# runs all 3 failure configurations for all 3 models
if __name__ == "__main__":
accuracy = accuracy("Health")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data, num_vars = init_data()
load_for_inference = False
failout_survival_settings = [
[0.95, 0.95, 0.95],
[0.9, 0.9, 0.9],
[0.7, 0.7, 0.7],
[0.5, 0.5, 0.5],
]
no_information_flow_map = make_no_information_flow_map(
"Health", default_skip_hyperconnection_config)
# file name with the experiments accuracy output
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = "results/health_failout_rate" + mux_adds_str + ".txt"
verbose = 2
# keep track of output so that output is in order
output_list = []
output = make_output_dictionary_failout_rate(failout_survival_settings,
reliability_settings,
num_iterations)
dropout_like_failout = False
make_results_folder()
for iteration in range(1, num_iterations + 1):
output_list.append("ITERATION " + str(iteration) + "\n")
print("ITERATION ", iteration)
model_file,
data,
epochs,
batch_size,
verbose,
)
return model
# runs all 3 failure configurations for all 3 models
if __name__ == "__main__":
accuracy = accuracy("Camera")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data = init_data()
no_information_flow_map = make_no_information_flow_map(
"Camera", default_skip_hyperconnection_config)
load_for_inference = False
# file name with the experiments accuracy output
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = "results/camera_hyperconnection_weight" + mux_adds_str + ".txt"
verbose = 2
model_name = "ResiliNet Hyperconnection Weight"
hyperconnection_weightedbyreliability_config = 2
# keep track of output so that output is in order
output_list = []
output, weight_schemes = make_output_dictionary_hyperconnection_weight(
model_name, reliability_settings, num_iterations)
considered_weight_schemes = [1, 2, 3, 4]
default_reliability_setting = [1, 1, 1, 1, 1, 1, 1, 1]
skip_hyperconnection_configurations = [
# [f2,f3,f4,e1,e2,e3,e4]
[1, 1, 1, 1, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 1],
[1, 0, 0, 0, 1, 0, 0],
[0, 0, 0, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 0, 1],
[0, 0, 0, 0, 0, 1, 1],
[1, 1, 1, 1, 0, 0, 1],
[1, 1, 1, 1, 1, 1, 1],
]
no_information_flow_map = {}
for skip_hyperconnection_configuration in skip_hyperconnection_configurations:
no_information_flow_map[tuple(skip_hyperconnection_configuration
)] = make_no_information_flow_map(
"Camera",
skip_hyperconnection_configuration)
default_reliability_setting = [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
load_for_inference = False
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = ("results/camera_skiphyperconnection_sensitivity" +
mux_adds_str + ".txt")
verbose = 2
# keep track of output so that output is in order
output_list = []
model_name = "ResiliNet Hyperconnection Weight Sensitivity"
output = make_output_dictionary(
model_name,
reliability_settings,
1] = calc_expected_accuracy(
model,
no_information_flow_map,
reliability_setting,
output_list,
data=data,
)
# runs all 3 failure configurations for all 3 models
if __name__ == "__main__":
accuracy = accuracy("Health")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data, num_vars = init_data()
ResiliNet_no_information_flow_map = make_no_information_flow_map(
"Health", default_skip_hyperconnection_config)
DFG_no_information_flow_map = make_no_information_flow_map(
"Health", default_skip_hyperconnection_config)
Vanilla_no_information_flow_map = make_no_information_flow_map("Health")
load_for_inference = False
# file name with the experiments accuracy output
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = "results/health_average_accuracy" + mux_adds_str + ".txt"
verbose = 2
# keep track of output so that output is in order
output_list = []
# [f2,e1,g1]
[0, 0, 0],
[1, 0, 0],
[0, 1, 0],
[0, 0, 1],
[1, 1, 0],
[1, 0, 1],
[0, 1, 1],
[1, 1, 1],
]
default_reliability_setting = [1.0, 1.0, 1.0]
no_information_flow_map = {}
for skip_hyperconnection_configuration in skip_hyperconnection_configurations:
no_information_flow_map[
tuple(skip_hyperconnection_configuration)
] = make_no_information_flow_map("Health", skip_hyperconnection_configuration)
load_for_inference = False
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = (
"results/health_skiphyperconnection_sensitivity" + mux_adds_str + ".txt"
)
verbose = 2
# keep track of output so that output is in order
output_list = []
model_name = "ResiliNet Hyperconnection Weight Sensitivity"
output = make_output_dictionary(
model_name,
reliability_settings,
num_iterations,
epochs,
progress_verbose,
checkpoint_verbose,
train_steps_per_epoch,
val_steps_per_epoch,
num_gpus,
)
return model
if __name__ == "__main__":
accuracy = accuracy("CIFAR")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data = init_data()
ResiliNet_no_information_flow_map = make_no_information_flow_map(
"CIFAR/Imagenet", default_skip_hyperconnection_config)
DFG_no_information_flow_map = make_no_information_flow_map(
"CIFAR/Imagenet", default_skip_hyperconnection_config)
Vanilla_no_information_flow_map = make_no_information_flow_map(
"CIFAR/Imagenet")
train_steps_per_epoch = math.ceil(len(data.train) / batch_size)
val_steps_per_epoch = math.ceil(len(data.val) / batch_size)
output = make_output_dictionary_average_accuracy(reliability_settings,
num_iterations)
load_for_inference = False
make_results_folder()
output_name = "results" + "/cifar_average_accuracy_results.txt"
output_list = []
)
return model
# hyperconnection weight experiment
if __name__ == "__main__":
accuracy = accuracy("ResNet")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data = init_data()
model_name = "ResiliNet Hyperconnection Weight"
output, weight_schemes = make_output_dictionary_hyperconnection_weight(
model_name, reliability_settings, num_iterations)
considered_weight_schemes = [1, 2, 3, 4]
no_information_flow_map = make_no_information_flow_map(
"ResNet", default_skip_hyperconnection_config)
load_for_inference = False
train_steps_per_epoch = math.ceil(len(data.train) / batch_size)
val_steps_per_epoch = math.ceil(len(data.val) / batch_size)
make_results_folder()
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = ("results/cifar_resnet_hyperconnection_weight_results" +
mux_adds_str + ".txt")
output_list = []
default_reliability_setting = [1, 1]
for iteration in range(1, num_iterations + 1):
print("iteration:", iteration)
for weight_scheme in considered_weight_schemes:
if (weight_scheme == 2 or weight_scheme
[1, 1],
]
model_name = "ResiliNet Hyperconnection Weight Sensitivity"
default_reliability_setting = [1.0, 1.0, 1.0]
output = make_output_dictionary(
model_name,
reliability_settings,
num_iterations,
skip_hyperconnection_configurations,
)
no_information_flow_map = {}
for skip_hyperconnection_configuration in skip_hyperconnection_configurations:
no_information_flow_map[tuple(skip_hyperconnection_configuration
)] = make_no_information_flow_map(
"CIFAR/Imagenet",
skip_hyperconnection_configuration)
load_for_inference = False
train_steps_per_epoch = math.ceil(len(data.train) / batch_size)
val_steps_per_epoch = math.ceil(len(data.val) / batch_size)
make_results_folder()
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = ("results/cifar_skiphyperconnection_sensitivity_results" +
mux_adds_str + ".txt")
output_list = []
for iteration in range(1, num_iterations + 1):
print("iteration:", iteration)
for skip_hyperconnection_configuration in skip_hyperconnection_configurations:
layer.set_weights([new_weights])
# ResiliNet variable failout experiment
if __name__ == "__main__":
accuracy = accuracy("CIFAR")
calc_expected_accuracy = accuracy.calc_expected_accuracy
data = init_data()
output_list = []
load_for_inference = False
train_steps_per_epoch = math.ceil(len(data.train) / batch_size)
val_steps_per_epoch = math.ceil(len(data.val) / batch_size)
no_information_flow_map = make_no_information_flow_map(
"CIFAR/Imagenet", default_skip_hyperconnection_config
)
failout_survival_settings = [[0.95, 0.95], [0.9, 0.9], [0.7, 0.7], [0.5, 0.5]]
dropout_like_failout = False
output = make_output_dictionary_failout_rate(
failout_survival_settings, reliability_settings, num_iterations
)
make_results_folder()
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = "results/cifar_failout_results" + mux_adds_str + ".txt"
for iteration in range(1, num_iterations + 1):
output_list.append("ITERATION " + str(iteration) + "\n")
print("iteration:", iteration)
output_list.append("ResiliNet" + "\n")
# variable failout rate
for reliability_setting in reliability_settings:
[1, 1],
]
model_name = "ResiliNet Hyperconnection Weight Sensitivity"
default_reliability_setting = [1.0, 1.0, 1.0]
output = make_output_dictionary(
model_name,
reliability_settings,
num_iterations,
skip_hyperconnection_configurations,
)
no_information_flow_map = {}
for skip_hyperconnection_configuration in skip_hyperconnection_configurations:
no_information_flow_map[
tuple(skip_hyperconnection_configuration)
] = make_no_information_flow_map("ResNet", skip_hyperconnection_configuration)
load_for_inference = False
train_steps_per_epoch = math.ceil(len(data.train) / batch_size)
val_steps_per_epoch = math.ceil(len(data.val) / batch_size)
make_results_folder()
mux_adds_str = "mux_adds" if ResiliNetPlus else ""
output_name = (
"results/cifar_resnet_skiphyperconnection_sensitivity_results"
+ mux_adds_str
+ ".txt"
)
output_list = []
for iteration in range(1, num_iterations + 1):
print("iteration:", iteration)