Fix name for hpobench

README.md

@@ -21,7 +21,7 @@ You can try FederatedScope via [FederatedScope Playground](https://try.federated
 - ![new](https://img.alicdn.com/imgextra/i4/O1CN01kUiDtl1HVxN6G56vN_!!6000000000764-2-tps-43-19.png) [08-18-2022] Our KDD 2022 [paper](https://arxiv.org/abs/2204.05562) on federated graph learning receives the KDD Best Paper Award for ADS track!
 - ![new](https://img.alicdn.com/imgextra/i4/O1CN01kUiDtl1HVxN6G56vN_!!6000000000764-2-tps-43-19.png) [07-30-2022] We release FederatedScope v0.2.0! 
 - [06-17-2022] We release **pFL-Bench**, a comprehensive benchmark for personalized Federated Learning (pFL), containing 10+ datasets and 20+ baselines. [[code](https://github.com/alibaba/FederatedScope/tree/master/benchmark/pFL-Bench), [pdf](https://arxiv.org/abs/2206.03655)]
-- [06-17-2022] We release **FedHPO-B**, a benchmark suite for studying federated hyperparameter optimization. [[code](https://github.com/alibaba/FederatedScope/tree/master/benchmark/FedHPOB), [pdf](https://arxiv.org/abs/2206.03966)]
+- [06-17-2022] We release **FedHPO-Bench**, a benchmark suite for studying federated hyperparameter optimization. [[code](https://github.com/alibaba/FederatedScope/tree/master/benchmark/FedHPOBench), [pdf](https://arxiv.org/abs/2206.03966)]
 - [06-17-2022] We release **B-FHTL**, a benchmark suit for studying federated hetero-task learning. [[code](https://github.com/alibaba/FederatedScope/tree/master/benchmark/B-FHTL), [pdf](https://arxiv.org/abs/2206.03436)]
 - [06-13-2022] Our project was receiving an attack, which has been resolved. [More details](https://github.com/alibaba/FederatedScope/blob/master/doc/news/06-13-2022_Declaration_of_Emergency.txt).
 - [05-25-2022] Our paper [FederatedScope-GNN](https://arxiv.org/abs/2204.05562) has been accepted by KDD'2022!

benchmark/FedHPOBench/README.md

@@ -1,14 +1,14 @@
-# FedHPO-B
+# FedHPO-Bench
 
-A benchmark suite for studying federated hyperparameter optimization. FedHPO-B incorporates comprehensive FL tasks, enables efficient function evaluations, and eases continuing extensions. We also conduct extensive experiments based on FedHPO-B to benchmark a few HPO methods.
+A benchmark suite for studying federated hyperparameter optimization. FedHPO-Bench incorporates comprehensive FL tasks, enables efficient function evaluations, and eases continuing extensions. We also conduct extensive experiments based on FedHPO-B to benchmark a few HPO methods.
 
 ## Quick Start
 
-We highly recommend running FedHPO-B with conda.
+We highly recommend running FedHPO-Bench with conda.
 
 ### Step 0. Dependency
 
-* FedHPO-B is built on a stable [FederatedScope](https://github.com/alibaba/FederatedScope), please see [Installation](https://github.com/alibaba/FederatedScope#step-1-installation) for install FederatedScope.
+* FedHPO-Bench is built on a stable [FederatedScope](https://github.com/alibaba/FederatedScope), please see [Installation](https://github.com/alibaba/FederatedScope#step-1-installation) for install FederatedScope.
 
   ```bash
   git clone https://github.com/alibaba/FederatedScope.git
@@ -33,7 +33,7 @@ We highly recommend running FedHPO-B with conda.
 
 ### Step 1. Installation
 
-We recommend installing FedHPOB directly using git by:
+We recommend installing FedHPOBench directly using git by:
 
 ```bash
 git clone https://github.com/alibaba/FederatedScope.git
@@ -43,9 +43,9 @@ export PYTHONPATH=~/FedHPOBench:$PYTHONPATH
 
 ### Step 2. Prepare data files
 
-**Note**: If you only want to use FedHPO-B with raw mode, you can skip to **Step3**.
+**Note**: If you only want to use FedHPO-Bench with raw mode, you can skip to **Step3**.
 
-All data files are available on AliyunOSS, you need to download the data files and place them in the `~/data/tabular_data/` or `~/data/surrogate_model/`  before using FedHPO-B.
+All data files are available on AliyunOSS, you need to download the data files and place them in the `~/data/tabular_data/` or `~/data/surrogate_model/`  before using FedHPO-Bench.
 
 The naming pattern of the url of data files obeys the rule:
 
@@ -62,8 +62,8 @@ Fortunately, we provide tools to automatically convert from tabular data to surr
 ### Step3. Start running
 
 ```python
-from fedhpob.config import fhb_cfg
-from fedhpob.benchmarks import TabularBenchmark
+from fedhpobench.config import fhb_cfg
+from fedhpobench.benchmarks import TabularBenchmark
 
 benchmark = TabularBenchmark('cnn', 'femnist', 'avg')
 
@@ -99,7 +99,7 @@ We take Figure 11 as an example.
 * Then draw the figure with tools we provide, the figures will be saved in `~/figures`.
 
   ```python
-  from fedhpob.utils.draw import rank_over_time
+  from fedhpobench.utils.draw import rank_over_time
 
   rank_over_time('exp_results', 'gcn', algo='avg', loss=False)
   ```
@@ -130,11 +130,11 @@ How to use:
 
 ## Publications
 
-If you find FedHPO-B useful for your research or development, please cite the following [paper](https://arxiv.org/abs/2206.03966):
+If you find FedHPO-Bench useful for your research or development, please cite the following [paper](https://arxiv.org/abs/2206.03966):
 
 ```tex
 @article{Wang2022FedHPOBAB,
-  title={FedHPO-B: A Benchmark Suite for Federated Hyperparameter Optimization},
+  title={FedHPO-Bench: A Benchmark Suite for Federated Hyperparameter Optimization},
   author={Zhen Wang and Weirui Kuang and Ce Zhang and Bolin Ding and Yaliang Li},
   journal={ArXiv},
   year={2022},

benchmark/FedHPOBench/demo/femnist_surrogate_benchmark.py

@@ -112,7 +112,7 @@ def get_meta_information() -> Dict:
                 'journal = {arXiv preprint arXiv:2206.03966},'
                 'year    = {2022}}', 'https://arxiv.org/pdf/2206.03966v4.pdf',
                 'https://github.com/alibaba/FederatedScope/tree/master'
-                '/benchmark/FedHPOB'
+                '/benchmark/FedHPOBench'
             ],
             'code': 'https://github.com/alibaba/FederatedScope/tree/master'
             '/benchmark/FedHPOBench',

benchmark/FedHPOBench/demo/femnist_tabular_benchmark.py

@@ -24,7 +24,7 @@ def __init__(self,
         Optimization",
         url: https://arxiv.org/pdf/2206.03966v4.pdf
         Source: https://github.com/alibaba/FederatedScope/tree/master
-        /benchmark/FedHPOB
+        /benchmark/FedHPOBench
         Parameters
         ----------
         data_path : str, Path
@@ -128,7 +128,7 @@ def get_meta_information() -> Dict:
                 'journal = {arXiv preprint arXiv:2206.03966},'
                 'year    = {2022}}', 'https://arxiv.org/pdf/2206.03966v4.pdf',
                 'https://github.com/alibaba/FederatedScope/tree/master'
-                '/benchmark/FedHPOB'
+                '/benchmark/FedHPOBench'
             ],
             'code': 'https://github.com/alibaba/FederatedScope/tree/master'
             '/benchmark/FedHPOBench',

benchmark/FedHPOBench/fedhpobench/benchmarks/__init__.py

@@ -1,5 +1,5 @@
-from fedhpob.benchmarks.raw_benchmark import RawBenchmark
-from fedhpob.benchmarks.tabular_benchmark import TabularBenchmark
-from fedhpob.benchmarks.surrogate_benchmark import SurrogateBenchmark
+from fedhpobench.benchmarks.raw_benchmark import RawBenchmark
+from fedhpobench.benchmarks.tabular_benchmark import TabularBenchmark
+from fedhpobench.benchmarks.surrogate_benchmark import SurrogateBenchmark
 
 __all__ = ['RawBenchmark', 'TabularBenchmark', 'SurrogateBenchmark']

benchmark/FedHPOBench/fedhpobench/benchmarks/base_benchmark.py

@@ -5,9 +5,9 @@
 import numpy as np
 from federatedscope.core.configs.config import global_cfg
 from federatedscope.core.auxiliaries.data_builder import get_data
-from fedhpob.utils.tabular_dataloader import load_data
-from fedhpob.utils.util import disable_fs_logger
-from fedhpob.utils.cost_model import get_cost_model
+from fedhpobench.utils.tabular_dataloader import load_data
+from fedhpobench.utils.util import disable_fs_logger
+from fedhpobench.utils.cost_model import get_cost_model
 
 
 class BaseBenchmark(abc.ABC):
@@ -79,14 +79,14 @@ def get_lamba_from_df(self, configuration, fidelity):
                 filterd_result['eval_time'])
             return c.total_seconds() / float(client_num)
         else:
-            from fedhpob.config import fhb_cfg
+            from fedhpobench.config import fhb_cfg
             return fhb_cfg.cost.c
 
     def _cost(self, configuration, fidelity, **kwargs):
         try:
             kwargs['const'] = self.get_lamba_from_df(configuration, fidelity)
         except:
-            from fedhpob.config import fhb_cfg
+            from fedhpobench.config import fhb_cfg
             kwargs['const'] = fhb_cfg.cost.c
         cost_model = get_cost_model(mode=self.cost_mode)
         t = cost_model(self.cfg, configuration, fidelity, self.data, **kwargs)

benchmark/FedHPOBench/fedhpobench/benchmarks/raw_benchmark.py

@@ -5,9 +5,9 @@
     get_server_cls
 from federatedscope.core.fed_runner import FedRunner
 
-from fedhpob.benchmarks.base_benchmark import BaseBenchmark
-from fedhpob.utils.util import disable_fs_logger
-from fedhpob.utils.cost_model import merge_cfg
+from fedhpobench.benchmarks.base_benchmark import BaseBenchmark
+from fedhpobench.utils.util import disable_fs_logger
+from fedhpobench.utils.cost_model import merge_cfg
 
 
 class RawBenchmark(BaseBenchmark):

benchmark/FedHPOBench/fedhpobench/benchmarks/surrogate_benchmark.py

@@ -1,7 +1,7 @@
 import os
 
-from fedhpob.benchmarks.base_benchmark import BaseBenchmark
-from fedhpob.utils.surrogate_dataloader import build_surrogate_model, \
+from fedhpobench.benchmarks.base_benchmark import BaseBenchmark
+from fedhpobench.utils.surrogate_dataloader import build_surrogate_model, \
     load_surrogate_model
 
 

benchmark/FedHPOBench/fedhpobench/benchmarks/tabular_benchmark.py

@@ -2,9 +2,9 @@
 import logging
 import numpy as np
 
-from fedhpob.utils.util import dict2cfg
-from fedhpob.utils.tabular_dataloader import load_data
-from fedhpob.benchmarks.base_benchmark import BaseBenchmark
+from fedhpobench.utils.util import dict2cfg
+from fedhpobench.utils.tabular_dataloader import load_data
+from fedhpobench.benchmarks.base_benchmark import BaseBenchmark
 
 
 class TabularBenchmark(BaseBenchmark):

benchmark/FedHPOBench/fedhpobench/config.py

@@ -1,8 +1,8 @@
 import ConfigSpace as CS
 from federatedscope.core.configs.config import CN
-from fedhpob.benchmarks import TabularBenchmark
-from fedhpob.benchmarks import RawBenchmark
-from fedhpob.benchmarks import SurrogateBenchmark
+from fedhpobench.benchmarks import TabularBenchmark
+from fedhpobench.benchmarks import RawBenchmark
+from fedhpobench.benchmarks import SurrogateBenchmark
 
 fhb_cfg = CN()
 

benchmark/FedHPOBench/fedhpobench/optimizers/__init__.py

@@ -1,8 +1,8 @@
-from fedhpob.optimizers.dehb_optimizer import run_dehb
-from fedhpob.optimizers.hpbandster_optimizer import run_hpbandster
-from fedhpob.optimizers.optuna_optimizer import run_optuna
-from fedhpob.optimizers.smac_optimizer import run_smac
-from fedhpob.optimizers.grid_search import run_grid_search
+from fedhpobench.optimizers.dehb_optimizer import run_dehb
+from fedhpobench.optimizers.hpbandster_optimizer import run_hpbandster
+from fedhpobench.optimizers.optuna_optimizer import run_optuna
+from fedhpobench.optimizers.smac_optimizer import run_smac
+from fedhpobench.optimizers.grid_search import run_grid_search
 
 __all__ = [
     'run_dehb', 'run_hpbandster', 'run_optuna', 'run_smac', 'run_grid_search'

benchmark/FedHPOBench/fedhpobench/optimizers/dehb_optimizer.py

@@ -27,8 +27,8 @@
 import random
 import logging
 from dehb.optimizers import DE, DEHB
-from fedhpob.config import fhb_cfg
-from fedhpob.utils.monitor import Monitor
+from fedhpobench.config import fhb_cfg
+from fedhpobench.utils.monitor import Monitor
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/optimizers/grid_search.py

@@ -6,8 +6,8 @@
 import ConfigSpace as CS
 from ConfigSpace.util import generate_grid
 
-from fedhpob.config import fhb_cfg
-from fedhpob.utils.monitor import Monitor
+from fedhpobench.config import fhb_cfg
+from fedhpobench.utils.monitor import Monitor
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/optimizers/hpbandster_optimizer.py

@@ -7,8 +7,8 @@
 from hpbandster.core.worker import Worker
 from hpbandster.optimizers import BOHB, HyperBand, RandomSearch
 
-from fedhpob.config import fhb_cfg
-from fedhpob.utils.monitor import Monitor
+from fedhpobench.config import fhb_cfg
+from fedhpobench.utils.monitor import Monitor
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/optimizers/optuna_optimizer.py

@@ -13,8 +13,8 @@
 from optuna.samplers import TPESampler
 from optuna.trial import Trial
 
-from fedhpob.config import fhb_cfg
-from fedhpob.utils.monitor import Monitor
+from fedhpobench.config import fhb_cfg
+from fedhpobench.utils.monitor import Monitor
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/optimizers/smac_optimizer.py

@@ -7,8 +7,8 @@
 from smac.facade.smac_hpo_facade import SMAC4HPO
 from smac.scenario.scenario import Scenario
 
-from fedhpob.config import fhb_cfg
-from fedhpob.utils.monitor import Monitor
+from fedhpobench.config import fhb_cfg
+from fedhpobench.utils.monitor import Monitor
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/utils/draw.py

@@ -24,7 +24,7 @@ def logloader(file):
 
 
 def ecdf(model, data_list, algo, sample_client=None, key='test_acc'):
-    from fedhpob.benchmarks import TabularBenchmark
+    from fedhpobench.benchmarks import TabularBenchmark
 
     # Draw ECDF from target data_list
     plt.figure(figsize=(10, 7.5))
@@ -294,8 +294,8 @@ def landscape(model='cnn',
               sample_client=None,
               key='test_acc'):
     import plotly.graph_objects as go
-    from fedhpob.config import fhb_cfg
-    from fedhpob.benchmarks import TabularBenchmark
+    from fedhpobench.config import fhb_cfg
+    from fedhpobench.benchmarks import TabularBenchmark
 
     z = []
     benchmark = TabularBenchmark(model, dname, algo, device=-1)

benchmark/FedHPOBench/fedhpobench/utils/monitor.py

@@ -5,7 +5,7 @@
 
 import numpy as np
 
-from fedhpob.utils.util import cfg2name
+from fedhpobench.utils.util import cfg2name
 
 logging.basicConfig(level=logging.WARNING)
 

benchmark/FedHPOBench/fedhpobench/utils/runner.py

@@ -1,6 +1,6 @@
 from federatedscope.core.cmd_args import parse_args
-from fedhpob.config import fhb_cfg, add_configs
-from fedhpob.optimizers import run_dehb, run_hpbandster, run_optuna, \
+from fedhpobench.config import fhb_cfg, add_configs
+from fedhpobench.optimizers import run_dehb, run_hpbandster, run_optuna, \
     run_smac, run_grid_search
 
 

benchmark/FedHPOBench/fedhpobench/utils/surrogate_dataloader.py

@@ -7,7 +7,7 @@
 from sklearn.model_selection import cross_validate as sk_cross_validate
 from tqdm import tqdm
 
-from fedhpob.utils.tabular_dataloader import load_data
+from fedhpobench.utils.tabular_dataloader import load_data
 
 
 def sampling(X, Y, over_rate=1, down_rate=1.0, cvg_score=0.5):

benchmark/FedHPOBench/scripts/cross_device/run_hpo_twitter_lr.sh

@@ -31,7 +31,7 @@ do
             do
                 for k in {1..3}
                 do
-                    python federatedscope/main.py --cfg benchmark/FedHPOB/scripts/lr/twitter.yaml device $cudaid train.optimizer.lr $lr train.optimizer.weight_decay ${wds[$w]} train.local_update_steps ${steps[$s]} data.batch_size ${batch_sizes[$b]} federate.sample_client_rate ${sample_rates[$sr]} seed $k outdir lr/${out_dir}_${sample_rates[$sr]} expname lr${lr}_wd${wds[$w]}_dropout0_step${steps[$s]}_batch${batch_sizes[$b]}_seed${k}
+                    python federatedscope/main.py --cfg benchmark/FedHPOBench/scripts/lr/twitter.yaml device $cudaid train.optimizer.lr $lr train.optimizer.weight_decay ${wds[$w]} train.local_update_steps ${steps[$s]} data.batch_size ${batch_sizes[$b]} federate.sample_client_rate ${sample_rates[$sr]} seed $k outdir lr/${out_dir}_${sample_rates[$sr]} expname lr${lr}_wd${wds[$w]}_dropout0_step${steps[$s]}_batch${batch_sizes[$b]}_seed${k}
                 done
             done
         done

benchmark/FedHPOBench/scripts/exp/run_mode.sh

@@ -7,7 +7,7 @@ device=$4
 algo=$5
 
 cd ../..
-cp fedhpob/utils/runner.py . || echo "File exists."
+cp fedhpobench/utils/runner.py . || echo "File exists."
 
 for k in {1..5}; do
   python runner.py --cfg scripts/exp/${dataset}.yaml benchmark.device ${device} benchmark.model ${model} benchmark.type ${mode} benchmark.data ${dataset} benchmark.algo ${algo} optimizer.type rs || echo "continue"

benchmark/FedHPOBench/scripts/gcn/run_hpo_cora_dp.sh

@@ -24,7 +24,7 @@ for ((l = 0; l < ${#lrs[@]}; l++)); do
     for ((d = 0; d < ${#dps[@]}; d++)); do
       for ((s = 0; s < ${#steps[@]}; s++)); do
         for k in {1..3}; do
-          python federatedscope/main.py --cfg benchmark/FedHPOB/scripts/gcn/cora_dp.yaml device $cudaid train.optimizer.lr ${lrs[$l]} train.optimizer.weight_decay ${wds[$w]} model.dropout ${dps[$d]} train.local_update_steps ${steps[$s]} federate.sample_client_num $sample_num seed $k outdir ${out_dir}/${sample_num} expname lr${lrs[$l]}_wd${wds[$w]}_dropout${dps[$d]}_step${steps[$s]}_seed${k} >/dev/null 2>&1
+          python federatedscope/main.py --cfg benchmark/FedHPOBench/scripts/gcn/cora_dp.yaml device $cudaid train.optimizer.lr ${lrs[$l]} train.optimizer.weight_decay ${wds[$w]} model.dropout ${dps[$d]} train.local_update_steps ${steps[$s]} federate.sample_client_num $sample_num seed $k outdir ${out_dir}/${sample_num} expname lr${lrs[$l]}_wd${wds[$w]}_dropout${dps[$d]}_step${steps[$s]}_seed${k} >/dev/null 2>&1
         done
       done
     done

benchmark/FedHPOBench/scripts/gcn/run_prox_cora.sh

@@ -24,7 +24,7 @@ for ((l = 0; l < ${#lrs[@]}; l++)); do
     for ((d = 0; d < ${#dps[@]}; d++)); do
       for ((s = 0; s < ${#steps[@]}; s++)); do
         for k in {1..3}; do
-          python federatedscope/main.py --cfg benchmark/FedHPOB/scripts/gcn/cora_prox.yaml device $cudaid train.optimizer.lr ${lrs[$l]} fedprox.use True fedprox.mu ${mu} train.optimizer.weight_decay ${wds[$w]} model.dropout ${dps[$d]} train.local_update_steps ${steps[$s]} federate.sample_client_num $sample_num seed $k outdir ${out_dir}/${sample_num} expname lr${lrs[$l]}_wd${wds[$w]}_dropout${dps[$d]}_step${steps[$s]}_mu${mu}_seed${k} >/dev/null 2>&1
+          python federatedscope/main.py --cfg benchmark/FedHPOBench/scripts/gcn/cora_prox.yaml device $cudaid train.optimizer.lr ${lrs[$l]} fedprox.use True fedprox.mu ${mu} train.optimizer.weight_decay ${wds[$w]} model.dropout ${dps[$d]} train.local_update_steps ${steps[$s]} federate.sample_client_num $sample_num seed $k outdir ${out_dir}/${sample_num} expname lr${lrs[$l]}_wd${wds[$w]}_dropout${dps[$d]}_step${steps[$s]}_mu${mu}_seed${k} >/dev/null 2>&1
         done
       done
     done