TensorFlow Examples for Beginners with Latest APIs

github.com

196 points·aymericdamien··10 comments

Thanks this is cool. I just quickly glanced but you are using Python 2.7 ... maybe make a note of that in the readme ?

The examples I've tried so far appear to be portable, and run fine with Python 3.6.1.

I would pay good money for a service where I upload a folder of images, gets an interface where I can classify each image one by one and then let ML classify the rest. Does this exist?

If you don't mind me asking, why do you want to do this?

Do you mean like this? https://azure.microsoft.com/en-us/services/cognitive-service...

Or you can use amazon mechanical turk to get everything classified by people

Interesting, we built a product similar to this and I'd be happy to have you try it out. Check it out here: http://symph-geoprocessing.appspot.com/ I don't think it has all the functions you are looking for yet, but we are happy to learn more about what specific features you are looking for and see if we can build them out.

Cool to see. I find the Estimator framework makes it much easier to write tests for training pipelines because of the way it modularizes your pipeline at a good granularity for testing.

Having tests for training pipelines is nice because you can catch a fair amount of bugs with speedy unit tests instead of having to launch them and wait for everything to spin up.

Another very helpful tutorial on new Experiment, Estimator an Dataset APIs: https://medium.com/onfido-tech/higher-level-apis-in-tensorfl...

Absolutely amazing, very neatly laid out.

Thank you so much for your efforts.

Reply on news.ycombinator.com