None, device: str | None = None, name: str | None = None) → Tensor[Layer]

convert to tensor

static full(dims: Sequence[Dim], fill_value: int | float | complex | number | ndarray | bool | str | Tensor, *, dtype: str, device: str | None = None, sparse_dim: Dim | None = None, feature_dim: Dim | None = None) → Tensor

classmethod compare(a: Tensor | int | float | complex | number | ndarray | bool | str, kind: str, b: Tensor | int | float | complex | number | ndarray | bool | str, *, allow_broadcast_all_sources: bool | None = None, dim_order: Sequence[Dim] | None = None) → Tensor

classmethod combine(a: Tensor | int | float | complex | number | ndarray | bool | str, kind: str, b: Tensor | int | float | complex | number | ndarray | bool | str, *, allow_broadcast_all_sources: bool | None = None, dim_order: Sequence[Dim] | None = None) → Tensor

static gather(source: Tensor, *, indices: Tensor | int, axis: Dim, clip_to_valid: bool = False) → Tensor

static slice(source: Tensor, *, axis: Dim, start: int | Tensor | None = None, end: int | Tensor | None = None, step: int | Tensor | None = None, size: int | Tensor | Dim | None = None, out_dim: Dim) → Tensor

static flip_no_mask(source: Tensor, *, axis: Dim) → Tensor

flip

static where(cond: Tensor, true_: Tensor | int | float | complex | number | ndarray | bool | str, false_: Tensor | int | float | complex | number | ndarray | bool | str, *, allow_broadcast_all_sources: bool = False) → Tensor

static clip_by_value(x: Tensor, clip_value_min: Tensor | int | float | complex | number | ndarray | bool | str, clip_value_max: Tensor | int | float | complex | number | ndarray | bool | str, *, allow_broadcast_all_sources: bool = False) → Tensor

clip by value

static cumsum(source: Tensor, *, spatial_dim: Dim) → Tensor

static matmul(a: Tensor, b: Tensor, *, reduce: Dim | Sequence[Dim], use_mask: bool = True) → Tensor

static range_over_dim(dim: Dim, *, dtype: str | None = None, device: str | None = None) → Tensor

range over dim

static replace_dim(source: Tensor, *, in_dim: Dim, out_dim: Dim) → Tensor

Parameters:

• source

• in_dim

• out_dim

Returns:

source with in_dim replaced by out_dim.

static set_sparse_dim(source: Tensor, sparse_dim: Dim) → Tensor

set sparse dim

static reduce(source: Tensor, *, mode: str, axis: Dim | Sequence[Dim], use_mask: bool = True) → Tensor

Reduce

static top_k(source: Tensor, *, axis: Dim | Sequence[Dim], k: int | Tensor, k_dim: Dim | None = None, sorted: bool = True) → Tuple[Tensor, Tensor | Sequence[Tensor], Dim]

static random_journal_replay(journal: RandomJournal)

Replays the journal. At exit, the journal is cleared, and we check that we replayed everything.

static random(*, dims: Sequence[Dim], dtype: str, device: str | None = None, sparse_dim: Dim | None = None, feature_dim: Dim | None = None, distribution: str, mean: int | float | Tensor | None = None, stddev: int | float | Tensor | None = None, bound: int | float | Tensor | None = None, minval: int | float | Tensor | None = None, maxval: int | float | Tensor | None = None, seed: int | Sequence[int] | ndarray | None = None, algorithm: str | None = None, explicit_state: Tensor | None = None, auto_update_state: bool | None = None, static: bool | None = None, out: Tensor | None = None) → Tensor

static masked_select(tensor: Tensor, *, mask: Tensor, dims: Sequence[Dim], out_dim: Dim | None = None) → Tuple[Tensor, Dim]

Parameters:

• tensor

• mask

• dims – the order of the dims defines the format. those dims should be exactly the dims of the mask.

• out_dim

Returns:

tensor where all dims in mask/dims are removed and replaced by a new dim. the new dim is also returned. if mask==True for all elements, the returned tensor would be simply the flattened input tensor.

static batch_norm(source: Tensor, *, in_dim: Dim | Sequence[Dim], running_mean: Tensor, running_variance: Tensor, gamma: Tensor | None, beta: Tensor | None, epsilon: float, momentum: float, affine: bool, use_mask: bool) → Tensor

batch norm

static conv(source: Tensor, *, in_dim: Dim, out_dim: Dim, in_spatial_dims: Sequence[Dim], out_spatial_dims: Sequence[Dim] | None = None, filter: Tensor, filter_size: Sequence[Dim], padding: str | int | Sequence[int], strides: int | Sequence[int] | None = None, dilation_rate: int | Sequence[int] | None = None, groups: int | None = None, bias: Tensor | None = None) → Tuple[Tensor, Sequence[Dim]]

static transposed_conv(source: Tensor, *, in_dim: Dim, out_dim: Dim, in_spatial_dims: Sequence[Dim], out_spatial_dims: Sequence[Dim] | None = None, filter: Tensor, filter_size: Sequence[Dim], padding: str, remove_padding: Sequence[int] | int = 0, output_padding: Sequence[int | None] | int | None = None, strides: Sequence[int] | None = None, bias: Tensor | None = None) → Tuple[Tensor, Sequence[Dim]]

transposed conv

static pool(source: Tensor, *, mode: str, pool_size: Sequence[int], padding: str | int | Sequence[int] = 'valid', dilation_rate: Sequence[int] | int = 1, strides: Sequence[int], in_spatial_dims: Sequence[Dim], out_spatial_dims: Sequence[Dim] | None = None) → Tuple[Tensor, Sequence[Dim]]

static stft(x: Tensor, *, in_spatial_dim: Dim, frame_step: int, frame_length: int, fft_length: int, window_use_frame_length: bool = True, align_window_left: bool = True, window_enforce_even: bool = True, out_spatial_dim: Dim, out_dim: Dim) → Tensor

static lstm(source: Tensor, *, state_h: Tensor, state_c: Tensor, ff_weight: Tensor, rec_weight: Tensor, bias: Tensor, spatial_dim: Dim, in_dim: Dim, out_dim: Dim) → Tuple[Tensor, Tuple[Tensor, Tensor]]

Returns:

output, (h, c)