sfquickstarts
sfquickstarts copied to clipboard
The RAG LLM Assistant Breaks out of the box on indexing
Describe the bug The RAG LLM Assistant Breaks out of the box on indexing
URL of where you see the bug https://quickstarts.snowflake.com/guide/build_rag_based_blog_ai_assistant_using_streamlit_openai_and_llamaindex/index.html?index=..%2F..index#3
To Reproduce Steps to reproduce the behavior:
- Follow the steps in the linked quickstart
- Upon running
build_index.py
the operation fails
(venv) ~/git/sfguide-blog-ai-assistant/.content/blogs (main) $ ls
cloud_native_data_engineering_with_matillion_and_snowflake.md data_engineering_with_snowpark_python_and_dbt.md
data_engineering_pipelines_with_snowpark_python.md getting_started_with_dataengineering_ml_using_snowpark_python.md
data_engineering_with_apache_airflow.md
(venv) ~/git/sfguide-blog-ai-assistant (main) $ python build_index.py
Building vector index...
Traceback (most recent call last):
File "/Users/vinooganesh/git/sfguide-blog-ai-assistant/build_index.py", line 40, in <module>
main()
File "/Users/vinooganesh/git/sfguide-blog-ai-assistant/build_index.py", line 36, in main
build_index(data_dir, knowledge_base_dir)
File "/Users/vinooganesh/git/sfguide-blog-ai-assistant/build_index.py", line 21, in build_index
documents = SimpleDirectoryReader(data_dir).load_data()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/vinooganesh/git/sfguide-blog-ai-assistant/venv/lib/python3.11/site-packages/llama_index/readers/file/base.py", line 149, in __init__
self.input_files = self._add_files(self.input_dir)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/vinooganesh/git/sfguide-blog-ai-assistant/venv/lib/python3.11/site-packages/llama_index/readers/file/base.py", line 209, in _add_files
raise ValueError(f"No files found in {input_dir}.")
ValueError: No files found in .content/blogs.
Expected behavior
The blogs exist in the .content/blogs
directory, I would expect files to be found.
Desktop (please complete the following information): MacOS Sonoma 14.2.1
Additional context Add any other context about the problem here.
same