search

TheFoundryVisionmongers / nuke-ML-server

A Nuke client plug-in which connects to a Python server to allow Machine Learning inference in Nuke.
Apache License 2.0
134 stars 36 forks source link

Dense Pose crashes #12

Open samhodge opened 5 years ago

samhodge commented 5 years ago

Server -> Receiving message of size: 24883378 Server -> 24883378 bytes read Server -> Message parsed Server -> Received inference request Server -> Requesting inference on model: densepose Server -> Starting inference WARNING:root:[====DEPRECATE WARNING====]: you are creating an object from CNNModelHelper class which will be deprecated soon. Please use ModelHelper object with brew module. For more information, please refer to caffe2.ai and python/brew.py, python/brew_test.py for more information. Server -> Exception caught on inference on model: Server -> Serializing message Server -> Sending response message of size: 18 Server -> ----------------------------------------------

But in the good news pile, MaskRCNN works.

Nuke-ML

ringdk commented 5 years ago

Great to hear Mask R-CNN is running!

As for DensePose not running, from the console output it looks like it should have returned an error to Nuke. Is any error displayed in the Viewer?

One thing to check is to start with a clean Nuke script, load MLClient, make sure that DensePose is the first model you choose to run. Another thing to check is that you have enough GPU memory available (but you normally get a big error when this happens so it's unlikely.)

samhodge commented 5 years ago

There was a Red error message, I will boot into CentOS 7.4 in about 12 hours from now, right now I have other priorities.

I am very interested in getting a lot more familiar with DensePose with regard to closing the loop and returning temporally sparse Roto shapes to Nuke.

It will probably take a block of time in the hundreds of hours to perfect.

samhodge commented 5 years ago

This is all I get as an error

ERROR: MLClient1: