-
TypeError: in converted code:
/home/jupyter-tide1994/tide/embedding/GraphNeuralNetwork-master/gnn/gcn.py:73 call *
TypeError: Input 'pred' of 'Switch' Op has type float32 that does not match…
-
We currently have the following definitions
```julia
# DataLoader compatibility passing a vector of graphs and
# effectively using `batch` as a collated function.
MLUtils.numobs(data::Vector{ shuf…
-
Related to https://github.com/FluxML/NNlib.jl/issues/411
```julia
using GraphNeuralNetworks
using CUDA
using Flux: gpu
g = rand_graph(10, 30) |> gpu
g = add_edges(g,[10],[10])
x = CUDA.rand(3, …
-
We currently have `SimpleGraph`/`SimpleDiGraph` -> `GNNGraph`. We should add the other direction.
-
There are quite a few useful datasets, benchmarks and leaderboards in [ogb](https://ogb.stanford.edu/). The paper is [here](https://arxiv.org/abs/2005.00687).
From what I can see it plays nicely w…
-
1. Node level tasks. We can use `Withgraph` to embed the graph. However, if I want to save the learned parameters and use them for another graph (transfer learning), it would be a problem since the gr…
-
After updating to the latest version, GPU executing stopped working for me. I have not created a minimal reproducible example yet, but this is the error I'm facing.
```julia
Failed to compile PTX c…
-
```julia
using GraphNeuralNetworks
using Flux
function test_nn(nn, x)
@show nn
println("CPU")
@show nn(x)
println("GPU")
x = Flux.gpu(x)
nn = Flux.gpu(nn)
@show…
-
Hello! Nice work on the library; it is very usable. I'm trying to calculate the hessian-vector product of a loss function involving `GNNGraph` datapoints and a `GNNChain` model. I've been using the [S…
-
When using Flux.DataLoader loading graph batches is slower than expected. It really slows everything down when dealing with large training sets.
A comparison to using vectors with the same amount of …