search

gabrieleilertsen / hdrcnn

HDR image reconstruction from a single exposure using deep CNNs
https://computergraphics.on.liu.se/hdrcnn/
BSD 3-Clause "New" or "Revised" License
502 stars 101 forks source link

Requirements file? #48

Closed rleaver152 closed 3 years ago

rleaver152 commented 3 years ago

Is there any possibility of a requirements.txt file (non gpu tensor flow)

I'm trying every conceivable combination of 1.x tensorflow, tensorlayer, wrapt and numpy etc but cannot get pip to find a working package version combination between them all. Python3.6 and 3.7 both failing. If there's a working requirements.txt file that could be generated, that would be great!!

gabrieleilertsen commented 3 years ago

Hi. It seems like it's starting to be a bit of a challenge with the requirements for tensorflow 1 code. It would probably be a good idea to update the code to tensorflow 2, but I haven't been able to find the time for it yet. I'm currently running the scripts with:

tensorflow-gpu 1.15.0 tensorlayer 1.11.1 wrapt 1.12.1 python 3.6

The problem here is that tensorflow 1.15.0 depends on wrapt>=1.11.1, while tensorlayer 1.11.1 depends on wrapt<1.11 and >=1.10. However, I it is still possible to run tensorflow with wrapt 1.10, or tensorlayer with wrapt 1.12.1. For this to work, you could for example install tensorlayer first, with wrapt 1.10, and then tensorflow, which will update wrapt to 1.12.1.

I think this will make it problematic to have a requirements file for installing the necessary dependencies, but I hope the above information can help you.

rleaver152 commented 3 years ago

Had problems but looking like pip install numpy==1.16.2 is working alongside your earlier solutions.

On 26 Mar 2021, at 19:07, Gabriel Eilertsen @.***> wrote:

Hi. It seems like it's starting to be a bit of a challenge with the requirements for tensorflow 1 code. It would probably be a good idea to update the code to tensorflow 2, but I haven't been able to find the time for it yet. I'm currently running the scripts with:

tensorflow-gpu 1.15.0 tensorlayer 1.11.1 wrapt 1.12.1 python 3.6

The problem here is that tensorflow 1.15.0 depends on wrapt>=1.11.1, while tensorlayer 1.11.1 depends on wrapt<1.11 and >=1.10. However, I it is still possible to run tensorflow with wrapt 1.10, or tensorlayer with wrapt 1.12.1. For this to work, you could for example install tensorlayer first, with wrapt 1.10, and then tensorflow, which will update wrapt to 1.12.1.

I think this will make it problematic to have a requirements file for installing the necessary dependencies, but I hope the above information can help you.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.