Merge pull request #342 from vespa-engine/tgm/remove-forced-ml-install

Remove ml module dependency from application module

vespa/application.py

@@ -19,7 +19,6 @@
 from time import sleep
 
 from vespa.io import VespaQueryResponse, VespaResponse
-from vespa.ml import TextTask
 from vespa.query import QueryModel
 from vespa.evaluation import EvalMetric
 from vespa.package import ApplicationPackage
@@ -1166,7 +1165,7 @@ def predict(self, x, model_id, function_name="output_0"):
         model = self.get_model_from_application_package(model_id)
         encoded_tokens = model.create_url_encoded_tokens(x=x)
         with VespaSync(self) as sync_app:
-            return TextTask.parse_vespa_prediction(
+            return model.parse_vespa_prediction(
                 sync_app.predict(
                     model_id=model_id,
                     function_name=function_name,

vespa/ml.py

@@ -90,7 +90,7 @@ def export_to_onnx(self, output_path: str) -> None:
             pipeline, opset=11, output=Path(output_path), use_external_format=False
         )
 
-    def predict(self, text: str):
+    def predict(self, text: str) -> List:
         """
         Predict using a local instance of the model
 
@@ -102,7 +102,7 @@ def predict(self, text: str):
         return [x["score"] for x in predictions]
 
     @staticmethod
-    def parse_vespa_prediction(prediction):
+    def parse_vespa_prediction(prediction) -> List:
         return [cell["value"] for cell in prediction["cells"]]
 
     def create_url_encoded_tokens(self, x):