3 ms·
> Also, what does "fully converged" signify? In machine learning/deep learning, the decrease in training loss has major diminishing returns as training continu
by minimaxir 10y ago
> Also, what does "fully converged" signify?
In machine learning/deep learning, the decrease in training loss has major diminishing returns as training continues. Eventually, training the model hits a point where the loss barely improves each epoch/iteration. (fun visualization from one of my projects: http://minimaxir.com/img/char-embeddings/epoch-losses.png http://minimaxir.com/img/char-embeddings/epoch-losses.png)
In some cases, the loss can stop improving entirely, or increase.
- carbocation 10y agoAny recs on choosing TensorFlow vs Keras-on-TensorFlow?
- minimaxir 10y agoKeras does not add much overhead, if any, and yes, it is as easy as everyone claims.
- jacquesm 10y agoStart with Keras, if you run into something you want to do that is not supported by Keras drop into TensorFlow, they are not mutually exclusive and all of TensorFlow is availble.
- carbocation 10y agoFantastic. Thanks to you and minimaxir for the guidance.