Afaq Ahmad

icon indicating a website link https://www.linkedin.com/in/afaqahmadalvi/ icon indicating an email [email protected]

icon company aihunt.com icon location Islamabad icon location Researcher in the fields of deep learning, data science and Robotics

Results 7 comments of Afaq Ahmad

nan train loss while training SSD

same problem

Translation to Tf.Keras?

> he can anyone help me set this up, I can pay. If you can please email me at [[email protected]](mailto:[email protected]) Hi, I have added few up-gradation in it, now you...

Pretrained model has fewer layers than actual model

facing the same problem, did anybody know how to solve it

Pretrained model has fewer layers than actual model

So i was using 2 n_local_enhancers after adding this option in test.py file like this solved the problem opt.n_local_enhancers=2 Thanks

How to reproduce the bleu score in 2 GPU cards?

Hi, someone can please guide me how to start the training correctly. I have created the dataset the same way. I have around 24millions sentences, but when I start training...

Reproducibility issue when training on a smaller dataset and fewer GPUs

@sanxing-chen Hi, can you please guide me about full dataset (~20m examples), from where I can get it. Thanks

Reproducibility issue when training on a smaller dataset and fewer GPUs

> Hi @afaq-ahmad : > > after half year of research and trial and error, I think if you got (~20m examples) then train a regular transformer is total cool...