remodel volumes #845

- unify `endpoint` and `credentials` volumes in `config`
- add `project` volume to train model
- adapt documentation

Author: wochinge

Dockerfile

@@ -31,7 +31,7 @@ COPY . .
 
 RUN pip install -e . --no-cache-dir
 
-VOLUME ["/app/dialogue", "/app/endpoints", "/app/credentials", "/app/out"]
+VOLUME ["/app/model", "/app/config", "/app/project"]
 
 EXPOSE 5005
 

docker/docker-compose.yml

@@ -7,19 +7,19 @@ services:
       ports:
       - "5005:5005"
       volumes:
-      - "./rasa-app-data/dialogue:/app/dialogue"
-      - "./rasa-app-data/endpoints:/app/endpoints"
-      - "./rasa-app-data/credentials:/app/credentials"
+      - "./rasa-app-data/core-model:/app/model"
+      - "./rasa-app-data/config:/app/config"
+      - "./rasa-app-data/project:/app/project"
       command: 
         - start
         - -d
-        - ./dialogue
+        - ./model
         - -c 
         - rest
         - -u
         - current/nlu
         - --endpoints
-        - endpoints/endpoints.yml
+        - config/endpoints.yml
   action_server:
       image: rasa/rasa_core_sdk:latest
       networks: ['rasa-network']
@@ -33,7 +33,7 @@ services:
     ports:
     - "5000:5000"
     volumes:
-    - "./rasa-app-data/projects/:/app/projects"
+    - "./rasa-app-data/nlu-models/:/app/projects"
     - "./rasa-app-data/logs:/app/logs"
   duckling:
     image: rasa/duckling:latest

docs/installation.rst

@@ -129,38 +129,42 @@ For example:
             --data '{"message": "Hi Bot"}' \
             http://localhost:5005/webhooks/rest/webhook
 
-To run commands inside a specific container, use ``docker-compose exec <container name>``.
-For example to retrain the core model:
+To run commands inside a specific container, use ``docker-compose run <container name>``.
+For example to train the core model:
 
     .. code-block:: bash
 
-         docker-compose exec rasacore ./entrypoint.sh train
+         docker-compose run rasacore train
 
 Volume Explanation
 ~~~~~~~~~~~~~~~~~~
-- **./rasa-app-data/dialogue**: This directory contains the trained Rasa Core models.
-  You can also move previously trained models to directoy to load them within the Docker container.
-- **./rasa-app-data/endpoints**: If you are using Rasa Core together with other components,
-  this directory should contain a file ``endpoints.yml`` which specifies how to reach this components.
-  For the shown example the file should look like this:
-
-    .. code-block:: yaml
-
-        action_endpoint:
-            url: 'http://action_server:5055/webhook'
-        nlu:
-            url: 'http://rasanlu:5000'
-
-- **./rasa-app-data/projects/**: This directory contains the trained Rasa NLU models.
-  You can also move previously trained models to directoy to load them within the Docker container.
-- **./rasa-app-data/credentials**: This directory is optionally and may contain credentials for various
-  `chat & voice platforms <https://rasa.com/docs/core/connectors/>`_ you want to connect your bot to.
-  Use the provided credentials by adding :code:`--credentials <path to your credentials file>``
-  to the run command of Rasa Core.
+- **./rasa-app-data/core-model**: This directory contains the trained Rasa Core models.
+  You can also move previously trained models to this directory to load them within the Docker container.
+- **./rasa-app-data/config**: This directory is for the configuration of the endpoints and of the
+  different `chat & voice platforms <https://rasa.com/docs/core/connectors/>`_ you can use Rasa Core with.
+    - To connect other components with Rasa Core this directory should contain a file ``endpoints.yml``,
+      which specifies how to reach these components.
+      For the shown docker-compose example the file should look like this:
+
+        .. code-block:: yaml
+
+            action_endpoint:
+                url: 'http://action_server:5055/webhook'
+            nlu:
+                url: 'http://rasanlu:5000'
+
+    - If you use a connector to a `chat & voice platforms <https://rasa.com/docs/core/connectors/>`_
+      you have to configure the required credentials for these in a file `credentials.yml`.
+      Use the provided credentials by adding :code:`--credentials <path to your credentials file>``
+      to the run command of Rasa Core.
+
+- **./rasa-app-data/project**: This directory contains your Rasa project and may be used to train a model.
+- **./rasa-app-data/nlu-models/**: This directory contains the trained Rasa NLU models.
+  You can also move previously trained models to this directory to load them within the Docker container.
 
 .. note::
 
-    You can always use custom directory structures or port mappings.
+    You can also use custom directory structures or port mappings.
     But don't forget to reflect this changes in the docker compose file and in your endpoint configuration.
 
 .. include:: feedback.inc

entrypoint.sh

@@ -19,7 +19,7 @@ case ${1} in
         exec "${@:2}"
         ;;
     train)
-        exec python -m rasa_core.train -s ./stories.md -d ./domain.yml -o ./out "${@:2}"
+        exec python -m rasa_core.train -s project/stories.md -d project/domain.yml -o ./model "${@:2}"
         ;;
     *)
         print_help