[Diff since v2.0.0](https://github.com/SciML/DiffEqFlux.jl/compare/v2.0.0...v2.1.0)

**Closed issues:**
- MethodError when running with NeuralODE with custom Lux layer (#825)
- Running NODEs on GPU while keeping them Float64? (#826)
- UODE with neural networks. Beginner questions (#831)
- Scalar indexing when running UDE on GPU (#832)

**Merged pull requests:**
- Update tests for Float32 (#824) (@ChrisRackauckas)
- OptimizationFlux (#828) (@anandijain)
- Updates to HNN (#829) (@avik-pal)
- Migrate GPU tests to Lux (and Fix them!!) (#830) (@avik-pal)

[Diff since v1.53.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.53.0...v2.0.0)

**Closed issues:**
- DiffEqFlux Layers don't satisfy Lux API (#727)
- SDEs gradients issues with EnsembleProblems (#765)
- DiffEqFlux.jl and NeuralPDE.jl keep precompiling and never stop (#783)
- Training Neural SDEs with Mutating Arrays (#785)
- Normalizing flow example does not produce accurate estimates (#786)
- Wrong order in collocations (#788)
- StackOverflow Error when trying to train a NSDE with no neural network on the diffusion (#811)
- Including unknown parameter values of known dynamics in UDE using Lux.jl (#812)
- Documentation about Flux.jl vs Lux.jl (#815)

**Merged pull requests:**
- Lux compatible layers (#750) (@Abhishek-1Bhatt)
- Fix FFJORD Example (#787) (@prbzrg)
- Fix flipped collocation (#789) (@ChrisRackauckas)
- [skip ci] LanguageTool (#792) (@ArnoStrouwen)
- CompatHelper: add new compat entry for Functors at version 0.4, (keep existing compat) (#793) (@github-actions[bot])
- Deprecate FastChain and sciml_train for v2.0 (#794) (@ChrisRackauckas)
- One more spot to mark with Flux (#796) (@Abhishek-1Bhatt)
- Flux.cpu (#797) (@Abhishek-1Bhatt)
- Specify Flux in all layers (#799) (@Abhishek-1Bhatt)
- Remove unused packages and undefined exports (#803) (@prbzrg)
- Switch to LuxCore (#804) (@Abhishek-1Bhatt)
- fix `NeuralCDDE` error (#805) (@vpuri3)
- Add neural ODE weather forecasting example (#807) (@SebastianCallh)
- CompatHelper: add new compat entry for DataDeps at version 0.7 for package docs, (keep existing compat) (#808) (@github-actions[bot])
- CompatHelper: add new compat entry for DataFrames at version 1 for package docs, (keep existing compat) (#809) (@github-actions[bot])
- CompatHelper: add new compat entry for CSV at version 0.10 for package docs, (keep existing compat) (#810) (@github-actions[bot])
- Use CA directly instead of using it via Lux (#813) (@avik-pal)
- Add a description of Flux vs Lux (#819) (@ChrisRackauckas)
- enable dependabot for GitHub actions (#821) (@ranocha)
- CompatHelper: bump compat for DataInterpolations to 4, (keep existing compat) (#823) (@github-actions[bot])

[Diff since v1.52.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.52.0...v1.53.0)

**Closed issues:**
- Performance drop when enabling bias in FastDense (#757)
- Specifying adjoint methods for gradient calculations? (#766)
- UndefVarError: FastDense not defined (#767)
- ERROR: MethodError: no method matching length(::Nothing) (#768)
- ERROR: MethodError: no method matching copy (#770)

**Merged pull requests:**
- Fix a typo (#754) (@prbzrg)
- Flux -> Lux (#756) (@YichengDWu)
- Add Dependabot (#758) (@prbzrg)
- Bump codecov/codecov-action from 1 to 3 (#759) (@dependabot[bot])
- Bump actions/checkout from 2 to 3 (#760) (@dependabot[bot])
- Bump actions/cache from 1 to 3 (#761) (@dependabot[bot])
- [skip ci] badges (#763) (@ArnoStrouwen)
- MassInstallAction: Install the Invalidations workflow on this repository (#764) (@devmotion)
- Update CompatHelper.yml (#769) (@devmotion)
- updated README and make.jl as per sciml template (#774) (@00krishna)
- canonize docs (#775) (@ArnoStrouwen)
- added reproducibility to index.md and make.jl (#777) (@00krishna)
- CompatHelper: bump compat for LoggingExtras to 1, (keep existing compat) (#778) (@github-actions[bot])
- Update Links (#779) (@prbzrg)
- doc compat (#782) (@ArnoStrouwen)

[Diff since v1.51.2](https://github.com/SciML/DiffEqFlux.jl/compare/v1.51.2...v1.52.0)

**Closed issues:**
- Suggestion:  Change dimension order on NeuralODE to match convention of ML batch (#111)
- ERROR: LoadError: MethodError: no method matching getindex(::Ptr{typeof(DiffEqBase.ODE_DEFAULT_NORM)}) (#714)
- Warning: `ignore(f)` is deprecated, use `ChainRulesCore.ignore_derivatives(f)` instead. (#745)
- Problems defined by parametric and neural network functions (#746)
- Tutorials Copy-Pasteable Code Error: LoadError: MethodError: objects of type Nothing are not callable (#748)

**Merged pull requests:**
- don't create a plot if not displaying it (#747) (@chriselrod)
- update Zygote.@ignore to ChainRulesCore.@ignore_derivatives (#751) (@ChrisRackauckas)
- CompatHelper: add new compat entry for ChainRulesCore at version 1, (keep existing compat) (#752) (@github-actions[bot])
- Dimension Mover Layer (#753) (@prbzrg)

[Diff since v1.51.1](https://github.com/SciML/DiffEqFlux.jl/compare/v1.51.1...v1.51.2)

[Diff since v1.51.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.51.0...v1.51.1)

[Diff since v1.50.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.50.0...v1.51.0)

**Closed issues:**
- Neural Graph Differential Equations example not working (#684)
- Incorrect use `Lux.ActivationFunction` (#737)

**Merged pull requests:**
- add cnf to the sidebar (#734) (@ChrisRackauckas)
- Neural SDE tutorial (#735) (@ChrisRackauckas)
- neural_sde example in Flux (#738) (@Abhishek-1Bhatt)
- Fix latex typos in Hamiltonian Neural Network Tutorial (#741) (@Algopaul)
- Lux.ActivationFunction not required (#742) (@Abhishek-1Bhatt)
- update to SciMLSensitivity (#743) (@ChrisRackauckas)

[Diff since v1.49.1](https://github.com/SciML/DiffEqFlux.jl/compare/v1.49.1...v1.50.0)

**Closed issues:**
- Add doctests (#566)
- Missing normalizing flows example (#726)
- `NeuralODE` is not a subtype of `Lux.AbstractExplicitLayer` (#728)
- Neural ODE example is not working correctly (#731)

**Merged pull requests:**
- Enable doctesting (#721) (@ChrisRackauckas)
- minor fixes (#729) (@Abhishek-1Bhatt)
- just remove staticdudt tests (#730) (@ChrisRackauckas)

[Diff since v1.49.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.49.0...v1.49.1)

**Closed issues:**
- NeuralDSDE Adjoint Fails on GPU (#138)
- Which types work as container of parameters? (#178)
- NeuralODE: modeling dynamics of system with (discontinuous) control inputs (#190)
- Nested reverse modes with fast layers (#207)
- BFGS Errors with GPU (#238)
- BoundsError: attempt to access () at index [0] (#253)
- Code in LV-Jump.md blog post no longer runs (#272)
- Reproducible TypeError in sciml_train (#278)
- Why the need for Fast* layers? (#328)
- Rename DiffEqFlux? (#354)
- The MNIST Neural ODE Classifier tutorial is broken  (#368)
- sciml_train crashes when fed Fluxs ::Params type (#379)
- Neural Graph Differential Equation tests disabled (#405)
- Link to stable documentation broken (#407)
- neural_ode_sciml example fails when Dense layer replaced by GRU (#432)
- Indexing issue caused by applychain (#436)
- Mutating Arrays not Allowed (#443)
- Optimal Control ref in exogenous input page is broken (#445)
- Support for sparse matrix ODEs (#462)
- implicit layers with neuralODEs ? (#478)
- Disable events tests (#483)
- PresetTimeCallback errors with Zygote differentiation (#502)
- Suggestion for Memory Limitations (#527)
- Unstable `FastDense` `hash` and comparison. (#555)
- SteadyStateAdjoint + GPUs: scalar getindex is disallowed (#557)
- sample code running slow on Atom (#560)
- ReverseDiff tape generation over newer Zygote versions segfaults (#586)
- ERROR: UndefVarError: eJ not defined  (#588)
- Neural Graph Tutorial not working (#609)
- Note on hybrid sensitivity analysis methods needs an update? (#612)
- ERROR: LoadError: MethodError: no method matching Float32(::ReverseDiff.TrackedReal{Float64, Float32, Nothing}) (#615)
- type HMCKernel has no field integrator (#618)
- Reverse-mode VJPs with mixed scalar+vector GPU code (#632)
- Error in Neural Graph Ordinary Differential Equations example (#640)
- Best practice for taking backpropagation with logging tools for DiffEq.jl (#661)
- Add batch support to FastDense layers, mimicking Flux (#672)
- Missing optimizations in preallocation form (#680)
- Error in the UODEs training when the network enlarges (#689)
- initial_params() always identical (#690)
- PINN with TensorLayer example fails - maybe Enzyme internal error (#694)
- MethodError: no method matching Flux.Dense(::Pair{Int64, Int64}, ::typeof(NNlib.elu)) (#700)
- `sciml_train` output API documentation (#702)
- LoadError with sciml_train (#713)
- Training a NODE using multiple shooting.  (#717)

**Merged pull requests:**
- Corrections in keyword arguments for NeuralODE example (#720) (@Abhishek-1Bhatt)
- Removal of deprecations (#722) (@Abhishek-1Bhatt)
- Fix Some Doctests (#723) (@DhairyaLGandhi)
- Fixed NeuralDSDE (#724) (@Abhishek-1Bhatt)
- Update neural_de tests (#725) (@frankschae)

[Diff since v1.48.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.48.0...v1.49.0)

**Closed issues:**
- Fast vjp form (#681)

**Merged pull requests:**
- swapped sciml_train calls with Optimization.solve() (#711) (@Abhishek-1Bhatt)
- CompatHelper: add new compat entry for Lux at version 0.4, (keep existing compat) (#716) (@github-actions[bot])
- Neural ODE tutorial (#718) (@Abhishek-1Bhatt)

[Diff since v1.47.1](https://github.com/SciML/DiffEqFlux.jl/compare/v1.47.1...v1.48.0)

**Closed issues:**
- Scaling and memory consumption on GPU (#687)

**Merged pull requests:**
- Generalize `input_params` to use `Flux.destructure` (#697) (@DhairyaLGandhi)
- Overhaul the documentation to only focus on the DiffEqFlux components (#709) (@ChrisRackauckas)
- add deprecation warnings to FastChain and sciml_train (#710) (@ChrisRackauckas)
- Switch from GalacticOptim, Galactic* to Optimization, Optimization* (#712) (@Vaibhavdixit02)

[Diff since v1.47.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.47.0...v1.47.1)

**Merged pull requests:**
- Keep support for `cb` to avoid breakage (#705) (@Vaibhavdixit02)

[Diff since v1.46.1](https://github.com/SciML/DiffEqFlux.jl/compare/v1.46.1...v1.47.0)

**Merged pull requests:**
- CompatHelper: bump compat for GalacticOptim to 3, (keep existing compat) (#704) (@github-actions[bot])

[Diff since v1.46.0](https://github.com/SciML/DiffEqFlux.jl/compare/v1.46.0...v1.46.1)

[Diff since v1.45.3](https://github.com/SciML/DiffEqFlux.jl/compare/v1.45.3...v1.46.0)

**Closed issues:**
- Enzyme segfault when using FastChain and AutoZygote (#691)
- DimensionMismatch: arrays could not be broadcast to a common size; (#701)

**Merged pull requests:**
- test Flux 0.13 (#699) (@ChrisRackauckas)

[Diff since v1.45.2](https://github.com/SciML/DiffEqFlux.jl/compare/v1.45.2...v1.45.3)

**Merged pull requests:**
- add prediction-error method tutorial (#688) (@baggepinnen)
- Disable Graph Conv test (#693) (@ChrisRackauckas)