-
请问训练了上百万次，没改动数据源，Accuracy和Cost一直上下波动是什么原因？总是附近徘徊，训练了几天都是这样
 create _n_ copies of the input. At test time the logit streams are averaged together. However, this is buggy under CTC training, as the blank la…
-
Hi all,
I've seen the new tf_clean branch is available, so be trying to use it. I'm using ```swbd/v1-tf``` recipe and was able to train it successfully. However, I cannot find any script for single w…
-
Error while make in cmake fashion. Complains
> cuda_compile_generated_ctc_loss_layer.cu.o' failed
cmake Log
```
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identificati…
-
The hparams.py says `n_frames_per_step=1, # currently only 1 is supported`, but reduction window is very important for them model to pick up alignment. Using a reduction window can be considered as d…
-
-
Can you please tell me what layers are different from original caffe？
I think this addition add CtcLoss layer and ContinuationIndicator layer. Is that right?
Where is the cpp file of the added layer…
-
- https://arxiv.org/abs/2104.07787
- 2021
本論文では、テキストライン認識の問題を研究しています。
シーンテキストや手書き文書のような特定のドメインを対象とした多くのアプローチとは異なり、ソースや入力モダリティに関わらず、あらゆる画像からテキストを抽出できる普遍的なアーキテクチャを開発するという一般的な問題を調査する。
本研究では、2つのデコーダ…
e4exp updated
3 years ago
-
Hi,
I am new to Icefall. I would like to extract framewise alignment information like what is shown in #39 with the pretrained model from: https://huggingface.co/csukuangfj/icefall-asr-librispeech…