Jordi Bolibar
Jordi Bolibar
As discussed on Slack, this is just to track the fix for the performance issue using `EnsembleDistributed`. There are performance gains on the forward solve when using `EnsembleDistributed`, but there...
I'm having an issue while saving a `PyObject` together with PyCallJLD.jl. I'm doing: `JLD.save(joinpath(ODINN.root_dir, "test/data/gdirs.jld"), "gdirs", gdirs)` Which returns the following error: ``` Error encountered while save FileIO.File{FileIO.DataFormat{:JLD}, String}("/home/jovyan/Julia/fork/ODINN/test/data/gdirs.jld"). Fatal...
After a long time hunting a bug with @facusapienza21, we have realized that Zygote fails to provide a gradient for the basic `sqrt` function. This has been discussed at length...
Julia just released the highly awaited 1.9 version. This release contains some major performance improvements, so I would be super keen to try it on the JupyterHub. Would it be...
This is a little bit Julia specific, but the behaviour is not normal, so I'm wondering if this might come from the JupyterHub. Whenever one closes a Julia session, in...
These last days I've been having issues with my connection to JupyterHub servers. Today I've had my server stopped twice while running simulations. All of the sudden my connection dropped...
To check if a variable is equal to one of many values, combine the values into a tuple and check if the variable is contained in it instead of checking...
Following the latest developments and what has been discussed in the review paper, we should try and benchmark the new `GaussAdjoint` to see if it makes things faster.
Saving the `Results` struct in `.jl2` is not efficient, as it takes too much space and time. We should save results in netCDF for large simulations, and probably agree on...
To accompany the preprint and the first release, we should create a nice Jupyter notebook based on `ODINN.jl` (i.e. a high-level interface), showcasing how ODINN works and how it can...