updating examples and documentation

SuperCowPowers · Dec 30, 2023 · 5189b08 · 5189b08
1 parent a6ebcd6
commit 5189b08
Show file tree

Hide file tree

Showing 6 changed files with 153 additions and 29 deletions.
diff --git a/docs/api_classes/endpoint.md b/docs/api_classes/endpoint.md
@@ -7,18 +7,139 @@
 
 
 ## Examples
-Example Description 1
 
+**Run Inference on an Endpoint**
+
+```py title="endpoint_inference.py"
+from sageworks.api.feature_set import FeatureSet
+from sageworks.api.model import Model
+from sageworks.api.endpoint import Endpoint
+
+# Grab an existing Endpoint
+endpoint = Endpoint("abalone-regression-end")
+
+# SageWorks has full ML Pipeline provenance, so we can backtrack the inputs,
+# get a DataFrame of data (not used for training) and run inference
+model_name = endpoint.get_input()
+fs_name = Model(model_name).get_input()
+fs = FeatureSet(fs_name)
+athena_table = fs.get_training_view_table()
+df = fs.query(f"SELECT * FROM {athena_table} where training = 0")
+
+# Run inference/predictions on the Endpoint
+results = endpoint.predict(df)
+print(results[["class_number_of_rings", "prediction"]])
+```
+
+**Output**
+
+```py
+Processing...
+     class_number_of_rings  prediction
+0                       13   11.477922
+1                       12   12.316887
+2                        8    7.612847
+3                        8    9.663341
+4                        9    9.075263
+..                     ...         ...
+839                      8    8.069856
+840                     15   14.915502
+841                     11   10.977605
+842                     10   10.173433
+843                      7    7.297976
+```
+**Endpoint Details**
+
+!!!tip inline end "The details() method"
+    The `detail()` method on the Endpoint class provides a lot of useful information. All of the SageWorks classes have a `details()` method try it out!
+
+```py title="endpoint_details.py"
+from sageworks.api.endpoint import Endpoint
+from pprint import pprint
+
+# Get Endpoint and print out it's details
+endpoint = Endpoint("abalone-regression-end")
+pprint(endpoint.details())
 ```
-example code 1
+
+**Output**
+
+```py
+{
+ 'input': 'abalone-regression',
+ 'instance': 'Serverless (2GB/5)',
+ 'model_metrics':   metric_name  value
+			0        RMSE  2.190
+			1         MAE  1.544
+			2          R2  0.504,
+ 'model_name': 'abalone-regression',
+ 'model_type': 'regressor',
+ 'modified': datetime.datetime(2023, 12, 29, 17, 48, 35, 115000, tzinfo=datetime.timezone.utc),
+     class_number_of_rings  prediction
+0                        9    8.648378
+1                       11    9.717787
+2                       11   10.933070
+3                       10    9.899738
+4                        9   10.014504
+..                     ...         ...
+495                     10   10.261657
+496                      9   10.788254
+497                     13    7.779886
+498                     12   14.718514
+499                     13   10.637320
+ 'sageworks_tags': ['abalone', 'regression'],
+ 'status': 'InService',
+ 'uuid': 'abalone-regression-end',
+ 'variant': 'AllTraffic'}
 ```
 
-Example Description 2
+**Endpoint Metrics**
+
+```py title="endpoint_metrics.py"
+from sageworks.api.endpoint import Endpoint
 
+# Grab an existing Endpoint
+endpoint = Endpoint("abalone-regression-end")
+
+# SageWorks tracks both Model performance and Endpoint Metrics
+model_metrics = endpoint.details()["model_metrics"]
+endpoint_metrics = endpoint.endpoint_metrics()
+print(model_metrics)
+print(endpoint_metrics)
 ```
-example code 2
+
+**Output**
+
+```py
+  metric_name  value
+0        RMSE  2.190
+1         MAE  1.544
+2          R2  0.504
+
+    Invocations  ModelLatency  OverheadLatency  ModelSetupTime  Invocation5XXErrors
+29          0.0          0.00             0.00            0.00                  0.0
+30          1.0          1.11            23.73           23.34                  0.0
+31          0.0          0.00             0.00            0.00                  0.0
+48          0.0          0.00             0.00            0.00                  0.0
+49          5.0          0.45             9.64           23.57                  0.0
+50          2.0          0.57             0.08            0.00                  0.0
+51          0.0          0.00             0.00            0.00                  0.0
+60          4.0          0.33             5.80           22.65                  0.0
+61          1.0          1.11            23.35           23.10                  0.0
+62          0.0          0.00             0.00            0.00                  0.0
+...
 ```
 
+
+## SageWorks UI
+Running these few lines of code creates and deploys an AWS Endpoint. The Endpoint artifacts can be viewed in the Sagemaker Console/Notebook interfaces or in the SageWorks Dashboard UI. SageWorks will monitor the endpoint, plot invocations, latencies, and tracks error metrics.
+
+<figure>
+<img alt="sageworks_new_light" src="https://github.com/SuperCowPowers/sageworks/assets/4806709/50687ae1-8a5b-4faa-b45e-18c9e6271af7">
+<figcaption>SageWorks Dashboard: Endpoints</figcaption>
+</figure>
+
+
 !!! note "Not Finding a particular method?"
     The SageWorks API Classes use 'Core' Classes Internally, so you can lookup all the methods in
 [SageWorks Core Classes](../core_classes/overview.md)
diff --git a/examples/endpoint_details.py b/examples/endpoint_details.py
@@ -0,0 +1,6 @@
+from sageworks.api.endpoint import Endpoint
+from pprint import pprint
+
+# Grab an existing Endpoint and print out it's details
+endpoint = Endpoint("abalone-regression-end")
+pprint(endpoint.details())
diff --git a/examples/endpoint_inference.py b/examples/endpoint_inference.py
@@ -2,21 +2,17 @@
 from sageworks.api.model import Model
 from sageworks.api.endpoint import Endpoint
 
-# Grab the abalone regression Model
-model = Model("abalone-regression")
-
-# By default, an Endpoint is serverless, but you can make it non-serverless
-serverless = True
-model.to_endpoint(name="abalone-regression-end", tags=["abalone", "regression"], serverless=serverless)
-
-# Now we'll run inference on the endpoint
+# Grab an existing Endpoint
 endpoint = Endpoint("abalone-regression-end")
 
 # SageWorks has full ML Pipeline provenance, so we can backtrack the inputs,
 # get a DataFrame of data (not used for training) and run inference
-fs_name = model.get_input()
+model_name = endpoint.get_input()
+fs_name = Model(model_name).get_input()
 fs = FeatureSet(fs_name)
 athena_table = fs.get_training_view_table()
 df = fs.query(f"SELECT * FROM {athena_table} where training = 0")
+
+# Run inference/predictions on the Endpoint
 results = endpoint.predict(df)
 print(results[["class_number_of_rings", "prediction"]])
diff --git a/examples/endpoint_metrics.py b/examples/endpoint_metrics.py
@@ -1,18 +1,15 @@
-from sageworks.api.model import Model
+from sageworks.api.endpoint import Endpoint
+import pandas as pd
 
-# Grab the abalone-regression Model
-model = Model("abalone-regression")
+pd.set_option("display.max_columns", None)
+pd.set_option("display.max_rows", None)
+pd.set_option("display.width", 1000)
 
-# Perform a health check on the model
-# Note: The health_check() method returns 'issues' if there are any
-#       problems, so if there are no issues, the model is healthy
-health_issues = model.health_check()
-if not health_issues:
-    print("Model is Healthy")
-else:
-    print("Model has issues")
-    print(health_issues)
+# Grab an existing Endpoint
+endpoint = Endpoint("abalone-regression-end")
 
-# Get the model metrics and regression predictions
-print(model.model_metrics())
-print(model.regression_predictions())
+# SageWorks tracks both Model performance and Endpoint Metrics
+model_metrics = endpoint.details()["model_metrics"]
+endpoint_metrics = endpoint.endpoint_metrics()
+print(model_metrics)
+print(endpoint_metrics.iloc[:, 1:6])
diff --git a/mkdocs.yml b/mkdocs.yml
@@ -28,6 +28,7 @@ nav:
 theme:
   name: material
   palette:
+  - scheme: slate  # Dark mode default
 
   # Palette toggle for light mode
   - scheme: default

diff --git a/src/sageworks/api/endpoint.py b/src/sageworks/api/endpoint.py
@@ -1,4 +1,7 @@
-"""Endpoint: SageWorks Endpoint API Class"""
+"""Endpoint: Manages AWS Endpoint creation and deployment.
+Endpoints are automatically set up and provisioned for deployment into AWS.
+Endpoints can be viewed in the AWS Sagemaker interfaces or in the SageWorks
+Dashboard UI, which provides additional model details and performance metrics"""
 
 # SageWorks Imports
 from sageworks.core.artifacts.endpoint_core import EndpointCore