- perf(artifacts): reuse session for file upload requests by @speezepearson in wandb#4708
- feat(artifacts): expose aliases list endpoint for artifact collections by @ibindlish in wandb#4809
- feat(launch): include the username of the run's author in the environment variables by @TimH98 in wandb#4851
- feat(launch): add support for local-container resource args by @KyleGoyette in wandb#4846
- feat(sdk): add the ability to append to a run with
wandb sync --append
by @raubitsj in wandb#4848 - feat(launch): add an escape hatch (
disable_job_creation
) to disable automatic job creation by @KyleGoyette in wandb#4901
- fix(launch): remove underscores from generated job name in kubernetes runner by @TimH98 in wandb#4752
- fix(sweeps): sweep command args can once again be int type by @gtarpenning in wandb#4728
- fix(artifacts): ensure prepared artifacts have the
latest
alias by @moredatarequired in wandb#4828 - fix(artifacts): catch FileNotFoundError and PermissionError during cache.cleanup() by @moredatarequired in wandb#4868
- fix(sdk): fix order of python executable resolves by @kptkin in wandb#4839
- fix(sdk): fix console handling when forking and setting stdout==stderr by @raubitsj in wandb#4877
- fix(launch): Fix issue where job artifacts are being logged without latest alias by @KyleGoyette in wandb#4884
- fix(launch): Ensure job names do not exceed maximum allowable for artifacts by @KyleGoyette in wandb#4889
- docs(sdk): fix broken reference link to W&B Settings page in Sweeps by @ngrayluna in wandb#4820
- docs(sdk): Docodoile autogen docs by @ngrayluna in wandb#4734
- test(artifacts): ensure manifest version is verified by @moredatarequired in wandb#4691
- test(sdk): add tests for custom SSL certs and disabling SSL by @speezepearson in wandb#4692
- test(sdk): fix nightly docker builds by @dmitryduev in wandb#4787
- chore(sdk): dont create universal py2/py3 package by @raubitsj in wandb#4797
- chore(sdk): fix flake8-bugbear B028 and ignore B017 by @kptkin in wandb#4799
- test(sdk): fix gcloud sdk version requested in nightly tests by @dmitryduev in wandb#4802
- chore(artifacts): remove unused parameters in StorageHandler.load_{path,file,reference} by @moredatarequired in wandb#4678
- chore(sdk): split unit tests to system tests and proper unit tests by @kptkin in wandb#4811
- test(sdk): address fixture server move from port 9010 to 9015 in local-testcontainer by @dmitryduev in wandb#4814
- chore(sdk): add aliases to ac query response by @ibindlish in wandb#4813
- test(sdk): run regression suite nightly by @dmitryduev in wandb#4788
- test(sdk): fix broken lightning test by @kptkin in wandb#4823
- chore(sdk): enable type checking for wandb_init.py by @dmitryduev in wandb#4784
- chore(launch): deprecate defaulting to default queue in launch-agent command by @gtarpenning in wandb#4801
- test(launch): add unit test for kubernetes runner with annotations by @TimH98 in wandb#4800
- test(integrations): fix train_gpu_ddp test by @dmitryduev in wandb#4831
- chore(sdk): fix docker testimage to pull amd64 version by @raubitsj in wandb#4838
- chore(sdk): fix codeowners after test restructure by @raubitsj in wandb#4843
- test(sdk): fix md5 test failures on Windows by @moredatarequired in wandb#4840
- chore(sdk): split out relay server so it can be shared with yea-wandb by @raubitsj in wandb#4837
- chore(sdk): fix a flake8 complaint in a test by @speezepearson in wandb#4806
- test(integrations): fix several import tests by @dmitryduev in wandb#4849
- test(sdk): don't use symlinks for SSL test assets, because Windows by @speezepearson in wandb#4847
- test(sdk): add unit tests for filesync.Stats by @speezepearson in wandb#4855
- chore(sdk): add async retry logic by @speezepearson in wandb#4738
- test(artifacts): strengthen tests for ArtifactSaver, StepUpload by @speezepearson in wandb#4808
- chore(launch): Agent logs full stack trace when catching exception by @TimH98 in wandb#4861
- chore(sdk): swallow warning printed by neuron-ls by @dmitryduev in wandb#4835
- build(sdk): pin pip and tox in development environments by @moredatarequired in wandb#4871
- refactor(sdk): strengthen StepUpload tests; make exception-handling more thorough in upload/commit by @speezepearson in wandb#4677
- refactor(artifacts): refactor Artifact query to fetch entity and project by @vwrj in wandb#4775
- refactor(sdk): replace more communicate calls with deliver by @raubitsj in wandb#4841
- refactor(artifacts): internally use Future to communicate success/failure of commit, not threading.Event by @speezepearson in wandb#4859
- refactor(sdk): use stdlib ThreadPoolExecutor in StepUpload instead of managing our own by @speezepearson in wandb#4860
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.9...v0.13.10
- fix(sdk): exercise extra caution when checking if AWS Trainium is available in the system by @dmitryduev in wandb#4769
- fix(sdk): restore 'util.generate_id' for legacy / user code by @moredatarequired in wandb#4776
- fix(sdk): replace
release
withabandon
when releasing mailbox handle during init by @kptkin in wandb#4766
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.8...v0.13.9
- feat(artifacts): keep uncommitted uploads in separate staging area by @moredatarequired in wandb#4505
- perf(sdk): improve file descriptor management by @dmitryduev in wandb#4617
- feat(launch): default to using model-registry project for agent and launch_add by @KyleGoyette in wandb#4613
- feat(sdk): add
exist_ok=False
tofile.download()
by @janosh in wandb#4564 - feat(launch): auto create job artifacts from runs with required ingredients by @KyleGoyette in wandb#4660
- feat(sdk): add generalized response injection pattern for tests by @kptkin in wandb#4729
- perf(sdk): replace multiprocessing.Queue's with queue.Queue's by @dmitryduev in wandb#4672
- feat(sdk): use transaction log to cap memory usage by @raubitsj in wandb#4724
- feat(integrations): support system metrics for AWS Trainium by @dmitryduev in wandb#4671
- fix(sdk): correct the type hint for wandb.run by @edwag in wandb#4585
- fix(sdk): resume collecting system metrics on object restart by @dmitryduev in wandb#4572
- fix(launch): fix env handling and node_selector handling by @KyleGoyette in wandb#4555
- fix(public-api): fix Job.call() using the wrong keyword (queue vs queue_name) when calling launch_add. by @TimH98 in wandb#4625
- fix(sweeps): sweeps schedulers handles multi word parameters by @gtarpenning in wandb#4640
- fix(launch): allow spaces in requirements file, remove duplicate wandb bootstrap file by @TimH98 in wandb#4647
- fix(artifacts): correctly handle url-encoded local file references. by @moredatarequired in wandb#4665
- fix(artifacts): get digest directly instead of from the manifests' manifest by @moredatarequired in wandb#4681
- fix(artifacts): artifact.version should be the version index from the associated collection by @vwrj in wandb#4486
- fix(sdk): remove duplicate generate_id functions, replace shortuuid with secrets by @moredatarequired in wandb#4676
- fix(integrations): fix type check for jax.Array introduced in jax==0.4.1 by @dmitryduev in wandb#4718
- fix(sdk): fix hang after failed wandb.init (add cancel) by @raubitsj in wandb#4405
- fix(sdk): allow users to provide path to custom executables by @kptkin in wandb#4604
- fix(sdk): fix TypeError when trying to slice a Paginator object by @janosh in wandb#4575
- fix(integrations): add
AttributeError
to the list of handled exceptions when saving a keras model by @froody in wandb#4732 - fix(launch): remove args from jobs by @KyleGoyette in wandb#4750
- docs(sweeps): fix typo in docs by @gtarpenning in wandb#4627
- docs(sdk): fix typo in docstring for data_types.Objects3D by @ngrayluna in wandb#4543
- docs(sdk): remove less than, greater than characters from dosctrings… by @ngrayluna in wandb#4687
- docs(sdk): update SECURITY.md by @dmitryduev in wandb#4616
- docs(sdk): Update README.md by @ngrayluna in wandb#4468
- test(sdk): update t2_fix_error_cond_feature_importances to install scikit-learn by @dmitryduev in wandb#4573
- chore(sdk): update base Docker images for nightly testing by @dmitryduev in wandb#4566
- chore(sdk): change sklearn to scikit-learn in functional sacred test by @dmitryduev in wandb#4577
- chore(launch): add error check for
--build
when resource=local-process by @gtarpenning in wandb#4513 - chore(sweeps): update scheduler and agent resource handling to allow DRC override by @gtarpenning in wandb#4480
- chore(sdk): require sdk-team review for adding or removing high-level… by @dmitryduev in wandb#4594
- chore(launch): remove requirement to make target project match queue by @KyleGoyette in wandb#4612
- chore(sdk): enhance nightly cloud testing process by @dmitryduev in wandb#4602
- chore(sdk): update pull request template by @raubitsj in wandb#4633
- chore(launch): return updated runSpec after pushToRunQueue query by @gtarpenning in wandb#4516
- chore(launch): fix for run spec handling in sdk by @gtarpenning in wandb#4636
- chore(sdk): remove test dependency on old fastparquet package by @raubitsj in wandb#4656
- test(artifacts): fix dtype np.float (does not exist), set to python float by @moredatarequired in wandb#4661
- chore(sdk): correct 'exclude' to 'ignore-paths' in .pylintrc by @moredatarequired in wandb#4659
- chore(sdk): use pytest tmp_path so we can inspect failures by @raubitsj in wandb#4664
- chore(launch): reset build command after building by @gtarpenning in wandb#4626
- ci(sdk): rerun flaking tests in CI with pytest-rerunfailures by @dmitryduev in wandb#4430
- chore(sdk): remove dead code from filesync logic by @speezepearson in wandb#4638
- chore(sdk): remove unused fields from a filesync message by @speezepearson in wandb#4662
- chore(sdk): refactor retry logic to use globals instead of dependency-injecting them by @speezepearson in wandb#4588
- test(sdk): add unit tests for filesync.StepUpload by @speezepearson in wandb#4652
- test(sdk): add tests for Api.upload_file_retry by @speezepearson in wandb#4639
- chore(launch): remove fallback resource when not specified for a queue by @gtarpenning in wandb#4637
- test(artifacts): improve storage handler test coverage by @moredatarequired in wandb#4674
- test(integrations): fix import tests by @dmitryduev in wandb#4690
- chore(sdk): make MetricsMonitor less verbose on errors by @dmitryduev in wandb#4618
- test(sdk): address fixture server move from port 9003 to 9010 in local-testcontainer by @dmitryduev in wandb#4716
- chore(sdk): vendor promise==2.3.0 to unequivocally rm six dependency by @dmitryduev in wandb#4622
- chore(artifacts): allow setting artifact cache dir in wandb.init(...) by @dmitryduev in wandb#3644
- test(sdk): temporary lower network buffer for testing by @raubitsj in wandb#4737
- chore(sdk): add telemetry if the user running in pex environment by @kptkin in wandb#4747
- chore(sdk): add more flow control telemetry by @raubitsj in wandb#4739
- chore(sdk): add settings and debug for service startup issues (wait_for_ports) by @raubitsj in wandb#4749
- test(sdk): fix AWS Trainium test by @dmitryduev in wandb#4753
- chore(sdk): fix status checker thread issue when user process exits without finish() by @raubitsj in wandb#4761
- chore(sdk): add telemetry for service disabled usage by @kptkin in wandb#4762
- style(sdk): use the same syntax whenever raising exceptions by @moredatarequired in wandb#4559
- refactor(sdk): combine _safe_mkdirs with mkdir_exist_ok by @moredatarequired in wandb#4650
- refactor(artifacts): use a pytest fixture for the artifact cache by @moredatarequired in wandb#4648
- refactor(artifacts): use ArtifactEntry directly instead of subclassing by @moredatarequired in wandb#4649
- refactor(artifacts): consolidate hash utilities into lib.hashutil by @moredatarequired in wandb#4525
- style(public-api): format public file with proper formating by @kptkin in wandb#4697
- chore(sdk): install tox into proper env in dev env setup tool by @dmitryduev in wandb#4318
- refactor(sdk): clean up the init and run logic by @kptkin in wandb#4730
- @edwag made their first contribution in wandb#4585
- @TimH98 made their first contribution in wandb#4625
- @froody made their first contribution in wandb#4732
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.7...v0.13.8
- revert(artifacts): revert
Circular reference detected
change to resolveObject of type Tensor is not JSON serializable
by @raubitsj in wandb#4629
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.6...v0.13.7
- feat(sweeps): add
Sweep.expected_run_count
to public Api by @gtarpenning in wandb#4434 - feat(launch): support volume mounts and security contexts in kubernetes runner by @KyleGoyette in wandb#4475
- feat(launch): add a new
--build
flag for building and then pushing the image to a queue by @gtarpenning in wandb#4061 - feat(integrations): add ability to log learning rate using WandbMetricsLogger by @soumik12345 in wandb#4391
- feat(sdk): improve Report API in preparation for GA by @andrewtruong in wandb#4499
- fix(artifacts): add filter for
artifact_version
to only retrieve committed artifacts by @estellazx in wandb#4401 - fix(cli): deflake
wandb verify
by @vanpelt in wandb#4438 - fix(launch): fix the type of the override args passed through to a LaunchProject from a Job by @KyleGoyette in wandb#4416
- fix(launch): remove extra colon from log prefix by @jamie-rasmussen in wandb#4450
- fix(sdk): add support for service running in a pex based environment by @kptkin in wandb#4440
- fix(sdk): fix probing static IPU info by @dmitryduev in wandb#4464
- fix(public-api): change
artifactSequence
toartifactCollection
in public GQL requests by @tssweeney in wandb#4531 - fix(integrations): fix TF compatibility issues with
WandbModelCheckpoint
by @soumik12345 in wandb#4432 - fix(integrations): make Keras WandbCallback compatible with TF version >= 2.11.0 by @ayulockin in wandb#4533
- fix(integrations): update gym integration to match last version by @younik in wandb#4571
- fix(sdk): harden internal thread management in SystemMetrics by @dmitryduev in wandb#4439
- docs(sdk): remove non-existent argument
table_key
fromplot_table()
doc string by @janosh in wandb#4495 - docs(artifacts): correct parameter name in docstring example by @ngrayluna in wandb#4528
- chore(launch): improved git fetch time by specifying a
refspec
anddepth=1
by @gtarpenning in wandb#4459 - chore(sdk): fix linguist rule to ignore grpc generated files by @raubitsj in wandb#4470
- chore(launch): new shard for launch tests by @gtarpenning in wandb#4427
- chore(public-api): upgrade Node 12 based GitHub Actions by @moredatarequired in wandb#4506
- test(artifacts): skip flaky
artifact_metadata_save
test by @speezepearson in wandb#4463 - test(artifacts): replace sleeps with flush when waiting on a file to write by @moredatarequired in wandb#4523
- test(artifacts): use
tmp_path
fixture instead of writing local files during tests by @moredatarequired in wandb#4521 - chore(launch): fix broken queue test by @gtarpenning in wandb#4548
- test(artifacts):
skip
instead ofxfail
for testtest_artifact_metadata_save
by @speezepearson in wandb#4550 - test(sdk): add many tests for InternalApi.upload_file by @speezepearson in wandb#4539
- chore(artifacts): add artifact Sequence fallback for older servers by @tssweeney in wandb#4565
- test(sdk): make protobuf version requirements more granular by @dmitryduev in wandb#4479
- fix(artifacts): when committing artifacts, don't retry 409 Conflict errors by @speezepearson in wandb#4260
- refactor(artifacts): add programmatic alias addition/removal from SDK on artifacts by @vwrj in wandb#4429
- fix(integrations): remove
wandb.sklearn.plot_decision_boundaries
that contains dead logic by @kptkin in wandb#4348 - chore(sdk): adds an option to force pull the latest version of a test dev-container image by @kptkin in wandb#4352
- feat(launch): noop builder by @KyleGoyette in wandb#4275
- refactor(launch): remove unused attribute by @jamie-rasmussen in wandb#4497
- style(sdk): update
mypy
to 0.991 by @dmitryduev in wandb#4546 - refactor(launch): add more robust uri parsing by @jamie-rasmussen in wandb#4498
- style(sdk): turn on linting for internal_api.py by @speezepearson in wandb#4545
- build(sdk): remove dependency on six by modifying vendored libs by @dmitryduev in wandb#4280
- @moredatarequired made their first contribution in wandb#4508
- @soumik12345 made their first contribution in wandb#4391
- @younik made their first contribution in wandb#4571
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.5...v0.13.6
- feat(artifacts): add an option to upload image references by @estellazx in wandb#4303
- fix(launch): generate more readable image names by @jamie-rasmussen in wandb#4379
- fix(artifacts): use hash(
etag
+url
) instead of justetag
, as key, in artifacts cache by @speezepearson in wandb#4371 - fix(artifacts): wait for artifact to commit before telling the user it's ready when using
wandb artifact put
by @speezepearson in wandb#4381 - fix(sdk): prefix vendor watchdog library by @raubitsj in wandb#4389
- fix(artifacts): fix
Circular reference detected
error, when updating metadata with numpy array longer than 32 elements by @estellazx in wandb#4221 - fix(integrations): add a random string to run_id on SageMaker not to break DDP mode by @dmitryduev in wandb#4276
- ci(sdk): make sure we dont shutdown test cluster before grabbing results by @raubitsj in wandb#4361
- test(artifacts): add standalone artifact test to nightly cpu suite by @raubitsj in wandb#4360
- chore(sdk): rename default branch to
main
by @raubitsj in wandb#4374 - build(sdk): update mypy extension for protobuf type checking by @dmitryduev in wandb#4392
- chore(sdk): update codeql-analysis.yml branch name by @zythosec in wandb#4393
- ci(sdk): move functional import tests to nightly and expand python version coverage by @dmitryduev in wandb#4395
- ci(sdk): add Slack notification for failed nightly import tests by @dmitryduev in wandb#4403
- test(cli): fix broken CLI tests that attempt uploading non-existent artifacts by @dmitryduev in wandb#4426
- fix(launch): job creation through use_artifact instead of log_artifact by @KyleGoyette in wandb#4337
- ci(sdk): add a GH action to automate parts of the release process by @dmitryduev in wandb#4355
- fix(media): 3D Point Clouds now viewable in UI in all situations by @ssisk in wandb#4353
- fix(launch): Git URLs were failing if fsmonitor is enabled by @jamie-rasmussen in wandb#4333
- style(sdk): ignore new proto generated file directories by @raubitsj in wandb#4354
- chore(launch): fix a bug preventing Run Queue deletion in the SDK by @gtarpenning in wandb#4321
- chore(launch): add support for
pushToRunQueueByName
mutation by @gtarpenning in wandb#4292 - refactor(sdk): refactor system metrics monitoring and probing by @dmitryduev in wandb#4213
- style(sdk): fix gitattribute for protobuf generated files by @raubitsj in wandb#4400
- @ssisk made their first contribution in wandb#4353
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.4...v0.13.5
- feat(launch): show entity and project in k8s job names by @KyleGoyette in wandb#4216
- feat(sweeps): add environment variable sweep command macro by @hu-po in wandb#4200
- feat(media): add
from_*
constructors and scene camera and bounding box confidence scores toObject3D
data type by @dmitryduev in wandb#4319 - feat(artifacts): add simple progress indicator for artifact downloads by @speezepearson in wandb#4255
- feat(integrations): add
WandbMetricsLogger
callback - aKeras
dedicated metrics logger callback by @ayulockin in wandb#4244 - feat(integrations): add
WandbModelCheckpoint
callback - aKeras
model checkpointing callback by @ayulockin in wandb#4245 - feat(integrations): add
WandbEvalCallback
callback - aKeras
callback for logging model predictions as W&B tables by @ayulockin in wandb#4302
- fix(launch): cast agent's config max_jobs attribute to integer by @KyleGoyette in wandb#4262
- fix(cli): correct the displayed path to the
debug-cli.log
(debug log) by @jamie-rasmussen in wandb#4271 - fix(artifacts): catch retry-able request timeout when uploading artifacts to AWS by @nickpenaranda in wandb#4304
- fix(sdk): improve user feedback for long running calls: summary, finish by @raubitsj in wandb#4169
- fix(integrations): fix RuntimeError when using
keras.WandbCallback
withtf.MirroredStrategy
by @ayulockin in wandb#4310
- ci(sdk): add code analysis/scanning with
codeql
by @dmitryduev in wandb#4250 - ci(sdk): validate PR titles to ensure compliance with Conventional Commits guidelines by @dmitryduev in wandb#4268
- chore(launch): harden launch by pining the build versions of
kaniko
andlaunch-agent-dev
by @KyleGoyette in wandb#4194 - test(sdk): add telemetry for the
mmengine
package by @manangoel99 in wandb#4273 - chore(sdk): add the
build
type to our conventional commits setup by @dmitryduev in wandb#4282 - test(sdk): add
tensorflow_datasets
requirement toimports12
shard by @dmitryduev in wandb#4316 - test(integrations): fix sb3 test by pinning upstream requirement by @dmitryduev in wandb#4346
- build(sdk): make the SDK compatible with protobuf v4 by @dmitryduev in wandb#4279
- chore(sdk): fix flake8 output coloring by @dmitryduev in wandb#4347
- test(artifacts): fix artifact reference test asset directory by @raubitsj in wandb#4350
- style(sdk): fix type hint for
filters
argument inpublic_api.runs
by @epwalsh in wandb#4256 - style(artifacts): improve type annotations around artifact-file-creation by @speezepearson in wandb#4259
- style(sdk): improve type annotations and VSCode config for public API by @speezepearson in wandb#4252
- style(sdk): make type annotations more easily navigable in VSCode by @speezepearson in wandb#4005
- style(artifacts): introduce str NewTypes and use them for various Artifact fields by @speezepearson in wandb#4326
- style(artifacts): add type annotations to get better IDE hints for boto3 usage by @speezepearson in wandb#4338
- @epwalsh made their first contribution in wandb#4256
- @mjvanderboon made their first contribution in wandb#4309
- @jamie-rasmussen made their first contribution in wandb#4271
- @nickpenaranda made their first contribution in wandb#4304
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.3...v0.13.4
- Adds
raytune
examples / tests by @raubitsj in wandb#4053 - Refactors
pytest
unit tests to run against realwandb server
by @kptkin in wandb#4066 - Adds Launch
kubernetes
support of taints and tolerations by @KyleGoyette in wandb#4086 - Adds Sweeps on Launch on Kubernetes by @hu-po in wandb#4035
- Adds parallelism to functional testing by @raubitsj in wandb#4096
- Upgrades
mypy
to version0.971
by @dmitryduev in wandb#3952 - Adds Mailbox async internal process communication by @raubitsj in wandb#3568
- Implements searching launch job in sweep config by @hu-po in wandb#4120
- Improves performance when sending large messages by @raubitsj in wandb#4119
- Vendors the latest
nvidia-ml-py-11.515.48
by @dmitryduev in wandb#4109 - Improves performance by increase recv size on service socket by @raubitsj in wandb#4122
- Adds isort support with black profile by @kptkin in wandb#4136
- Implements pushing test-results to CircleCI for nightly tests by @raubitsj in wandb#4153
- Adds debug mode for
pytest
unit tests by @dmitryduev in wandb#4145 - Adds support for arguments in Launch Jobs by @KyleGoyette in wandb#4129
- Adds FetchRunQueueItemById query by @gtarpenning in wandb#4106
- Adds telemetry for keras-cv by @manangoel99 in wandb#4196
- Adds sentry session tracking by @raubitsj in wandb#4157
- Adds the ability to log artifact while linking to registered model by @ibindlish in wandb#4233
- Breaks gradient and parameters hooks by @kptkin in wandb#3509
- Adds explicit error message for double uri/docker-image by @gtarpenning in wandb#4069
- Tests that the wandb_init fixture args are in sync with wandb.init() by @dmitryduev in wandb#4079
- Upgrades the GKE cluster used for nightly tests to
n1-standard-8
by @dmitryduev in wandb#4065 - Moves service teardown to the end of tests by @kptkin in wandb#4083
- Reduce the
pytest
job parallelism from 10 to 6 by @kptkin in wandb#4085 - Removes service user doc by @kptkin in wandb#4088
- Move
_timestamp
logic to the internal process by @kptkin in wandb#4087 - Adds Launch
gitversion
error message by @gtarpenning in wandb#4028 - Updates KFP machine VM image in CircleCI by @dmitryduev in wandb#4094
- Upgrades sweeps to latest version by @hu-po in wandb#4104
- Implements Sweep scheduler cleanup and better tests by @hu-po in wandb#4100
- Adds a requirement for the sdk-team to approve API changes by @raubitsj in wandb#4128
- Adds additional time for artifact commit by @raubitsj in wandb#4133
- Implements tox configuration with dynamic resolution by @kptkin in wandb#4138
- Removes
buildx
version pin for nightly builds by @dmitryduev in wandb#4144 - Moves Launch run configs from entrypoint into params by @hu-po in wandb#4164
- Removes Slack orb usage from Win job on CircleCI by @dmitryduev in wandb#4171
- Adds heartbeat parsing for Launch run args using legacy agent by @hu-po in wandb#4180
- Add better error handling when tearing down service by @kptkin in wandb#4161
- Cleans up Launch job creation pipeline by @KyleGoyette in wandb#4183
- Adds detail to error message when uploading an artifact with the wrong type by @speezepearson in wandb#4184
- Adds optional timeout parameter to artifacts wait() by @estellazx in wandb#4181
- Sanitizes numpy generics in keys by @raubitsj in wandb#4146
- Removes reassignment of run function in public api by @martinabeleda in wandb#4115
- Makes pulling sweeps optional when using public api to query for runs by @kptkin in wandb#4186
- Updates ref docs for
wandb.init
to give more info on special characters by @scottire in wandb#4191
- Fixes Sweeps on Launch Jobs requirement by @hu-po in wandb#3947
- Fixes Artifact metadata JSON-encoding to accept more types by @speezepearson in wandb#4038
- Adjusts
root_dir
setting processing logic by @dmitryduev in wandb#4049 - Prevents run.log() from mutating passed in arguments by @kptkin in wandb#4058
- Fixes
05-batch5.py
test by @dmitryduev in wandb#4074 - Allows users to control the
run_id
through the launch spec by @gtarpenning in wandb#4070 - Fixes accidental overwrite in
config.yml
by @dmitryduev in wandb#4081 - Ensures propagating overridden
base_url
when initializing public API by @dmitryduev in wandb#4026 - Fixes Sweeps on Launch CLI launch config, relpath by @hu-po in wandb#4073
- Fixes broken Launch apikey error message by @gtarpenning in wandb#4071
- Marks flakey sweeps test xfail by @hu-po in wandb#4095
- Fixes Launch
gitversion
error message by @gtarpenning in wandb#4103 - Fixes
yea-wandb
dev release -> release by @raubitsj in wandb#4098 - Cleans up outstanding issues after the client->wandb rename by @kptkin in wandb#4105
- Fixes test precision recall by @kptkin in wandb#4108
- Fixes functional sklearn test by @raubitsj in wandb#4107
- Fixes hang caused by keyboard interrupt on windows by @kptkin in wandb#4116
- Fixes default test container tag by @kptkin in wandb#4137
- Fixes summary handling in conftest.py by @dmitryduev in wandb#4140
- Fixes some small typos in cli output by @lukas in wandb#4126
- Fixes issue triggered by colab update by using default file and catching exceptions by @raubitsj in wandb#4156
- Fixes mailbox locking issue by @raubitsj in wandb#4214
- Fixes variable inclusion in log string by @klieret in wandb#4219
- Corrects
wandb.Artifacts.artifact.version
attribute by @ngrayluna in wandb#4199 - Fixes piping of docker args by Launch Agent by @KyleGoyette in wandb#4215
- Fixes RecursionError when printing public API User object without email fetched by @speezepearson in wandb#4193
- Fixes deserialization of numeric column names by @tssweeney in wandb#4241
- @gtarpenning made their first contribution in wandb#4069
- @estellazx made their first contribution in wandb#4181
- @klieret made their first contribution in wandb#4219
- @ngrayluna made their first contribution in wandb#4199
- @martinabeleda made their first contribution in wandb#4115
- @ibindlish made their first contribution in wandb#4233
- @scottire made their first contribution in wandb#4191
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.2...v0.13.3
- Fix issue triggered by colab update by using default file and catching exceptions by @raubitsj in wandb#4156
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.1...v0.13.2
- Prevents run.log() from mutating passed in arguments by @kptkin in wandb#4058
Full Changelog: https://github.com/wandb/wandb/compare/v0.13.0...v0.13.1
- Turns service on by default by @kptkin in wandb#3895
- Adds support logic for handling server provided messages by @kptkin in wandb#3706
- Allows runs to produce jobs on finish by @KyleGoyette in wandb#3810
- Adds Job, QueuedRun and job handling in launch by @KyleGoyette in wandb#3809
- Supports in launch agent of instance roles in ec2 and eks by @KyleGoyette in wandb#3596
- Adds default behavior to the Keras Callback: always save model checkpoints as artifacts by @vwrj in wandb#3909
- Sanitizes the artifact name in the KerasCallback for model artifact saving by @vwrj in wandb#3927
- Improves console logging by moving emulator to the service process by @raubitsj in wandb#3828
- Fixes data corruption issue when logging large sizes of data by @kptkin in wandb#3920
- Adds the state to the Sweep repr in the Public API by @hu-po in wandb#3948
- Adds an option to specify different root dir for git using settings or environment variables by @bcsherma in wandb#3250
- Adds an option to pass
remote url
andcommit hash
as arguments to settings or as environment variables by @kptkin in wandb#3934 - Improves time resolution for tracked metrics and for system metrics by @raubitsj in wandb#3918
- Defaults to project name from the sweep config when project is not specified in the
wandb.sweep()
call by @hu-po in wandb#3919 - Adds support to use namespace set user by the the launch agent by @KyleGoyette in wandb#3950
- Adds telemetry to track when a run might be overwritten by @raubitsj in wandb#3998
- Adds a tool to export
wandb
's history intosqlite
by @raubitsj in wandb#3999 - Replaces some
Mapping[str, ...]
types withNamedTuples
by @speezepearson in wandb#3996 - Adds import hook for run telemetry by @kptkin in wandb#3988
- Implements profiling support for IPUs by @cameron-martin in wandb#3897
- Fixes sweep agent with service by @raubitsj in wandb#3899
- Fixes an empty type equals invalid type and how artifact dictionaries are handled by @KyleGoyette in wandb#3904
- Fixes
wandb.Config
object to support default values when getting an attribute by @farizrahman4u in wandb#3820 - Removes default config from jobs by @KyleGoyette in wandb#3973
- Fixes an issue where patch is
None
by @KyleGoyette in wandb#4003 - Fixes requirements.txt parsing in nightly SDK installation checks by @dmitryduev in wandb#4012
- Fixes 409 Conflict handling when GraphQL requests timeout by @raubitsj in wandb#4000
- Fixes service teardown handling if user process has been terminated by @raubitsj in wandb#4024
- Adds
storage_path
and fixedartifact.files
by @vanpelt in wandb#3969 - Fixes performance issue syncing runs with a large number of media files by @vanpelt in wandb#3941
- Adds an escape hatch logic to disable service by @kptkin in wandb#3829
- Annotates
wandb/docker
and reverts change in the docker fixture by @dmitryduev in wandb#3871 - Fixes GFLOPS to GFLOPs in the Keras
WandbCallback
by @ayulockin in wandb#3913 - Adds type-annotate for
file_stream.py
by @dmitryduev in wandb#3907 - Renames repository from
client
towandb
by @dmitryduev in wandb#3977 - Updates documentation: adding
--report_to wandb
for HuggingFace Trainer by @ayulockin in wandb#3959 - Makes aliases optional in link_artifact by @vwrj in wandb#3986
- Renames
wandb local
towandb server
by @jsbroks in wandb#3793 - Updates README badges by @raubitsj in wandb#4023
- @bcsherma made their first contribution in wandb#3250
- @cameron-martin made their first contribution in wandb#3897
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.21...v0.13.0
- Fixes config not showing up until the run finish by @KyleGoyette in wandb#3734
- Adds new types to the TypeRegistry to handling artifact objects in jobs and run configs by @KyleGoyette in wandb#3806
- Adds new query to the the internal api getting the state of the run by @hu-po in wandb#3799
- Replaces unsafe yaml loaders with yaml.safe_load by @zythosec in wandb#3753
- Improves testing tooling by allowing to specify shards in manual testing by @dmitryduev in wandb#3826
- Fixes ROC and PR curves in the sklearn integration by stratifying sampling by @tylerganter in wandb#3757
- Fixes input box in notebooks exceeding cell space by @dmitryduev in wandb#3849
- Allows string to be passed as alias to link_model by @tssweeney in wandb#3834
- Adds Support for FLOPS Calculation in
keras
'sWandbCallback
by @dmitryduev in wandb#3869 - Extends python report editing by @andrewtruong in wandb#3732
- Fixes stats logger so it can find all the correct GPUs in child processes by @raubitsj in wandb#3727
- Fixes regression in s3 reference upload for folders by @jlzhao27 in wandb#3825
- Fixes artifact commit logic to handle collision in the backend by @speezepearson in wandb#3843
- Checks for
None
response in the retry logic (safety check) by @raubitsj in wandb#3863 - Adds sweeps on top of launch (currently in MVP) by @hu-po in wandb#3669
- Renames functional tests dir and files by @raubitsj in wandb#3879
- Fixes conditions order of
_to_dict
helper by @dmitryduev in wandb#3772 - Fixes changelog broken link to PR 3709 by @janosh in wandb#3786
- Fixes public api query (QueuedJob Api ) by @KyleGoyette in wandb#3798
- Renames local runners to local-container and local-process by @hu-po in wandb#3800
- Adds type annotations to files in the wandb/filesync directory by @speezepearson in wandb#3774
- Re-organizes all the testing directories to have common root dir by @dmitryduev in wandb#3740
- Fixes testing configuration and add bigger machine on
CircleCi
by @dmitryduev in wandb#3836 - Fixes typo in the
wandb-service-user
readme file by @Co1lin in wandb#3847 - Fixes broken artifact test for regression by @dmitryduev in wandb#3857
- Removes unused files (relating to
py27
) and emptysubmodules
declaration by @dmitryduev in wandb#3850 - Adds extra for model reg dependency on cloudpickle by @tssweeney in wandb#3866
- Replaces deprecated threading aliases by @hugovk in wandb#3794
- Updates the
sdk
readme to the renamed (local -> server) commands by @sephmard in wandb#3771
- @janosh made their first contribution in wandb#3786
- @Co1lin made their first contribution in wandb#3847
- @tylerganter made their first contribution in wandb#3757
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.20...v0.12.21
- Retry
commit_artifact
on conflict-error by @speezepearson in wandb#3843
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.19...v0.12.20
- Fix regression in s3 reference upload for folders by @jlzhao27 in wandb#3825
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.18...v0.12.19
- Launch: BareRunner based on LocalRunner by @hu-po in wandb#3577
- Add ability to specify api key to public api by @dannygoldstein in wandb#3657
- Add support in artifacts for files with unicode on windows by @kptkin in wandb#3650
- Added telemetry for new packages by @manangoel99 in wandb#3713
- Improve API key management by @vanpelt in wandb#3718
- Add information about
wandb server
during login by @raubitsj in wandb#3754
- fix(weave): Natively support timestamps in Python Table Types by @dannygoldstein in wandb#3606
- Add support for magic with service by @kptkin in wandb#3623
- Add unit tests for DirWatcher and supporting classes by @speezepearson in wandb#3589
- Improve
DirWatcher.update_policy
O(1) instead of O(num files uploaded) by @speezepearson in wandb#3613 - Add argument to control what to log in SB3 callback by @astariul in wandb#3643
- Improve parameter naming in sb3 integration by @dmitryduev in wandb#3647
- Adjust the requirements for the dev environment setup on an M1 Mac by @dmitryduev in wandb#3627
- Launch: Fix NVIDIA base image Linux keys by @KyleGoyette in wandb#3637
- Fix launch run queue handling from config file by @KyleGoyette in wandb#3636
- Fix issue where tfevents were not always consumed by @minyoung in wandb#3673
- [Snyk] Fix for 8 vulnerabilities by @snyk-bot in wandb#3695
- Fix s3 storage handler to upload folders when key names collide by @jlzhao27 in wandb#3699
- Correctly load timestamps from tables in artifacts by @dannygoldstein in wandb#3691
- Require
protobuf<4
by @dmitryduev in wandb#3709 - Make Containers created through launch re-runnable as container jobs by @KyleGoyette in wandb#3642
- Fix tensorboard integration skipping steps at finish() by @KyleGoyette in wandb#3626
- Rename
wandb local
towandb server
by @jsbroks in wandb#3716 - Fix busted docker inspect command by @vanpelt in wandb#3742
- Add dedicated sentry wandb by @dmitryduev in wandb#3724
- Image Type should gracefully handle older type params by @tssweeney in wandb#3731
- Inline FileEventHandler.synced into the only method where it's used by @speezepearson in wandb#3594
- Use passed size argument to make
PolicyLive.min_wait_for_size
a classmethod by @speezepearson in wandb#3593 - Make FileEventHandler an ABC, remove some "default" method impls which were only used once by @speezepearson in wandb#3595
- Remove unused field from DirWatcher by @speezepearson in wandb#3592
- Make sweeps an extra instead of vendoring by @dmitryduev in wandb#3628
- Add nightly CI testing by @dmitryduev in wandb#3580
- Improve keras and data type Reference Docs by @ramit-wandb in wandb#3676
- Update
pytorch
version requirements in dev environments by @dmitryduev in wandb#3683 - Clean up CircleCI config by @dmitryduev in wandb#3722
- Add
py310
testing in CI by @dmitryduev in wandb#3730 - Ditch
dateutil
from the requirements by @dmitryduev in wandb#3738 - Add deprecated string to
Table.add_row
by @nate-wandb in wandb#3739
- @sephmard made their first contribution in wandb#3610
- @astariul made their first contribution in wandb#3643
- @manangoel99 made their first contribution in wandb#3713
- @nate-wandb made their first contribution in wandb#3739
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.17...v0.12.18
- Update requirements to fix incompatibility with protobuf >= 4 by @dmitryduev in wandb#3709
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.16...v0.12.17
- Improve W&B footer by aligning summary/history in notebook env by @kptkin in wandb#3479
- Enable experimental history step logging in artifacts by @raubitsj in wandb#3502
- Add
args_no_boolean_flags
macro to sweep configuration by @hu-po in wandb#3489 - Add logging support for
jax.bfloat.bfloat16
by @dmitryduev in wandb#3528 - Raise exception when Table size exceeds limit by @dannygoldstein in wandb#3511
- Add kaniko k8s builder for wandb launch by @KyleGoyette in wandb#3492
- Add wandb.init() timeout setting by @kptkin in wandb#3579
- Do not assume executable for given entrypoints with wandb launch by @KyleGoyette in wandb#3461
- Jupyter environments no longer collect command arguments by @KyleGoyette in wandb#3456
- Add support for TensorFlow/Keras SavedModel format by @ayulockin in wandb#3276
- Support version IDs in artifact refs, fix s3/gcs references in Windows by @annirudh in wandb#3529
- Fix support for multiple finish for single run using wandb-service by @kptkin in wandb#3560
- Fix duplicate backtrace when using wandb-service by @kptkin in wandb#3575
- Fix wrong entity displayed in login message by @kptkin in wandb#3490
- Fix hang when
wandb.init
is interrupted mid setup using wandb-service by @kptkin in wandb#3569 - Fix handling keyboard interrupt to avoid hangs with wandb-service enabled by @kptkin in wandb#3566
- Fix console logging with very long print out when using wandb-service by @kptkin in wandb#3574
- Fix broken artifact string in launch init config by @KyleGoyette in wandb#3582
- Fix typo in wandb.log() docstring by @RobRomijnders in wandb#3520
- Cleanup custom chart code and add type annotations to plot functions by @kptkin in wandb#3407
- Improve
wandb.init(settings=)
to handleSettings
object similarly todict
parameter by @dmitryduev in wandb#3510 - Add documentation note about api.viewer in api.user() and api.users() by @ramit-wandb in wandb#3552
- Be explicit about us being py3+ only in setup.py by @dmitryduev in wandb#3549
- Add type annotations to DirWatcher by @speezepearson in wandb#3557
- Improve wandb.log() docstring to use the correct argument name by @idaho777 in wandb#3585
- @RobRomijnders made their first contribution in wandb#3520
- @ramit-wandb made their first contribution in wandb#3552
- @idaho777 made their first contribution in wandb#3585
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.15...v0.12.16
- Optimize wandb.Image logging when linked to an artifact by @tssweeney in wandb#3418
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.14...v0.12.15
- Fix regression: disable saving history step in artifacts by @vwrj in wandb#3495
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.13...v0.12.14
- Revert strictened api_key validation by @dmitryduev in wandb#3485
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.12...v0.12.13
- Allow run objects to be passed to other processes when using wandb-service by @kptkin in wandb#3308
- Add create user to public api by @vanpelt in wandb#3438
- Support logging from multiple processes with wandb-service by @kptkin in wandb#3285
- Add gpus flag for local launch runner with cuda by @KyleGoyette in wandb#3417
- Improve Launch deployable agent by @KyleGoyette in wandb#3388
- Add Launch kubernetes integration by @KyleGoyette in wandb#3393
- KFP: Add wandb visualization helper by @andrewtruong in wandb#3439
- KFP: Link back to Kubeflow UI by @andrewtruong in wandb#3427
- Add boolean flag arg macro by @hugo.ponte in wandb#3489
- Improve host / WANDB_BASE_URL validation by @dmitryduev in wandb#3314
- Fix/insecure tempfile by @dmitryduev in wandb#3360
- Fix excess warning span if requested WANDB_DIR/root_dir is not writable by @dmitryduev in wandb#3304
- Fix line_series to plot array of strings by @kptkin in wandb#3385
- Properly handle command line args with service by @kptkin in wandb#3371
- Improve api_key validation by @dmitryduev in wandb#3384
- Fix multiple performance issues caused by not using defaultdict by @dmitryduev in wandb#3406
- Enable inf max jobs on launch agent by @stephchen in wandb#3412
- fix colab command to work with launch by @stephchen in wandb#3422
- fix typo in Config docstring by @hu-po in wandb#3416
- Make code saving not a policy, keep previous custom logic by @dmitryduev in wandb#3395
- Fix logging sequence images with service by @kptkin in wandb#3339
- Add username to debug-cli log file to prevent conflicts of multiple users by @zythosec in wandb#3301
- Fix python sweep agent for users of wandb service / pytorch-lightning by @raubitsj in wandb#3465
- Remove unnecessary launch reqs checks by @KyleGoyette in wandb#3457
- Workaround for MoviePy's Unclosed Writer by @tssweeney in wandb#3471
- Improve handling of Run objects when service is not enabled by @kptkin in wandb#3362
- @hu-po made their first contribution in wandb#3416
- @zythosec made their first contribution in wandb#3301
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.11...v0.12.12
- Add captions to Molecules by @dmitryduev in wandb#3173
- Add CatBoost Integration by @ayulockin in wandb#2975
- Launch: AWS Sagemaker integration by @KyleGoyette in wandb#3007
- Launch: Remove repo2docker and add gpu support by @stephchen in wandb#3161
- Adds Timestamp inference from Python for Weave by @tssweeney in wandb#3212
- Launch GCP vertex integration by @stephchen in wandb#3040
- Use Artifacts when put into run config. Accept a string to represent an artifact in the run config by @KyleGoyette in wandb#3203
- Improve xgboost
wandb_callback
(#2929) by @ayulockin in wandb#3025 - Add initial kubeflow pipeline support by @andrewtruong in wandb#3206
- Fix logging of images with special characters in the key by @speezepearson in wandb#3187
- Fix azure blob upload retry logic by @vanpelt in wandb#3218
- Fix program field for scripts run as a python module by @dmitryduev in wandb#3228
- Fix issue where
sync_tensorboard
could die on large histograms by @KyleGoyette in wandb#3019 - Fix wandb service performance issue during run shutdown by @raubitsj in wandb#3262
- Fix vendoring of gql and graphql by @raubitsj in wandb#3266
- Flush log data without finish with service by @kptkin in wandb#3137
- Fix wandb service hang when the service crashes by @raubitsj in wandb#3280
- Fix issue logging images with "/" on Windows by @KyleGoyette in wandb#3146
- Add image filenames to images/separated media by @KyleGoyette in wandb#3041
- Add setproctitle to requirements.txt by @raubitsj in wandb#3289
- Fix issue where sagemaker run ids break run queues by @KyleGoyette in wandb#3290
- Fix encoding exception when using %%capture magic by @raubitsj in wandb#3310
- @speezepearson made their first contribution in wandb#3188
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.10...v0.12.11
- Improve validation when creating Tables with invalid columns from dataframes by @tssweeney in wandb#3113
- Enable digest deduplication for
use_artifact()
calls by @annirudh in wandb#3109 - Initial prototype of azure blob upload support by @vanpelt in wandb#3089
- Fix wandb launch using python dev versions by @stephchen in wandb#3036
- Fix loading table saved with mixed types by @vwrj in wandb#3120
- Fix ResourceWarning when calling wandb.log by @vwrj in wandb#3130
- Fix missing cursor in ProjectArtifactCollections by @KyleGoyette in wandb#3108
- Fix windows table logging classes issue by @vwrj in wandb#3145
- Gracefully handle string labels in wandb.sklearn.plot.classifier.calibration_curve by @acrellin in wandb#3159
- Do not display login warning when calling wandb.sweep() by @acrellin in wandb#3162
- Drop python2 backport deps (enum34, subprocess32, configparser) by @jbylund in wandb#3004
- Settings refactor by @dmitryduev in wandb#3083
- @jbylund made their first contribution in wandb#3004
- @acrellin made their first contribution in wandb#3159
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.9...v0.12.10
- Fix regression in
upload_file()
exception handler by @raubitsj in wandb#3059
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.8...v0.12.9
- Update contributing guide and dev env setup tool by @dmitryduev in wandb#2968
- Improve
wandb_callback
for LightGBM (#2945) by @ayulockin in wandb#3024
- Reduce GPU memory usage when generating histogram of model weights by @TOsborn in wandb#2927
- Support mixed classes in bounding box and image mask annotation layers by @tssweeney in wandb#2914
- Add max-jobs and launch async args by @stephchen in wandb#2925
- Support lists of Summary objects encoded as strings to wandb.tensorboard.log by @dmitryduev in wandb#2934
- Fix handling of 0 dim np arrays by @rpitonak in wandb#2954
- Fix handling of empty default config file by @vwrj in wandb#2957
- Add service backend using sockets (support fork) by @raubitsj in wandb#2892
- Send git port along with url when sending git repo by @KyleGoyette in wandb#2959
- Add support raw ip addresses for launch by @KyleGoyette in wandb#2950
- Tables no longer serialize and hide 1d NDArrays by @tssweeney in wandb#2976
- Fix artifact file uploads to S3 stores by @annirudh in wandb#2999
- Send uploaded file list on file stream heartbeats by @annirudh in wandb#2978
- Add support for keras experimental layers by @KyleGoyette in wandb#2776
- Fix
from wandb import magic
to not require tensorflow by @raubitsj in wandb#3021 - Fix launch permission error by @KyleGoyette in wandb#3038
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.7...v0.12.8
- Fix issue where console log streaming was causing excessive network traffic by @vwrj in wandb#2786
- Metaflow: Make optional dependencies actually optional by @andrewtruong in wandb#2842
- Fix docstrings for wandb.watch and ValidationDataLogger by @charlesfrye in wandb#2849
- Prevent launch agent from sending runs to a different project or entity by @KyleGoyette in wandb#2872
- Fix logging pr_curves through tensorboard by @KyleGoyette in wandb#2876
- Prevent TPU monitoring from reporting invalid metrics when not available by @kptkin in wandb#2753
- Make import order dependencies for WandbCallback more robust by @kptkin in wandb#2807
- Fix a bug in feature importance plotting to handle matrices of different shapes by @dannygoldstein in wandb#2811
- Fix base url handling to allow trailing / by @kptkin in wandb#2910
- Prevent wandb.agent() from sending too many heartbeats impacting rate limits by @dannygoldstein in wandb#2923
- Redact sensitive information from debug logs by @raubitsj in wandb#2931
- Add wandb.Molecule support for rdkit supported formats by @dmitryduev in wandb#2902
- Add module-level docstrings for reference doc modules. by @charlesfrye in wandb#2847
- Store launch metadata in file by @KyleGoyette in wandb#2582
- Add Project.sweeps() public API call to view all sweeps in a project by @stephchen in wandb#2729
- Ensures API key prompt remains captive when user enters nothing by @dannygoldstein in wandb#2721
- Refactors wandb.sklearn into submodules by @charlesfrye in wandb#2869
- Support code artifacts in wandb launch by @KyleGoyette in wandb#2860
- Improve launch agent (async, stop, heartbeat updates) by @stephchen in wandb#2871
- Improve usage and error messages for anonymous mode by @kimjyhello in wandb#2823
- Add example on how to find runs with wandb.Api().runs(...) matching a regex by @dmitryduev in wandb#2926
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.6...v0.12.7
- Fix sklearn
plot_calibration_curve()
issue breaking the provided model by @vwrj in wandb#2791 - Fix CondaEnvExportError by redirecting stderr by @charlesfrye in wandb#2814
- Fix
use_artifact()
when specifying an artifact from a different project by @KyleGoyette in wandb#2832
- Add metric names to pr curve charts in tensorboard by @vanpelt in wandb#2822
Full Changelog: https://github.com/wandb/wandb/compare/v0.12.5...v0.12.6
- Report errors for invalid characters in logged media keys on windows
- Handle errors when probing for TPUs in unsupported environments
- Fixed bug where
%%wandb
followed by wandb.init() does not display run links - Fixed api.runs() to correctly return all runs for the current entity/project
- Add
wandb.require(experiment="service")
to improve multiprocessing support - Add support for swappable artifacts in launch context
- Add
wandb.login(timeout=)
support for jupyter environments - Add ability to disable git ref saving with
WANDB_DISABLE_GIT
- Support newer versions of pytest-mock and PyYAML
- Add ability to delete artifacts with aliases:
artifact.delete(delete_aliases=True)
- Add
unwatch()
method to the Run object
- Fix regression introduced in 0.12.2 causing network access when
WANDB_MODE=offline
- Fixes the grid search stopping condition in the local controller
- New jupyter magic for displaying runs, sweeps, and projects
%wandb path/to/run -h 1024
- We no longer display run iframe by default in jupyter, add
%%wandb
to a cell to display a run - Makes api key prompting retry indefinitely on malformed input
- Invite users to teams via the api
api.team("team_name").invite("username_or_email")
- Remove users from a team via the api
api.team("team_name").members[0].delete()
- Create service accounts via the api
api.team("team_name").create_service_account("Description")
- Manage api keys via the api
api.user("username_or_email").generate_api_key()
- Add pytorch profiling trace support with
wandb.profiler.torch_trace_handler()
- Fix tensorboard_sync to handle ephemeral Sagemaker tfevents files
- Fix Reports query from the public api (broken pagination and report path)
- Fix
wandb.login()
when relogin is specified (only force login once)
- Clean up footer output of summary and history metrics
- Clean up error message from
wandb sweep --update
- Add warning for
wandb local
users to update their docker - Add optional argument log_learning_curve to wandb.sklearn.plot_classifier()
- Restore frozen pip package versions when using
wandb launch
- Add support for jupyter notebooks in launch
- Add
wandb.login()
timeout option
- Fix tensorflow/keras 2.6 not logging validation examples
- Fix metrics logged through tensorboard not supporting time on x-axis
- Fix
WANDB_IGNORE_GLOBS
environment variable handling - Fix handling when sys.stdout is configured to a custom logger
- Fix sklearn feature importance plots not matching feature names properly
- Fix an issue where colab urls were not being captured
- Save program commandline if run executable was outside cwd
- Add Prodigy integration to upload annotated datasets to W&B Tables
- Add initial Metaflow support
- Add experimental wandb launch support
- Add warnings that public API requests are timing out and allow override
- Improve error handling in local controller sweeps engine
- Remove Python 3.5 support
- Fix issue that could cause artifact uploads to fail if artifact files are being modified
- Fix issue where
wandb.restore()
wouldn't work with runs from a sweep
- Improve run execution time calculation
- Restore vendored graphql-core library because of network regression
- Python 3.5 will not be supported as of
wandb==0.12.0
- Reduce Memory Footprint of Images In Tables
- Added a dependency on graphql-core>=2.3.0
- Removed urllib3 pin to avoid conflicts, if you see urllib3 related errors run
pip install --upgrade urllib3
- Improved Public API HTTP error messages
- Set run.dir to the generated directory name in disabled mode
- Adds support for native Jax array logging
- Tables now support Molecule data type
- Improve Stable-Baselines3 API by auto log model's name and always upload models at the end of training
- Implements the sweep local controller using wandb/sweeps
- Remove Python 2.7 support
- Fix issue where
wandb.watch()
broke model saving in pytorch - Fix issue where uniform sweep parameters were parsed as int_uniform
- Fix issue where file_stream thread was killed on 4xx errors
- Improve performance of artifact logging by making it non-blocking
- Add wandb integration for Stable-Baselines3
- Improve keras callback validation logging inference logic
- Expose sweep state via the public API
- Improve performance of sweep run fetches via the API
- Fix issue where wandb restore 404ed if the run did not have a diff.patch file
- Fix issue where wandb.log raised an Exception after trying to log a pandas dataframe
- Fix issue where runs could be marked finished before files were finished uploading
- Disable reloading of run metadata (such as command) in resumed runs
- Allow logging of pandas dataframes by automatically converting them to W&B tables
- Fix up
log_code()
exclude fn to handle .wandb dir - Improve handling of PyTorch model topology
- Increase config debounce interval to 30s to reduce load on WB/backend
- Improve reliability of CLI in generating sweeps with names, programs, and settings
- Make
log_artifact()
more resilient to network errors - Removed Duplicate Artifact Dependencies
- Workaround urlib3 issue on windows
- Fix regression where ipython was hanging
- Allow logging of numpy high precision floating point values
- Reduce liklyhood of collisions for file backed media or artifact objects
- Fix wandb.watch() regression when logging pytorch graphs
- Add support for logging joined and partitioned table
- Handle schema validation warnings for sweep configs
- Improve wandb sync to handle errors
- Add ability to label scripts and repositories who use wandb
- wandb.login() did not properly persist the host parameter
- Fix issue where step information was not synced properly when syncing tensorboard directories
- Fix some unicode issues with python2.7
- Fixed bug in
plot_calibration_curve
for ComplementNB - Fall back to not using SendFile on some linux systems
- Fix console issues where lines were truncated
- Fix console issues where console logging could block
- Add support for preemptible sweeps
- Add command line for sweep control
- Add support to load artifact collection properties
- Found and fixed the remaining issues causing runs to be marked crashed during outages
- Improved performance for users of
define_metric
, pytorch-lightning, and aggressive config saving - Fix issue when trying to log a cuda tensor to config or summary
- Remove dependancy on torch
backward_hooks
to compute graph - Fix an issue preventing the ability to resume runs on sagemaker
- Fix issues preventing pdb from working reliably with wandb
- Fix deprecation warning in vendored library (user submission)
- Fix logging behavior where the library was accidently outputting logs to the console
- Fix disabled mode to not create wandb dir and log files
- Renamed types to prep for Tables launch
- Allow renaming groups with public api
- Fix more network handling issues causing runs to be marked crashed (wandb sync to recover)
- Improve logging and exception handling to improve reporting and logging of crashed processes
- Fix network handling issue causing runs to be marked crashed (wandb sync to recover)
- Use
register_full_backward_hook
to support models with Dict outputs - Allow periods in table columns
- Fix artifact cache collisions when using forked processes
- Fix issue where custom charts do not display properly with pytorch-lightning
- Add experimental incremental artifact support
- Improve warnings when logging is being rate limited
- Fix tensorboard_sync condition where metrics at end of short run are dropped
- Fix
wandb sync
when tensorboard files are detected - Fix api key prompt in databricks notebook
- Integrate DSViz into Keras WandbCallback
- Add support for conda dependencies (user submit)
- Fix network handling issue where syncing stopped (use wandb sync to recover)
- Fix auth problem when using sagemaker and hugginface integrations together
- Fix handling of NaN values in tables with non floats
- Lazy load API object to prevent unnessary file access on module load
- Improve error messages when using public api history accessors
- Fix possible artifact cache race when using parallel artifact reads
- Fix artifact reference when
checksum=False
- Release
run.define_metric()
to simplify custom x-axis and more - Add column operators
add_column
,get_column
,get_index
towandb.Table()
- Significant fixes to stdout/stderr console logging
- Prevent excessive network when saving files with policy=
live
- Fix errors when trying to send large updates (most common with
wandb sync
)
- Automatically generate
run_table
artifact for logged tables - Add bracket notation to artifacts
- Improve URL validation when specifying server url to
wandb login
- Fix logged artifacts to be accessible after wait()
- Fix spell.run integration
- Performance fix syncing console logs with carriage returns
- Fix confusion matrix with class names and unlabeled data
- Add the ability to save artifacts without creating a run
- Add Foreign Table References to wandb.Table
- Allow the same runtime object to be logged to multiple artifacts
- Add experimental
run._define_metric()
support - Warn and ignore unsupported multiprocess
wandb.log()
calls
- Fix system metric logging rate in 0.10.x
- Fix Audio external reference issue
- Fix short runs with tensorboard_sync
- Ignore
wandb.init(id=)
when running a sweep - Sanitize artifact metadata if needed
- Allow syncing of tfevents with
wandb sync --sync-tensorboard
- Fix artifact.get() regression since 0.10.18
- Allow 0 byte artifacts
- Fix codesaving and program name reporting
- Added support for glb files for
wandb.Object3D()
- Added support for external references for
wandb.Audio()
- Custom chart support tensorboard
pr_curves
plugin - Support saving entire code directory in an artifact
- wandb.login() now respects disabled mode
- handle exception when trying to log TPUs in colab
- Add
WANDB_START_METHOD=thread
to support non-multiprocessing - Add
group
andjob_type
to Run object in the export API - Improve artifact docstrings
- Fix artifact manifest files incorrectly named with patch suffix
- Add run delete and file delete to the public API
- Align steps between
tensorboard_sync
and wandb.log() history - Add
WANDB_START_METHOD
to allow POSIX systems to use fork - Support mixed types in wandb.Table() with
allow_mixed_types
- Fix potential leaked file due to log not being closed properly
- Improve
wandb verify
to better handle network issues and report errors - Made file downloads more deterministic with respect to filesystem caches
- Fix regression seen with python 3.5
- Silence vendored watchdog warnings on mac
- Artifacts now support parallel writers for large distributed workflows.
- Artifacts support distributed tables for dataset visualization.
- Improvements to PR templates
- Added more type annotations
- Vendored watchdog 0.9.0 removing it as a dependency
- New documentation generator
- Public api now has
file.direct_url
to avoid redirects for signed urls.
- Allow
config-defaults.yaml
to be overwritten when running sweeps - General bug fixes and improvements to
wandb verify
- Disabled widgets in Spyder IDE
- Fixed WANDB_SILENT in Spyder IDE
- Reference file:// artifacts respect the
name
attribute.
- Add
wandb verify
to troubleshoot local installs
- Fix tensorboard_sync issue writing to s3
- Prevent git secrets from being stored
- Disable verbose console messages when using moviepy
- Fix artifacts with checkpoints to be more robust when overwriting files
- Fix artifacts recycled id issue
- Add wandb.Audio support to Artifacts
- Fix wandb config regressions introduced in 0.10.13
- Rollback changes supporting media with slashes in keys
- Add support for Mac M1 GPU monitoring
- Add support for TPU monitoring
- Add setting to disable sagemaker integration
- Fix tensorboard_sync with tensorboardX and tf1
- Fix issues logging images with slashes
- Fix custom charts issues
- Improve error messages using
wandb pull
- Improve error messages with
wandb.Table()
- Make sure silent mode is silent
- Fix
wandb online
to renable logging - Multiple artifact fixes
- Add Artifact.used_by and Artifact.logged_by
- Validate type consistency when logging Artifacts
- Enhance JoinedTable to not require downloaded assets
- Add ability to recursively download dependent artifacts
- Enable gradient logging with keras and tf2+
- Validate pytorch models are passed to wandb.watch()
- Improved docstrings for public methods / objects
- Warn when image sequences are logged with different sizes
- Fix incorrectly generated filenames in summary
- Fix anonymous mode to include the api key in URLs
- Fix pickle issue with disabled mode
- Fix artifact from_id query
- Fix handling of Tables with different image paths
- Disable wandb logging with
wandb disabled
orwandb.init(mode="disabled")
- Support cloning an artifact when logging wandb.Image()
- Multiple media artifact improvements and internal refactor
- Improve handling of artifact errors
- Fix issue where notebook name was ignored
- Extend silent mode for jupyter logging
- Fix issue where vendored libraries interfered with python path
- Fix various exceptions (divide by zero, int conversion, TypeError)
- Added confusion matrix plot
- Better jupyter messages with wandb.init()/reinit/finish
- Fix for fastai 2.1.5 (removed log_args)
- Fixed media logging when directories are changed
- Added artifact media logging (alpha)
- Add scriptable alerts
- Add url attribute for sweep public api
- Update docstrings for wandb sdk functions
- Fix cases where offline mode was making network connections
- Fix issues with python sweeps and run stopping
- Fix logging issue where we could accidently display an api key
- Fix wandb login issues with malformed hosts
- Allow wandb.restore() to be called without wandb.init()
- Fix resuming (reusing run_id) with empty summary
- Fix artitifact download issue
- Add missing wandb.unwatch() function
- Avoid creating spurious wandb directories
- Fix collections import issue when using an old version of six
- Allow callables to be serialized
- Fix compatibility issue with python 3.9
- Fix
wandb sync
failure introduced in 0.10.6 - Improve python agent handling of failing runs
- Fix rare condition where resuming runs does not work
- Improve symlink handling when called in thread context
- Fix issues when changing directories before calling wandb.init()
- Fix issue when checking for updated releases on pypi
- Make sure code saving is enabled in jupyter environments after login
- Sweep agents have extended timeout for large sweep configs
- Support WANDB_SILENT environment variable
- Warn about missing python package when logging images
- Fix wandb.restore() to apply diff patch
- Improve artifact error messages
- Fix loading of config-defaults.yaml and specified list of yaml config files
- Add new custom plots:
wandb.plot.*
- Add new python based sweep agent:
wandb.agent()
- Console log fixes (tqdm on windows, fix close exceptions)
- Add more attributes to the Run object (group, job_type, urls)
- Fix sagemaker login issues
- Fix issue where plots were not uploaded until the end of run
- Fix an issue where wandb.init(allow_val_change=) throws exception
- Added warning when trying to sync pre 0.10.0 run dirs
- Improved jupyter support for wandb run syncing information
- Fix artifact download issues
- Fix multiple issues with tensorboard_sync
- Fix multiple issues with juypter/python sweeps
- Fix issue where login was timing out
- Fix issue where config was overwritten when resuming runs
- Ported sacred observer to 0.10.x release
- Fix predicted bounding boxes overwritten by ground truth boxes
- Add missing save_code parameter to wandb.init()
- Added upload_file to API
- wandb.finish() can be called without matching wandb.init()
- Fix issue where files were being logged to wrong parallel runs
- Fix missing properties/methods -- as_dict(), sweep_id
- Fix wandb.summary.update() not updating all keys
- Code saving was not properly enabled based on UI settings
- Tensorboard now logging images before end of program
- Fix resume issues dealing with config and summary metrics
- Added sync_tensorboard ability to handle S3 and GCS files
- Added ability to specify host with login
- Improved artifact API to allow modifying attributes
- Fix codesaving to respect the server settings
- Fix issue runing wandb.init() on restricted networks
- Fix issue where we were ignoring settings changes
- Fix artifact download issues
- Added history sparklines at end of run
- Artifact improvements and API for linking
- Improved offline support and syncing
- Basic noop mode support to simplify testing
- Improved windows/pycharm support
- Run object has more modifiable properties
- Public API supports attaching artifacts to historic runs
- Many bugs fixed due to simplifying logic
- New sacred observer available at wandb.sacred.WandbObserver
- Improved artifact reference tracking for HTTP urls
- Print meaningful error message when runs are queried with
summary
instead ofsummary_metrics
- Sub paths of artifacts now expose an optional root directory argument to download()
- Artifact.new_file accepts an optional mode argument
- Removed legacy fastai docs as we're now packaged with fastai v2!
- Fix yaml parsing error handling logic
- Bad spelling in torch docstring, thanks @mkkb473
- Remove unused y_probas in sklearn plots, thanks @dreamflasher
- New deletion apis for artifacts
- Fix
wandb restore
when not logged in - Fix artifact download paths on Windows
- Retry 408 errors on upload
- Fix mask numeric types, thanks @numpee
- Fix artifact reference naming mixup
- Default pytorch histogram logging frequency from 100 -> 1000 steps
- Fix multiple prompts for login when using the command line
- Fix "no method rename_file" error
- Fixed edgecase histogram calculation in PyTorch
- Fix error in jupyter when saving session history
- Correctly return artifact metadata in public api
- Fix matplotlib / plotly rendering error
- New artifact cli commands!
wandb artifact put path_file_or_ref
wandb artifact get artifact:version
wandb artifact ls project_name
- New artifact api commands!
wandb.log_artifact()
wandb.use_artifact()
wandb.Api().artifact_versions()
wandb.Api().run.used_artifacts()
wandb.Api().run.logged_artifacts()
wandb.Api().Artifact().file()
- Improved syncing of large wandb-history.jsonl files for wandb sync
- New Artifact.verify method to ensure the integrity of local artifacts
- Better testing harness for api commands
- Run directory now store local time instead of utc time in the name, thanks @aiyolo!
- Improvements to our doc strings across the board.
- wandb.Table now supports a
dataframe
argument for logging dataframes as tables!
- Artifacts work in python2
- Artifacts default download locations work in Windows
- GCS references now properly cache / download, thanks @yoks!
- Fix encoding of numpy arrays to JSON
- Fix string comparison error message
- Major overhaul of artifact caching
- Configurable cache directory for artifacts
- Configurable download directory for artifacts
- New Artifact.verify method to ensure the integrity of local artifacts
- use_artifact no longer requires
type
- Deleted artifacts can now be be recommitted
- Lidar scenes now support vectors
- Fix issue with artifact downloads returning errors.
- Segmentation masks now handle non-unint8 data
- Fixed path parsing logic in
api.runs()
- Fix issue where files were always logged to latest run in a project.
- Fix issue where url was not display url on first call to wandb.init
- Handle multiple inits in Jupyter
- Handle ValueError's when capturing signals, thanks @jsbroks
- wandb agent handles rate limiting properly
- wandb.Artifact is now generally available!
- feature_importances now supports CatBoost, thanks @neomatrix369
- Catch all exceptions when saving Jupyter sessions
- validation_data automatically set in TF >= 2.2
- implements* hooks now implemented in keras callback for TF >= 2.2
- Raw source code saving now disabled by default
- We now support global settings on boot to enable code saving on the server
- New
code_save=True
argument to wandb.init to enable code saving manually
- Ensure cells don't hang on completion
- Fixed jupyter integration in PyCharm shells
- Made session history saving handle None metadata in outputs
- Save session history in jupyter notebooks
- Kaggle internet enable notification
- Extend wandb.plots.feature_importances to work with more model types, thanks @neomatrix369!
- Code saving for jupyter notebooks restored
- Fixed thread errors in jupyter
- Ensure final history rows aren't dropped in jupyter
- Add default class labels for semantic segmentation
- Enhance bounding box API to be similar to semantic segmentation API
- Increase media table rows to improve ROC/PR curve logging
- Fix issue where pre binned histograms were not being handled properly
- Handle nan values in pytorch histograms
- Fix handling of binary image masks
- Improve semantic segmentation image mask logging
- Close all open files to avoice ResourceWarnings, thanks @CrafterKolyan!
- Parse "tensor" protobufs, fixing issues with tensorboard syncing in 2.1
- Add ROC, precision_recall, HeatMap, explainText, POS, and NER to wandb.plots
- Add wandb.Molecule() logging
- Capture kaggle runs for metrics
- Add ability to watch from run object
- Avoid accidently picking up global debugging logs
- Improve bounding box annotations
- Log active GPU system metrics
- Only writing wandb/settings file if wandb init is called
- Improvements to wandb local command
- Fix GPU logging on some devices without power metrics
- Fix sweep config command handling
- Fix tensorflow string logging
- Added code saving of main python module
- Added ability to specify metadata for bounding boxes and segmentation masks
- Fix situations where uncommited data from wandb.log() is not persisted
- Fix dependency conflict with new versions of six package
- Add best metric and epoch to run summary with Keras callback
- Added wandb.run.config_static for environments required pickled config
- Fixed regression causing failures with wandb.watch() and DataParallel
- Improved compatibility with python 3.8
- Fix model logging under windows
- Fix exception when using wandb.watch() in a notebook
- Improve support for sparse tensor gradient logging on GPUs
- Relax version dependancy for PyYAML for users with old environments
- Added scikit-learn support
- Added ability to specify/exclude specific keys when building wandb.config
- Fix wandb.watch() on sparse tensors
- Fix incompatibilty with ray 0.8.1
- Fix missing pyyaml requirement
- Fix "W&B process failed to launch" problems
- Improved ability to log large model graphs and plots
- Added ability to configure agent commandline from sweep config
- Fix fast.ai prediction logging
- Fix logging of eager tensorflow tensors
- Fix jupyter issues with logging notebook name and wandb.watch()
- Ignore wandb.init() specified project and entity when running a sweep
- Fix agent "flapping" detection
- Fix local controller not starting when sweep is pending
- Added support for LightGBM
- Added local board support (Experimental)
- Added ability to modify sweep configuration
- Added GPU power logging to system metrics
- Prevent sweep agent from failing continously when misconfigured
- Added beta support for ray/tune hyperopt search strategy
- Added ability to specify max runs per agent
- Improve experience starting a sweep without a project already created
- Fix repeated wandb.Api().Run(id).scan_history() calls get updated data
- Fix early_terminate/hyperband in notebook/python environments
- Added min_step and max_step to run.scan_history for grabbing sub-sections of metrics
- wandb.init(reinit=True) now automatically calls wandb.join() to better support multiple runs per process
- wandb.init(sync_tensorboard=True) works again for TensorFlow 2.0
- Handle tags being passed in as a string
- Pin graphql-core < 3.0.0 to fix install errors
- TQDM progress bars update logs properly
- Oversized summary or history logs are now dropped which prevents retry hanging
- Fix regression syncing some versions of Tensorboard since 0.8.13
- Fix network error in Jupyter
- Fix calling wandb.init with sync_tensorboard multiple times in Jupyter
- Fix RuntimeError race when using threads and calling wandb.log
- Don't initialize Sentry when error reporting is disabled
- Added best_run() to wandb.sweep() public Api objects
- Remove internal tracking keys from wandb.config objects in the public Api
- Improve large object warning when values reach maximum size
- Warn when wandb.save isn't passed a string
- Run stopping from the UI works since regressing in 0.8.12
- Restoring a file that already exists locally works
- Fixed TensorBoard incorrectly placing some keys in the wrong step since 0.8.10
- wandb.Video only accepts uint8 instead of incorrectly converting to floats
- SageMaker environment detection is now more robust
- Resuming correctly populates config
- wandb.restore respects root when run.dir is set #658
- Calling wandb.watch multiple times properly namespaces histograms and graphs
- Sweeps now work in Windows!
- Added sweep attribute to Run in the public api
- Added sweep link to Jupyter and terminal output
- TensorBoard logging now stores proper timestamps when importing historic results
- TensorBoard logging now supports configuring rate_limits and filtering event types
- Use simple output mirroring stdout doesn't have a file descriptor
- Write wandb meta files to the system temp directory if the local directory isn't writable
- Added beta api.reports to the public API
- Added wandb.unwatch to remove hooks from pytorch models
- Store the framework used in config._wandb
- Create nested directory when videos are logged from tensorboard namespaces
- Fix race when using wandb.log
async=True
- run.summary acts like a proper dictionary
- run.summary sub dictionaries properly render
- handle None when passing class_colors for segmentation masks
- handle tensorflow2 not having a SessionHook
- properly escape args in windows
- fix hanging login when in anonymode
- tf2 keras patch now handles missing callbacks args
- Updates documentation autogenerated from docstrings in /docs
- wandb.init(config=config_dict) does not update sweep specified parameters
- wandb.config object now has a setdefaults method enabling improved sweep support
- Improved terminal and jupyter message incorporating 🚀 emojii!
- Allow wandb.watch to be called multiple times on different models
- Improved support for watching multple tfevent files
- Windows no longer requires
wandb run
simply runpython script_name.py
wandb agent
now works on windows.- Nice error message when wandb.log is called without a dict
- Keras callback has a new
log_batch_frequency
for logging metrics every N batches
- Fix compatibility issue with python 2.7 and old pip dependencies
- Improved onboarding flow when creating new accounts and entering api_key
- Fix public api returning incorrect data when config value is 0 or False
- Resumed runs no longer overwrite run names with run id
- Added recording of spell.run id in config
- wandb magic handles the case of tf.keras and keras being loaded
- tensorboard logging won't drop steps if multiple loggers have different global_steps
- keras gradient logging works in the latest tf.keras
- keras validation_data is properly set in tensorflow 2
- wandb pull command creates directories if they don't exist, thanks @chmod644
- file upload batching now asserts a minimum size
- sweeps works in python2 again
- scan_history now iterates the full set of points
- jupyter will run local mode if credentials can't be obtained
- Sweeps can now be run from within jupyter / directly from python! https://docs.wandb.com/sweeps/python
- New openai gym integration will automatically log videos, enabled with the monitor_gym keyword argument to wandb.init
- Ray Tune logging callback in wandb.ray.WandbLogger
- New global config file in ~/.config/wandb for global settings
- Added tests for fastai, thanks @borisdayma
- Public api performance enhancements
- Deprecated username in favor of enitity in the public api for consistency
- Anonymous login support enabled by default
- New wandb.login method to be used in jupyter enabling anonymous logins
- Better dependency error messages for data frames
- Initial integration with spell.run
- All images are now rendered as PNG to avoid JPEG artifacts
- Public api now has a projects field
- run.summary updates work in jupyter before log is called
- don't require numpy to be installed
- Setting nested keys in summary works
- notebooks in nested directories are properly saved
- Don't retry 404's / better error messaging from the server
- Strip leading slashes when loading paths in the public api
- Small files are batch uploaded as gzipped tarballs
- TensorBoardX gifs are logged to wandb
- wandb.init properly handles network failures on startup
- Keras callback only logs examples if data_type or input_type is set
- Fix edge case PyTorch model logging bug
- Handle patching tensorboard multiple times in jupyter
- Sweep picks up config.yaml from the run directory
- Dataframes handle integer labels
- Handle invalid JSON when querying jupyter servers
- fastai uses a fixed seed for example logging
- increased the max number of images for fastai callback
- new wandb.Video tag for logging video
- sync=False argument to wandb.log moves logging to a thread
- New local sweep controller for custom search logic
- Anonymous login support for easier onboarding
- Calling wandb.init multiple times in jupyter doesn't error out
- keras callback no longer guesses input_type for 2D data
- wandb.Image handles images with 1px height
- wandb Public API now has
run.scan_history
to return all history rows - wandb.config prints helpful errors if used before calling init
- wandb.summary prints helpful errors if used before calling init
- filestream api points to new url on the backend
- fastai callback uses the default monitor instead of assuming val_loss
- notebook introspections handles error cases and doesn't print stacktrace on failure
- Don't print description warning when setting name
- Fixed dataframe logging error with the keras callback
- Fixed line offsets in logs when resuming runs
- wandb.config casts non-builtins before writing to yaml
- vendored backports.tempfile to address missing package on install
- Added
api.sweep
to the python export api for querying sweeps - Added
WANDB_NOTEBOOK_NAME
for specifying the notebook name in cases we can't infer it - Added
WANDB_HOST
to override hostnames - Store if a run was run within jupyter
- wandb now supports stopping runs from the web ui
- Handle floats passed as step to
wandb.log
- wandb.config has full unicode support
- sync the main file to wandb if code saving is enabled and it's untracked by git
- XGBoost callback: wandb.xgboost.wandb_callback()
- Fixed plotly charts with large numpy arrays not rendering
wandb docker
works when nvidia is present- Better error when non string keys are sent to log
- Relaxed pyyaml dependency to fix AMI installs
- Magic works in jupyter notebooks.
- New preview release of auto-dataframes for Keras
- Added input_type and output_type to the Keras callback for simpler config
- public api supports retrieving specific keys and custom xaxis
- WANDB_IGNORE_GLOBS is respected on the final scan of files
- Unified run.id, run.name, and run.notes across all apis
- Handle funky terminal sizes when setting up our psuedo tty
- Fixed Jupyter notebook introspection logic
- run.summary.update() persists changes to the server
- tensorboard syncing is robust to invalid histograms and truncated files
- preview release of magic, calling wandb.init(magic=True) should automatically track config and metrics when possible
- cli now supports local installs of the backend
- fastai callback supports logging example images
- image logging works in Windows
- wandb sync handles tfevents with a single timestep
- fix incorrect command in overview page for running runs
- handle histograms with > 512 bins when streaming tensorboard
- better error message when calling wandb sync on a file instead of a directory
- new helper function for handling hyperparameters in sweeps
wandb.config.user_items()
- better mocking for improved testing
- entity is persisted on wandb.run when queried from the server
- tmp files always use the temporary directory to avoid syncing
- raise error if file shrinks while uploading
- images log properly in windows
- upgraded pyyaml requirement to address CVE
- no longer store a history of rows to prevent memory leak
- summary now supports new dataframe format
- WANDB_SILENT environment variable writes all wandb messages to debug.log
- Improved error messages for windows and tensorboard logging
- output.log is uploaded at the end of each run
- metadata, requirements, and patches are uploaded at the beginning of a run
- when not running from a git repository, store the main python file
- added WANDB_DISABLE_CODE to prevent diffing and code saving
- when running in jupyter store the name of the notebook
- auto-login support for colab
- store url to colab notebook
- store the version of this library in config
- store sys.executable in metadata
- fastai callback no longer requires path
- wandb.init now accepts a notes argument
- The cli replaced the message argument with notes and name
- wandb sync handles tensorboard embeddings
- wandb sync correctly handles images in tensorboard
- tf.keras correctly handles single input functional models
- wandb.Api().runs returns an iterator that's reusable
- WANDB_DIR within a hidden directory doesn't prevent syncing
- run.files() iterates over all files
- pytorch recurssion too deep error
- wandb sync accepts an --ignore argument with globs to skip files
- run.summary now has an items() method for iterating over all keys
- Better error messages on access denied
- Better error messages when optional packages aren't installed
- Urls printed to the termial are url-escaped
- Namespaced tensorboard events work with histograms
- Public API now retries on failures and re-uses connection pool
- Catch git errors when remotes aren't pushed to origin
- Moved keras graph collection to on_train_begin to handle unbuilt models
- Handle more cases of not being able to save weights
- Updates to summary after resuming are persisted
- PyTorch histc logging fixed in 0.4.1
- Fixed
wandb sync
tensorboard import
- wandb.init(tensorboard=True) works with Tensorflow 2 and Eager Execution
- wandb.init(tensorboard=True) now works with tb-nightly and PyTorch
- Automatically log examples with tf.keras by adding missing validation_data
- Socket only binds to localhost for improved security and prevents firewall warnings in OSX
- Added user object to public api for getting the source user
- Added run.display_name to the public api
- Show display name in console output
- Added --tags, --job_group, and --job_type to
wandb run
- Added environment variable for minimum time to run before considering crashed
- Added flake8 tests to CI, thanks @cclauss!
- wandb-docker-run accepts image digests
- keras callback works in tensorflow2-alpha0
- keras model graph now puts input layer first
- PyTorch log frequency added for gradients and weights
- PyTorch logging performance enhancements
- wandb.init now accepts a name parameter for naming runs
- wandb.run.name reflects custom display names
- Improvements to nested summary values
- Deprecated wandb.Table.add_row in favor of wandb.Table.add_data
- Initial support for a fast.ai callback thanks to @borisdayma!
- run.get_url resolves the default entity if one wasn't specified
- wandb restore accepts run paths with only slashes
- Fixed PyYaml deprecation warnings
- Added entrypoint shell script to manifest
- Strip newlines from cuda version
- handle case insensitive docker credentials
- fix app_url for private cloud login flow
- don't retry 404's when starting sweep agents
- ensure DNS lookup failures can't prevent startup
- centralized debug logging
- wandb agent waits longer to send a SIGKILL after sending SIGINT
- support for logging docker images with the WANDB_DOCKER env var
- WANDB_DOCKER automatically set when run in kubernetes
- new wandb-docker-run command to automatically set env vars and mount code
- wandb.restore supports launching docker for runs that ran with it
- python packages are now recorded and saved in a requirements.txt file
- cpu_count, gpu_count, gpu, os, and python version stored in wandb-metadata.json
- the export api now supports docker-like paths, i.e. username/project:run_id
- better first time user messages and login info
- Improve error reporting for sweeps
- fixed Jupyter logging, don't change logger level
- fixed resuming in Jupyter
- wandb.init now degrades gracefully if a user hasn't logged in to wandb
- added a force flag to wandb.init to require a machine to be logged in
- Tensorboard and TensorboardX logging is now automatically instrumented when enabled
- added a tensorboard to wandb.init which patches tensorboard for logging
- wandb.save handles now accepts a base path to files in sub directories
- wandb.tensorflow and wandb.tensorboard can now be accessed without directly importing
wandb sync
will now traverse a wandb run directory and sync all runs
- Fixed race where wandb process could hang at the end of a run
- Fix resuming in Jupyter on kernel restart
- wandb.save ensures files are pushed regardless of growth
- Added replace=True keyword to init for auto-resuming
- New run.resumed property that can be used to detect if we're resuming
- New run.step property to use for setting an initial epoch on resuming
- Made Keras callback save the best model as it improves
- Really don't require numpy
- Better error message if wandb.log is called before wandb.init
- Prevent calling wandb.watch multiple times
- Handle datetime attributes in logs / plotly
- Add environment to sweeps
- Enable tagging in the public API and in wandb.init
- New media type wandb.Html for logging arbitrary html
- Add Public api.create_run method for custom integrations
- Added glob support to wandb.save, files save as they're written to
- Added wandb.restore for pulling files on resume
- Added a timeout for generating diffs on large repos
- Fixed edge case where file syncing could hang
- Ensure all file changes are captured before exit
- Handle cases of sys.exit where code isn't passed
- Don't require numpy
- New
wandb sync
command that pushes a local directory to the cloud - Support for syncing tfevents file during training
- Detect when running as TFJob and auto group
- New Kubeflow module with initial helpers for pipelines
- Fixed history / summary bug
- Initial support for AWS SageMaker
hook_torch
renamed towatch
with a deprecation warning- Projects are automatically created if they don't exist
- Additional GPU memory_allocated metric added
- Keras Graph stores edges
- PyTorch graph parsing is more robust
- Fixed PyTorch 0.3 support
- File download API supports WANDB_API_KEY authentication
- Sweeps work with new backend (early release).
- Summary tracks all history metrics unless they're overridden by directly writing to summary.
- Files support in data API.
- Show ongoing media file uploads in final upload progress.
- wandb.Audio supports duration
- Pass username header in filestream API
- New wandb.Audio data type.
- New step keyword argument when logging metrics
- Ability to specify run group and job type when calling wandb.init() or via environment variables. This enables automatic grouping of distributed training runs in the UI
- Ability to override username when using a service account API key
- Handle non-tty environments in Python2
- Handle non-existing git binary
- Fix issue where sometimes the same image was logged twice during a Keras step
- PyTorch
- Added a new
wandb.hook_torch
method which records the graph and logs gradients & parameters of pytorch models wandb.Image
detects pytorch tensors and uses torchvision.utils.make_grid to render the image.
- Added a new
wandb restore
handles the case of not being run from within a git repo.
- We now open stdout and stderr in raw mode in Python 2 ensuring tools like bpdb work.
- Catastrophic errors are now reported to Sentry unless WANDB_ERROR_REPORTING is set to false
- Improved error handling and messaging on startup
- The first image when calling wandb.log was not being written, now it is
wandb.log
andrun.summary
now remove whitespace from keys
- Vendored prompt_toolkit < 1.0.15 because the latest ipython is pinned > 2.0
- Lazy load wandb.h5 only if
summary
is accessed to improve Data API performance
- Jupyter
- Deprecated
wandb.monitor
in favor of automatically starting system metrics after the first wandb.log call - Added new %%wandb jupyter magic method to display live results
- Removed jupyter description iframe
- Deprecated
- The Data API now supports
per_page
andorder
options to theapi.runs
method - Initial support for wandb.Table logging
- Initial support for matplotlib logging