Open suman1209 opened 1 month ago
You can now generate plots by running main.py in the 'main' branch, Please adjust the timesteps in the inference and the epochs to play around and experiment with results. I noticed that it takes a very long time for sampling / recommendations when you use T = 1000
I created a plotting.py to compare against all three baselines using the .json inside /results
There are path related things that need to be checked when running in your local pc, but running main.py should give you some results like the one below
Task Details
the task is to generate the performance plots after training the diffusion model
TODOs
[ ] inside main.py , use irene.methods.MyAlgorithm and get the relevant final plots
[x] update the code to use comapare against all three baselines of "RandomSearch", "Gaussian Process" and "Deep Gaussian Process"