How to introduce the manual weight and bias?

1

I want to use the same weight and bias for the two model respectively at 1st epoch and need to be updated accordingly.

def init_normal(m):
    if type(m) == nn.Linear:
        nn.init.kaiming_uniform_(m.weight)
        if m.bias is not None:
            nn.init.constant_(m.bias.data, 0.001)
model = MLP(INPUT_DIM, OUTPUT_DIM)
model=model.apply(init_normal)

As i am doing the iteration ,the weight and bias is updated .And affecting the overall accuracy of the next model.
I tried the following peice of code

model1=copy.deepcopy(model)

this is working fine for first epoch but at the second epoch it is using the same weight and bias.

I want to use the updated one at the second epoch.

Any reference code will be highly appriciated…
I am new to pytorch…please support…

thanks with best regards,
pramesh