DiffEqFlux.jl icon indicating copy to clipboard operation
DiffEqFlux.jl copied to clipboard

Pre-built implicit layer architectures with O(1) backprop, GPUs, and stiff+non-stiff DE solvers, demonstrating scientific machine learning (SciML) and physics-informed machine learning methods

Results 90 DiffEqFlux.jl issues
Sort by recently updated
recently updated
newest added

* https://pubs.acs.org/doi/10.1021/acs.jpca.0c09316 * https://www.frontiersin.org/articles/10.3389/fsysb.2024.1338518/full

**Questionโ“** What do we feel about moving some of the non-diffeq specific layers to `Boltz.jl` and re-exporting them from here? I was working on implementing KANs and we might be...

good first issue
question

This pull request changes the compat entry for the `LuxCore` package from `0.1` to `0.1, 1`. This keeps the compat entries for earlier versions. Note: I have not tested your...

fixes #929 needs https://github.com/LuxDL/Boltz.jl/pull/54

- removes dependencies that aren't being used - loads DataInterpolations (fixes #935) - removes duplicated tests

**Describe the bug ๐Ÿž** Collocation with interpolators throws an error since `DataInterpolations` is not imported. **Minimal Reproducible Example ๐Ÿ‘‡** ```julia x = randn(1001) |> cumsum t = collect(0.0:1000) t2 =...

bug

Update the workflows in this repository to use SciML's reusable workflows. This is part of a larger effort to standardize the SciML's CI workflows for more generic and common requirements,...

Hello, I tried [the continuous normalizing flows example in the documentation](https://docs.sciml.ai/DiffEqFlux/stable/examples/normalizing_flows/), although the code kept running, it threw exceptions at ```julia res1 = Optimization.solve( optprob, OptimizationOptimisers.Adam(0.01); maxiters = 20, callback...

bug

**Describe the bug ๐Ÿž** MNIST example throws exception at solve. https://docs.sciml.ai/DiffEqFlux/stable/examples/mnist_neural_ode/ **Expected behavior** A clear and concise description of what you expected to happen. **Minimal Reproducible Example ๐Ÿ‘‡** Without MRE,...

bug

Hi! Just wondering how the RNN could be mixed into the `ODEProblem` In flux times, it seems a Recur layer need to be created. However there is already a `Recurrence`...

question