mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-08-08 09:32:20 +00:00
1d44da845b
* Begin to add RNN-T training for librispeech. * Copy files from conformer_ctc. Will edit it. * Use conformer/transformer model as encoder. * Begin to add training script. * Add training code. * Remove long utterances to avoid OOM when a large max_duraiton is used. * Begin to add decoding script. * Add decoding script. * Minor fixes. * Add beam search. * Use LSTM layers for the encoder. Need more tunings. * Use stateless decoder. * Minor fixes to make it ready for merge. * Fix README. * Update RESULT.md to include RNN-T Conformer. * Minor fixes. * Fix tests. * Minor fixes. * Minor fixes. * Fix tests.
56 lines
1.8 KiB
Python
56 lines
1.8 KiB
Python
# Copyright 2021 Xiaomi Corp. (authors: Fangjun Kuang)
|
|
#
|
|
# See ../../../../LICENSE for clarification regarding multiple authors
|
|
#
|
|
# 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 torch.nn.functional as F
|
|
|
|
|
|
class Joiner(nn.Module):
|
|
def __init__(self, input_dim: int, output_dim: int):
|
|
super().__init__()
|
|
|
|
self.output_linear = nn.Linear(input_dim, output_dim)
|
|
|
|
def forward(
|
|
self, encoder_out: torch.Tensor, decoder_out: torch.Tensor
|
|
) -> torch.Tensor:
|
|
"""
|
|
Args:
|
|
encoder_out:
|
|
Output from the encoder. Its shape is (N, T, C).
|
|
decoder_out:
|
|
Output from the decoder. Its shape is (N, U, C).
|
|
Returns:
|
|
Return a tensor of shape (N, T, U, C).
|
|
"""
|
|
assert encoder_out.ndim == decoder_out.ndim == 3
|
|
assert encoder_out.size(0) == decoder_out.size(0)
|
|
assert encoder_out.size(2) == decoder_out.size(2)
|
|
|
|
encoder_out = encoder_out.unsqueeze(2)
|
|
# Now encoder_out is (N, T, 1, C)
|
|
|
|
decoder_out = decoder_out.unsqueeze(1)
|
|
# Now decoder_out is (N, 1, U, C)
|
|
|
|
logit = encoder_out + decoder_out
|
|
logit = F.relu(logit)
|
|
|
|
output = self.output_linear(logit)
|
|
|
|
return output
|