snowfall
snowfall copied to clipboard
k2-fsa
Moved to https://github.com/k2-fsa/icefall
LSTM model
Right now decode.py is giving very terrible WER, over 90%, even after tuning bias_penalty. Would be nice if someone can try using an LSTM model (needed to give spiky outputs...
I follow `librispeech/asr/simple_v1/` recipe example, and rebuild k2 newest master(2021-1-4). When I run `python3 asg_train.py` , It gives the following error: ``` scores: tensor([ 1.9269, 1.4873, 0.9007, ..., 0.0990, -1.0238,...
We are currently adding `ASG` training scripts, which have a large overlap with the existing CTC training scripts. Later I think we will add training scripts for other approaches. The...
I follow the librispeech recipe of snowfall, I make our own private data. NOTE: The private data is running successfully in the Kaldi platform. When I use the below command...
Some issues and doubts about the current method to generate L.fst: - ~Since there is no G, do we still need the disambiguation symbols?~ - What is the purpose of...
There exist shell scripts and Perl scripts for data preparation, most of which are modified from Kaldi. They provide extra functionalities that are actually not needed in snowfall. I would...
I've been playing around with the recipe to get to know k2 better. I trained the model for 10 epochs and then tried to perform decoding with the following code...
I have been running some experiments to debug why prepare.py is so slow. I'll post the results here in a bit, but let me just give you a preview of...
I met the error when run stage 6 ` python train.py` ` $ python train.py /data/app/anaconda3/envs/k2/lib/python3.7/site-packages/lhotse/augmentation/torchaudio.py:13: UserWarning: Torchaudio SoX effects chains are only introduced in version 0.7 - please upgrade...
Hello, I'm learning using k2 to implement CTC. In the "create_decoding_graph" function, it seems like merging a batch of label text into one decoding graph, is it reasonable? https://github.com/k2-fsa/snowfall/blob/26398e8e98a3ff83664a43b60884853514ba94e2/egs/librispeech/asr/simple_v1/train.py#L34 ```...
