Machine Learning in Elixir:294 Axon.Loop.log/4 is undefined or private. Did you mean: * log/2 * log/3

Pragmatic Bookshelf PragProg Customers
,
2

Thanks for pointing out. The correct function is:

|> Axon.Loop.log(fn
  %Axon.Loop.State{epoch: epoch, iteration: iter, step_state: state} ->
    d_loss = state[:loss]["discriminator"]
    g_loss = state[:loss]["generator"]
    "\rEpoch: #{epoch}, batch: #{iter},"
      <> " d_loss: #{Nx.to_number(d_loss)},"
      <> " g_loss: #{Nx.to_number(g_loss)}"
end, event: :iteration_completed, device: :stdio)
1 Like