fix text version and add precision in readme

Author: amaleelhamri

README.md

@@ -281,9 +281,37 @@ gcloud beta artifacts repositories create $ARTIFACT_REGISTRY_NAME \
  --repository-format=docker \
  --location=$REGION
  ```
-3. Create AI Platform version model using the docker image of the torchserve API package.
 ```bash
-gcloud beta ai-platform versions create $VERSION_NAME  --region=$REGION --model=$MODEL_NAME   --machine-type=n1-standard-4 --image=$DOCKER_IMAGE_PATH  --ports=$PORT   --health-route=$HEALTH_ROUTE   --predict-route=$PREDICT_ROUTE
+ docker build -t $REGION-docker.pkg.dev/$GCP_PROJECT_ID/$ARTIFACT_REGISTRY_NAME/$DOCKER_NAME:$DOCKER_TAG . -f TextDockerfile
+```
+```bash
+ docker push $REGION-docker.pkg.dev/$GCP_PROJECT_ID/$ARTIFACT_REGISTRY_NAME/$DOCKER_NAME:$DOCKER_TAG
+```
+
+3. [Optional] Run your docker locally and try to send a prediction
+
+```bash
+docker run -d -p 8080:8080 --name local_imdb  $REGION-docker.pkg.dev/$GCP_PROJECT_ID/$ARTIFACT_REGISTRY_NAME/$DOCKER_NAME:$DOCKER_TAG
+```
+```bash
+curl -X POST -H "Content-Type: application/json" -d '["this was a bad movie"]' 127.0.0.1:8080/predictions/fastai_model
+```
+If everything is working okay, you should receive a response from the server in your console
+```bash
+[
+  {
+    "Categories": "0",
+    "Tensor": [
+      0.9999990463256836,
+      9.371918849865324e-07
+    ]
+  }
+]
+```
+
+4. Create AI Platform version model using the docker image of the torchserve API package.
+```bash
+gcloud beta ai-platform versions create $VERSION_NAME  --region=$REGION --model=$MODEL_NAME --image=$REGION-docker.pkg.dev/$GCP_PROJECT_ID/$ARTIFACT_REGISTRY_NAME/$DOCKER_NAME:$DOCKER_TAG  --ports=8080   --health-route=/ping   --predict-route=/predictions/fastai_model
 ```
 
 

TextDockerfile

@@ -15,16 +15,16 @@ COPY requirements.txt /home/model-server/
 RUN pip install --upgrade pip
 RUN pip install -r /home/model-server/requirements.txt
 
-COPY torch_serve/text/handler.py torch_serve/text/config.py torch_serve/text/vocab.json /home/model-server/
+ADD torch_serve/ /home/model-server/torch_serve/
 
 COPY model/text/fastai_cls_weights.pth /home/model-server/
 
 RUN torch-model-archiver \
   --model-name=fastai_model \
   --version=1.0 \
   --serialized-file=/home/model-server/fastai_cls_weights.pth \
-  --extra-files=/home/model-server/config.py,/home/model-server/vocab.json \
-  --handler=/home/model-server/handler.py \
+  --extra-files=/home/model-server/torch_serve/text/config.py,/home/model-server/torch_serve/text/vocab.json \
+  --handler=/home/model-server/torch_serve/text/handler.py \
   --export-path=/home/model-server/model-store/
 
 

torch_serve/text/handler.py

@@ -77,14 +77,12 @@ def _load_preprocessor(self):
         self.preprocessor = preprocessor
 
     def preprocess(self, data):
-        logger.debug(f"input data : {data}")
         text = data[0].get("data")
         if text is None:
             text = data[0].get("body")
-        logger.info(f"text : {text} of type {type(text)}")
         if not isinstance(text, list):
             text = [text]
-        logger.info(f"text : {text}")
+        logger.info(f"text: {text}")
         text_preprocessed = [self.preprocessor(el) for el in text]
         max_len = find_max_list(text_preprocessed)
         text_preprocessed = [
@@ -132,8 +130,12 @@ def handle(data, context):
     if data is None:
         return None
 
+    logger.info(f"raw data: {data}")
     data = _service.preprocess(data)
+    logger.info(f"preprocessed data: {data}")
     data = _service.inference(data)
+    logger.info(f"infered data: {data}")
     data = _service.postprocess(data)
+    logger.info(f"postprocessed data: {data}")
 
     return data