sketch-rnn-experiments

KO | EN

gitlite — search

#python #java #python3 #arduino #golang #machine-learning #rust #html #flask #javascript #seismology #nodejs

sketch-rnn-experiments

★ 11 Open GitHub ↗

Experiments with Sketch-RNN and the Quick Draw dataset

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

A playground for experiments with the [Quick Draw dataset](https://github.com/googlecreativelab/quickdraw-dataset) and [Sketch-RNN](https://github.com/tensorflow/magenta/tree/master/magenta/models/sketch_rnn). Sorry about the mess.

`get_perplexities`, the function in the iPython notebook for calculating the loss per sketch in the dataset relies on a one-line local modification I made to Magenta to make it possible to grab the pre-aggregated loss of a batch. I added the following line after the call to `get_lossfunc` in sketch_rnn/model.py: `self.lossfunc = lossfunc`.

If you don't want to bother building Magenta from source, you can use `_get_perplexities` with a model having `hps.batch_size = 1`, but it's an order of magnitude slower.

To run the notebook code, you'll need at least one pre-trained Sketch-RNN model, saved locally to `models/` (see `sketch_rnn_train.download_pretrained_models`), and at least one slice of Quick Draw data as an `npz` file saved locally to `data/`. Links [here](https://github.com/googlecreativelab/quickdraw-dataset#sketch-rnn-quickdraw-dataset). I used flamingo/owl models and datasets, plus a few datasets for unrelated categories.

sketch-rnn-experiments

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README