Yajie Miao

You can open nnet.mdl directly. Depending on the model size, loading the file may take some time. nnet.cfg is pickle-formatted dump of network configuration (a class). You can load it...

you can do it with cmds/run_Extract_Feats.py

Word-based language model built on TIMIT is relatively weak. I recommend you to compose a phone language model. You plug a fake dictionary which simply contains the duplicates of phones:...

yep My very first verification of EESEN was done on TIMIT. I was able to get reasonable (if not state-of-the-art) phone error rates

In general, CTC highly depends on BiLSTM for reasonable performance. If you refer to http://www.cs.cmu.edu/~ymiao/pub/icassp2016_ctc.pdf, on Switchboard, Uni-directional models perform >15% worse than Bi-directional models, with the same number of...

If it still fails, try to reduce --frame-num-limit as it affects the deltas of parameters.

We are studying the validity of the comparisons they presented, specifically the linear correlation between time and N for Eesen.

The previous tedlium char recipe does not work. I just made changes. It should work now. I am validating it on my side.

You can submit the running of train_ctc_parallel.sh (for example https://github.com/srvk/eesen/blob/master/asr_egs/wsj/run_ctc_phn.sh#L75) to the scheduler Alternatively, in train_ctc_parallel.sh, you can modify it by following https://github.com/srvk/eesen/blob/master/asr_egs/wsj/steps/train_ctc_parallel_h.sh#L141

You should set it to the number of jobs (in your case just 1), instead of the number of GPUs. When you set it to 3, the script will submit...