lab: exit program after catching and reporting exceptions (NVIDIA#336)

* cli: use non-zero exit status for error scenarios

Invoking sys.exit() with no args results in an exit status
of zero, which traditionally indicates success. This is not
appropriate for error scenarios.

Signed-off-by: Daniel P. Berrangé <[email protected]>

* lab: exit program after catching and reporting exceptions

Creating the app failed with an exception raised due to a missing
model file. Execution carried on, however, leading to use of an
undefined 'app' variable. Fix this, and a couple of other places
which also caught (unrecoverable) exceptions and forgot to exit.

Signed-off-by: Daniel P. Berrangé <[email protected]>

---------

Signed-off-by: Daniel P. Berrangé <[email protected]>

cli/lab.py

@@ -150,6 +150,7 @@ def init(ctx, interactive, model_path, taxonomy_path, repository, min_taxonomy):
                         f"Cloning {repository} failed with the following error: {exc}",
                         fg="red",
                     )
+                    sys.exit(1)
 
         # check if models dir exists, and if so ask for which model to use
         models_dir = dirname(model_path)
@@ -241,6 +242,7 @@ def serve(ctx, model_path, gpu_layers):
             f"Creating App using model failed with following value error: {err}",
             fg="red",
         )
+        sys.exit(1)
     try:
         llama_app._llama_proxy._current_model.chat_handler = llama_chat_format.Jinja2ChatFormatter(
             template="{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
@@ -253,6 +255,7 @@ def serve(ctx, model_path, gpu_layers):
             f"Error creating chat handler: {e}",
             fg="red",
         )
+        sys.exit(1)
     click.echo("Starting server process")
     click.echo(
         "After application startup complete see http://127.0.0.1:8000/docs for API."
@@ -334,6 +337,7 @@ def generate(
             f"Error connecting to the server: {exc.__cause__}",
             fg="red",
         )
+        sys.exit(1)
 
 
 @cli.command()
@@ -380,6 +384,7 @@ def chat(ctx, question, model, context, session, quick_question):
         )
     except ChatException as exc:
         click.secho(f"Executing chat failed with: {exc}", fg="red")
+        sys.exit(1)
 
 
 @cli.command()
@@ -420,6 +425,7 @@ def download(ctx, repository, release, filename, model_dir):
             f"Downloading model failed with the following Hugging Face Hub error: {exc}",
             fg="red",
         )
+        sys.exit(1)
 
 
 @cli.command()
@@ -482,19 +488,19 @@ def train(
                 f"Could not read taxonomy directory: {exc}",
                 fg="red",
             )
-            sys.exit()
+            sys.exit(1)
         except OSError as exc:
             click.secho(
                 f"Could not create data dir: {exc}",
                 fg="red",
             )
-            sys.exit()
+            sys.exit(1)
         except IndexError as exc:
             click.secho(
                 f"Could not copy into data directory: {exc}",
                 fg="red",
             )
-            sys.exit()
+            sys.exit(1)
 
     if not skip_preprocessing:
         script = os.path.join(cli_dir, "train/lora-mlx/make_data.py")