-
In Neural_ode.py, we're using the tf automatic differentiation API for calculating the differential, so do we use the adjoint method to calculate the gradient? Sorry but getting a little mixed up here…
-
Dear all,
I'm trying to work with toy ODE that is predator prey.
```
using DifferentialEquations, Plots, ParameterizedFunctions, Random, Distributions, Turing, StatsPlots
predator_prey = @ode…
-
First off, awesome paper, thank you!
When I try to run the NeuralODE example on CPU, it works great. However, when I try to switch it over to the GPU using the approach described in the paper i.e.,…
-
Hi! First of all, thank you very much for this package! I've been working in the connection between dynamical systems and machine learning for a couple of years and this package will be extremely help…
-
I accidentally stopped the tracking in the neural_ode layer and of course it can't backprop through since what comes out is just a float (@MikeInnes told me this would happen). So an MWE of trying to …
-
```
WARNING: The config value `html_theme_path' has type `str', defaults to `list'.
building [mo]: targets for 0 po files that are out of date
building [html]: targets for 98 source files that are …
-
One small problem is that `cur_pred` inside the `cb` function is tracked and has to be wrapped in `Flux.data`. After doing this, calling `cb()` works fine and correctly shows the plot.
A bigger pro…
-
Hey Ricky,
I'm running out of memory during the backward pass on a 16gb gpu when running the adjoint method with rtol 1e-5, atol 1e-5, and a network with 631058 parameters.
I'm not sure why this…
-
Hello
I have a 1 issue for a looong time.
When I enable DKIM threw mail-cow administration, create DNS record and so on, everything is working fine... But everytime on second day, DKIM is not work…
-
I am using a 1070 and 8GB graphic card, and I still get the OOM resource exhausted method in between training a mnist for default configuration. I even Reduced the batch sizefrom 32->16->8 , can you h…