flwr-nightly 1.22.0.dev20250916__tar.gz → 1.22.0.dev20250918__tar.gz

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: flwr-nightly
-Version: 1.22.0.dev20250916
+Version: 1.22.0.dev20250918
 Summary: Flower: A Friendly Federated AI Framework
 License: Apache-2.0
 Keywords: Artificial Intelligence,Federated AI,Federated Analytics,Federated Evaluation,Federated Learning,Flower,Machine Learning

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/py/flwr/cli/app.py

@@ -25,6 +25,7 @@ from .log import log
 from .login import login
 from .ls import ls
 from .new import new
+from .pull import pull
 from .run import run
 from .stop import stop
 
@@ -46,6 +47,7 @@ app.command()(log)
 app.command()(ls)
 app.command()(stop)
 app.command()(login)
+app.command()(pull)
 
 typer_click_object = get_command(app)
 

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/py/flwr/cli/new/new.py

@@ -41,6 +41,7 @@ class MlFramework(str, Enum):
     JAX = "JAX"
     MLX = "MLX"
     NUMPY = "NumPy"
+    XGBOOST = "XGBoost"
     FLOWERTUNE = "FlowerTune"
     BASELINE = "Flower Baseline"
     PYTORCH_LEGACY_API = "PyTorch (Legacy API, deprecated)"
@@ -201,7 +202,7 @@ def new(
         }
 
         # Challenge specific context
-        fraction_fit = "0.2" if llm_challenge_str == "code" else "0.1"
+        fraction_train = "0.2" if llm_challenge_str == "code" else "0.1"
         if llm_challenge_str == "generalnlp":
             challenge_name = "General NLP"
             num_clients = "20"
@@ -220,7 +221,7 @@ def new(
             dataset_name = "flwrlabs/code-alpaca-20k"
 
         context["llm_challenge_str"] = llm_challenge_str
-        context["fraction_fit"] = fraction_fit
+        context["fraction_train"] = fraction_train
         context["challenge_name"] = challenge_name
         context["num_clients"] = num_clients
         context["dataset_name"] = dataset_name
@@ -247,6 +248,7 @@ def new(
             MlFramework.TENSORFLOW.value,
             MlFramework.SKLEARN.value,
             MlFramework.NUMPY.value,
+            MlFramework.XGBOOST.value,
             "pytorch_legacy_api",
         ]
         if framework_str in frameworks_with_tasks:

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/py/flwr/cli/new/templates/app/README.flowertune.md.tpl

@@ -26,7 +26,7 @@ pip install -e .
 ## Experimental setup
 
 The dataset is divided into $num_clients partitions in an IID fashion, a partition is assigned to each ClientApp.
-We randomly sample a fraction ($fraction_fit) of the total nodes to participate in each round, for a total of `200` rounds.
+We randomly sample a fraction ($fraction_train) of the total nodes to participate in each round, for a total of `200` rounds.
 All settings are defined in `pyproject.toml`.
 
 > [!IMPORTANT]

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/client.baseline.py.tpl

@@ -0,0 +1,75 @@
+"""$project_name: A Flower Baseline."""
+
+import torch
+from flwr.app import ArrayRecord, Context, Message, MetricRecord, RecordDict
+from flwr.clientapp import ClientApp
+
+from $import_name.dataset import load_data
+from $import_name.model import Net
+from $import_name.model import test as test_fn
+from $import_name.model import train as train_fn
+
+# Flower ClientApp
+app = ClientApp()
+
+
+@app.train()
+def train(msg: Message, context: Context):
+    """Train the model on local data."""
+
+    # Load the model and initialize it with the received weights
+    model = Net()
+    model.load_state_dict(msg.content["arrays"].to_torch_state_dict())
+    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+
+    # Load the data
+    partition_id = int(context.node_config["partition-id"])
+    num_partitions = int(context.node_config["num-partitions"])
+    trainloader, _ = load_data(partition_id, num_partitions)
+    local_epochs = context.run_config["local-epochs"]
+
+    # Call the training function
+    train_loss = train_fn(
+        model,
+        trainloader,
+        local_epochs,
+        device,
+    )
+
+    # Construct and return reply Message
+    model_record = ArrayRecord(model.state_dict())
+    metrics = {
+        "train_loss": train_loss,
+        "num-examples": len(trainloader.dataset),
+    }
+    metric_record = MetricRecord(metrics)
+    content = RecordDict({"arrays": model_record, "metrics": metric_record})
+    return Message(content=content, reply_to=msg)
+
+
+@app.evaluate()
+def evaluate(msg: Message, context: Context):
+    """Evaluate the model on local data."""
+
+    # Load the model and initialize it with the received weights
+    model = Net()
+    model.load_state_dict(msg.content["arrays"].to_torch_state_dict())
+    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+
+    # Load the data
+    partition_id = int(context.node_config["partition-id"])
+    num_partitions = int(context.node_config["num-partitions"])
+    _, valloader = load_data(partition_id, num_partitions)
+
+    # Call the evaluation function
+    eval_loss, eval_acc = test_fn(model, valloader, device)
+
+    # Construct and return reply Message
+    metrics = {
+        "eval_loss": eval_loss,
+        "eval_acc": eval_acc,
+        "num-examples": len(valloader.dataset),
+    }
+    metric_record = MetricRecord(metrics)
+    content = RecordDict({"metrics": metric_record})
+    return Message(content=content, reply_to=msg)

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/client.xgboost.py.tpl

@@ -0,0 +1,110 @@
+"""$project_name: A Flower / $framework_str app."""
+
+import warnings
+
+import numpy as np
+import xgboost as xgb
+from flwr.app import ArrayRecord, Context, Message, MetricRecord, RecordDict
+from flwr.clientapp import ClientApp
+from flwr.common.config import unflatten_dict
+
+from $import_name.task import load_data, replace_keys
+
+warnings.filterwarnings("ignore", category=UserWarning)
+
+
+# Flower ClientApp
+app = ClientApp()
+
+
+def _local_boost(bst_input, num_local_round, train_dmatrix):
+    # Update trees based on local training data.
+    for i in range(num_local_round):
+        bst_input.update(train_dmatrix, bst_input.num_boosted_rounds())
+
+    # Bagging: extract the last N=num_local_round trees for sever aggregation
+    bst = bst_input[
+        bst_input.num_boosted_rounds()
+        - num_local_round : bst_input.num_boosted_rounds()
+    ]
+    return bst
+
+
+@app.train()
+def train(msg: Message, context: Context) -> Message:
+    # Load model and data
+    partition_id = context.node_config["partition-id"]
+    num_partitions = context.node_config["num-partitions"]
+    train_dmatrix, _, num_train, _ = load_data(partition_id, num_partitions)
+
+    # Read from run config
+    num_local_round = context.run_config["local-epochs"]
+    # Flatted config dict and replace "-" with "_"
+    cfg = replace_keys(unflatten_dict(context.run_config))
+    params = cfg["params"]
+
+    global_round = msg.content["config"]["server-round"]
+    if global_round == 1:
+        # First round local training
+        bst = xgb.train(
+            params,
+            train_dmatrix,
+            num_boost_round=num_local_round,
+        )
+    else:
+        bst = xgb.Booster(params=params)
+        global_model = bytearray(msg.content["arrays"]["0"].numpy().tobytes())
+
+        # Load global model into booster
+        bst.load_model(global_model)
+
+        # Local training
+        bst = _local_boost(bst, num_local_round, train_dmatrix)
+
+    # Save model
+    local_model = bst.save_raw("json")
+    model_np = np.frombuffer(local_model, dtype=np.uint8)
+
+    # Construct reply message
+    # Note: we store the model as the first item in a list into ArrayRecord,
+    # which can be accessed using index ["0"].
+    model_record = ArrayRecord([model_np])
+    metrics = {
+        "num-examples": num_train,
+    }
+    metric_record = MetricRecord(metrics)
+    content = RecordDict({"arrays": model_record, "metrics": metric_record})
+    return Message(content=content, reply_to=msg)
+
+
+@app.evaluate()
+def evaluate(msg: Message, context: Context) -> Message:
+    # Load model and data
+    partition_id = context.node_config["partition-id"]
+    num_partitions = context.node_config["num-partitions"]
+    _, valid_dmatrix, _, num_val = load_data(partition_id, num_partitions)
+
+    # Load config
+    cfg = replace_keys(unflatten_dict(context.run_config))
+    params = cfg["params"]
+
+    # Load global model
+    bst = xgb.Booster(params=params)
+    global_model = bytearray(msg.content["arrays"]["0"].numpy().tobytes())
+    bst.load_model(global_model)
+
+    # Run evaluation
+    eval_results = bst.eval_set(
+        evals=[(valid_dmatrix, "valid")],
+        iteration=bst.num_boosted_rounds() - 1,
+    )
+    auc = float(eval_results.split("\t")[1].split(":")[1])
+
+    # Construct and return reply Message
+    metrics = {
+        "auc": auc,
+        "num-examples": num_val,
+    }
+    metric_record = MetricRecord(metrics)
+    content = RecordDict({"metrics": metric_record})
+    return Message(content=content, reply_to=msg)

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/flwr_tune/client_app.py.tpl

@@ -0,0 +1,92 @@
+"""$project_name: A Flower / FlowerTune app."""
+
+import os
+import warnings
+
+from flwr.app import ArrayRecord, Context, Message, MetricRecord, RecordDict
+from flwr.clientapp import ClientApp
+from flwr.common.config import unflatten_dict
+from omegaconf import DictConfig
+from peft import get_peft_model_state_dict, set_peft_model_state_dict
+from transformers import TrainingArguments
+from trl import SFTTrainer
+
+from $import_name.dataset import (
+    get_tokenizer_and_data_collator_and_propt_formatting,
+    load_data,
+    replace_keys,
+)
+from $import_name.models import cosine_annealing, get_model
+
+# Avoid warnings
+os.environ["TOKENIZERS_PARALLELISM"] = "true"
+os.environ["RAY_DISABLE_DOCKER_CPU_WARNING"] = "1"
+warnings.filterwarnings("ignore", category=UserWarning)
+
+
+# Avoid warnings
+os.environ["TOKENIZERS_PARALLELISM"] = "true"
+os.environ["RAY_DISABLE_DOCKER_CPU_WARNING"] = "1"
+warnings.filterwarnings("ignore", category=UserWarning)
+
+
+# Flower ClientApp
+app = ClientApp()
+
+
+@app.train()
+def train(msg: Message, context: Context):
+    """Train the model on local data."""
+    # Parse config
+    partition_id = context.node_config["partition-id"]
+    num_partitions = context.node_config["num-partitions"]
+    num_rounds = context.run_config["num-server-rounds"]
+    cfg = DictConfig(replace_keys(unflatten_dict(context.run_config)))
+    training_arguments = TrainingArguments(**cfg.train.training_arguments)
+
+    # Let's get the client partition
+    trainset = load_data(partition_id, num_partitions, cfg.static.dataset.name)
+    (
+        tokenizer,
+        data_collator,
+        formatting_prompts_func,
+    ) = get_tokenizer_and_data_collator_and_propt_formatting(cfg.model.name)
+
+    # Load the model and initialize it with the received weights
+    model = get_model(cfg.model)
+    set_peft_model_state_dict(model, msg.content["arrays"].to_torch_state_dict())
+
+    # Set learning rate for current round
+    new_lr = cosine_annealing(
+        msg.content["config"]["server-round"],
+        num_rounds,
+        cfg.train.learning_rate_max,
+        cfg.train.learning_rate_min,
+    )
+
+    training_arguments.learning_rate = new_lr
+    training_arguments.output_dir = msg.content["config"]["save_path"]
+
+    # Construct trainer
+    trainer = SFTTrainer(
+        model=model,
+        tokenizer=tokenizer,
+        args=training_arguments,
+        max_seq_length=cfg.train.seq_length,
+        train_dataset=trainset,
+        formatting_func=formatting_prompts_func,
+        data_collator=data_collator,
+    )
+
+    # Do local training
+    results = trainer.train()
+
+    # Construct and return reply Message
+    model_record = ArrayRecord(get_peft_model_state_dict(model))
+    metrics = {
+        "train_loss": results.training_loss,
+        "num-examples": len(trainset),
+    }
+    metric_record = MetricRecord(metrics)
+    content = RecordDict({"arrays": model_record, "metrics": metric_record})
+    return Message(content=content, reply_to=msg)

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/py/flwr/cli/new/templates/app/code/flwr_tune/models.py.tpl

@@ -4,18 +4,10 @@ import math
 
 import torch
 from omegaconf import DictConfig
-from collections import OrderedDict
-from peft import (
-    LoraConfig,
-    get_peft_model,
-    get_peft_model_state_dict,
-    set_peft_model_state_dict,
-)
+from peft import LoraConfig, get_peft_model
 from peft.utils import prepare_model_for_kbit_training
 from transformers import AutoModelForCausalLM, BitsAndBytesConfig
 
-from flwr.common.typing import NDArrays
-
 
 def cosine_annealing(
     current_round: int,
@@ -62,17 +54,3 @@ def get_model(model_cfg: DictConfig):
         model.config.use_cache = False
 
     return get_peft_model(model, peft_config)
-
-
-def set_parameters(model, parameters: NDArrays) -> None:
-    """Change the parameters of the model using the given ones."""
-    peft_state_dict_keys = get_peft_model_state_dict(model).keys()
-    params_dict = zip(peft_state_dict_keys, parameters)
-    state_dict = OrderedDict({k: torch.Tensor(v) for k, v in params_dict})
-    set_peft_model_state_dict(model, state_dict)
-
-
-def get_parameters(model) -> NDArrays:
-    """Return the parameters of the current net."""
-    state_dict = get_peft_model_state_dict(model)
-    return [val.cpu().numpy() for _, val in state_dict.items()]

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/flwr_tune/server_app.py.tpl

@@ -0,0 +1,73 @@
+"""$project_name: A Flower / FlowerTune app."""
+
+import os
+from datetime import datetime
+
+from flwr.app import ArrayRecord, ConfigRecord, Context, MetricRecord
+from flwr.common.config import unflatten_dict
+from flwr.serverapp import Grid, ServerApp
+from omegaconf import DictConfig
+from peft import get_peft_model_state_dict, set_peft_model_state_dict
+
+from $import_name.dataset import replace_keys
+from $import_name.models import get_model
+from $import_name.strategy import FlowerTuneLlm
+
+# Create ServerApp
+app = ServerApp()
+
+
+@app.main()
+def main(grid: Grid, context: Context) -> None:
+    """Main entry point for the ServerApp."""
+    # Create output directory given current timestamp
+    current_time = datetime.now()
+    folder_name = current_time.strftime("%Y-%m-%d_%H-%M-%S")
+    save_path = os.path.join(os.getcwd(), f"results/{folder_name}")
+    os.makedirs(save_path, exist_ok=True)
+
+    # Read from config
+    num_rounds = context.run_config["num-server-rounds"]
+    cfg = DictConfig(replace_keys(unflatten_dict(context.run_config)))
+
+    # Get initial model weights
+    init_model = get_model(cfg.model)
+    arrays = ArrayRecord(get_peft_model_state_dict(init_model))
+
+    # Define strategy
+    strategy = FlowerTuneLlm(
+        fraction_train=cfg.strategy.fraction_train,
+        fraction_evaluate=cfg.strategy.fraction_evaluate,
+    )
+
+    # Start strategy, run FedAvg for `num_rounds`
+    strategy.start(
+        grid=grid,
+        initial_arrays=arrays,
+        train_config=ConfigRecord({"save_path": save_path}),
+        num_rounds=num_rounds,
+        evaluate_fn=get_evaluate_fn(
+            cfg.model, cfg.train.save_every_round, num_rounds, save_path
+        ),
+    )
+
+
+# Get function that will be executed by the strategy
+# Here we use it to save global model checkpoints
+def get_evaluate_fn(model_cfg, save_every_round, total_round, save_path):
+    """Return an evaluation function for saving global model."""
+
+    def evaluate(server_round: int, arrays: ArrayRecord) -> MetricRecord:
+        # Save model
+        if server_round != 0 and (
+            server_round == total_round or server_round % save_every_round == 0
+        ):
+            # Init model
+            model = get_model(model_cfg)
+            set_peft_model_state_dict(model, arrays.to_torch_state_dict())
+
+            model.save_pretrained(f"{save_path}/peft_{server_round}")
+
+        return MetricRecord()
+
+    return evaluate

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/flwr_tune/strategy.py.tpl

@@ -0,0 +1,78 @@
+"""$project_name: A Flower / FlowerTune app."""
+
+from collections.abc import Iterable
+from logging import INFO, WARN
+from typing import Optional
+
+from flwr.app import ArrayRecord, ConfigRecord, Message, MetricRecord
+from flwr.common import log
+from flwr.serverapp import Grid
+from flwr.serverapp.strategy import FedAvg
+
+
+class FlowerTuneLlm(FedAvg):
+    """Customised FedAvg strategy implementation.
+
+    This class behaves just like FedAvg but also tracks the communication
+    costs associated with `train` over FL rounds.
+    """
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.comm_tracker = CommunicationTracker()
+
+    def configure_train(
+            self, server_round: int, arrays: ArrayRecord, config: ConfigRecord, grid: Grid
+    ) -> Iterable[Message]:
+        """Configure the next round of training."""
+        messages = super().configure_train(server_round, arrays, config, grid)
+
+        # Track communication costs
+        self.comm_tracker.track(messages)
+
+        return messages
+
+    def aggregate_train(
+            self,
+            server_round: int,
+            replies: Iterable[Message],
+    ) -> tuple[Optional[ArrayRecord], Optional[MetricRecord]]:
+        """Aggregate ArrayRecords and MetricRecords in the received Messages."""
+        # Track communication costs
+        self.comm_tracker.track(replies)
+
+        arrays, metrics = super().aggregate_train(server_round, replies)
+
+        return arrays, metrics
+
+
+class CommunicationTracker:
+    """Communication costs tracker over FL rounds."""
+    def __init__(self):
+        self.curr_comm_cost = 0.0
+
+    def track(self, messages: Iterable[Message]):
+        comm_cost = (
+            sum(
+                record.count_bytes()
+                for msg in messages
+                if msg.has_content()
+                for record in msg.content.array_records.values()
+            )
+            / 1024**2
+        )
+
+        self.curr_comm_cost += comm_cost
+        log(
+            INFO,
+            "Communication budget: used %.2f MB (+%.2f MB this round) / 200,000 MB",
+            self.curr_comm_cost,
+            comm_cost,
+        )
+
+        if self.curr_comm_cost > 2e5:
+            log(
+                WARN,
+                "The accumulated communication cost has exceeded 200,000 MB. "
+                "Please consider reducing it if you plan to participate "
+                "FlowerTune LLM Leaderboard.",
+            )

{flwr_nightly-1.22.0.dev20250916 → flwr_nightly-1.22.0.dev20250918}/py/flwr/cli/new/templates/app/code/model.baseline.py.tpl

@@ -1,7 +1,5 @@
 """$project_name: A Flower Baseline."""
 
-from collections import OrderedDict
-
 import torch
 import torch.nn.functional as F
 from torch import nn
@@ -66,15 +64,3 @@ def test(net, testloader, device):
     accuracy = correct / len(testloader.dataset)
     loss = loss / len(testloader)
     return loss, accuracy
-
-
-def get_weights(net):
-    """Extract model parameters as numpy arrays from state_dict."""
-    return [val.cpu().numpy() for _, val in net.state_dict().items()]
-
-
-def set_weights(net, parameters):
-    """Apply parameters to an existing model."""
-    params_dict = zip(net.state_dict().keys(), parameters)
-    state_dict = OrderedDict({k: torch.from_numpy(v) for k, v in params_dict})
-    net.load_state_dict(state_dict, strict=True)

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/server.baseline.py.tpl

@@ -0,0 +1,43 @@
+"""$project_name: A Flower Baseline."""
+
+import torch
+from flwr.app import ArrayRecord, Context
+from flwr.serverapp import Grid, ServerApp
+from flwr.serverapp.strategy import FedAvg
+
+from $import_name.model import Net
+
+# Create ServerApp
+app = ServerApp()
+
+
+@app.main()
+def main(grid: Grid, context: Context) -> None:
+    """Main entry point for the ServerApp."""
+
+    # Read from config
+    num_rounds = context.run_config["num-server-rounds"]
+    fraction_train = context.run_config["fraction-train"]
+
+    # Load global model
+    global_model = Net()
+    arrays = ArrayRecord(global_model.state_dict())
+
+    # Initialize FedAvg strategy
+    strategy = FedAvg(
+        fraction_train=fraction_train,
+        fraction_evaluate=1.0,
+        min_available_nodes=2,
+    )
+
+    # Start strategy, run FedAvg for `num_rounds`
+    result = strategy.start(
+        grid=grid,
+        initial_arrays=arrays,
+        num_rounds=num_rounds,
+    )
+
+    # Save final model to disk
+    print("\nSaving final model to disk...")
+    state_dict = result.arrays.to_torch_state_dict()
+    torch.save(state_dict, "final_model.pt")

flwr_nightly-1.22.0.dev20250918/py/flwr/cli/new/templates/app/code/server.xgboost.py.tpl

@@ -0,0 +1,56 @@
+"""$project_name: A Flower / $framework_str app."""
+
+import numpy as np
+import xgboost as xgb
+from flwr.app import ArrayRecord, Context
+from flwr.common.config import unflatten_dict
+from flwr.serverapp import Grid, ServerApp
+from flwr.serverapp.strategy import FedXgbBagging
+
+from $import_name.task import replace_keys
+
+# Create ServerApp
+app = ServerApp()
+
+
+@app.main()
+def main(grid: Grid, context: Context) -> None:
+    # Read run config
+    num_rounds = context.run_config["num-server-rounds"]
+    fraction_train = context.run_config["fraction-train"]
+    fraction_evaluate = context.run_config["fraction-evaluate"]
+    # Flatted config dict and replace "-" with "_"
+    cfg = replace_keys(unflatten_dict(context.run_config))
+    params = cfg["params"]
+
+    # Init global model
+    # Init with an empty object; the XGBooster will be created
+    # and trained on the client side.
+    global_model = b""
+    # Note: we store the model as the first item in a list into ArrayRecord,
+    # which can be accessed using index ["0"].
+    arrays = ArrayRecord([np.frombuffer(global_model, dtype=np.uint8)])
+
+    # Initialize FedXgbBagging strategy
+    strategy = FedXgbBagging(
+        fraction_train=fraction_train,
+        fraction_evaluate=fraction_evaluate,
+    )
+
+    # Start strategy, run FedXgbBagging for `num_rounds`
+    result = strategy.start(
+        grid=grid,
+        initial_arrays=arrays,
+        num_rounds=num_rounds,
+    )
+
+    # Save final model to disk
+    bst = xgb.Booster(params=params)
+    global_model = bytearray(result.arrays["0"].numpy().tobytes())
+
+    # Load global model into booster
+    bst.load_model(global_model)
+
+    # Save model
+    print("\nSaving final model to disk...")
+    bst.save_model("final_model.json")