PaddleSpeech/paddlespeech/s2t/models/wav2vec2/modules/modeling_wav2vec2.py

# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
# Copyright 2021 The Fairseq Authors and the HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
""" Paddle Wav2Vec2 model."""
from dataclasses import dataclass
from typing import Optional
from typing import Tuple
from typing import Union

import numpy as np
import paddle
from paddle import nn

from paddlespeech.s2t.models.wav2vec2.modules.activations import ACT2FN
from paddlespeech.s2t.models.wav2vec2.modules.modeling_outputs import BaseModelOutput
from paddlespeech.s2t.models.wav2vec2.modules.modeling_outputs import ModelOutput
from paddlespeech.s2t.models.wav2vec2.modules.modeling_outputs import Wav2Vec2BaseModelOutput
from paddlespeech.s2t.utils.log import Log
logger = Log(__name__).getlog()


@dataclass
class Wav2Vec2ForPreTrainingOutput(ModelOutput):
    """
    Output type of [`Wav2Vec2ForPreTraining`], with potential hidden states and attentions.

    Args:
        loss (*optional*, returned when `sample_negative_indices` are passed, `paddle.Tensor` of shape `(1,)`):
            Total loss as the sum of the contrastive loss (L_m) and the diversity loss (L_d) as stated in the [official
            paper](https://arxiv.org/pdf/2006.11477.pdf) . (classification) loss.
        projected_states (`paddle.Tensor` of shape `(batch_size, sequence_length, config.proj_codevector_dim)`):
            Hidden-states of the model projected to *config.proj_codevector_dim* that can be used to predict the masked
            projected quantized states.
        projected_quantized_states (`paddle.Tensor` of shape `(batch_size, sequence_length, config.proj_codevector_dim)`):
            Quantized extracted feature vectors projected to *config.proj_codevector_dim* representing the positive
            target vectors for contrastive loss.
        hidden_states (`tuple(paddle.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `paddle.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the initial embedding outputs.
        attentions (`tuple(paddle.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `paddle.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
        contrastive_loss (*optional*, returned when `sample_negative_indices` are passed, `paddle.Tensor` of shape `(1,)`):
            The contrastive loss (L_m) as stated in the [official paper](https://arxiv.org/pdf/2006.11477.pdf) .
        diversity_loss (*optional*, returned when `sample_negative_indices` are passed, `paddle.Tensor` of shape `(1,)`):
            The diversity loss (L_d) as stated in the [official paper](https://arxiv.org/pdf/2006.11477.pdf) .
    """

    loss: Optional[paddle.Tensor] = None
    projected_states: paddle.Tensor = None
    projected_quantized_states: paddle.Tensor = None
    codevector_perplexity: paddle.Tensor = None
    hidden_states: Optional[Tuple[paddle.Tensor]] = None
    attentions: Optional[Tuple[paddle.Tensor]] = None
    contrastive_loss: Optional[paddle.Tensor] = None
    diversity_loss: Optional[paddle.Tensor] = None


def _compute_mask_indices(
        shape: Tuple[int, int],
        mask_prob: float,
        mask_length: int,
        attention_mask: Optional[paddle.Tensor]=None,
        min_masks: int=0, ) -> np.ndarray:
    """
    Computes random mask spans for a given shape. Used to implement [SpecAugment: A Simple Data Augmentation Method for
    ASR](https://arxiv.org/abs/1904.08779). Note that this method is not optimized to run on TPU and should be run on
    CPU as part of the preprocessing during training.

    Args:
        shape: The shape for which to compute masks. This should be of a tuple of size 2 where
               the first element is the batch size and the second element is the length of the axis to span.
        mask_prob:  The percentage of the whole axis (between 0 and 1) which will be masked. The number of
                    independently generated mask spans of length `mask_length` is computed by
                    `mask_prob*shape[1]/mask_length`. Note that due to overlaps, `mask_prob` is an upper bound and the
                    actual percentage will be smaller.
        mask_length: size of the mask
        min_masks: minimum number of masked spans
        attention_mask: A (right-padded) attention mask which independently shortens the feature axis of
                        each batch dimension.
    """
    batch_size, sequence_length = shape

    if mask_length < 1:
        raise ValueError("`mask_length` has to be bigger than 0.")

    if mask_length > sequence_length:
        raise ValueError(
            f"`mask_length` has to be smaller than `sequence_length`, but got `mask_length`: {mask_length}"
            f" and `sequence_length`: {sequence_length}`")

    # epsilon is used for probabilistic rounding
    epsilon = np.random.rand(1).item()

    def compute_num_masked_span(input_length):
        """Given input length, compute how many spans should be masked"""
        num_masked_span = int(mask_prob * input_length / mask_length + epsilon)
        num_masked_span = max(num_masked_span, min_masks)

        # make sure num masked span <= sequence_length
        if num_masked_span * mask_length > sequence_length:
            num_masked_span = sequence_length // mask_length

        # make sure num_masked span is also <= input_length - (mask_length - 1)
        if input_length - (mask_length - 1) < num_masked_span:
            num_masked_span = max(input_length - (mask_length - 1), 0)

        return num_masked_span

    # compute number of masked spans in batch
    input_lengths = (attention_mask.sum(-1).detach().tolist()
                     if attention_mask is not None else
                     [sequence_length for _ in range(batch_size)])

    # SpecAugment mask to fill
    spec_aug_mask = np.zeros((batch_size, sequence_length), dtype=np.bool)
    spec_aug_mask_idxs = []

    max_num_masked_span = compute_num_masked_span(sequence_length)

    if max_num_masked_span == 0:
        return spec_aug_mask

    for input_length in input_lengths:
        # compute num of masked spans for this input
        num_masked_span = compute_num_masked_span(input_length)

        # get random indices to mask
        spec_aug_mask_idx = np.random.choice(
            np.arange(input_length - (mask_length - 1)),
            num_masked_span,
            replace=False)

        # pick first sampled index that will serve as a dummy index to pad vector
        # to ensure same dimension for all batches due to probabilistic rounding
        # Picking first sample just pads those vectors twice.
        if len(spec_aug_mask_idx) == 0:
            # this case can only happen if `input_length` is strictly smaller then
            # `sequence_length` in which case the last token has to be a padding
            # token which we can use as a dummy mask id
            dummy_mask_idx = sequence_length - 1
        else:
            dummy_mask_idx = spec_aug_mask_idx[0]

        spec_aug_mask_idx = np.concatenate([
            spec_aug_mask_idx,
            np.ones(max_num_masked_span - num_masked_span, dtype=np.int32) *
            dummy_mask_idx
        ])
        spec_aug_mask_idxs.append(spec_aug_mask_idx)

    spec_aug_mask_idxs = np.array(spec_aug_mask_idxs)

    # expand masked indices to masked spans
    spec_aug_mask_idxs = np.broadcast_to(
        spec_aug_mask_idxs[:, :, None],
        (batch_size, max_num_masked_span, mask_length))
    spec_aug_mask_idxs = spec_aug_mask_idxs.reshape(
        (batch_size, max_num_masked_span * mask_length))

    # add offset to the starting indexes so that indexes now create a span
    offsets = np.arange(mask_length)[None, None, :]
    offsets = np.broadcast_to(offsets, (
        batch_size, max_num_masked_span, mask_length)).reshape(
            (batch_size, max_num_masked_span * mask_length))
    spec_aug_mask_idxs = spec_aug_mask_idxs + offsets

    # ensure that we cannot have indices larger than sequence_length
    if spec_aug_mask_idxs.max() > sequence_length - 1:
        spec_aug_mask_idxs[spec_aug_mask_idxs >
                           sequence_length - 1] = sequence_length - 1

    # scatter indices to mask
    np.put_along_axis(spec_aug_mask, spec_aug_mask_idxs, 1, -1)

    return spec_aug_mask


def _sample_negative_indices(features_shape: Tuple,
                             num_negatives: int,
                             mask_time_indices: Optional[np.ndarray]=None):
    """
    Sample `num_negatives` vectors from feature vectors.
    """
    batch_size, sequence_length = features_shape

    # generate indices of the positive vectors themselves, repeat them `num_negatives` times
    sequence_length_range = np.arange(sequence_length)

    # get `num_negatives` random vector indices from the same utterance
    sampled_negative_indices = np.zeros(
        shape=(batch_size, sequence_length, num_negatives), dtype=np.int32)

    mask_time_indices = (mask_time_indices.astype(np.bool)
                         if mask_time_indices is not None else
                         np.ones(features_shape, dtype=np.bool))

    for batch_idx in range(batch_size):
        high = mask_time_indices[batch_idx].sum() - 1
        mapped_masked_indices = sequence_length_range[mask_time_indices[
            batch_idx]]

        feature_indices = np.broadcast_to(
            np.arange(high + 1)[:, None], (high + 1, num_negatives))
        sampled_indices = np.random.randint(
            0, high, size=(high + 1, num_negatives))
        # avoid sampling the same positive vector, but keep the distribution uniform
        sampled_indices[sampled_indices >= feature_indices] += 1

        # remap to actual indices
        sampled_negative_indices[batch_idx][mask_time_indices[
            batch_idx]] = mapped_masked_indices[sampled_indices]

        # correct for batch size
        sampled_negative_indices[batch_idx] += batch_idx * sequence_length

    return sampled_negative_indices


class Wav2Vec2NoLayerNormConvLayer(nn.Layer):
    def __init__(self, config, layer_id=0):
        super().__init__()
        self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1
        self.out_conv_dim = config.conv_dim[layer_id]

        self.conv = nn.Conv1D(
            self.in_conv_dim,
            self.out_conv_dim,
            kernel_size=config.conv_kernel[layer_id],
            stride=config.conv_stride[layer_id],
            bias_attr=config.conv_bias, )
        self.activation = ACT2FN[config.feat_extract_activation]

    def forward(self, hidden_states):
        hidden_states = self.conv(hidden_states)
        hidden_states = self.activation(hidden_states)
        return hidden_states


class Wav2Vec2LayerNormConvLayer(nn.Layer):
    def __init__(self, config, layer_id=0):
        super().__init__()
        self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1
        self.out_conv_dim = config.conv_dim[layer_id]

        self.conv = nn.Conv1D(
            self.in_conv_dim,
            self.out_conv_dim,
            kernel_size=config.conv_kernel[layer_id],
            stride=config.conv_stride[layer_id],
            bias_attr=config.conv_bias, )
        self.layer_norm = nn.LayerNorm(self.out_conv_dim)
        self.activation = ACT2FN[config.feat_extract_activation]

    def forward(self, hidden_states):
        hidden_states = self.conv(hidden_states)
        hidden_states = hidden_states.transpose([0, 2, 1])
        hidden_states = self.layer_norm(hidden_states)
        hidden_states = hidden_states.transpose([0, 2, 1])

        hidden_states = self.activation(hidden_states)
        return hidden_states


class Wav2Vec2GroupNormConvLayer(nn.Layer):
    def __init__(self, config, layer_id=0):
        super().__init__()
        self.in_conv_dim = config.conv_dim[layer_id - 1] if layer_id > 0 else 1
        self.out_conv_dim = config.conv_dim[layer_id]

        self.conv = nn.Conv1D(
            self.in_conv_dim,
            self.out_conv_dim,
            kernel_size=config.conv_kernel[layer_id],
            stride=config.conv_stride[layer_id],
            bias_attr=config.conv_bias, )
        self.activation = ACT2FN[config.feat_extract_activation]

        self.layer_norm = nn.GroupNorm(
            num_groups=self.out_conv_dim, num_channels=self.out_conv_dim)

    def forward(self, hidden_states):
        hidden_states = self.conv(hidden_states)
        hidden_states = self.layer_norm(hidden_states)
        hidden_states = self.activation(hidden_states)
        return hidden_states


class Wav2Vec2PositionalConvEmbedding(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.conv = nn.Conv1D(
            config.hidden_size,
            config.hidden_size,
            kernel_size=config.num_conv_pos_embeddings,
            padding=config.num_conv_pos_embeddings // 2,
            groups=config.num_conv_pos_embedding_groups, )

        self.conv = nn.utils.weight_norm(self.conv, name="weight", dim=2)

        self.padding = Wav2Vec2SamePadLayer(config.num_conv_pos_embeddings)
        self.activation = ACT2FN[config.feat_extract_activation]

    def forward(self, hidden_states):
        hidden_states = hidden_states.transpose([0, 2, 1])

        hidden_states = self.conv(hidden_states)
        hidden_states = self.padding(hidden_states)
        hidden_states = self.activation(hidden_states)

        hidden_states = hidden_states.transpose([0, 2, 1])
        return hidden_states


class Wav2Vec2SamePadLayer(nn.Layer):
    def __init__(self, num_conv_pos_embeddings):
        super().__init__()
        self.num_pad_remove = 1 if num_conv_pos_embeddings % 2 == 0 else 0

    def forward(self, hidden_states):
        if self.num_pad_remove > 0:
            hidden_states = hidden_states[:, :, :-self.num_pad_remove]
        return hidden_states


class Wav2Vec2FeatureEncoder(nn.Layer):
    """Construct the features from raw audio waveform"""

    def __init__(self, config):
        super().__init__()

        if config.feat_extract_norm == "group":
            conv_layers = [Wav2Vec2GroupNormConvLayer(config, layer_id=0)] + [
                Wav2Vec2NoLayerNormConvLayer(config, layer_id=i + 1)
                for i in range(config.num_feat_extract_layers - 1)
            ]
        elif config.feat_extract_norm == "layer":
            conv_layers = [
                Wav2Vec2LayerNormConvLayer(config, layer_id=i)
                for i in range(config.num_feat_extract_layers)
            ]
        else:
            raise ValueError(
                f"`config.feat_extract_norm` is {config.feat_extract_norm}, but has to be one of ['group', 'layer']"
            )
        self.conv_layers = nn.LayerList(conv_layers)
        self.gradient_checkpointing = False

    def _freeze_parameters(self):
        for param in self.parameters():
            param.trainable = False

    def forward(self, input_values):
        hidden_states = input_values[:, None]
        for conv_layer in self.conv_layers:
            hidden_states = conv_layer(hidden_states)

        return hidden_states


class Wav2Vec2FeatureProjection(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.layer_norm = nn.LayerNorm(
            config.conv_dim[-1], epsilon=config.layer_norm_eps)
        self.projection = nn.Linear(config.conv_dim[-1], config.hidden_size)
        self.dropout = nn.Dropout(config.feat_proj_dropout)

    def forward(self, hidden_states):
        # non-projected hidden states are needed for quantization
        norm_hidden_states = self.layer_norm(hidden_states)
        hidden_states = self.projection(norm_hidden_states)
        hidden_states = self.dropout(hidden_states)
        return hidden_states, norm_hidden_states


# Copied from transformers.models.bart.modeling_bart.BartAttention with Bart->Wav2Vec2
class Wav2Vec2Attention(nn.Layer):
    """Multi-headed attention from 'Attention Is All You Need' paper"""

    def __init__(
            self,
            embed_dim: int,
            num_heads: int,
            dropout: float=0.0,
            is_decoder: bool=False,
            bias: bool=True, ):
        super().__init__()
        self.embed_dim = embed_dim
        self.num_heads = num_heads
        self.dropout = dropout
        self.head_dim = embed_dim // num_heads

        if (self.head_dim * num_heads) != self.embed_dim:
            raise ValueError(
                f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim}"
                f" and `num_heads`: {num_heads}).")
        self.scaling = self.head_dim**-0.5
        self.is_decoder = is_decoder

        self.k_proj = nn.Linear(embed_dim, embed_dim, bias_attr=bias)
        self.v_proj = nn.Linear(embed_dim, embed_dim, bias_attr=bias)
        self.q_proj = nn.Linear(embed_dim, embed_dim, bias_attr=bias)
        self.out_proj = nn.Linear(embed_dim, embed_dim, bias_attr=bias)

    def _shape(self, tensor: paddle.Tensor, seq_len: int, bsz: int):
        return paddle.reshape(tensor, (bsz, seq_len, self.num_heads,
                                       self.head_dim)).transpose([0, 2, 1, 3])

    def forward(
            self,
            hidden_states: paddle.Tensor,
            key_value_states: Optional[paddle.Tensor]=None,
            past_key_value: Optional[Tuple[paddle.Tensor]]=None,
            attention_mask: Optional[paddle.Tensor]=None,
            layer_head_mask: Optional[paddle.Tensor]=None,
            output_attentions: bool=False, ) -> Tuple[paddle.Tensor, Optional[
                paddle.Tensor], Optional[Tuple[paddle.Tensor]]]:
        """Input shape: Batch x Time x Channel"""

        # if key_value_states are provided this layer is used as a cross-attention layer
        # for the decoder
        is_cross_attention = key_value_states is not None

        bsz, tgt_len, _ = hidden_states.shape

        # get query proj
        query_states = self.q_proj(hidden_states) * self.scaling
        # get key, value proj
        if is_cross_attention and past_key_value is not None:
            # reuse k,v, cross_attentions
            key_states = past_key_value[0]
            value_states = past_key_value[1]
        elif is_cross_attention:
            # cross_attentions
            key_states = self._shape(self.k_proj(key_value_states), -1, bsz)
            value_states = self._shape(self.v_proj(key_value_states), -1, bsz)
        elif past_key_value is not None:
            # reuse k, v, self_attention
            key_states = self._shape(self.k_proj(hidden_states), -1, bsz)
            value_states = self._shape(self.v_proj(hidden_states), -1, bsz)
            key_states = paddle.concat([past_key_value[0], key_states], axis=2)
            value_states = paddle.concat(
                [past_key_value[1], value_states], axis=2)
        else:
            # self_attention
            key_states = self._shape(self.k_proj(hidden_states), -1, bsz)
            value_states = self._shape(self.v_proj(hidden_states), -1, bsz)

        if self.is_decoder:
            # if cross_attention save Tuple(paddle.Tensor, paddle.Tensor) of all cross attention key/value_states.
            # Further calls to cross_attention layer can then reuse all cross-attention
            # key/value_states (first "if" case)
            # if uni-directional self-attention (decoder) save Tuple(paddle.Tensor, paddle.Tensor) of
            # all previous decoder key/value_states. Further calls to uni-directional self-attention
            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
            # if encoder bi-directional self-attention `past_key_value` is always `None`
            past_key_value = (key_states, value_states)

        proj_shape = (bsz * self.num_heads, -1, self.head_dim)
        query_states = self._shape(query_states, tgt_len,
                                   bsz).reshape(proj_shape)
        key_states = key_states.reshape(proj_shape)
        value_states = value_states.reshape(proj_shape)

        src_len = key_states.shape[1]
        attn_weights = paddle.bmm(query_states, key_states.transpose([0, 2, 1]))

        if attn_weights.shape != [bsz * self.num_heads, tgt_len, src_len]:
            raise ValueError(
                f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is"
                f" {attn_weights.shape}")

        if attention_mask is not None:
            if attention_mask.shape != [bsz, 1, tgt_len, src_len]:
                raise ValueError(
                    f"Attention mask should be of size {[bsz, 1, tgt_len, src_len]}, but is {attention_mask.shape}"
                )
            attn_weights = attn_weights.reshape(bsz, self.num_heads, tgt_len,
                                                src_len) + attention_mask
            attn_weights = attn_weights.reshape(bsz * self.num_heads, tgt_len,
                                                src_len)

        attn_weights = nn.functional.softmax(attn_weights, axis=-1)

        if layer_head_mask is not None:
            if layer_head_mask.shape != [
                    self.num_heads,
            ]:
                raise ValueError(
                    f"Head mask for a single layer should be of size {[self.num_heads,]}, but is"
                    f" {layer_head_mask.shape}")
            attn_weights = layer_head_mask.reshape(
                (1, -1, 1, 1)) * attn_weights.reshape(
                    (bsz, self.num_heads, tgt_len, src_len))
            attn_weights = attn_weights.reshape(
                (bsz * self.num_heads, tgt_len, src_len))

        if output_attentions:
            # this operation is a bit awkward, but it's required to
            # make sure that attn_weights keeps its gradient.
            # In order to do so, attn_weights have to be reshaped
            # twice and have to be reused in the following
            attn_weights_reshaped = attn_weights.reshape(
                (bsz, self.num_heads, tgt_len, src_len))
            attn_weights = attn_weights_reshaped.reshape(
                (bsz * self.num_heads, tgt_len, src_len))
        else:
            attn_weights_reshaped = None

        attn_probs = nn.functional.dropout(
            attn_weights, p=self.dropout, training=self.training)

        attn_output = paddle.bmm(attn_probs, value_states)

        if attn_output.shape != [bsz * self.num_heads, tgt_len, self.head_dim]:
            raise ValueError(
                f"`attn_output` should be of size {[bsz, self.num_heads, tgt_len, self.head_dim]}, but is"
                f" {attn_output.shape}")

        attn_output = attn_output.reshape(
            (bsz, self.num_heads, tgt_len, self.head_dim))
        attn_output = attn_output.transpose([0, 2, 1, 3])

        # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be
        # partitioned aross GPUs when using tensor-parallelism.
        attn_output = attn_output.reshape((bsz, tgt_len, self.embed_dim))

        attn_output = self.out_proj(attn_output)

        return attn_output, attn_weights_reshaped, past_key_value


class Wav2Vec2FeedForward(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.intermediate_dropout = nn.Dropout(config.activation_dropout)

        self.intermediate_dense = nn.Linear(config.hidden_size,
                                            config.intermediate_size)
        if isinstance(config.hidden_act, str):
            self.intermediate_act_fn = ACT2FN[config.hidden_act]
        else:
            self.intermediate_act_fn = config.hidden_act

        self.output_dense = nn.Linear(config.intermediate_size,
                                      config.hidden_size)
        self.output_dropout = nn.Dropout(config.hidden_dropout)

    def forward(self, hidden_states):
        hidden_states = self.intermediate_dense(hidden_states)
        hidden_states = self.intermediate_act_fn(hidden_states)
        hidden_states = self.intermediate_dropout(hidden_states)

        hidden_states = self.output_dense(hidden_states)
        hidden_states = self.output_dropout(hidden_states)
        return hidden_states


class Wav2Vec2EncoderLayer(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.attention = Wav2Vec2Attention(
            embed_dim=config.hidden_size,
            num_heads=config.num_attention_heads,
            dropout=config.attention_dropout,
            is_decoder=False, )
        self.dropout = nn.Dropout(config.hidden_dropout)
        self.layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)
        self.feed_forward = Wav2Vec2FeedForward(config)
        self.final_layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)

    def forward(self,
                hidden_states,
                attention_mask=None,
                output_attentions=False):
        attn_residual = hidden_states
        hidden_states, attn_weights, _ = self.attention(
            hidden_states,
            attention_mask=attention_mask,
            output_attentions=output_attentions)
        hidden_states = self.dropout(hidden_states)
        hidden_states = attn_residual + hidden_states

        hidden_states = self.layer_norm(hidden_states)
        hidden_states = hidden_states + self.feed_forward(hidden_states)
        hidden_states = self.final_layer_norm(hidden_states)

        outputs = (hidden_states, )

        if output_attentions:
            outputs += (attn_weights, )

        return outputs


class Wav2Vec2EncoderLayerStableLayerNorm(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.attention = Wav2Vec2Attention(
            embed_dim=config.hidden_size,
            num_heads=config.num_attention_heads,
            dropout=config.attention_dropout,
            is_decoder=False, )
        self.dropout = nn.Dropout(config.hidden_dropout)
        self.layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)
        self.feed_forward = Wav2Vec2FeedForward(config)
        self.final_layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)

    def forward(self,
                hidden_states,
                attention_mask=None,
                output_attentions=False):
        attn_residual = hidden_states
        hidden_states = self.layer_norm(hidden_states)
        hidden_states, attn_weights, _ = self.attention(
            hidden_states,
            attention_mask=attention_mask,
            output_attentions=output_attentions)
        hidden_states = self.dropout(hidden_states)
        hidden_states = attn_residual + hidden_states
        hidden_states = hidden_states + self.feed_forward(
            self.final_layer_norm(hidden_states))

        outputs = (hidden_states, )

        if output_attentions:
            outputs += (attn_weights, )

        return outputs


class Wav2Vec2Encoder(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.config = config
        self.pos_conv_embed = Wav2Vec2PositionalConvEmbedding(config)
        self.layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)
        self.dropout = nn.Dropout(config.hidden_dropout)
        self.layers = nn.LayerList([
            Wav2Vec2EncoderLayer(config)
            for _ in range(config.num_hidden_layers)
        ])
        self.gradient_checkpointing = False

    def forward(
            self,
            hidden_states,
            attention_mask=None,
            output_attentions=False,
            output_hidden_states=False,
            return_dict=True, ):
        all_hidden_states = () if output_hidden_states else None
        all_self_attentions = () if output_attentions else None

        if attention_mask is not None:
            # make sure padded tokens output 0
            expand_attention_mask = attention_mask.unsqueeze(-1).repeat(
                1, 1, hidden_states.shape[2])
            hidden_states[~expand_attention_mask] = 0

            # extend attention_mask
            attention_mask = 1.0 - attention_mask[:, None, None, :].to(
                dtype=hidden_states.dtype)
            attention_mask = attention_mask * np.iinfo(np.float32).min
            attention_mask = attention_mask.expand(attention_mask.shape[0], 1,
                                                   attention_mask.shape[-1],
                                                   attention_mask.shape[-1])

        position_embeddings = self.pos_conv_embed(hidden_states)
        hidden_states = hidden_states + position_embeddings
        hidden_states = self.layer_norm(hidden_states)
        hidden_states = self.dropout(hidden_states)

        #deepspeed_zero3_is_enabled = is_deepspeed_zero3_enabled()

        for layer in self.layers:
            if output_hidden_states:
                all_hidden_states = all_hidden_states + (hidden_states, )

            # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description)
            dropout_probability = np.random.uniform(0, 1)

            skip_the_layer = True if self.training and (
                dropout_probability < self.config.layerdrop) else False
            if not skip_the_layer:  # or deepspeed_zero3_is_enabled:
                # under deepspeed zero3 all gpus must run in sync
                if self.gradient_checkpointing and self.training:
                    # create gradient checkpointing function
                    def create_custom_forward(module):
                        def custom_forward(*inputs):
                            return module(*inputs, output_attentions)

                        return custom_forward
                else:
                    layer_outputs = layer(
                        hidden_states,
                        attention_mask=attention_mask,
                        output_attentions=output_attentions)
                hidden_states = layer_outputs[0]

            if skip_the_layer:
                layer_outputs = (None, None)

            if output_attentions:
                all_self_attentions = all_self_attentions + (layer_outputs[1], )

        if output_hidden_states:
            all_hidden_states = all_hidden_states + (hidden_states, )

        if not return_dict:
            return tuple(
                v
                for v in
                [hidden_states, all_hidden_states, all_self_attentions]
                if v is not None)
        return BaseModelOutput(
            last_hidden_state=hidden_states,
            hidden_states=all_hidden_states,
            attentions=all_self_attentions, )


class Wav2Vec2EncoderStableLayerNorm(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.config = config
        self.pos_conv_embed = Wav2Vec2PositionalConvEmbedding(config)
        self.layer_norm = nn.LayerNorm(
            config.hidden_size, epsilon=config.layer_norm_eps)
        self.dropout = nn.Dropout(config.hidden_dropout)
        self.layers = nn.LayerList([
            Wav2Vec2EncoderLayerStableLayerNorm(config)
            for _ in range(config.num_hidden_layers)
        ])
        self.gradient_checkpointing = False

    def forward(
            self,
            hidden_states,
            attention_mask=None,
            output_attentions=False,
            output_hidden_states=False,
            return_dict=True, ):
        all_hidden_states = () if output_hidden_states else None
        all_self_attentions = () if output_attentions else None

        if attention_mask is not None:
            # make sure padded tokens are not attended to
            expand_attention_mask = attention_mask.unsqueeze(
                -1).repeat_interleave(
                    hidden_states.shape[2], axis=2)
            hidden_states[~expand_attention_mask] = 0

            # extend attention_mask
            attention_mask = 1.0 - attention_mask[:, None, None, :].to(
                dtype=hidden_states.dtype)
            attention_mask = attention_mask * np.iinfo(np.float32).min
            attention_mask = attention_mask.expand(attention_mask.shape[0], 1,
                                                   attention_mask.shape[-1],
                                                   attention_mask.shape[-1])

        position_embeddings = self.pos_conv_embed(hidden_states)
        hidden_states = hidden_states + position_embeddings
        hidden_states = self.dropout(hidden_states)

        for layer in self.layers:
            if output_hidden_states:
                all_hidden_states = all_hidden_states + (hidden_states, )

            # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description)
            dropout_probability = np.random.uniform(0, 1)

            skip_the_layer = True if self.training and (
                dropout_probability < self.config.layerdrop) else False
            if not skip_the_layer:  # or deepspeed_zero3_is_enabled:
                # under deepspeed zero3 all gpus must run in sync
                # XXX: could optimize this like synced_gpus in generate_utils but not sure if it's worth the code complication
                if self.gradient_checkpointing and self.training:
                    # create gradient checkpointing function
                    def create_custom_forward(module):
                        def custom_forward(*inputs):
                            return module(*inputs, output_attentions)

                        return custom_forward
                else:
                    layer_outputs = layer(
                        hidden_states,
                        attention_mask=attention_mask,
                        output_attentions=output_attentions)
                hidden_states = layer_outputs[0]

            if skip_the_layer:
                layer_outputs = (None, None)

            if output_attentions:
                all_self_attentions = all_self_attentions + (layer_outputs[1], )

        hidden_states = self.layer_norm(hidden_states)

        if output_hidden_states:
            all_hidden_states = all_hidden_states + (hidden_states, )

        if not return_dict:
            return tuple(
                v
                for v in
                [hidden_states, all_hidden_states, all_self_attentions]
                if v is not None)
        return BaseModelOutput(
            last_hidden_state=hidden_states,
            hidden_states=all_hidden_states,
            attentions=all_self_attentions, )


class Wav2Vec2GumbelVectorQuantizer(nn.Layer):
    """
    Vector quantization using gumbel softmax. See `[CATEGORICAL REPARAMETERIZATION WITH
    GUMBEL-SOFTMAX](https://arxiv.org/pdf/1611.01144.pdf) for more information.
    """

    def __init__(self, config):
        super().__init__()
        self.num_groups = config.num_codevector_groups
        self.num_vars = config.num_codevectors_per_group

        if config.codevector_dim % self.num_groups != 0:
            raise ValueError(
                f"`config.codevector_dim {config.codevector_dim} must be divisible "
                f"by `config.num_codevector_groups` {self.num_groups} for concatenation"
            )

        # storage for codebook variables (codewords)
        self.codevectors = paddle.static.create_parameter(
            shape=[
                1, self.num_groups * self.num_vars,
                config.codevector_dim // self.num_groups
            ],
            dtype='float32')
        self.weight_proj = nn.Linear(config.conv_dim[-1],
                                     self.num_groups * self.num_vars)

        # can be decayed for training
        self.temperature = 2

    @staticmethod
    def _compute_perplexity(probs, mask=None):
        if mask is not None:
            mask_extended = mask.flatten()[:, None, None].expand(probs.shape)
            probs = paddle.where(mask_extended, probs, paddle.zeros_like(probs))
            marginal_probs = probs.sum(dim=0) / mask.sum()
        else:
            marginal_probs = probs.mean(dim=0)

        perplexity = paddle.exp(-paddle.sum(
            marginal_probs * paddle.log(marginal_probs + 1e-7), dim=-1)).sum()
        return perplexity

    def forward(self, hidden_states, mask_time_indices=None):
        batch_size, sequence_length, hidden_size = hidden_states.shape

        # project to codevector dim
        hidden_states = self.weight_proj(hidden_states)
        hidden_states = hidden_states.reshape(
            (batch_size * sequence_length * self.num_groups, -1))

        if self.training:
            # sample code vector probs via gumbel in differentiateable way
            codevector_probs = nn.functional.gumbel_softmax(
                hidden_states.float(), tau=self.temperature,
                hard=True).type_as(hidden_states)

            # compute perplexity
            codevector_soft_dist = paddle.softmax(
                hidden_states.reshape((batch_size * sequence_length,
                                       self.num_groups, -1)).float(),
                axis=-1)
            perplexity = self._compute_perplexity(codevector_soft_dist,
                                                  mask_time_indices)
        else:
            # take argmax in non-differentiable way
            # comptute hard codevector distribution (one hot)
            codevector_idx = hidden_states.argmax(dim=-1)
            codevector_probs = hidden_states.new_zeros(
                *hidden_states.shape).scatter_(-1,
                                               codevector_idx.reshape((-1, 1)),
                                               1.0)
            codevector_probs = codevector_probs.reshape(
                (batch_size * sequence_length, self.num_groups, -1))

            perplexity = self._compute_perplexity(codevector_probs,
                                                  mask_time_indices)

        codevector_probs = codevector_probs.reshape(
            (batch_size * sequence_length, -1))
        # use probs to retrieve codevectors
        codevectors_per_group = codevector_probs.unsqueeze(
            -1) * self.codevectors
        codevectors = codevectors_per_group.reshape(
            (batch_size * sequence_length, self.num_groups, self.num_vars, -1))
        codevectors = codevectors.sum(-2).reshape(
            (batch_size, sequence_length, -1))

        return codevectors, perplexity


class Wav2Vec2Adapter(nn.Layer):
    def __init__(self, config):
        super().__init__()

        # feature dim might need to be down-projected
        if config.output_hidden_size != config.hidden_size:
            self.proj = nn.Linear(config.hidden_size, config.output_hidden_size)
            self.proj_layer_norm = nn.LayerNorm(config.output_hidden_size)
        else:
            self.proj = self.proj_layer_norm = None

        self.layers = nn.LayerList(
            Wav2Vec2AdapterLayer(config)
            for _ in range(config.num_adapter_layers))
        self.layerdrop = config.layerdrop

    def forward(self, hidden_states):
        # down project hidden_states if necessary
        if self.proj is not None and self.proj_layer_norm is not None:
            hidden_states = self.proj(hidden_states)
            hidden_states = self.proj_layer_norm(hidden_states)

        hidden_states = hidden_states.transpose([0, 2, 1])

        for layer in self.layers:
            layerdrop_prob = np.random.random()
            if not self.training or (layerdrop_prob > self.layerdrop):
                hidden_states = layer(hidden_states)

        hidden_states = hidden_states.transpose([0, 2, 1])
        return hidden_states


class Wav2Vec2AdapterLayer(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.conv = nn.Conv1D(
            config.output_hidden_size,
            2 * config.output_hidden_size,
            config.adapter_kernel_size,
            stride=config.adapter_stride,
            padding=1, )

    def forward(self, hidden_states):
        hidden_states = self.conv(hidden_states)
        hidden_states = nn.functional.glu(hidden_states, axis=1)

        return hidden_states


class Wav2Vec2Model(nn.Layer):
    def __init__(self, config):
        super().__init__()
        self.config = config
        self.feature_extractor = Wav2Vec2FeatureEncoder(config)
        self.feature_projection = Wav2Vec2FeatureProjection(config)

        # model only needs masking vector if mask prob is > 0.0
        if config.mask_time_prob > 0.0 or config.mask_feature_prob > 0.0:
            #  self.masked_spec_embed = nn.Parameter(paddle.Tensor(config.hidden_size).uniform_())
            #self.masked_spec_embed = paddle.uniform([config.hidden_size])
            self.masked_spec_embed = paddle.static.create_parameter(
                shape=[config.hidden_size],
                dtype='float32',
                default_initializer=paddle.nn.initializer.Uniform(
                    low=0, high=1.0))
        if config.do_stable_layer_norm:
            self.encoder = Wav2Vec2EncoderStableLayerNorm(config)
        else:
            self.encoder = Wav2Vec2Encoder(config)

        self.adapter = Wav2Vec2Adapter(config) if config.add_adapter else None

        # Initialize weights and apply final processing
        self.post_init()

    def freeze_feature_encoder(self):
        """
        Calling this function will disable the gradient computation for the feature encoder so that its parameter will
        not be updated during training.
        """
        self.feature_extractor._freeze_parameters()

    def _mask_hidden_states(
            self,
            hidden_states: paddle.Tensor,
            mask_time_indices: Optional[paddle.Tensor]=None,
            attention_mask: Optional[paddle.Tensor]=None, ):
        """
        Masks extracted features along time axis and/or along feature axis according to
        [SpecAugment](https://arxiv.org/abs/1904.08779).
        """
        # `config.apply_spec_augment` can set masking to False
        if not getattr(self.config, "apply_spec_augment", True):
            return hidden_states

        # generate indices & apply SpecAugment along time axis
        batch_size, sequence_length, hidden_size = hidden_states.shape
        if mask_time_indices is not None:
            # apply SpecAugment along time axis with given mask_time_indices
            hidden_states[mask_time_indices] = self.masked_spec_embed.to(
                hidden_states.dtype)
        elif self.config.mask_time_prob > 0 and self.training:
            mask_time_indices = _compute_mask_indices(
                (batch_size, sequence_length),
                mask_prob=self.config.mask_time_prob,
                mask_length=self.config.mask_time_length,
                attention_mask=attention_mask,
                min_masks=self.config.mask_time_min_masks, )
            mask_time_indices = paddle.to_tensor(
                mask_time_indices, dtype=paddle.bool)
            hidden_states[mask_time_indices] = self.masked_spec_embed.to(
                hidden_states.dtype)

        if self.config.mask_feature_prob > 0 and self.training:
            # generate indices & apply SpecAugment along feature axis
            mask_feature_indices = _compute_mask_indices(
                (batch_size, hidden_size),
                mask_prob=self.config.mask_feature_prob,
                mask_length=self.config.mask_feature_length,
                min_masks=self.config.mask_feature_min_masks, )
            mask_feature_indices = paddle.to_tensor(
                mask_feature_indices, dtype=paddle.bool)
            mask_feature_indices = mask_feature_indices[:, None].expand(
                -1, sequence_length, -1)
            hidden_states[mask_feature_indices] = 0

        return hidden_states

    def forward(
            self,
            input_values: Optional[paddle.Tensor],
            attention_mask: Optional[paddle.Tensor]=None,
            mask_time_indices: Optional[paddle.Tensor]=None,
            output_attentions: Optional[bool]=None,
            output_hidden_states: Optional[bool]=None,
            return_dict: Optional[bool]=None,
    ) -> Union[Tuple, Wav2Vec2BaseModelOutput]:
        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
        output_hidden_states = (output_hidden_states
                                if output_hidden_states is not None else
                                self.config.output_hidden_states)
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
        extract_features = self.feature_extractor(input_values)
        extract_features = extract_features.transpose([0, 2, 1])

        if attention_mask is not None:
            # compute reduced attention_mask corresponding to feature vectors
            attention_mask = self._get_feature_vector_attention_mask(
                extract_features.shape[1], attention_mask, add_adapter=False)
        hidden_states, extract_features = self.feature_projection(
            extract_features)
        hidden_states = self._mask_hidden_states(
            hidden_states,
            mask_time_indices=mask_time_indices,
            attention_mask=attention_mask)

        encoder_outputs = self.encoder(
            hidden_states,
            attention_mask=attention_mask,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict, )

        hidden_states = encoder_outputs[0]

        if self.adapter is not None:
            hidden_states = self.adapter(hidden_states)

        if not return_dict:
            return (hidden_states, extract_features) + encoder_outputs[1:]

        return Wav2Vec2BaseModelOutput(
            last_hidden_state=hidden_states,
            extract_features=extract_features,
            hidden_states=encoder_outputs.hidden_states,
            attentions=encoder_outputs.attentions, )

    def post_init(self):
        """
        A method executed at the end of each Transformer model initialization, to execute code that needs the model's
        modules properly initialized (such as weight initialization).
        """
        #   self.init_weights()
        #   self._backward_compatibility_gradient_checkpointing()
        pass


class Wav2Vec2ConfigPure():
    model_type = "wav2vec2"

    def __init__(self, config):
        self.output_attentions = False
        self.output_hidden_states = False
        self.use_return_dict = True

        self.hidden_size = config.hidden_size
        self.feat_extract_norm = config.feat_extract_norm
        self.feat_extract_activation = config.feat_extract_activation
        self.conv_dim = config.conv_dim
        self.conv_stride = config.conv_stride
        self.conv_kernel = config.conv_kernel
        self.conv_bias = config.conv_bias
        self.num_conv_pos_embeddings = config.num_conv_pos_embeddings
        self.num_conv_pos_embedding_groups = config.num_conv_pos_embedding_groups
        self.num_feat_extract_layers = len(self.conv_dim)
        self.num_hidden_layers = config.num_hidden_layers
        self.intermediate_size = config.intermediate_size
        self.hidden_act = config.hidden_act
        self.num_attention_heads = config.num_attention_heads
        self.hidden_dropout = config.hidden_dropout
        self.attention_dropout = config.attention_dropout
        self.activation_dropout = config.activation_dropout
        self.feat_proj_dropout = config.feat_proj_dropout
        self.final_dropout = config.final_dropout
        self.layerdrop = config.layerdrop
        self.layer_norm_eps = config.layer_norm_eps
        self.initializer_range = config.initializer_range
        self.do_stable_layer_norm = config.do_stable_layer_norm
        self.use_weighted_layer_sum = config.use_weighted_layer_sum

        if ((len(self.conv_stride) != self.num_feat_extract_layers) or
            (len(self.conv_kernel) != self.num_feat_extract_layers) or
            (len(self.conv_dim) != self.num_feat_extract_layers)):
            raise ValueError(
                "Configuration for convolutional layers is incorrect. It is required that `len(config.conv_dim)` =="
                " `len(config.conv_stride)` == `len(config.conv_kernel)`, but is `len(config.conv_dim) ="
                f" {len(self.conv_dim)}`, `len(config.conv_stride) = {len(self.conv_stride)}`,"
                f" `len(config.conv_kernel) = {len(self.conv_kernel)}`.")

        # fine-tuning config parameters for SpecAugment: https://arxiv.org/abs/1904.08779
        self.apply_spec_augment = config.apply_spec_augment
        self.mask_time_prob = config.mask_time_prob
        self.mask_time_length = config.mask_time_length
        self.mask_time_min_masks = config.mask_time_min_masks
        self.mask_feature_prob = config.mask_feature_prob
        self.mask_feature_length = config.mask_feature_length
        self.mask_feature_min_masks = config.mask_feature_min_masks

        # parameters for pretraining with codevector quantized representations
        self.num_codevectors_per_group = config.num_codevectors_per_group
        self.num_codevector_groups = config.num_codevector_groups
        self.contrastive_logits_temperature = config.contrastive_logits_temperature
        self.feat_quantizer_dropout = config.feat_quantizer_dropout
        self.num_negatives = config.num_negatives
        self.codevector_dim = config.codevector_dim
        self.proj_codevector_dim = config.proj_codevector_dim
        self.diversity_loss_weight = config.diversity_loss_weight

        # adapter
        self.add_adapter = config.add_adapter
        self.adapter_kernel_size = config.adapter_kernel_size
        self.adapter_stride = config.adapter_stride
        self.num_adapter_layers = config.num_adapter_layers
        self.output_hidden_size = config.output_hidden_size or config.hidden_size

    @property
    def inputs_to_logits_ratio(self):
        return functools.reduce(operator.mul, self.conv_stride, 1)