Small edits.

fnocera · May 23, 2019 · 789a5f1 · 789a5f1
1 parent db00666
commit 789a5f1
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 2 deletions.
diff --git a/README.md b/README.md
@@ -18,4 +18,13 @@ You can develop your model with training pipeline and run with whatever variable
 You can then package it as an MLProject and run with this command (replacing variables with whatever you prefer using). 
 ```
 mlflow run model/wine_model_train.py -P model_name="model_1" alpha=0.8 l1_ratio=0.4
-```
+```
+
+TODO: Add details of ML Flow and AML details.
+
+To deploy a server using an existing Azure Blob storage run: 
+```
+mlflow server --file-store /tmp/mlflow --default-artifact-root "$MLFLOW_SERVER_DEFAULT_ARTIFACT_ROOT" --host 0.0.0.0
+``` 
+Then you can link to the server/where it is hosted for the deployment part. `MLFLOW_SERVER_DEFAULT_ARTIFACT_ROOT=wasbs://container@blob)NAME.blob.core.windows.net
+
diff --git a/requirements.txt b/requirements.txt
@@ -7,4 +7,11 @@ pandas==0.24.1
 python-dotenv==0.10.1
 scipy==1.2.1
 scikit-learn==0.20.1
-azure-storage
+azure-storage
+papermill==0.14.1
+numba==0.42.0
+python-dotenv==0.10.1
+Pillow==5.4.1
+azure-cli==2.0.50
+tensorflow
+keras==2.2.4