Chiranshu Adik

@ArEnSc @golsun _python src/generation.py play -pg=restore/medium_ft.pkl -pr=restore/updown.pth --sampling_ This command worked perfectly, got awesome results. I'm going to play around with this some more. But I think this looks perfect...

@golsun @ArEnSc Since this model is huge and requires GPU to compute inferences. Do you know if there is any simpler way to generate responses? Can we make it run...

@ArEnSc 1. Yes, I tried with my cpu (3700x) and it was still slow. It's going to be slower on a server then. Hence a rest api won't be an...

@ArEnSc with GPU it took around 1-2 seconds, while on CPU it took approximately 8 seconds. First I tried with the large model that you had linked to but had...

> > @ArEnSc @golsun > > > > _python src/generation.py play -pg=restore/medium_ft.pkl -pr=restore/updown.pth --sampling_ > > > > This command worked perfectly, got awesome results. I'm going to play around...

The same issue for me, the training never starts on my desktop PC. Never faced any issues like this when directly working with transformers. 

Is this feature available in 2021?

Sounds good, it's unfortunate that we can't just retrain google's existing model with more data to improve it. Your suggestion about BERT sounds good to me, I'll also check if...

@Narasimha1997 Can you please tell me how did you get the header files for the custom ops used in the tflite model? Are they from the tensorflow git repo? or...

Thanks, I'm currently trying to use the tflite model from the ML kit website. Tried their android sample, it is much better than the one from the TensorFlow site. Surprisingly...