torchdiffeq
torchdiffeq copied to clipboard
rtqichen
Differentiable ODE solvers with full GPU support and O(1)-memory backpropagation.
Hi, I've been trying to train a drift and event function but the parameters of my event function are not changing. Here's a simplified portion of the code I've been...
Hi! I've got an optimization question related to speeding up or re-organizing odeint calls for performance purposes. My current setup is that I have a hypernetwork that outputs the parameters...
Hi! I'm currently working on an experiment with ffjord, and Issues in the repository for ffjord seems not available at the moment, so I'm posting my issue here. The following...
Thanks a lot for your work! However, it seems that `backward` after `grad` is not supported yet. Here is a minimal example: ```python # dy/dt = a b y, t...
Hi, I'm trying to migrate my code solving ode from scipy to torchdiffeq, believing it is the solution. I have the ode function like ``` def rhs_unit_vector(z,t,J,K, n,omega): # several...
I was wondering how batching would work when using an event function. #122 describes a way of handling batching. But I can't figure out how to do this while being...
Here's a sample script that solves the 1D heat equation discretized using the method of lines. `torchdiffeq` turned out to be ~7x slower, which is not what I expected. I...
Hi, in situations where you'd be happy with a partial result but other considerations such as total time spent are more important, you'd want to be able to stop the...
Hi, great work, and thanks for the code! I was wondering if the following is possible. I have a system of ODEs (two ODEs): `x_state = f(x, t, theta)` `0...
Hi, I have instantiated a class like in issue #129 where I am passing in w and updating it. class BNM(nn.Module): def __init__(self, w): super(BNM, self).__init__() self.w = nn.Parameter(w) def...
