torchdyn

torchdyn copied to clipboard

defunc is not defined in this notebook. Could you include code for this function? tutorials/module4-model/m4e_lagrangian_nets.ipynb [https://github.com/DiffEqML/torchdyn/blob/310c3750e35115a9e98fab78ff302440c8634843/tutorials/module4-model/m4e_lagrangian_nets.ipynb](url)

breesebj

When run the `pip install torchdyn`, it automatically installs the torch and uninstalls the old version installed in the environment. This makes touble if we have already installed a good...

cauchy7203

In torchdiffeq, odeint has a adjoint option to turn on seminorm? how should i pass parameters to the seminorm option in torchdyn? Merely passing a True gives an error. `dyn_odeint(...

chansigit

Example of controlled dynamical system?

2

I'm trying to fit a Neural ODE to some observations of a dynamical system, which has some control inputs. That is, the control inputs are not part of the state...

rlrs

In torchdyn -> numerics -> sensitivity.py function **_gather_odefunc_adjoint()**, line **71**: ``` dμ = torch.cat([el.flatten() if el is not None else torch.zeros(1) for el in dμ], dim=-1) ``` should be fixed...

ljxw88

__init__() got an unexpected keyword argument 'func'

2

Hi, I am trying to implement neuralSDE using torchdyn, my code is: ``` # drift_func f = nn.Sequential(...) # diffusion function g = nn.Sequential(...) self.func = NeuralSDE(f, g, solver=args.solver, rtol=args.rtol,...

ClaudiaShu

I am just running the lagrangian net tutorial and getting this error Any idea ?

interXadmin

Model require `forward` to have `*args, **kwargs` in 1.0.6

2

`torchdyn>=1.0.3` breaks if a model does not have *args, **kwargs in the `forward` method. This breaks downstream code which has custom models compatible with previous torchdyn versions. I propose to...

atong01

Thanks for this really amazing package! I would like to know if the feature associated with parallel integration routines with different t_spans for each sample in a batch is being...

YouvenZ

**There are mismatched arguments in `problems.ODEProblem.odeint`** My torchdyn version is 1.0.3 **Step to Reproduce** I want to see how many steps did the adaptive dopri5 solver take, so I sought...

cantabile-kwok