SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2102.13451 · NeurIPS · 2021

FjORD: Fair and Accurate Federated Learning under heterogeneous targets with Ordered Dropout

Stylianos Venieris, Samuel Horváth, Nicholas Lane, Stefanos Laskaridis, Mario Almeida

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2102.13451.json · badge: /api/ran/2102.13451.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
k1l1/SLT community
Papers with Code archive
2 ran · 1 no recorded run
FunctionStatusWhere it lives
FjordDevice Ran
ran on a synthesized input · constructed an object; no result was computed
k1l1/SLT/algorithms/fjord.py
code served (permissive licence) · get_code("0b51c806bdc788bc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, FjordDevice: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from torch.utils.data import TensorDataset

torch.manual_seed(0)

# Build a minimal nn.Module that satisfies the function's usage
class SimpleModel(nn.Module):
    def __init__(self, input_dim=4, hidden_dim=8, output_dim=3, subset_factor=1.0):
        super().__init__()
        self.subset_factor = subset_factor
        self.fc1 = nn.Linear(input_dim, hidden_dim)
        self.fc2 = nn.Linear(hidden_dim, output_dim)
    
    def forward(self, x):
        x = torch.relu(self.fc1(x))
        x = self.fc2(x)
        return x

# Create minimal training and test datasets
# Using 2 samples, 4 input features, 3 output classes
train_x = torch.randn(2, 4)
train_y = torch.randint(0, 3, (2,))
train_dataset = TensorDataset(train_x, train_y)

test_x = torch.randn(2, 4)
test_y = torch.randint(0, 3, (2,))
test_dataset = TensorDataset(test_x, test_y)

# Build ARGS dict with all parameters for FjordDevice
ARGS = {
    'device_id': 0,
    'storage_path': '/tmp/fedavg',
}

# Additional setup needed for FjordDevice to work
# We'll set up an instance and configure it
device = None

def setup_fjord_device():
    from fedavg import FjordDevice
    device = FjordDevice(device_id=0, storage_path='/tmp/fedavg')
    
    # Set model
    device.set_model(
        model_list=SimpleModel,
        kwargs_list={'input_dim': 4, 'hidden_dim': 8, 'output_dim': 3, 'subset_factor': 1.0}
    )
    
    # Initialize model
    device.init_model()
    
    # Set datasets
    device.set_train_data(train_dataset)
    device.set_test_data(test_dataset)
    
    # Set torch device
    device.set_torch_device(torch.device('cpu'))
    
    # Set optimizer and args
    device.set_optimizer(torch.optim.SGD, {})
    
    # Set loss function
    device.set_loss_function(nn.CrossEntropyLoss)
    
    # Set learning rate
    device.lr = 0.01
    
    # Set FJORD-specific attributes
    device.fjord_p = [0.5, 0.75, 1.0]
    device.fjord_rng = torch.Generator()
    device.fjord_rng.manual_seed(0)
    
    # Set subset_factor class attribute
    FjordDevice.subset_factor = 1.0
    
    return device

# For the ARGS requirement, we need to map constructor parameters
ARGS = {
    'device_id': 0,
    'storage_path': '/tmp/fedavg',
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
fedavg, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type FjordDevice — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a FjordDevice output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
SubsetDevice Ran
ran on a synthesized input · constructed an object; no result was computed
k1l1/SLT/algorithms/fjord.py
code served (permissive licence) · get_code("e37de38d95d7cf93")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, SubsetDevice: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from torch.utils.data import TensorDataset

torch.manual_seed(0)

# Create a minimal model
class SimpleModel(nn.Module):
    def __init__(self, input_dim=4, num_classes=3):
        super().__init__()
        self.linear = nn.Linear(input_dim, num_classes)
    
    def forward(self, x):
        return self.linear(x)

# Create minimal train and test datasets
train_inputs = torch.randn(8, 4)
train_labels = torch.randint(0, 3, (8,))
train_dataset = TensorDataset(train_inputs, train_labels)

test_inputs = torch.randn(4, 4)
test_labels = torch.randint(0, 3, (4,))
test_dataset = TensorDataset(test_inputs, test_labels)

ARGS = {
    'device_id': 0,
    'storage_path': '/tmp/test_model',
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SubsetDevice — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a SubsetDevice output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
FedAvgDevice Not yet run
unverified (harvested, no recorded run)
k1l1/SLT/algorithms/fjord.py
code served (permissive licence) · get_code("ed316958b64c1cb2")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 later) and 1 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
FedAvgDevice
brick of this paper's assembly: helper
3 other papers
1 dated earlier, 2 later · 1 only as an assembly brick
2010.01264 (2020-10)
HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients
2212.01548 (2022-12)
FedRolex: Model-Heterogeneous Federated Learning with Rolling Sub-Model Extraction
2305.17005 (2023-05)
Aggregating Capacity in FL through Successive Layer Training for Computationally-Constrained Devices
SubsetDevice
brick of this paper's assembly: helper
3 other papers
1 dated earlier, 2 later · 1 only as an assembly brick
2010.01264 (2020-10)
HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients
2212.01548 (2022-12)
FedRolex: Model-Heterogeneous Federated Learning with Rolling Sub-Model Extraction
2305.17005 (2023-05)
Aggregating Capacity in FL through Successive Layer Training for Computationally-Constrained Devices

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Federated Learning (FL) has been gaining significant traction across different ML tasks, ranging from vision to keyboard predictions. In large-scale deployments, client heterogeneity is a fact and constitutes a primary problem for fairness, training performance and accuracy. Although significant efforts have been made into tackling statistical data heterogeneity, the diversity in the processing capabilities and network bandwidth of clients, termed as system heterogeneity, has remained largely unexplored. Current solutions either disregard a large portion of available devices or set a uniform limit on the model's capacity, restricted by the least capable participants. In this work, we introduce Ordered Dropout, a mechanism that achieves an ordered, nested representation of knowledge in deep neural networks (DNNs) and enables the extraction of lower footprint submodels without the need of retraining. We further show that for linear maps our Ordered Dropout is equivalent to SVD. We employ this technique, along with a self-distillation methodology, in the realm of FL in a framework called FjORD. FjORD alleviates the problem of client system heterogeneity by tailoring the model width to the client's capabilities. Extensive evaluation on both CNNs and RNNs across diverse modalities shows that FjORD consistently leads to significant performance gains over state-of-the-art baselines, while maintaining its nested structure. * Indicates equal contribution. † Work while intern at Samsung AI Center. Preprint. Under review.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2102.13451")
get_code_for_paper("2102.13451")
have("2102.13451")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2102.13451.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2102.13451.svg)](https://syntology.ai/paper/2102.13451)

Connect an agent — have() is free.