-
Am I the only person to have some negative losses with TF2.4 and CUDA11.0 ? I just upated TF from 2.3 to 2.4, recompile warprnnt and always got negative losses (even with the tests of the library).
…
-
Hi,
Thank you for sharing this great framework.
I found it very useful for several ASR experiments.
Yet I had trouble to make the LibriSpeech recipes converging when using auto_mix_prec.
I obser…
-
Hello,
I'm trying to test possible enhancements based on ContextNet, but I've noticed the current model file has now deviated from Quartznet. I was wondering how you were able to arrive on the arch…
-
Traceback (most recent call last):
File "./tools/train.py", line 125, in
main()
File "./tools/train.py", line 121, in main
timestamp=timestamp)
File "/home/hs/hao/contextnet/ctfpnr…
-
Hi, I still got the issue after I ran `jupyter labextension install @jupyter-widgets/jupyterlab-manager`.
Initially, I came across some problems using `tqdm_notebook` in `jupyterlab`. After googlin…
-
Traceback (most recent call last):
File "train.py", line 138, in
main()
File "train.py", line 116, in main
cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg)
File "/home/hs…
-
Congratulations for nice work.
Have you also seen BMVC 2018 paper, ContextNet with 0.8 million parameter yield 66.1% accuracy and runs in 41.9 fps in the oldest Titan X gpu?