In the Quickstart Jupyter Notebook, under hypersolvers, running this block:
net = nn.Sequential(nn.Linear(3, 64), nn.Softplus(), nn.Linear(64, 64), nn.Softplus(), nn.Linear(64, 3))
hypersolver = HyperEuler(VanillaHyperNet(net))
t_eval, sol = odeint(sys, x0, t_span, solver=hypersolver) # note: this has to be trained!
Produces this error:
File torchdyn/numerics/odeint.py:428, in _fixed_odeint(f, x, t_span, solver, save_at, args)
426 steps = 1
427 while steps <= len(t_span) - 1:
--> 428 _, x, _ = solver.step(f, x, t, dt, k1=None, args=args)
429 t = t + dt
431 if torch.isclose(t, save_at).sum():
TypeError: HyperEuler.step() got an unexpected keyword argument 'args'
Looks like this was changed for the "normal" ODE solvers in cb83c8c, but not in hyper.py. I can put up a PR to fix.
In the Quickstart Jupyter Notebook, under hypersolvers, running this block:
Produces this error:
Looks like this was changed for the "normal" ODE solvers in cb83c8c, but not in
hyper.py. I can put up a PR to fix.