problem with set_weights

General
1

Hi all,
I’m new with Flower, and I’m using an autoencoder with TensorFlow for federated learning, but I’m encountering this error: ValueError: You called s) oset_weights(weightn layer 'sequential_1' with a weight list of length 12, but the layer was expecting 0 weights.

I initialize the global model and compile it without fit so it has 0 weights
can you suggest me a solution?
Thanks

2 Likes
2

Hi @sinda1980 , let me point you to the quickstart-tensorflow example. There, the ServerApp initializes the model (and therefore fit() has never been called).

Please take a look and let us know if this helps!

1 Like