# About training strategy
**URL:** https://discourse.julialang.org/t/about-training-strategy/70528
**Category:** Optimization (Mathematical)
**Tags:** question, pde
**Created:** [October 28, 2021, 2:08am UTC](https://discourse.julialang.org/t/about-training-strategy/70528 "2021-10-28T02:08:17Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@iiisoo](https://discourse.julialang.org/u/iiisoo)
#### Post date: [October 28, 2021, 2:08am UTC](https://discourse.julialang.org/t/about-training-strategy/70528/1 "2021-10-28T02:08:17Z")
I am trying to run the example of PINN soling PDE( [https://github.com/SciML/NeuralPDE.jl](https://github.com/SciML/NeuralPDE.jl) )
But when I use the:
discretization = PhysicsInformedNN(chain, QuadratureTraining(),init\_params =initθ)
in later part of:
it is having some error here.
But if I change the training strategy to GridTraining:
discretization = PhysicsInformedNN(chain, GridTraining(dx),init\_params =initθ)
That error part is solved.
I am confused since the datatype of this : prob = discretize(pde\_system,discretization)
is always OptimizationProblem for both training strategy why I cannot use the QuadratureTraining here
---
### Author:  [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)
#### Post date: [October 30, 2021, 2:00pm UTC](https://discourse.julialang.org/t/about-training-strategy/70528/2 "2021-10-30T14:00:03Z")
This is a failure on master due to type piracy in DistributionsAD.jl. See [https://github.com/SciML/NeuralPDE.jl/issues/412](https://github.com/SciML/NeuralPDE.jl/issues/412). I want to get this fixed ASAP as well.
---
### Author:  [@iiisoo](https://discourse.julialang.org/u/iiisoo)
#### Post date: [November 1, 2021, 1:43am UTC](https://discourse.julialang.org/t/about-training-strategy/70528/3 "2021-11-01T01:43:55Z")
Thx