harupy
Follow
harupy
Follow
Achievements
Achievements
Popular repositories
-
-
-
-
-
keras-flask-classifier Public
A minimum implementation of image classifier web app with Keras & Flask
1,656 contributions in the last year
Less
More
Contribution activity
October 2021
Created 21 commits in 1 repository
Created 1 repository
- harupy/amazon-sagemaker-mlflow-fargate Jupyter Notebook
Created a pull request in mlflow/mlflow that received 8 comments
Use from_password and from_token when constructing DatabricksConfig
Signed-off-by: harupy 17039389+harupy@users.noreply.github.com
What changes are proposed in this pull request?
Use from_password and from_token whe…
+16
−16
•
8
comments
Opened 24 other pull requests in 1 repository
mlflow/mlflow
18
merged
1
open
5
closed
-
Avoid using
inspect.getfullargspecthat doesn't work properly for a wrapped function -
Avoid using
conda searchwhen setting python version - [WIP][Proxied artifact operations] Implement REST API endpoints and artifact repository
- Fix contour plot on run comparison page
- Fix incorrect parameter descriptions of model-registry APIs
-
Unpin
pytest - Avoid using conda when running tests
- Cache python dependencies to make CI build faster
-
Specify
datamodulein pytorch lightning tests -
Persist
preSwitchCategorizedUncheckedKeysandpostSwitchCategorizedUncheckedKeysto preserve the diff swtich state when switching experiments -
Set
orderByKeytoDEFAULT_ORDER_BY_KEYwhen switching experiments -
Support filtering runs by
start_time - Avoid using defaults conda channel and use conda-forge instead
-
Run
test_databricks_utils.pyagainst the latest master version ofdatabricks-cli -
Fix
test_spark_udf_autofills_no_arguments - [DO NOT MERGE] Run CI checks
- Try scikit-learn 1.0
-
Set
applyColumnDefOrderforAgGridReactto make sure columns order matches column definitions order - Remove example for tensorflow 1.X
- Normalize package names using rule defined in PEP 503
- Drop support for keras < 2.3.0 and tensorflow < 2.0.0
- [Example] Stream uploads/downloads
- [Example] Artifact uploads/downloads via presigned URL
-
Avoid creating a temporary file in
infer_pip_requirments
Reviewed 27 pull requests in 1 repository
mlflow/mlflow
27 pull requests
- [WIP][Proxied artifact operations] Implement REST API endpoints and artifact repository
- Add debugging statement to parameter collision exception
- Improve statsmodels autologging metrics
- Autologging functionality for scikit-learn integration with XGBoost (Part 1)
- Fix PyTorch Lightning dev version compatibility
-
Avoid using
inspect.getfullargspecthat doesn't work properly for a wrapped function -
Unpin
pytest - PyTorch Examples - latest pytorch and pytorch-lightning update
- Projects cli logging statements add for large project status reporting [ML-11215]
- Fix Kubernetes example
- Avoid using conda when running tests
- Fix contour plot on run comparison page
- Set globally_configured flag correctly for spark and pyspark.ml autologging integration
- MLflow UI updates
-
Support filtering runs by
start_time -
Run
test_databricks_utils.pyagainst the latest master version ofdatabricks-cli -
Use
from_passwordandfrom_tokenwhen constructingDatabricksConfig - Added diff-view switch
- [WIP] [Draft] Autologging functionality for scikit-learn integration with XGBoost and LightGBM
- Print warning when inferred requirements are not available on PyPI
- Improve sklearn autolog model pickling error handling
- Don't use pip in R install_mlflow function
- Normalize package names using rule defined in PEP 503
- [Bugfix] Alter_column on metrics table only called for mysql database engine
- Drop support for keras < 2.3.0 and tensorflow < 2.0.0
- Some pull request reviews not shown.
Created an issue in harupy/mlflow that received 2 comments
19
contributions
in private repositories
Oct 4 – Oct 27