import torch import torch.optim as op import torch.nn as nn network = nn.Sequential( nn.Linear(2, 5), nn.Tanh(), nn.Linear(5,2) ) X = torch.tensor([[1.0,1.0],[2.0,1.0]]) y = torch.tensor([[1.0,1.0],[3.0,2.0]]) test = network(X) print(test) #args = torch.tensor([1.0,1.0], requires_grad=True) opt = op.SGD(network.parameters(), lr=0.001) mse = nn.MSELoss() def train(model, opt, loss, X,y, epochs=1): for epoch in range(epochs): y2 = model(X) loss1 = loss(y2, y) loss1.backward() opt.step() opt.zero_grad() pass pass train(network, opt, mse, X, y, 1000) test2 = network(X) print(test2) loss_tensor = mse(network(X),y) #loss_value is the total loss loss_value = loss_tensor.item() x = torch.randn(1, 2) # Dictionary to store the intermediate outputs layer_outputs = {} current_input = x for i, layer in enumerate(network): # Pass the data sequentially through each layer current_input = layer(current_input) # Save the output of the current layer layer_outputs[f"layer_{i}_{layer.__class__.__name__}"] = current_input # Print results for layer_name, output in layer_outputs.items(): print(f"{layer_name} output shape: {output.shape}") #splitting the network into two models # 1. Split into two sub-models # Model 1 contains layers up to the Tanh activation (indices 0 and 1) model_part1 = network[:2] # Model 2 contains the final linear layer (index 2 onwards) model_part2 = network[2:] # 2. Test the split models with a dummy input x = torch.randn(1, 2) # Pass through Part 1 hidden_features = model_part1(x) print("Part 1 Output Shape:", hidden_features.shape) # Expected: [1, 5] # Pass the intermediate result through Part 2 final_output = model_part2(hidden_features) print("Part 2 Output Shape:", final_output.shape) # Extract weights as a list of nested lists linear_weights = [] for layer in network: if isinstance(layer, nn.Linear): # .detach() removes it from the computational graph # .tolist() automatically converts the PyTorch tensor into standard Python lists layer_weight_matrix = layer.weight.detach().tolist() linear_weights.append(layer_weight_matrix) # Print the final result print(linear_weights) pass