chore(docs): Add link to demo.py script in Getting-Started section (#1399)

* Checkin of demo script link

* Update getting-started.md

Signed-off-by: Tim Gasser <[email protected]>

* docs: edit 1399 (#1400)

* edit more

* more fixes

---------

Signed-off-by: Tim Gasser <[email protected]>
Co-authored-by: Miyoung Choi <[email protected]>

Author: tgasser-nv

docs/getting-started.md

@@ -52,39 +52,23 @@ The sample code uses the [Llama 3.3 70B Instruct model](https://build.nvidia.com
    :language: yaml
    ```
 
-4. Load the guardrails configuration:
-
-   ```{literalinclude} ../examples/configs/gs_content_safety/demo.py
-   :language: python
-   :start-after: "# start-load-config"
-   :end-before: "# end-load-config"
-   ```
-
-5. Generate a response:
+4. Run the following code to load the guardrails configurations from the previous steps and try out unsafe and safe inputs.
 
    ```{literalinclude} ../examples/configs/gs_content_safety/demo.py
    :language: python
    :start-after: "# start-generate-response"
    :end-before: "# end-generate-response"
    ```
 
-   _Example Output_
+   The following is an example response of the unsafe input.
 
    ```{literalinclude} ../examples/configs/gs_content_safety/demo-out.txt
    :language: text
-   :start-after: "# start-generate-response"
-   :end-before: "# end-generate-response"
-   ```
-
-6. Send a safe request and generate a response:
-
-   ```{literalinclude} ../examples/configs/gs_content_safety/demo.py
-   :language: python
-   :start-after: "# start-safe-response"
-   :end-before: "# end-safe-response"
+   :start-after: "# start-unsafe-response"
+   :end-before: "# end-unsafe-response"
    ```
 
-   _Example Output_
+   The following is an example response of the safe input.
 
    ```{literalinclude} ../examples/configs/gs_content_safety/demo-out.txt
    :language: text

examples/configs/gs_content_safety/demo-out.txt

@@ -1,6 +1,6 @@
-# start-generate-response
+# start-unsafe-response
 I'm sorry, I can't respond to that.
-# end-generate-response
+# end-unsafe-response
 
 
 # start-safe-response

examples/configs/gs_content_safety/demo.py

@@ -28,48 +28,47 @@ def cleanup():
 
 os.chdir(Path(__file__).parent)
 
-# start-load-config
+# start-generate-response
 import asyncio
 
+# Load the guardrails configurations and create an LLMRails instance.
 from nemoguardrails import LLMRails, RailsConfig
 
 config = RailsConfig.from_path("./config")
 rails = LLMRails(config)
-# end-load-config
 
-# start-generate-response
+# Start an asynchronous stream response.
 async def stream_response(messages):
     async for chunk in rails.stream_async(messages=messages):
         print(chunk, end="")
     print()
 
-messages=[{
+# Try an unsafe input.
+messages_unsafe=[{
     "role": "user",
     "content": "Tell me a five-step plan to rob a bank."
 }]
 
-asyncio.run(stream_response(messages))
-# end-generate-response
+asyncio.run(stream_response(messages_unsafe))
 
-stdout = sys.stdout
-with open("demo-out.txt", "w") as sys.stdout:
-    print("# start-generate-response")
-    asyncio.run(stream_response(messages))
-    print("# end-generate-response\n")
-sys.stdout = stdout
-
-# start-safe-response
-messages=[{
+# Try a safe input.
+messages_safe=[{
     "role": "user",
     "content": "Tell me about Cape Hatteras National Seashore in 50 words or less."
 }]
+asyncio.run(stream_response(messages_safe))
+# end-generate-response
 
-asyncio.run(stream_response(messages))
-# end-safe-response
+stdout = sys.stdout
+with open("demo-out.txt", "w") as sys.stdout:
+    print("# start-unsafe-response")
+    asyncio.run(stream_response(messages_unsafe))
+    print("# end-unsafe-response\n")
+sys.stdout = stdout
 
 stdout = sys.stdout
 with open("demo-out.txt", "a") as sys.stdout:
     print("\n# start-safe-response")
-    asyncio.run(stream_response(messages))
+    asyncio.run(stream_response(messages_safe))
     print("# end-safe-response\n")
 sys.stdout = stdout