Wei Wang comments

Results 28 comments of


                                            Wei Wang

Major memory leak on somewhat advanced models using cpu backend

Thanks @miketout for digging into the above issue and coming up a workaround. We will try to take a look and try to reproduce after you release the open source...

Can I install neon on 32-bit computer?

Thanks @anonymous-u for trying neon. MKL is not available for 32-bit systems, so it will not be possible to install neon with MKL support on a 32-bit system.

Can I install neon on 32-bit computer?

Theoretically you can remove/disable the MKL component and try installing neon (make sure GPU requirement/dependencies like cuda driver etc are met). This configuration (32 bit + GPU) is less tested,...

how can i add my own collection of images on neon to train a model ?

@campos537 Can you please start with examples/binary ? It uses MNIST dataset. For advanced data ingestion, please also refer to aeon (which is part of neon and is automatically installed)....

Update visualization docs

Thanks @chonbas Do you mind resolving the conflict and update this PR? Thanks!

why num_batches changed ?

Thanks for reporting this! We will be fixing it in one of the future releases.

Any converted pretrained weights of UCF101-C3D for neon v2?

Thanks for reporting, we will address this gap.

modified makefile for properly supporting python 3 in system install …

Thanks for your PR, we will look into bringing your PR in.

'DirectoryIterator' object has no attribute 'nbatches'

Hi @mrysztow Could you please take a look?

Masked data iterator

Thank you @Bam4d It has been quite a while and sorry for the delay. Could you please resolve the conflicts?