|
1 | 1 | # Changelog
|
2 | 2 |
|
| 3 | +## v1.45.0 (2019-12-04) |
| 4 | + |
| 5 | +### Features |
| 6 | + |
| 7 | + * add support for TF 1.15.0, PyTorch 1.3.1 and MXNet 1.6rc0. |
| 8 | + * add S3Downloader.list(s3_uri) functionality |
| 9 | + * introduce SageMaker AutoML |
| 10 | + * wrap up Processing feature |
| 11 | + * add a few minor features to Model Monitoring |
| 12 | + * add enable_sagemaker_metrics flag |
| 13 | + * Amazon SageMaker Model Monitoring |
| 14 | + * add utils.generate_tensorboard_url function |
| 15 | + * Add jobs list to Estimator |
| 16 | + |
| 17 | +### Bug fixes and other changes |
| 18 | + |
| 19 | + * remove unnecessary boto model files |
| 20 | + * update boto version to >=1.10.32 |
| 21 | + * correct Debugger tests |
| 22 | + * fix bug in monitor.attach() for empty network_config |
| 23 | + * Import smdebug_rulesconfig from PyPI |
| 24 | + * bump the version to 1.45.0 (publishes 1.46.0) for re:Invent-2019 |
| 25 | + * correct AutoML imports and expose current_job_name |
| 26 | + * correct Model Monitor eu-west-3 image name. |
| 27 | + * use DLC prod images |
| 28 | + * remove unused env variable for Model Monitoring |
| 29 | + * aws model update |
| 30 | + * rename get_debugger_artifacts to latest_job_debugger_artifacts |
| 31 | + * remove retain flag from update_endpoint |
| 32 | + * correct S3Downloader behavior |
| 33 | + * consume smdebug_ruleconfig .whl for ITs |
| 34 | + * disable DebuggerHook and Rules for TF distributions |
| 35 | + * incorporate smdebug_ruleconfigs pkg until availability in PyPI |
| 36 | + * remove pre/post scripts per latest validations |
| 37 | + * update rules_config .whl |
| 38 | + * remove py_version from SKLearnProcessor |
| 39 | + * AutoML improvements |
| 40 | + * stop overwriting custom rules volume and type |
| 41 | + * fix tests due to latest server-side validations |
| 42 | + * Minor processing changes |
| 43 | + * minor processing changes (instance_count + docs) |
| 44 | + * update api to latest |
| 45 | + * Eureka master |
| 46 | + * Add support for xgboost version 0.90-2 |
| 47 | + * SageMaker Debugger revision |
| 48 | + * Add support for SageMaker Debugger [WIP] |
| 49 | + * Fix linear learner crash when num_class is string and predict type is `multiclass_classifier` |
| 50 | + * Additional Processing Jobs integration tests |
| 51 | + * Migrate to updated Processing Jobs API |
| 52 | + * Processing Jobs revision round 2 |
| 53 | + * Processing Jobs revision |
| 54 | + * remove instance_pools parameter from tuner |
| 55 | + * Multi-Algorithm Hyperparameter Tuning Support |
| 56 | + * Import Processors in init files |
| 57 | + * Remove SparkML Processors and corresponding unit tests |
| 58 | + * Processing Jobs Python SDK support |
| 59 | + |
3 | 60 | ## v1.44.4 (2019-12-02)
|
4 | 61 |
|
5 | 62 | ### Bug fixes and other changes
|
|
0 commit comments