Ilya Kavalerov
Ilya Kavalerov
Do you think it is a good evaluation of a wake-word model assuming the model is not trained on any test samples? For example the alexa test is 25 hours...
Can you suggest a benchmarking methodology? I am pretty close to an independent party, I am looking to compare established wakeword detectors to the one from [this paper](https://arxiv.org/pdf/2102.00313.pdf). My motivation...
@euwern I can't get to 8.25 either. Did you ever do the hyper parameter search they mention in the [paper](https://arxiv.org/pdf/1610.09585.pdf)? i.e. "Performing a grid search across 27 hyperparameter configurations" where...
p.s. @euwern Actually after ensuring the GAN produced image is between 0,1 and then saving, I can get the IS to 6.2. Still quite a bit below the reported number...
👍 Worked for me. Led to 5x speedup on 50k CIFAR10 images while keeping the IS
I was able to train this for 100 epochs once and got a val loss of `-20.156` on wsj0-2mix
I trained this a second time for 100 epochs and got training loss down to `-21.109` and eval loss down to `-20.229` on wsj0-2mix
Nice repo you've got there. You'll notice I closed a bunch of issues here b/c your repo should be the future of sup. Any code that I may write in...
As you note in the comment on ambit, for this particular problem, there are so many bad triplets. If we take a look at a graph where connects indicate a...
Thanks for this! Fixed the following error for me: ``` Traceback (most recent call last): File "convert.py", line 68, in main() File "convert.py", line 64, in main args.phase) File "convert.py",...