mirkosavasta / ganetano Goto Github PK
View Code? Open in Web Editor NEWGenerative Adversarial Network to create synthetic time series
Generative Adversarial Network to create synthetic time series
Hi, Thank you for sharing this code!
I have a question about how one can use this as a datagenerator. I see that the trainer is the only part that can be used to generate samples. Should one save that using torch.save and then use it to generate several samples of the real data?
Another question. If I have time-series of size NxT, could one generate synthetic data of that size too? Right now the code generates data of size 16x50.
Do you have any tips of how to test whether the generated data share similar properties as the real data?
I really appreciate your help on this.
Hi,
After I laod 'g_state_dict' to model
, how can I use it to generate new data?
Hi,
if plot_training_samples
is set to False
, the variable noise_length
is not set and the main training loop (l. 147) will fail because it will try to sample dynamic_latents
with a shape that is only set above (l. 140) when the condition is met.
Hi,
I don't understand why the len method of your custom datasets is returning the number of datapoints.
Shouldn't it be returning the number of samples which here is the number of series n_series?
Thanks!
Hey, very nice repo. Thank you for making your work available.
Question: I have been studying your code and cannot for the life of me figure out how the Critic class dimensions are being established. The Generator seems fairly straightforward, the input dimension from the sines dataset is 100 datapoints. But after attempting to modify those values in both the datasets.py and wgangp.py, I am still getting a deeply nested runtime that indicates the Critic shape is wrong. But I'm not seeing how the Critic gets its input shape?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.