Training Pipeline Finalization

Simsso / NIPS-2018-Adversarial-Vision-Challenge

Code, documents, and deployment configuration files, related to our participation in the 2018 NIPS Adversarial Vision Challenge "Robust Model Track"

MIT License

10 stars 1 forks source link

Training Pipeline Finalization #62

Closed Simsso closed 6 years ago

Simsso commented 6 years ago

[x] Docker image with Python 3.6 (tricky for tensorflow-gpu) or make code compatible with Python 3.5 (@simsso)
[x] Installation of the vq_layer pip module
[x] Versioning of models with strings (not only numbers)
[x] Drop /models folder

ghost commented 6 years ago

@Simsso Fixed all problems mentioned above.

Still remaining one problem: Model doesn't use the given GPU (removed the line with \cpu:0).

Can you help me out here?

Simsso commented 6 years ago

Can you provide a link to the commit which contains that state you have tested things with? Then I can look into the code, and also I'm curious to see how you've positioned the vq-layer folder. Thanks.

ghost commented 6 years ago

@Simsso This commit is the very last.

Simsso commented 6 years ago

Maybe the reason is that we're installing tensorflow with pip (here)

ghost commented 6 years ago

I'll check this out again. Yesterday, it didn't seem like to work even with tensorflow-gpu.