search

waylonflinn / weblas

GPU Powered BLAS for Browsers :gem:
MIT License
702 stars 43 forks source link

Your example does not work for me. #30

Closed MariasStory closed 7 years ago

MariasStory commented 8 years ago

Hi @waylonflinn,

It all looks very interesting and I would like to see your solution combined with neocortex. I could not run your example in Chrome on Ubuntu. I have a build in I5 gpu and get: "Bound framebuffer is not complete." Error.

The error is in webgl.js: if( gl.checkFramebufferStatus(gl.FRAMEBUFFER) != gl.FRAMEBUFFER_COMPLETE) throw new Error("Bound framebuffer is not complete.");

Am I doing something wrong there?

MariasStory commented 8 years ago

Sorry, the demo page that did not work is: http://waylonflinn.github.io/DeepBeliefSDK/

This example works on windows 10 with I5. I guess some gpu functionality are disabled in chrome on Ubuntu. I will check my settings.

BZW: Please cache the demo database, so it is not loaded every time.

waylonflinn commented 8 years ago

Thanks for the feedback. Caching the database is an excellent idea. If you're thinking about porting neocortex, you might be interested in this project of mine: https://github.com/waylonflinn/webnn

It's mission is to create a convolutional neural network that uses weblas in the browser. I'm hoping to return to work on it in the next week or so.

MariasStory commented 7 years ago

Hi @waylonflinn,

In linux, I have this two options: Native GpuMemoryBuffers: Software only. Hardware acceleration disabled Rasterization: Software only. Hardware acceleration disabled

Hope this helps to find the problem.

MariasStory commented 7 years ago

Hi @waylonflinn, it looks like your web example is working on Win with I5.

waylonflinn commented 7 years ago

@MariasStory glad to hear it. I'll close this. :smile: