File size: 6,016 Bytes
300a419 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 |
####################################################################
## Built on top of Ultravox: https://github.com/fixie-ai/ultravox ##
####################################################################
# modified from https://github.com/huggingface/transformers/blob/main/src/transformers/models/whisper/modeling_whisper.py
# see this issue for the commentary: https://github.com/huggingface/transformers/issues/25744
#
# Copyright 2022 The OpenAI 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.
import torch
import torch.nn as nn
import transformers
import transformers.modeling_outputs
from transformers.models.whisper import modeling_whisper as whisper
class WhisperEncoder(whisper.WhisperEncoder):
"""
Encoder portion of OpenAI's Whisper model.
This implementation is a slightly modified version of HF Transformers' Whisper Encoder, with only a few fixes:
1. base_model_prefix updated to allow for doing `.from_pretrained` directly on the encoder
2. allow less than 30 second of audio padding to be passed in:
- relaxed ValueError check for `input_features` length to be less than or equal to `expected_seq_length` instead of strictly equal
- embed_pos is now sliced to match the length of `inputs_embeds`
Original: https://github.com/huggingface/transformers/blob/main/src/transformers/models/whisper/modeling_whisper.py
"""
base_model_prefix = "model.encoder"
def forward(
self,
input_features,
attention_mask=None,
head_mask=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
expected_seq_length = (
self.config.max_source_positions
* self.conv1.stride[0]
* self.conv2.stride[0]
)
if input_features.shape[-1] > expected_seq_length:
raise ValueError(
f"Whisper expects the mel input features to be of length {expected_seq_length} or less, but found {input_features.shape[-1]}. Make sure to pad the input mel features to {expected_seq_length}."
)
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
)
inputs_embeds = nn.functional.gelu(self.conv1(input_features))
inputs_embeds = nn.functional.gelu(self.conv2(inputs_embeds))
inputs_embeds = inputs_embeds.permute(0, 2, 1)
embed_pos = self.embed_positions.weight[: inputs_embeds.size(-2)]
hidden_states = inputs_embeds + embed_pos
hidden_states = nn.functional.dropout(
hidden_states, p=self.dropout, training=self.training
)
encoder_states = () if output_hidden_states else None
all_attentions = () if output_attentions else None
# check if head_mask has a correct number of layers specified if desired
if head_mask is not None:
assert head_mask.size()[0] == (
len(self.layers)
), f"The head_mask should be specified for {len(self.layers)} layers, but it is for {head_mask.size()[0]}."
for idx, encoder_layer in enumerate(self.layers):
if output_hidden_states:
encoder_states = encoder_states + (hidden_states,)
# add LayerDrop (see https://arxiv.org/abs/1909.11556 for description)
to_drop = False
if self.training:
dropout_probability = torch.rand([])
if dropout_probability < self.layerdrop: # skip the layer
to_drop = True
if to_drop:
layer_outputs = (None, None)
else:
if self.gradient_checkpointing and self.training:
layer_outputs = self._gradient_checkpointing_func(
encoder_layer.__call__,
hidden_states,
None,
(head_mask[idx] if head_mask is not None else None),
output_attentions,
)
else:
layer_outputs = encoder_layer(
hidden_states,
None,
layer_head_mask=(
head_mask[idx] if head_mask is not None else None
),
output_attentions=output_attentions,
)
hidden_states = layer_outputs[0]
if output_attentions:
all_attentions = all_attentions + (layer_outputs[1],)
hidden_states = self.layer_norm(hidden_states)
if output_hidden_states:
encoder_states = encoder_states + (hidden_states,)
if not return_dict:
return tuple(
v
for v in [hidden_states, encoder_states, all_attentions]
if v is not None
)
return transformers.modeling_outputs.BaseModelOutput(
last_hidden_state=hidden_states,
hidden_states=encoder_states,
attentions=all_attentions,
)
|