2.4.0RC1

What's Changed

• Add list of publication and talks [skip ci] by @holgerroth in #1089
• fix paper order [skip ci] by @holgerroth in #1093
• Fix docstring description [skip ci] by @YuanTingHsieh in #1071
• Clean up xgboost rst file and move to doc directory [skip ci] by @YuanTingHsieh in #1088
• Fixed the abort_job HA failed to remove snapshot issue. by @yhwen in #1091
• fix date and order [skip ci] by @holgerroth in #1098
• rename check_result to is_resource_enough by @YuanTingHsieh in #1096
• Fix example code format [skip ci] by @YuanTingHsieh in #1094
• Update ci remove snapshot by @YuanTingHsieh in #1101
• add Chinese blogs/talks [skip ci] by @holgerroth in #1100
• Move s3_storage [skip ci] by @YuanTingHsieh in #1095
• Upgrade for Python 3.10. by @yhwen in #1102
• add nvflare paper to citation file [skip ci] by @holgerroth in #1107
• Added field names to auth error by @nvidianz in #1099
• Update build and package steps by @YuanTingHsieh in #1104
• Add ResponseProcessors with example for client-side model init by @holgerroth in #1105
• Add ModelLearnable decomposers by @YuanTingHsieh in #1111
• update model selector by @holgerroth in #1116
• add links and improve introductions of some features for better flow [skip ci] by @nvkevlu in #1120
• Update requirements-dev.txt by @YuanTingHsieh in #1112
• Add feature request issue template [skip ci] by @YuanTingHsieh in #1127
• fix controller racing condition by @holgerroth in #1124
• make aggregation helper thread safe by @holgerroth in #1123
• Add question issue template [skip ci] by @YuanTingHsieh in #1126
• FOBS redesign with enum and data class support by @nvidianz in #1103
• clean up docs for authorization related to 2.2.1 updates [skip ci] by @nvkevlu in #1140
• update status returned when calling logout by @nvkevlu in #1141
• enhanced class_utils. by @yhwen in #1144
• Added warnings for vulnerable OpenSSL versions by @nvidianz in #1142
• Create quickstart page for outside links, fix provisioning section [skip ci] by @nvkevlu in #1145
• Dead clients handle by @yhwen in #1136
• Revert "Added warnings for vulnerable OpenSSL versions" by @nvidianz in #1157
• Enable header scan and fix unit tests by @YuanTingHsieh in #1137
• update cifar-10 sim example to use run script [skip ci] by @holgerroth in #1158
• Fixed a race condition and client privacy.json issue. by @yhwen in #1164
• Signing jobs by @IsaacYangSLA in #1152
• Enable the single gpu option for simulator run by @yhwen in #1162
• Set each container in docker compose with its friendly name by @IsaacYangSLA in #1167
• Add .env file to support variables in compose.yaml by @IsaacYangSLA in #1166
• Fix signing jobs in POC by @IsaacYangSLA in #1169
• Fix unit test filename pattern [skip ci] by @holgerroth in #1173
• add example unit tests by @chesterxgchen in #1170
• restore the changes to fix the bug introduced. by @chesterxgchen in #1175
• Simulartor hang oom by @yhwen in #1159
• Add missing poc_key in api.py by @YuanTingHsieh in #1183
• Scheduler enhancement by @yhwen in #1184
• update the simulator docs [skip ci] by @yhwen in #1186
• Improve XGBoost examples by @Jeffwan in #1181
• Fix type error in FedXGBHistogramExecutor by @Jeffwan in #1179
• Stats refactoring by @chesterxgchen in #1191
• Auto-registering decomposers for Enum by @nvidianz in #1192
• fix simulator missing client heartbeat call. by @yhwen in #1195
• add decorator to measure time for each method call by @chesterxgchen in #1198
• Add BroadcastAndWait Component by @chesterxgchen in #1199
• increase python version support to 3.10 [skip ci] by @chesterxgchen in #1203
• add CommonExecutor and ClientExecutor by @chesterxgchen in #1202
• MLFlowReceiver by @chesterxgchen in #1204
• add missing dependencies [skip ci] by @holgerroth in #1209
• Allow easier import of SimulatorRunner by @holgerroth in #1208
• fix broken link to operation guide [skip ci] by @holgerroth in #1218
• Backend and schema updated by @IsaacYangSLA in #1215
• 1214 svt_privacy adds an option for random.choice replacement by @wyli in #1224
• PSI initial PR by @chesterxgchen in #1207
• add flare_api initial changes by @nvkevlu in #1161
• PSI upgrade to openmined.psi = 1.1.0 by @chesterxgchen in #1226
• update missed requirement [skip ci] by @chesterxgchen in #1227
• Clean up controller logic by @YuanTingHsieh in #1228
• Add quantifying-data-leakage folder [skip ci] by @holgerroth in #1230
• Update monai example [skip ci] by @holgerroth in #1229
• PSI Update example [skip ci] by @chesterxgchen in #1231
• Fix preflight_check issue and add unit tests and integration tests by @YuanTingHsieh in #1234
• update psi readme [skip ci] by @holgerroth in #1235
• Update monai examples again [skip ci] by @holgerroth in #1237
• Only terminate the local session in the worker_process, not logout th… by @yhwen in #1236
• PSI Bug fix [skip ci] by @chesterxgchen in #1238
• Fix integration tests by @YuanTingHsieh in #1239
• remove key check in task get_prop method by @YuanTingHsieh in #1242
• Fix tb streaming by @YuanTingHsieh in #1241
• Update black/isort/flake8 and fix code style [skip ci] by @YuanTingHsieh in #1243
• Research: Quantifying Data Leakage in Federated Learning [skip ci] by @holgerroth in #1244
• Add notebook with FLARE API to example [skip ci] by @nvkevlu in #1232
• Revert flake8 and isort version by @YuanTingHsieh in #1250
• Revert "Backend and schema updated (#1215)" by @IsaacYangSLA in #1251
• Fix missing information in fed_admin.json by @IsaacYangSLA in #1253
• Drop python 3.7 support by @YuanTingHsieh in #1246
• Add kmeans clustering example based on scikit-learn by @ZiyueXu77 in #1163
• Fix joblib_model_param_persistor by @ZiyueXu77 in #1265
• Cloud launch scripts for both client and server by @IsaacYangSLA in #1259
• Script to launch dashboard on cloud by @IsaacYangSLA in #1269
• Add notebook with FLARE API and migration docs, minor improvements by @nvkevlu in #1257
• CIFAR-10: make local number GPUs configurable by @holgerroth in #1268
• Update cifar10 real world instructions by @holgerroth in #1275
• Update diagram on overview [skip ci] by @nvkevlu in #1281
• Add missing init.py by @YuanTingHsieh in #1284
• Allow metrics aggregation by @holgerroth in #1294
• Create pull_request_template.md [skip ci] by @YuanTingHsieh in #1286
• Add sklearn linear model example by @ZiyueXu77 in #1139
• Re-organize example folder structure by @YuanTingHsieh in #1295
• GradInv research: Make prior image filename configurable [skip ci] by @holgerroth in #1302
• Fix issue #1317 by @IsaacYangSLA in #1330
• Random forest algorithm based on xgboost lib by @ZiyueXu77 in #1086
• Add non-linear SVM example by @ZiyueXu77 in #1176
• Add cifar integration test by @nvkevlu in #1315
• PSI: upgrade openmined psi version, restore unit tests and move openmined PSI apt_opt by @chesterxgchen in #1340
• Update documentation by @YuanTingHsieh in #1314
• Update k-means example by @holgerroth in #1341
• Fix validation transform bug by @ZiyueXu77 in #1350
• Move pt from app_common to app_opt by @YuanTingHsieh in #1348
• Increase the wait to ensure the primary sp is updated by @YuanTingHsieh in #1303
• Add support for AWS cloud launch scripts (dashboard/server/client) by @IsaacYangSLA in #1344
• Fix app_opt pt import by @YuanTingHsieh in #1354
• SFM Implementation [skip ci] by @nvidianz in #1352
• Add warnings to ModuleScanner by @YuanTingHsieh in #1347
• Remove mlflow example on dev branch [skip ci] by @YuanTingHsieh in #1360
• FCI TCP Driver [skip ci] by @nvidianz in #1356
• Fix argument error in model_persistence_format_manager by @YuanTingHsieh in #1358
• Standardize tensorboard-streaming example by @YuanTingHsieh in #1359
• Refactor to remove sklearner by @ZiyueXu77 in #1361
• FCI UDS Driver [skip ci] by @nvidianz in #1362
• Fixed the issue could not locate the components due to app_opt move. by @yhwen in #1367
• Merging Cellnet to dev by @nvidianz in #1365
• Fix typo and improve message in dashboard scripts by @IsaacYangSLA in #1374
• move psi executor from app_common to app_opt by @chesterxgchen in #1372
• Merging MPM by @nvidianz in #1375
• Split learning with CIFAR-10 by @holgerroth in #1168
• move psi unit tests to the proper directory [skip ci] by @chesterxgchen in #1357
• FL server deploy integrate with cellnet by @yhwen in #1376
• sklearn linear update [skip ci] by @holgerroth in #1382
• add install dependencies command argument and allow install dependencies [skip ci] by @chesterxgchen in #1383
• FL Client integrate with FCI Cellnet by @yhwen in #1380
• Fixed a bug caused by refactor by @nvidianz in #1384
• disable unit tests again due to CI and other issues [skip ci] by @chesterxgchen in #1386
• Merging Async GRPC Driver [skip ci] by @nvidianz in #1378
• Job run integrate with FCI Cellnet by @yhwen in #1387
• Async HTTP Driver for FCI [skip ci] by @nvidianz in #1390
• Change warning to debug for ModuleScanner by @YuanTingHsieh in #1369
• Multi-process worker integration with FCI cellnet by @yhwen in #1393
• Quick start [skip ci] by @chesterxgchen in #1385
• Move HE from app_common to app_opt. Update app_opt requirements by @YuanTingHsieh in #1392
• Fix in time model selector by @YuanTingHsieh in #1401
• Simulator integration with FCI Cellnet by @yhwen in #1398
• Add limit to number of jobs in list_jobs and options to flare_api by @nvkevlu in #1381
• Fixed Close CB bug and added socket cleanup [skip ci] by @nvidianz in #1399
• Async TCP Driver [skip ci] by @nvidianz in #1397
• fix new_insecure_session by @nvkevlu in #1403
• Update SKLearn readmes and refactor SKLearnExecutor [skip ci] by @holgerroth in #1388
• Async UDS Driver [skip ci] by @nvidianz in #1404
• Add AUC & Homogeneity log to SK Learners [skip ci] by @holgerroth in #1406
• update README for hello-pt on model initialization [skip ci] by @chesterxgchen in #1402
• Graceful cell stop by @yanchengnv in #1405
• Ha fix by @yhwen in #1407
• update README.md [skip ci] by @chesterxgchen in #1408
• README Update for PSI [skip ci] by @chesterxgchen in #1409
• update README 3 [skip ci] by @chesterxgchen in #1410
• Add note for brats18 data access by @ZiyueXu77 in #1245
• Readme 4 [skip ci] by @chesterxgchen in #1413
• Fix integration tests and add preflight check test stage by @YuanTingHsieh in #1370
• remove_endpoint call and fixed bugs [skip ci] by @nvidianz in #1414
• RESTORE Old README before Release [skip ci] by @chesterxgchen in #1418
• Sync fl_context by @holgerroth in #1400
• Randomize azure client resource group by @IsaacYangSLA in #1419
• Enahce Simulator to avoid the Cell Error at end run. by @yhwen in #1421
• Hide cell cmds by @yanchengnv in #1420
• Changed the fetch_task fetch_again without delay. by @yhwen in #1423
• fix default order of jobs in list_jobs command by @nvkevlu in #1416
• Add back the SimulatorRunner by @YuanTingHsieh in #1425
• Move virtual env of all examples to main folder by @ZiyueXu77 in #1411
• Move CIFAR10 example and update CI tests by @YuanTingHsieh in #1415
• Add debug mode to ci by @YuanTingHsieh in #1428
• Restructure hello-world examples to standardize for tests by @nvkevlu in #1412
• Move TBReceiver to experiment tracking by @YuanTingHsieh in #1424
• Add init.py to decomposers folder so build system will include it. by @IsaacYangSLA in #1430
• Add messages at the end of cloud launch scripts by @IsaacYangSLA in #1432
• UPDATE PSI README.md [skip CI] by @chesterxgchen in #1434
• Avoid the simulator cell error after END_RUN. by @yhwen in #1431
• Log Cleanup [skip ci] by @nvidianz in #1426
• Enable Simulator to use resources.json. by @yhwen in #1435
• Fix list jobs command argument parsing bug by @YuanTingHsieh in #1427
• Fixed the simulator hang due to missing import. by @yhwen in #1436
• Switch to --use-device-code for all az login cases [skip ci] by @IsaacYangSLA in #1437
• update nightly build version [skip ci] by @chesterxgchen in #1439
• Enhance the job run process not to kill its own process, instead let … by @yhwen in #1440
• Remove unused codes [skip ci] by @IsaacYangSLA in #1442
• Fixed a few QA bugs. by @yhwen in #1445
• Random forest update by @ZiyueXu77 in #1441
• Improve POC shutdown by @chesterxgchen in #1438
• Randomize resource group to avoid duplicate resource group names [skip ci] by @IsaacYangSLA in #1450
• More detailed error for recursive data by @nvidianz in #1448
• Fixed the QA test recursive ref issue. by @yhwen in #1451
• Fixed the issue job status not updated to exception when controller e… by @yhwen in #1447
• Update integration tests; Add test config auto generation code by @YuanTingHsieh in #1446
• README redesign [skip ci] by @chesterxgchen in #1449
• Check if resource group exists. by @IsaacYangSLA in #1456
• Move split learning to advanced examples; update release notes [skip ci] by @holgerroth in #1457
• Fix admin API issues and support optional messages by @yanchengnv in #1458
• Restore executable scripts in examples [skip ci] by @holgerroth in #1460
• Fix jupyter notebook FLARE API path issue by @IsaacYangSLA in #1462
• Silent Reconnect by @nvidianz in #1463
• Fix shutdown log messages and speed up shutdown process by @yanchengnv in #1465
• Add FedSM example by @ZiyueXu77 in #562
• Randomize security group in AWS dashboard and client scripts by @IsaacYangSLA in #1467
• Fix cifar and auto generated integration tests by @YuanTingHsieh in #1455
• Update Federated Stats to follow the new example structure [skip ci] by @chesterxgchen in #1464
• Update research folders by @holgerroth in #1469
• fixed peer context handling in aux runner by @yanchengnv in #1470
• Convert PSI to the standard test structure [skip ci] by @chesterxgchen in #1468
• Update docs to have release notes in whats new, new glossary, fixes [skip ci] by @nvkevlu in #1461
• Server Listens on All Interfaces by @nvidianz in #1471
• Fix configuration for readthedocs to build docs with new requirements [skip ci] by @nvkevlu in #1472
• Fix fl context prop by @YuanTingHsieh in #1474
• Fixed exception in list_jobs by @nvidianz in #1473
• Fixed the simulator threads option for multi_gpu case. by @yhwen in #1476
• Change Azure VM create to remove warning by @IsaacYangSLA in #1477
• cleanup error msg; fix sag wait; fix get_task timeout by @yanchengnv in #1479
• Fix job runner multiple start issue by @YuanTingHsieh in #1466
• Get rid of hello-examples warnings by @YuanTingHsieh in #1475
• Update protobuf version by @YuanTingHsieh in #1478
• Configuration exception handling by @chesterxgchen in #1480
• fix api status and dead job message by @yanchengnv in #1484
• fix list_job in flare api by @yanchengnv in #1487
• protect server state against multiple state changes by @yanchengnv in #1489
• Fix loading conf in aws scripts by @IsaacYangSLA in #1488
• add wait_for_system_shutdown [skip ci] by @chesterxgchen in #1481
• Add Jupyter-Lab notebooks [skip ci] by @chesterxgchen in #1482
• Fix controller timing issue by @YuanTingHsieh in #1459
• Fix many build warning and issues, more documentation additions [skip ci] by @nvkevlu in #1486
• Splitnn fix by @holgerroth in #1485
• Random forest minor fixes by @ZiyueXu77 in #1495
• fix job listing by @yanchengnv in #1496
• Add user guide on cloud deployment by @IsaacYangSLA in #1497
• Add back sections for migrating that were removed by @nvkevlu in #1498
• simulator create the clients in parallel. by @yhwen in #1491
• Add notebooks for traditional ml examples by @ZiyueXu77 in #1483
• use common JupyterLab instructions by @holgerroth in #1499
• fix wf task exit status handling by @yanchengnv in #1494
• Update what's new to include data leakage tool research by @holgerroth in #1502
• MONAI example updates by @holgerroth in #1506
• Add check on az login exit code by @IsaacYangSLA in #1504
• silent abort message logging by @yanchengnv in #1505
• fix list_jobs detail handling by @yanchengnv in #1503
• not creating internal listener for the job cell. by @yhwen in #1507
• Update README, Notebook, Fed Stats fix by @chesterxgchen in #1501
• Remove unnecessary virtualenv folder [skip ci] by @holgerroth in #1512
• Ensure the start_run event for sub_worker_process. by @yhwen in #1514
• Remove things in init.py in app_opt by @YuanTingHsieh in #1508
• Add notebooks for other machine learning methods by @ZiyueXu77 in #1500
• More docs additions and fixes [skip ci] by @nvkevlu in #1510
• Hello World Notebook (New) [skip ci] by @nvidianz in #1518
• Fix auth test by @YuanTingHsieh in #1519
• CIFAR-10 Auto-FedRL example by @holgerroth in #1283
• Limit the ip address range of inbound ssh to creator's public ip only by @IsaacYangSLA in #1516
• Add one FAQ item to describe DNS cache/propagation and how to resolve it by @IsaacYangSLA in #1517
• Update what's new to include Auto-FedRL [skip ci] by @holgerroth in #1522
• Restore set_env.sh [skip ci] by @holgerroth in #1513
• Check that requirements are consistent through examples, update doc [skip ci] by @nvkevlu in #1521
• Doc & Talks updates [skip ci] by @holgerroth in #1525
• Add config_type to distinguish by @YuanTingHsieh in #1526
• Throw exception when connection monitor is not registered by @nvidianz in #1520
• Optimize the get_all_clients, move to the training process beginning. by @yhwen in #1524
• [CICD] add a new member to trigger list of blossom-ci [skip ci] by @guopengf in #1528
• fix job status and speed up fed event end_run by @yanchengnv in #1523
• [documentation] update logging config example[ski ci] by @chesterxgchen in #1530
• Fix a style issue on FAQ about server DNS propagation/caching. by @IsaacYangSLA in #1533
• Added Decomposers for HE Classes [skip ci] by @nvidianz in #1527
• Raise Exception when error code = EXECUTION_RESULT_ERROR in Controller error handling by @chesterxgchen in #1529
• Fix SAG typo by @YuanTingHsieh in #1536
• Notebooks update [skip ci] by @chesterxgchen in #1541
• Add new notebooks and some updates in docs [skip ci] by @nvkevlu in #1545
• update nvflare version by @chesterxgchen in #1546
• Support direct cell message by @yanchengnv in #1534
• Add a controller_lock to prevent racing condition by @YuanTingHsieh in #1537
• Add link to on-shot-vfl repo [skip ci] by @holgerroth in #1548
• Ha authentication fix by @yhwen in #1535
• Fixed the shared object issue in the controller task return. by @yhwen in #1549
• Remove unneeded cancel all task call by @YuanTingHsieh in #1540
• Add information about ssh source IP (skip ci) by @IsaacYangSLA in #1550
• Fix HA integration tests by @YuanTingHsieh in #1492
• Update scatter & gather messages by @holgerroth in #1552
• Limit the FOBS error log size by @nvidianz in #1544
• Enhance job meta validator by @YuanTingHsieh in #1555
• Use python3 command in examples by @holgerroth in #1551
• update comments and exception messages[skip ci] by @chesterxgchen in #1559
• Remove manual serialize/deserialize for HE components by @holgerroth in #1538
• Fix cell timing by @yanchengnv in #1558
• Restructure docs and notebooks as discussed [skip ci] by @nvkevlu in #1554
• update monai integration versions [skip_ci] by @chesterxgchen in #1560
• Enhance preflight check by @YuanTingHsieh in #1557
• Fixed -m option in list_jobs [skip ci] by @nvidianz in #1556
• Fix integration tests issues by @YuanTingHsieh in #1562
• Fix incorrect server status after job aborted and server restarted by @IsaacYangSLA in #1569
• Updated a re-activate client error message. by @yhwen in #1567
• Early stop on both AWS/Azure when duplicate servers are launched (by … by @IsaacYangSLA in #1553
• Fix abort job with only connected clients by @YuanTingHsieh in #1563
• Fix a typo by @IsaacYangSLA in #1573
• update notebooks based on feedback [skip ci] by @nvkevlu in #1570
• Fix max client in client_manager by @YuanTingHsieh in #1572
• Update fed policy example [skip ci] by @YuanTingHsieh in #1575
• Qa issues by @yhwen in #1568
• Fix a typo in a script [skip ci] by @nvkevlu in #1577
• Notebooks upgrade [skip ci] by @chesterxgchen in #1574
• fix RESULT_ERROR in FedStats by @chesterxgchen in #1579
• Fix SAG client result error handling by @YuanTingHsieh in #1571
• update POC and tutorial storage locations [skip ci] by @chesterxgchen in #1580
• Controller no deepcopy by @yhwen in #1565
• Ensure to end the simulator run after client exception. by @yhwen in #1582
• Update xgboost path in doc [skip ci] by @YuanTingHsieh in #1584
• Notebook and documentation fixes [skip ci] by @nvkevlu in #1581
• update notebook setup_poc [skip_ci] by @chesterxgchen in #1588
• add notebook links to example README.md [skip ci] by @chesterxgchen in #1585
• Update InitializeGlobalWeights workflow to not require clients by @holgerroth in #1576
• Additional reorganization of the linking for the documentation [skip ci] by @nvkevlu in #1591
• Fix provision notebook bugs [ski ci] by @chesterxgchen in #1589
• Minor fix to menu [skip ci] by @nvkevlu in #1594
• Change job_configs to jobs for consistency by @YuanTingHsieh in #1596
• Add example of fednlp for NER task using BERT model by @ZiyueXu77 in #1564
• Ignore unknown task result in SAG by @YuanTingHsieh in #1595
• Cell no executor pool by @yhwen in #1590
• test client-side model initialization by @holgerroth in #1593
• Fixed cell not been stopped properly when config error. by @yhwen in #1597
• fix bugs and cleanup notebooks for FLARE API tutorials [skip ci] by @chesterxgchen in #1598
• Ensure the daemon process that monitors the client root process will end if error happens by @YuanTingHsieh in #1578
• Add job submit success to CI by @YuanTingHsieh in #1601
• Fix typo in fuel communicator by @YuanTingHsieh in #1604
• Fix abort job command return message by @YuanTingHsieh in #1603
• Validate client name type in GlobalWeightsInitializer by @holgerroth in #1606
• Revert "Ignore unknown task result in SAG" by @YuanTingHsieh in #1607
• fix workspace bug in notebook [skip ci] by @chesterxgchen in #1605
• fix POC command bug by @chesterxgchen in #1609
• Update split learning readme [skip ci] by @holgerroth in #1610
• Re-factor PSI and add user email match to CI by @YuanTingHsieh in #1583
• Add section on run modes and fix description for list_jobs in notebook by @nvkevlu in #1600
• Fix various notebook bugs [skip ci] by @chesterxgchen in #1618
• Don't submit update with task data from old SSID by @YuanTingHsieh in #1611
• Job status management enhancement by @yhwen in #1613
• Add docstring to executor [skip ci] by @YuanTingHsieh in #1599
• fix controller dead client handling; added stats pool to_dict by @yanchengnv in #1617
• Make consistent the error message for shutdown_system without auth by @nvkevlu in #1614
• more notebooks bug fixes and updates [slip ci] by @chesterxgchen in #1624
• Fixes several shutdown related issues by @nvidianz in #1608
• fixed the job status for config error. by @yhwen in #1615
• Fixed job could not run when overseer is offline. by @yhwen in #1625
• add qat to repo by @yanchengnv in #1628
• Removing UDS by @nvidianz in #1616
• Add link and base readme to fed-ce repo [skip ci] by @ZiyueXu77 in #1623
• Add readme for one-shot VFL paper [skip ci] by @holgerroth in #1629
• update the abort_job status after the job complete by @yhwen in #1627
• Change default initial task fetch interval at client side from 0.1 to 0.5 by @YuanTingHsieh in #1621
• Add missing parent constructor by @YuanTingHsieh in #1612
• Fix POC stop exception by @chesterxgchen in #1620
• Reduced the non-meaningful logs. by @yhwen in #1630
• Clean up fed stats example by @YuanTingHsieh in #1602
• Add default_task_fetch_interval by @YuanTingHsieh in #1633
• Fixed a save_workspace error. by @yhwen in #1634
• delay the overseer agent start for client job worker process. by @yhwen in #1636
• [PSI] add fl_ctx to finalize() and fix bug by @chesterxgchen in #1638
• Scripts refactoring and notebooks bug fixes/update [skip ci] by @chesterxgchen in #1635
• Fix AIO task cancellation and improve abort_job by @yanchengnv in #1637
• Fix CI by @YuanTingHsieh in #1639
• update the aborted job status immediately by @yhwen in #1640
• remove the _ensure_daemon_process_shutdown which caused restart fail. by @yhwen in #1642
• Correction to xgboost requirements files [skip ci] by @ZiyueXu77 in #1641
• Add GPT-2 model by @ZiyueXu77 in #1626
• Print job schedule result by @YuanTingHsieh in #1631
• Do not shutdown job runner when server turn to cold state by @YuanTingHsieh in #1619
• Fix file license headers by @YuanTingHsieh in #1643
• Add to documentation [skip ci] by @nvkevlu in #1644
• Use secure logging for exceptions by @YuanTingHsieh in #1645
• Fixed the server_command_agent AUTHENTICATION_ERROR reply. by @yhwen in #1648
• Update the _turn_to_cold to set to ColdState first. by @yhwen in #1649
• Improvement on model diff computation by @ZiyueXu77 in #1647
• Fix description of list_jobs in FLARE API notebook [skip ci] by @nvkevlu in #1646
• Fix readme typos [skip ci] by @YuanTingHsieh in #1653
• Change abort_job command to return None by @yanchengnv in #1650
• notebooks tweaks [skip ci] by @chesterxgchen in #1651
• fix abort_job in old FLAdminAPI by @yanchengnv in #1657
• Update monai integration notebook [skip ci] by @YuanTingHsieh in #1652
• Update split nn notebook [skip ci] by @YuanTingHsieh in #1654
• Update xgboost notebooks [skip ci] by @YuanTingHsieh in #1655
• Update RF notebook [skip ci] by @YuanTingHsieh in #1656
• Add notebook info [skip ci] by @nvkevlu in #1658
• Improve example readme [skip ci] by @YuanTingHsieh in #1659
• Add markdown link check workflow [skip ci] by @YuanTingHsieh in #1660
• [2.3] remove apt_opt in the hello-world and tutorial nvflare install lskip ci] by @chesterxgchen in #1663
• [dev] remove apt_opt in the hello-world and tutorial nvflare install lskip ci] by @chesterxgchen in #1664
• Remove unused file [skip ci] by @YuanTingHsieh in #1670
• Update RC to real release by @YuanTingHsieh in #1667
• Add nvflare_getting_started.ipynb, make table for examples, minor fixes [skip ci] by @nvkevlu in #1665
• Add links to new nlp example [skip ci] by @nvkevlu in #1673
• Update overview in documetation [skip ci] by @nvkevlu in #1679
• Fix adhoc conn creation by @yanchengnv in #1676
• Add fed stats examples into CI by @YuanTingHsieh in #1592
• Fix image stats format and typos [skip ci] by @YuanTingHsieh in #1682
• Fix learners in examples by @YuanTingHsieh in #1688
• Support pool stats file creation by @yanchengnv in #1687
• Change sticky to False for SSID by @YuanTingHsieh in #1689
• Install all requirements first in auto test config by @YuanTingHsieh in #1683
• Clean up AdminAPI by @YuanTingHsieh in #1678
• Fix SKLearn executor by @YuanTingHsieh in #1691
• Make CsvRecordHandler thread safe by @yanchengnv in #1694
• Docs fixes links [skip ci] by @nvkevlu in #1685
• Split learning: remove need to set python path outside of notebook by @holgerroth in #1696
• Passing the PYTHONPATH to the child process. by @yhwen in #1695
• add overall stats to pool showing command by @yanchengnv in #1700
• Update to fit monai 1.2 release by @holgerroth in #1701
• Fix SAG docstring by @YuanTingHsieh in #1702
• Fix code format [skip ci] by @YuanTingHsieh in #1707
• Fix minor problems in simulator by @YuanTingHsieh in #1706
• Add cuml backend to SVM example by @ZiyueXu77 in #1709
• Add MONAI mlflow example for local provisioning by @holgerroth in #1710
• Fixed a race condition issue for HA when job is about to end_run. by @yhwen in #1708
• Protect learner against unsafe jobs by @yanchengnv in #1705
• Add FedCE example by @ZiyueXu77 in #1704
• Add scheme in fed_server.json and fed_client.json by @IsaacYangSLA in #1717
• provision refactoring by @chesterxgchen in #1713
• Update assembler of SVM and KMeans to use DXO by @ZiyueXu77 in #1719
• Add dev/prod container images in local and cloud by @IsaacYangSLA in #1720
• Api parity by @yanchengnv in #1722
• Add support to run container in cloud scripts by @IsaacYangSLA in #1728
• Update job path in getting started guide docs [skip ci] by @nvkevlu in #1733
• Fix HA tests by @YuanTingHsieh in #1737
• Add detached mode in docker.sh by @IsaacYangSLA in #1738
• Add codes to generate docker.sh by @IsaacYangSLA in #1741
• Fix readme typo [skip ci] by @YuanTingHsieh in #1746
• Improve client side docker.sh by @IsaacYangSLA in #1753
• Improve server side docker.sh by @IsaacYangSLA in #1754
• FL Hub by @yanchengnv in #1739
• Fix app_opt PTFileModelPersistor typo by @YuanTingHsieh in #1751
• NeMo integration folder & p-tuning example by @holgerroth in #1723
• Upgrade POC: changing POC mode to Local Prod Mode, but keep POC command name. by @chesterxgchen in #1755
• NeMo-NVFlare 20B p-tuning instructions by @holgerroth in #1762
• Move the common header of all cloud scripts to one section by @IsaacYangSLA in #1767
• Make changes for groundwork for experiment tracking by @nvkevlu in #1768
• Add mlflow components by @nvkevlu in #1771
• NeMo integration: Check to initialize learner by @holgerroth in #1775
• Improve pipe for general use by @yanchengnv in #1773
• Enhance DXO aggregator by @YuanTingHsieh in #1769
• NeMo integration: fix initialize logic by @holgerroth in #1776
• Add fl hub example by @holgerroth in #1143
• Revert "Add fl hub example" by @yhwen in #1777
• Re-add FL hub example by @holgerroth in #1778
• Add examples for MLflow by @nvkevlu in #1772
• Add DXO unit test by @holgerroth in #1779
• Fix experiment tracking and integration tests by @YuanTingHsieh in #1784
• Nemo: update license headers; update job paths [skip ci] by @holgerroth in #1780
• Remove dead link [skip ci] by @YuanTingHsieh in #1788
• NeMo: add fed p-tuning figures [skip ci] by @holgerroth in #1792
• Add missing cert chain verification on submitted job signatures by @IsaacYangSLA in #1791
• CodeQL ignore test folder by @YuanTingHsieh in #1794
• Fix deprecated by @YuanTingHsieh in #1793
• Fix tensorboard-streaming integration test by @YuanTingHsieh in #1795
• Add tests for resigning on signed folders by @IsaacYangSLA in #1798
• FLModel: Introduce a common model structure that can cover most of FL Application needs by @chesterxgchen in #1797
• Added GPU group support for simulator. by @yhwen in #1750
• add METRICS_ONLY Transfer Type to FLModel by @chesterxgchen in #1800
• Fix HIGGS dataset link by @YuanTingHsieh in #1801
• Add FLModelUtils by @YuanTingHsieh in #1799
• Enhance runtest.sh by @YuanTingHsieh in #1804
• Allow additional Config Options Part 1 : Config, ConfigLoader, ConfigFormat and ConfigFactory by @chesterxgchen in #1796
• Fix an issue in lighter utils CI tests by @IsaacYangSLA in #1806
• Add UCI website note by @YuanTingHsieh in #1805
• Use --dist loadgroup for pytest by @YuanTingHsieh in #1808
• Fix restart command in FLARE API by @nvkevlu in #1803
• Custom Comm Drivers [skip ci] by @nvidianz in #1632
• Upgrade flask to 2.2.5 to address CVE-2023-30861. by @IsaacYangSLA in #1817
• Refactor pipe and rename pipe monitor by @YuanTingHsieh in #1819
• Allow Alternative Config Option Part 2: Adding OmegaConf loader and Config by @chesterxgchen in #1816
• Make changes to fix MLflow experiment tracking by @nvkevlu in #1823
• Add One-shot VFL project to the "research" folder by @jeremy313 in #1807
• FCI Stream API by @nvidianz in #1758
• PIckup Custom Driver Changes to dev by @nvidianz in #1828

New Contributors

• @Jeffwan made their first contribution in #1181
• @guopengf made their first contribution in #1528
• @jeremy313 made their first contribution in #1807

Full Changelog: 2.2.1...2.4.0rc1