Update random alpha

eval/step_configs/config_celeba_topkacc_norm.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.Celeba
+dataset_class = Celeba
+model_class = CNN
+images_dir = /mnt/nfs/shared/leaf/data/celeba/data/raw/img_align_celeba
+train_dir = /mnt/nfs/shared/leaf/data/celeba/per_user_data/train
+test_dir = /mnt/nfs/shared/leaf/data/celeba/data/test
+; python list of fractions below
+sizes = 
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 4
+full_epochs = False
+batch_size = 16
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCP
+comm_class = TCP
+addresses_filepath = ip_addr_6Machines.json
+
+[SHARING]
+sharing_package = decentralizepy.sharing.TopKNormalized
+sharing_class = TopKNormalized
+alpha = 0.1
+accumulation = True
+accumulate_averaging_changes = True
\ No newline at end of file

eval/step_configs/config_celeba_topkaccbound.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.Celeba
+dataset_class = Celeba
+model_class = CNN
+images_dir = /mnt/nfs/shared/leaf/data/celeba/data/raw/img_align_celeba
+train_dir = /mnt/nfs/shared/leaf/data/celeba/per_user_data/train
+test_dir = /mnt/nfs/shared/leaf/data/celeba/data/test
+; python list of fractions below
+sizes = 
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 4
+full_epochs = False
+batch_size = 16
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCP
+comm_class = TCP
+addresses_filepath = ip_addr_6Machines.json
+
+[SHARING]
+sharing_package = decentralizepy.sharing.LowerBoundTopK
+sharing_class = LowerBoundTopK
+lower_bound = 0.1
+alpha = 0.1
+metro_hastings = False
+accumulation = True
+accumulate_averaging_changes = True
\ No newline at end of file

eval/step_configs/config_celeba_topkparam.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]

eval/step_configs/config_celeba_topkrandom.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]

eval/step_configs/config_celeba_wavelet.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]
@@ -36,3 +36,4 @@ alpha = 0.1
 wavelet=sym2
 level= 4
 accumulation = True
+accumulate_averaging_changes = True

eval/step_configs/config_cifar_SharingWithRWAsyncDynamic.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.CIFAR10
+dataset_class = CIFAR10
+model_class = LeNet
+train_dir = /mnt/nfs/shared/CIFAR
+test_dir = /mnt/nfs/shared/CIFAR
+; python list of fractions below
+sizes =
+random_seed = 99
+partition_niid = True
+shards = 1
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 65
+full_epochs = False
+batch_size = 8
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCPRandomWalkRouting
+comm_class = TCPRandomWalkRouting
+addresses_filepath = ip_addr_6Machines.json
+sampler = equi
+
+[SHARING]
+sharing_package = decentralizepy.sharing.SharingWithRWAsyncDynamic
+sharing_class = SharingWithRWAsyncDynamic
\ No newline at end of file

eval/step_configs/config_CIFAR_topk.ini → eval/step_configs/config_cifar_partialmodel.ini

@@ -5,19 +5,20 @@ model_class = LeNet
 train_dir = /mnt/nfs/shared/CIFAR
 test_dir = /mnt/nfs/shared/CIFAR
 ; python list of fractions below
-sizes = 
-random_seed = 91
+sizes =
+random_seed = 99
 partition_niid = True
+shards = 4
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]
 training_package = decentralizepy.training.Training
 training_class = Training
-rounds = 5
+rounds = 65
 full_epochs = False
 batch_size = 8
 shuffle = True
@@ -32,5 +33,4 @@ addresses_filepath = ip_addr_6Machines.json
 [SHARING]
 sharing_package = decentralizepy.sharing.PartialModel
 sharing_class = PartialModel
-alpha = 0.5
-accumulation = True
+alpha=0.5

eval/step_configs/config_CIFAR.ini → eval/step_configs/config_cifar_sharing.ini

@@ -5,20 +5,20 @@ model_class = LeNet
 train_dir = /mnt/nfs/shared/CIFAR
 test_dir = /mnt/nfs/shared/CIFAR
 ; python list of fractions below
-sizes = 
-random_seed = 91
+sizes =
+random_seed = 99
 partition_niid = True
 shards = 4
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]
 training_package = decentralizepy.training.Training
 training_class = Training
-rounds = 5
+rounds = 65
 full_epochs = False
 batch_size = 8
 shuffle = True

eval/step_configs/config_cifar_subsampling.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.CIFAR10
+dataset_class = CIFAR10
+model_class = LeNet
+train_dir = /mnt/nfs/shared/CIFAR
+test_dir = /mnt/nfs/shared/CIFAR
+; python list of fractions below
+sizes =
+random_seed = 99
+partition_niid = True
+shards = 4
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 65
+full_epochs = False
+batch_size = 8
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCP
+comm_class = TCP
+addresses_filepath = ip_addr_6Machines.json
+
+[SHARING]
+sharing_package = decentralizepy.sharing.SubSampling
+sharing_class = SubSampling
+alpha = 0.5

eval/step_configs/config_cifar_topkacc.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.CIFAR10
+dataset_class = CIFAR10
+model_class = LeNet
+train_dir = /mnt/nfs/shared/CIFAR
+test_dir = /mnt/nfs/shared/CIFAR
+; python list of fractions below
+sizes =
+random_seed = 99
+partition_niid = True
+shards = 4
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 65
+full_epochs = False
+batch_size = 8
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCP
+comm_class = TCP
+addresses_filepath = ip_addr_6Machines.json
+
+[SHARING]
+sharing_package = decentralizepy.sharing.PartialModel
+sharing_class = PartialModel
+alpha = 0.5
+accumulation = True
+accumulate_averaging_changes = True
\ No newline at end of file

eval/step_configs/config_CIFAR_wavelet.ini → eval/step_configs/config_cifar_wavelet.ini

@@ -5,20 +5,20 @@ model_class = LeNet
 train_dir = /mnt/nfs/shared/CIFAR
 test_dir = /mnt/nfs/shared/CIFAR
 ; python list of fractions below
-sizes = 
-random_seed = 91
+sizes =
+random_seed = 99
 partition_niid = True
 shards = 4
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]
 training_package = decentralizepy.training.Training
 training_class = Training
-rounds = 5
+rounds = 65
 full_epochs = False
 batch_size = 8
 shuffle = True
@@ -38,3 +38,4 @@ alpha = 0.5
 wavelet=sym2
 level= 4
 accumulation = True
+accumulate_averaging_changes = True

eval/step_configs/config_femnist_fft.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 # There are 734463 femnist samples

eval/step_configs/config_femnist_grow.ini

@@ -9,7 +9,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]

eval/step_configs/config_femnist_partialmodel.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]

eval/step_configs/config_femnist_sharing.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 [TRAIN_PARAMS]

eval/step_configs/config_femnist_subsampling.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 # There are 734463 femnist samples

eval/step_configs/config_femnist_topkacc.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 # There are 734463 femnist samples
@@ -33,4 +33,5 @@ addresses_filepath = ip_addr_6Machines.json
 sharing_package = decentralizepy.sharing.PartialModel
 sharing_class = PartialModel
 alpha = 0.1
-accumulation = True
+accumulation = True
\ No newline at end of file
+accumulate_averaging_changes = True
\ No newline at end of file

eval/step_configs/config_femnist_topkaccbound.ini

+[DATASET]
+dataset_package = decentralizepy.datasets.Femnist
+dataset_class = Femnist
+random_seed = 97
+model_class = CNN
+train_dir = /mnt/nfs/shared/leaf/data/femnist/per_user_data/train
+test_dir = /mnt/nfs/shared/leaf/data/femnist/data/test
+; python list of fractions below
+sizes =
+
+[OPTIMIZER_PARAMS]
+optimizer_package = torch.optim
+optimizer_class = SGD
+lr = 0.001
+
+# There are 734463 femnist samples
+[TRAIN_PARAMS]
+training_package = decentralizepy.training.Training
+training_class = Training
+rounds = 47
+full_epochs = False
+batch_size = 16
+shuffle = True
+loss_package = torch.nn
+loss_class = CrossEntropyLoss
+
+[COMMUNICATION]
+comm_package = decentralizepy.communication.TCP
+comm_class = TCP
+addresses_filepath = ip_addr_6Machines.json
+
+[SHARING]
+sharing_package = decentralizepy.sharing.LowerBoundTopK
+sharing_class = LowerBoundTopK
+lower_bound = 0.1
+alpha = 0.1
+accumulation = True
+accumulate_averaging_changes = True
\ No newline at end of file

eval/step_configs/config_femnist_topkparam.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 # There are 734463 femnist samples

eval/step_configs/config_femnist_wavelet.ini

@@ -10,7 +10,7 @@ sizes =
 
 [OPTIMIZER_PARAMS]
 optimizer_package = torch.optim
-optimizer_class = Adam
+optimizer_class = SGD
 lr = 0.001
 
 # There are 734463 femnist samples
@@ -37,3 +37,4 @@ alpha = 0.1
 wavelet=sym2
 level= 4
 accumulation = True
+accumulate_averaging_changes = True