Dockers for llm-app pipelines (#6711)

Co-authored-by: Olivier Ruas <olivier@pathway.com>
GitOrigin-RevId: 0ed6ad276de863574c6837860a8374a2c4263189

examples/pipelines/alert/Dockerfile

@@ -0,0 +1,6 @@
+FROM pathwaycom/pathway:latest
+WORKDIR /app
+COPY . .
+EXPOSE 8080
+
+CMD ["python", "app.py"]

examples/pipelines/alert/README.md

@@ -7,7 +7,7 @@
   </a>
 </p>
 
-# Alert Pipeline
+# Real-time alerting based on local documents: End-to-end template
 
 This example implements a pipeline that answers questions based on documents in a given folder. Additionally, in your prompts you can ask to be notified of any changes - in such case an alert will be sent to a Slack channel.
 
@@ -36,36 +36,54 @@ For this demo, Slack notifications are optional and notifications will be printe
 Your Slack application will need at least `chat:write.public` scope enabled.
 
 ### Setup environment:
-Set your env variables in the .env file placed in this directory or in the root of the repo.
+Set your env variables in the .env file placed in this directory.
 
 ```bash
 OPENAI_API_KEY=sk-...
-SLACK_ALERT_CHANNEL_ID=  # If unset, alerts will be printed to the terminal
+SLACK_ALERT_CHANNEL_ID=
 SLACK_ALERT_TOKEN=
-PATHWAY_DATA_DIR= # If unset, defaults to ./data/live/
+PATHWAY_DATA_DIR= # If unset, defaults to ./data/live/. If running with Docker, when you change this variable you may need to change the volume mount.
 PATHWAY_PERSISTENT_STORAGE= # Set this variable if you want to use caching
 ```
 
-### Run the project
+### Run with Docker
 
-Make sure you have installed poetry dependencies with `--extras unstructured`. 
+To run jointly the Alert pipeline and a simple UI execute:
 
 ```bash
-poetry install --with examples --extras unstructured
+docker compose up --build
 ```
 
-Run:
+Then, the UI will run at http://0.0.0.0:8501 by default. You can access it by following this URL in your web browser.
 
+The `docker-compose.yml` file declares a [volume bind mount](https://docs.docker.com/reference/cli/docker/container/run/#volume) that makes changes to files under `data/` made on your host computer visible inside the docker container. The files in `data/live` are indexed by the pipeline - you can paste new files there and they will impact the computations.
+
+### Run manually
+
+Alternatively, you can run each service separately.
+
+Make sure you have installed poetry dependencies. 
 ```bash
-poetry run python app.py
+poetry install --with examples
 ```
 
-If all dependencies are managed manually rather than using poetry, you can run:
+Then run:
+```bash
+poetry run python app.py
+```
 
+If all dependencies are managed manually rather than using poetry, you can alternatively use:
 ```bash
 python app.py
 ```
 
+To run the Streamlit UI, run:
+```bash
+streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
+```
+
+### Querying the pipeline
+
 To create alerts, you can call the REST API:
 
 ```bash
@@ -75,8 +93,4 @@ curl --data '{
 }' http://localhost:8080/ | jq
 ```
 
-or use the Streamlit UI. Run:
-```bash
-streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
-```
-and then you can access the UI at `0.0.0.0:8501`.
+or access the Streamlit UI at `0.0.0.0:8501`.

examples/pipelines/alert/docker-compose.yml

@@ -0,0 +1,21 @@
+version: "3.8"
+services:
+  pathway:
+    build:
+      context: .
+    ports:
+      - "8080:8080"
+    environment:
+      OPENAI_API_KEY:
+      PATHWAY_PERSISTENT_STORAGE:
+    volumes:
+      - "./data:/app/data"
+  streamlit_ui:
+    depends_on:
+      - pathway
+    build:
+      context: ./ui
+    ports:
+      - "8501:8501"
+    environment:
+      PATHWAY_REST_CONNECTOR_HOST: "pathway"

examples/pipelines/alert/ui/Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.11
+
+WORKDIR /app
+
+RUN pip install streamlit python-dotenv
+
+COPY . .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "server.py", "--server.port", "8501", "--server.address", "0.0.0.0"]

examples/pipelines/alert/ui/server.py

@@ -49,6 +49,4 @@
             st.markdown(response)
         st.session_state.messages.append({"role": "assistant", "content": response})
     else:
-        st.error(
-            f"Failed to send data to Discounts API. Status code: {response.status_code}"
-        )
+        st.error(f"Failed to send data. Status code: {response.status_code}")

examples/pipelines/contextful/Dockerfile

@@ -0,0 +1,6 @@
+FROM pathwaycom/pathway:latest
+WORKDIR /app
+COPY . .
+EXPOSE 8080
+
+CMD ["python", "app.py"]

examples/pipelines/contextful/README.md

@@ -7,7 +7,7 @@
   </a>
 </p>
 
-# Contextful Pipeline
+# RAG pipeline with up-to-date knowledge: get answers based on documents in local folder
 
 This example implements a simple pipeline that answers questions based on documents in a given folder.
 
@@ -19,32 +19,52 @@ and sent to the OpenAI chat service for processing.
 ## How to run the project
 
 ### Setup environment:
-Set your env variables in the .env file placed in this directory or in the root of the repo.
+Set your env variables in the .env file placed in this directory.
 
 ```bash
 OPENAI_API_KEY=sk-...
-PATHWAY_DATA_DIR= # If unset, defaults to ./data/
+PATHWAY_DATA_DIR= # If unset, defaults to ./data/. If running with Docker, when you change this variable you may need to change the volume mount.
 PATHWAY_PERSISTENT_STORAGE= # Set this variable if you want to use caching
 ```
 
-### Run the project
+### Run with Docker
+
+To run jointly the Alert pipeline and a simple UI execute:
 
 ```bash
-poetry install --with examples
+docker compose up --build
 ```
 
-Run:
+Then, the UI will run at http://0.0.0.0:8501 by default. You can access it by following this URL in your web browser.
+
+The `docker-compose.yml` file declares a [volume bind mount](https://docs.docker.com/reference/cli/docker/container/run/#volume) that makes changes to files under `data/` made on your host computer visible inside the docker container. The files in `data/live` are indexed by the pipeline - you can paste new files there and they will impact the computations.
+
+### Run manually
+
+Alternatively, you can run each service separately.
 
+Make sure you have installed poetry dependencies. 
 ```bash
-poetry run python app.py
+poetry install --with examples
 ```
 
-If all dependencies are managed manually rather than using poetry, you can run either:
+Then run:
+```bash
+poetry run python app.py
+```
 
+If all dependencies are managed manually rather than using poetry, you can alternatively use:
 ```bash
 python app.py
 ```
 
+To run the Streamlit UI, run:
+```bash
+streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
+```
+
+### Querying the pipeline
+
 To query the pipeline, you can call the REST API:
 
 ```bash
@@ -54,8 +74,4 @@ curl --data '{
 }' http://localhost:8080/ | jq
 ```
 
-or use the Streamlit UI. Run:
-```bash
-streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
-```
-and then you can access the UI at `0.0.0.0:8501`.
+or access the Streamlit UI at `0.0.0.0:8501`.

examples/pipelines/contextful/docker-compose.yml

@@ -0,0 +1,21 @@
+version: "3.8"
+services:
+  pathway:
+    build:
+      context: .
+    ports:
+      - "8080:8080"
+    environment:
+      OPENAI_API_KEY:
+      PATHWAY_PERSISTENT_STORAGE:
+    volumes:
+      - "./data:/app/data"
+  streamlit_ui:
+    depends_on:
+      - pathway
+    build:
+      context: ./ui
+    ports:
+      - "8501:8501"
+    environment:
+      PATHWAY_REST_CONNECTOR_HOST: "pathway"

examples/pipelines/contextful/ui/Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.11
+
+WORKDIR /app
+
+RUN pip install streamlit python-dotenv
+
+COPY . .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "server.py", "--server.port", "8501", "--server.address", "0.0.0.0"]

examples/pipelines/contextful/ui/server.py

@@ -49,6 +49,4 @@
             st.markdown(response)
         st.session_state.messages.append({"role": "assistant", "content": response})
     else:
-        st.error(
-            f"Failed to send data to Discounts API. Status code: {response.status_code}"
-        )
+        st.error(f"Failed to send data. Status code: {response.status_code}")

examples/pipelines/contextful_geometric/Dockerfile

@@ -0,0 +1,6 @@
+FROM pathwaycom/pathway:latest
+WORKDIR /app
+COPY . .
+EXPOSE 8080
+
+CMD ["python", "app.py"]

examples/pipelines/contextful_geometric/README.md

@@ -7,7 +7,7 @@
   </a>
 </p>
 
-# Contextful Geometric Pipeline
+# RAG pipeline with up-to-date knowledge: get answers based on increasing number of documents
 
 This example implements a pipeline that answers questions based on documents in a given folder. To get the answer it sends increasingly more documents to the LLM chat until it can find an answer. You can read more about the reasoning behind this approach [here](https://pathway.com/developers/showcases/adaptive-rag).
 
@@ -24,32 +24,52 @@ do so until either question is answered or a limit of iterations is reached.
 ## How to run the project
 
 ### Setup environment:
-Set your env variables in the .env file placed in this directory or in the root of the repo.
+Set your env variables in the .env file placed in this directory.
 
 ```bash
 OPENAI_API_KEY=sk-...
-PATHWAY_DATA_DIR= # If unset, defaults to ./data/
+PATHWAY_DATA_DIR= # If unset, defaults to ./data/. If running with Docker, when you change this variable you may need to change the volume mount.
 PATHWAY_PERSISTENT_STORAGE= # Set this variable if you want to use caching
 ```
 
-### Run the project
+### Run with Docker
+
+To run jointly the Alert pipeline and a simple UI execute:
 
 ```bash
-poetry install --with examples
+docker compose up --build
 ```
 
-Run:
+Then, the UI will run at http://0.0.0.0:8501 by default. You can access it by following this URL in your web browser.
+
+The `docker-compose.yml` file declares a [volume bind mount](https://docs.docker.com/reference/cli/docker/container/run/#volume) that makes changes to files under `data/` made on your host computer visible inside the docker container. The files in `data/live` are indexed by the pipeline - you can paste new files there and they will impact the computations.
+
+### Run manually
 
+Alternatively, you can run each service separately.
+
+Make sure you have installed poetry dependencies with `--extras unstructured`. 
 ```bash
-poetry run python app.py
+poetry install --with examples --extras unstructured
 ```
 
-If all dependencies are managed manually rather than using poetry, you can run either:
+Then run:
+```bash
+poetry run python app.py
+```
 
+If all dependencies are managed manually rather than using poetry, you can alternatively use:
 ```bash
 python app.py
 ```
 
+To run the Streamlit UI, run:
+```bash
+streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
+```
+
+### Querying the pipeline
+
 To query the pipeline, you can call the REST API:
 
 ```bash
@@ -59,8 +79,5 @@ curl --data '{
 }' http://localhost:8080/ | jq
 ```
 
-or use the Streamlit UI. Run:
-```bash
-streamlit run ui/server.py --server.port 8501 --server.address 0.0.0.0
-```
-and then you can access the UI at `0.0.0.0:8501`.
+or access the Streamlit UI at `0.0.0.0:8501`.
+

examples/pipelines/contextful_geometric/docker-compose.yml

@@ -0,0 +1,21 @@
+version: "3.8"
+services:
+  pathway:
+    build:
+      context: .
+    ports:
+      - "8080:8080"
+    environment:
+      OPENAI_API_KEY:
+      PATHWAY_PERSISTENT_STORAGE:
+    volumes:
+      - "./data:/app/data"
+  streamlit_ui:
+    depends_on:
+      - pathway
+    build:
+      context: ./ui
+    ports:
+      - "8501:8501"
+    environment:
+      PATHWAY_REST_CONNECTOR_HOST: "pathway"

examples/pipelines/contextful_geometric/ui/Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.11
+
+WORKDIR /app
+
+RUN pip install streamlit python-dotenv
+
+COPY . .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "server.py", "--server.port", "8501", "--server.address", "0.0.0.0"]

examples/pipelines/contextful_geometric/ui/server.py

@@ -49,6 +49,4 @@
             st.markdown(response)
         st.session_state.messages.append({"role": "assistant", "content": response})
     else:
-        st.error(
-            f"Failed to send data to Discounts API. Status code: {response.status_code}"
-        )
+        st.error(f"Failed to send data. Status code: {response.status_code}")