aws/sagemaker-python-sdk issues and pull requests

#5198 - AttributeError: 'Session' object has no attribute 'settings'

Issue - State: closed - Opened by vektaBastien 5 months ago - 1 comment
Labels: component: training

#5023 - Support of Hydra’s CLI override mechanism

Issue - State: closed - Opened by discort 9 months ago - 6 comments
Labels: type: question, component: training

#4967 - fit() function hangs indefinitely with no logging or debugging info

Issue - State: closed - Opened by OmpahDev 11 months ago - 5 comments
Labels: type: bug, component: training

#4451 - PyTorch Estimator `max_run` parameter not working at all

Issue - State: open - Opened by yiphei over 1 year ago - 1 comment
Labels: type: bug, PyTorch, component: training

#4144 - CodeArtifact login support for FrameworkProcessor Jobs

Issue - State: open - Opened by akuma12 about 2 years ago
Labels: type: feature request, component: training

#4125 - @ remote support for multi-instance training job

Issue - State: open - Opened by sateeshmannar about 2 years ago - 1 comment
Labels: type: feature request, component: training, remote-function

#4040 - Enable processing and/or memory optimized instances when using sagemaker.remote_function's @remote decorator

Issue - State: open - Opened by diegodebrito over 2 years ago - 1 comment
Labels: type: feature request, component: training, remote-function

#3893 - HyperparameterTuner and Experiments UI not showing data

Issue - State: open - Opened by fjpa121197 over 2 years ago - 2 comments
Labels: type: bug, component: training

#3633 - Estimator can’t take network configuration object as parameter. Instead, it asks for subnets, sg’s and network isolation flag for each

Issue - State: open - Opened by idanmoradarthas almost 3 years ago - 1 comment
Labels: type: feature request, component: training

#3518 - For `Estimator`, `entry_point` is unnecessarily required to be at the root of `source_dir`.

Issue - State: open - Opened by stevenmanton almost 3 years ago - 4 comments
Labels: type: bug, component: training

#3469 - create_model() method on HyperparameterTuner

Issue - State: open - Opened by athewsey about 3 years ago
Labels: type: feature request, component: training

#3466 - Calling Estimator.attach(...) does not populate `debugger_hook_config` so `latest_job_debugger_artifacts_path()` returns `None`

Issue - State: open - Opened by morelen17 about 3 years ago
Labels: type: bug, component: training

#3455 - Training job not saved to S3 despite providing S3 output location due to no model artifact saved under path /opt/ml/model

Issue - State: closed - Opened by yshen92 about 3 years ago - 3 comments
Labels: bug, component: training

#3454 - Estimator .fit() fails in Jakarta region with built-in algo, due to added "USE_SMDEBUG" environment variable when debugger is not supported

Issue - State: open - Opened by yudho about 3 years ago
Labels: bug, component: training

#3318 - incorrect py_version argument documentation

Issue - State: closed - Opened by surajkota over 3 years ago - 1 comment
Labels: type: documentation, component: training

#2988 - Support to override model_data when calling Estimator.create_model

Issue - State: closed - Opened by jonsnowseven over 3 years ago - 1 comment
Labels: type: feature request, component: training

#2915 - IP Insights: "AttributeError: 'str' object has no attribute 'feature_dim'"

Issue - State: open - Opened by jonrau-lightspin almost 4 years ago
Labels: bug, component: training

#2876 - Unable to deploy best HPO model trained using open-source XGBoost framework

Issue - State: open - Opened by brianloyal almost 4 years ago
Labels: type: bug, XGBoost, component: training

#2835 - [DOCS] HyperparameterTuner Random strategy not documented

Issue - State: closed - Opened by TremaMiguel almost 4 years ago - 1 comment
Labels: type: documentation, component: training

#2794 - custom_overrides flag in sdk

Issue - State: open - Opened by mathephysicist almost 4 years ago
Labels: type: feature request, component: training

#2776 - Opaque error when specifying only bucket for `TrainingInput`

Issue - State: open - Opened by mszep almost 4 years ago - 1 comment
Labels: type: bug, component: training

#2734 - Support sub-commands (sub-parsers) for PyTorch training scripts

Issue - State: closed - Opened by jraman about 4 years ago - 1 comment
Labels: type: feature request, PyTorch, component: training

#2726 - Framework estimator .transform() fails with 'NoneType' object has no attribute '_create_sagemaker_model'

Issue - State: open - Opened by Dan-Treacher about 4 years ago - 1 comment
Labels: type: bug, component: training, component: pysdk-team

#2702 - Network Isolation for TensorFlow Training Requires Setting `model_dir` to `False` when creating a TensorFlow Estimator Object

Issue - State: open - Opened by evakravi about 4 years ago
Labels: type: bug, Tensorflow, component: training

#2688 - UnexpectedStatusException: Error for Training job. Reason: AlgorithmError: ExecuteUserScriptError:

Issue - State: closed - Opened by Electric-Dragon about 4 years ago - 5 comments
Labels: bug, Tensorflow, component: training

#2683 - Unable to set inter-container traffic encryption to True for default profiler processing job generated with training job

Issue - State: closed - Opened by IamCarlMichael about 4 years ago - 1 comment
Labels: type: feature request, component: training

#2641 - Unable to exactly reproduce MAP scores reported on validation dataset for SSD object detection model from a deployed endpoint

Issue - State: open - Opened by MercyPrasanna about 4 years ago
Labels: type: bug, component: training, component: autogluon

GitHub / aws/sagemaker-python-sdk issues and pull requests