Merge pull request #1175 from SciML/ChrisRackauckas-patch-3

Check new compats

Author: ChrisRackauckas

Project.toml

@@ -52,14 +52,14 @@ SciMLSensitivityMooncakeExt = "Mooncake"
 ADTypes = "1.9"
 Accessors = "0.1.36"
 Adapt = "1.0, 2.0, 3.0, 4"
-AlgebraicMultigrid = "0.6.0"
+AlgebraicMultigrid = "1"
 Aqua = "0.8.4"
 ArrayInterface = "7"
 Calculus = "0.5.1"
 ChainRulesCore = "0.10.7, 1"
 ComponentArrays = "0.15.5"
 DelayDiffEq = "5.43.2"
-DiffEqBase = "6.151.1"
+DiffEqBase = "6.166.1"
 DiffEqCallbacks = "4"
 DiffEqNoiseProcess = "5.19"
 Distributed = "1"
@@ -92,7 +92,7 @@ RecursiveArrayTools = "3.27.2"
 Reexport = "1.0"
 ReverseDiff = "1.15.1"
 SafeTestsets = "0.1.0"
-SciMLBase = "2.51.4"
+SciMLBase = "2.79"
 SciMLJacobianOperators = "0.1"
 SciMLOperators = "0.3"
 SciMLStructures = "1.3"

test/sde_neural.jl

@@ -99,7 +99,7 @@ Random.seed!(238248735)
             sensealg = ReverseDiffAdjoint()))
         tmp_mean = mean(tmp_sol, dims = 3)[:, :]
         tmp_var = var(tmp_sol, dims = 3)[:, :]
-        sum(abs2, truemean - tmp_mean) + 0.1 * sum(abs2, truevar - tmp_var), tmp_mean
+        sum(abs2, truemean - tmp_mean) + 0.1 * sum(abs2, truevar - tmp_var)
     end
 
     function loss_op(θ)
@@ -112,11 +112,11 @@ Random.seed!(238248735)
             sensealg = ReverseDiffAdjoint()))
         tmp_mean = mean(tmp_sol, dims = 3)[:, :]
         tmp_var = var(tmp_sol, dims = 3)[:, :]
-        sum(abs2, truemean - tmp_mean) + 0.1 * sum(abs2, truevar - tmp_var), tmp_mean
+        sum(abs2, truemean - tmp_mean) + 0.1 * sum(abs2, truevar - tmp_var)
     end
 
     losses = []
-    function callback(θ, l, pred)
+    function callback(state, l)
         begin
             push!(losses, l)
             if length(losses) % 50 == 0
@@ -189,12 +189,12 @@ end
             sensealg = BacksolveAdjoint(autojacvec = ReverseDiffVJP()),
             saveat = ts, trajectories = 10, abstol = 1e-1, reltol = 1e-1)
         A = convert(Array, _sol)
-        sum(abs2, A .- 1), mean(A)
+        sum(abs2, A .- 1)
     end
 
     # Actually training/fitting the model
     losses = []
-    function callback(θ, l, pred)
+    function callback(state, l)
         begin
             push!(losses, l)
             if length(losses) % 1 == 0