TensorFlow JS: WebGL-Accelerated ML in the Browser

js.tensorflow.org

11 points·bufferoverflow··4 comments

Is there an active project that adds TensorFlow bindings to Node. I've found a few, but the latest commit on any of these is 5 months ago.

https://github.com/node-tensorflow/node-tensorflow

https://github.com/yorkie/tensorflow-nodejs

https://github.com/rchipka/node-tensorflow

What you want to check out is Propel ML. Creator is Ryan Dahl (ry) who invented Node.js.

- http://propelml.org/

- https://github.com/propelml/propel

"Propel runs both in the browser and natively from Node. In both environments Propel is able to use GPU hardware for computations. In the browser it utilizes WebGL through deeplearn.js and on Node it uses TensorFlow's C API."

Thanks a lot for this. I just visited https://deeplearnjs.org and read that "deeplearn.js has become TensorFlow.js", so I guess that just about brings everything full circle :).

Formerly deeplearn.js (https://deeplearnjs.org/). For those who were following deeplearn.js, from the TensorFlow.js FAQ:

> "What is the difference between TensorFlow.js and deeplearn.js?"

> TensorFlow.js is an ecosystem of JavaScript tools for machine learning that evolved from deeplearn.js. deeplearn.js is now called TensorFlow.js Core. TensorFlow.js also includes a Layers API—a higher level library for building machine learning models—as well as tools for automatically porting TensorFlow SavedModels and Keras HDF5 models.

Github repo has been redirected. Nikhil Thorat (nsthorat) and Daniel Smilkov (dsmilkov) look to still be leading. Formal announcement expected at 10:30am PST at the TensorFlow Dev Summit (livestream: https://www.youtube.com/watch?v=gplTc2F5Wvk)

Reply on news.ycombinator.com