TII_ObliviousNN

This README would normally document whatever steps are necessary to get your application up and running.

Matrix triples

By default, these sample programs use generated convolutional triples with 0 as values for simplicity. A program called matrix_triples_generation.mpc is provided as an example of how real triples can be generated and stored using the Input/Output system.

Running the Oblivious Neural Networks

Lenet and Pruned Resnet can be executed in the traditional FANNG-MPC way and using any available Input/Output system. However, some scripts have been created to make them easier to run using the Input/Output filesystem without having to answer each question about file paths manually.

Lenet and Pruned Resnet can be easily executed using the run-lenet.sh and run-pruned-resnet.sh in this way:

• ./run-lenet <FANNG-MPC main folder path> <path_to_lenet_relative_to_FANNG-MPC>.

  Example: ./run-lenet FANNG-MPC ../test_obliv_nn_lenet

The files generated from the pytorch code with the parameters, input and output files must be located inside the FANNG-MPC Data folder with LenetData and PrunedResnetData names respectively. generate-lenet-answers-for-io-fs.sh and generate-pruned-resnet-answers-for-io-fs.sh scripts are responsible for generating responses necessary to use the input/output filesystem when LeNet and Pruned Resnet are respectively executed using those files. If you want to use a custom data structure for these files, you have to change the paths of them inside these scripts.

Communications channels in use

ONN for MNIST

Parameters

Range of channels reserved: 100000 - 100019

Channels in use:

• 100000 → LenetData/Parameters/Default-parameters/P0-classifier.0.bias.txt
• 100001 → LenetData/Parameters/Default-parameters/P0-classifier.0.weight.txt
• 100002 → LenetData/Parameters/Default-parameters/P0-classifier.2.bias.txt
• 100003 → LenetData/Parameters/Default-parameters/P0-classifier.2.weight.txt
• 100004 → LenetData/Parameters/Default-parameters/P0-feature_extractor.0.bias.txt
• 100005 → LenetData/Parameters/Default-parameters/P0-feature_extractor.0.weight.txt
• 100006 → LenetData/Parameters/Default-parameters/P0-feature_extractor.3.bias.txt
• 100007 → LenetData/Parameters/Default-parameters/P0-feature_extractor.3.weight.txt
• 100008 → LenetData/Parameters/Default-parameters/P0-feature_extractor.6.bias.txt
• 100009 → LenetData/Parameters/Default-parameters/P0-feature_extractor.6.weight.txt

First layer input

Range of channels reserved: 90000 - 90049

• 90000 → execution 1 ./LenetData/Inputs/Execution_0/P0-feature_extractor_0.input.txt
• 90001 → execution 2 ./LenetData/Inputs/Execution_1/P0-feature_extractor_0.input.txt
• 90002 → execution 3 ./LenetData/Inputs/Execution_2/P0-feature_extractor_0.input.txt

ONN for CIFAR10

Parameters

Range of channels reserved: 100020 - 100049

Channels in use:

Bias

• 100020 → PrunedResnetData/all_parameters/bias/P0-conv_layer_1_bias.txt
• 100021 → PrunedResnetData/all_parameters/bias/P0-conv_layer_2_bias.txt
• 100022 → PrunedResnetData/all_parameters/bias/P0-conv_layer_3_bias.txt
• 100023 → PrunedResnetData/all_parameters/bias/P0-conv_layer_4_bias.txt
• 100024 → PrunedResnetData/all_parameters/bias/P0-conv_layer_5_bias.txt
• 100025 → PrunedResnetData/all_parameters/bias/P0-conv_layer_6_bias.txt
• 100026 → PrunedResnetData/all_parameters/bias/P0-linear_layer_1_bias.txt
• 100027 → PrunedResnetData/all_parameters/bias/P0-linear_layer_2_bias.txt

Beta

• 100028 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_1_new_beta.txt
• 100029 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_2_new_beta.txt
• 100030 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_3_new_beta.txt
• 100031 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_4_new_beta.txt
• 100032 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_5_new_beta.txt
• 100033 → PrunedResnetData/all_parameters/new_beta/P0-bn_layer_6_new_beta.txt

Gamma

• 100034 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_1_new_gamma.txt
• 100035 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_2_new_gamma.txt
• 100036 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_3_new_gamma.txt
• 100037 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_4_new_gamma.txt
• 100038 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_5_new_gamma.txt
• 100039 → PrunedResnetData/all_parameters/new_gamma/P0-bn_layer_6_new_gamma.txt

Weights

• 100040 → PrunedResnetData/all_parameters/weights/P0-conv_layer_1_weight.txt
• 100041 → PrunedResnetData/all_parameters/weights/P0-conv_layer_2_weight.txt
• 100042 → PrunedResnetData/all_parameters/weights/P0-conv_layer_3_weight.txt
• 100043 → PrunedResnetData/all_parameters/weights/P0-conv_layer_4_weight.txt
• 100044 → PrunedResnetData/all_parameters/weights/P0-conv_layer_5_weight.txt
• 100045 → PrunedResnetData/all_parameters/weights/P0-conv_layer_6_weight.txt
• 100046 → PrunedResnetData/all_parameters/weights/P0-linear_layer_1_weight.txt
• 100047 → PrunedResnetData/all_parameters/weights/P0-linear_layer_2_weight.txt

First layer input

Range of channels reserved: 90050 - 90059

• 90050 → execution 1 ./PrunedResnetData/Inputs/Execution_0/P0-Conv_layer_1.input.txt