Parameter containing nan

[longnguyenQB]

Hello, I’m new in deep learning.
Please help me fix this error, my model:

HybridGNN(
  (leaky_relu): LeakyReLU(negative_slope=0.01)
  (gnn_0): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_1): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_2): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_3): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_4): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_5): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_6): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (gnn_7): SpGAT(
    (attention_0): SpGATLayer (8 -> 12)
    (attention_1): SpGATLayer (8 -> 12)
    (attention_2): SpGATLayer (8 -> 12)
    (out_att): SpGATLayer (36 -> 72)
  )
  (fc): Sequential(
    (0): LazyLinear(in_features=0, out_features=72, bias=True)
    (1): LeakyReLU(negative_slope=0.01)
    (2): Linear(in_features=72, out_features=36, bias=True)
    (3): ReLU()
    (4): Linear(in_features=36, out_features=2, bias=True)
  )
)

This is the error I get:
![image](https://user-images.githubusercontent.com/76984368/161428197-cbae2af1-8ddc-43b5-a4ea-a881ddae5958.png)