Skip to content

Not working as expected #4

Description

@zdposter

I try this:

git clone https://github.com/imanoop7/Generative-Search-Engine-For-Local-Files --depth=1
cd Generative-Search-Engine-For-Local-Files
uv venv --python 3.11.9
.venv\Scripts\activate
uv pip install -r requirements.txt
mkdir files
cd files
wget https://www.gutenberg.org/cache/epub/74652/pg74652.txt
wget "https://ec.europa.eu/programmes/erasmus-plus/project-result-content/415e4859-ca57-404d-a5ea-2b897a8b3beb/ba9a1458.docx&ved=2ahUKEwjM8_T2nbSJAxUb7AIHHaR1M3cQFnoECA4QAQ&usg=AOvVaw1RUo4QRIAW7Yvwjt-KIB7S" -O ba9a1458.docx
streamlit run local_genai_search.py

Then I asked to index files in files/ and I put a simple question "What are the Arguments for Python as programming language for ev3?" (content of docx file). I got a reply (from ollama only?) but my documents were ignored (not mentioned in Referenced Documents). What am I doing wrong?
image

In terminal I see these messages:


  You can now view your Streamlit app in your browser.

  Local URL: http://localhost:8501
  Network URL: http://192.168.10.236:8501

Starting the application...
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Loading FAISS index and metadata
Loaded index with 715 vectors and 715 metadata entries
Application finished
Loading FAISS index and metadata
Loaded index with 715 vectors and 715 metadata entries
Application finished
2024-10-29 20:11:25.089 Examining the path of torch.classes raised: Tried to instantiate class '__path__._path', but it does not exist! Ensure that it is registered via torch::class_
2024-10-29 20:11:25.639 Examining the path of torch.classes raised: Tried to instantiate class '__path__._path', but it does not exist! Ensure that it is registered via torch::class_
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Loading FAISS index and metadata
Loaded index with 715 vectors and 715 metadata entries
Application finished
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Indexing documents in C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files
Indexing documents in directory: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files
Processing file: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\ba9a1458.docx
Reading DOCX: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\ba9a1458.docx
Chunking text of length 1574 with chunk size 500 and overlap 50
Created 1 chunks
Processing file: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Encoding 387 document chunks
Adding embeddings to FAISS index
Saving FAISS index and metadata
Indexed 387 document chunks.
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Loading FAISS index and metadata
Loaded index with 387 vectors and 387 metadata entries
Application finished
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Loading FAISS index and metadata
Loaded index with 387 vectors and 387 metadata entries
Application finished
Starting the application...
Initialized model and FAISS index with dimension 768
Starting Streamlit UI
Loading FAISS index and metadata
Loaded index with 387 vectors and 387 metadata entries
User asked: 'What are the arguments for Python as programming language for ev3?'
Performing semantic search for query: 'What are the arguments for Python as programming language for ev3?', k=10
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\ba9a1458.docx, chunk_id: 0
Reading DOCX: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\ba9a1458.docx
Chunking text of length 1574 with chunk size 500 and overlap 50
Created 1 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 56
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 317
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 83
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 64
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 251
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 96
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 50
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 254
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Reading document chunk: C:\Users\zdenko.podobny\Downloads\Generative-Search-Engine-For-Local-Files\files\pg74652.txt, chunk_id: 246
Chunking text of length 1038572 with chunk size 500 and overlap 50
Created 386 chunks
Found 10 search results
Generating answer for query: 'What are the arguments for Python as programming language for ev3?'
Sending prompt to Ollama
Received response from Ollama
Displaying 0 referenced documents
Application finished

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions