FileMind

Troubleshooting

Solutions for common issues. If your problem isn't listed here, contact support.

Ollama Not Connecting

  1. Verify Ollama is running: open a terminal and run ollama serve
  2. Check the port: FileMind expects Ollama at http://localhost:11434
  3. If you changed the default port, update [llm] base_url in your config
  4. Re-run the setup wizard from Settings → Setup Wizard

OCR Not Working

FileMind uses PaddleOCR by default for scanned PDFs. If OCR results are poor:

  • Check that scanned pages have reasonable DPI (300+ recommended)
  • Try enabling GPU acceleration: set [ocr] paddleocr_use_gpu = true
  • For non-English documents, change the language: [ocr] paddleocr_lang = "fr"
  • If OCR fails entirely, check the dashboard for files with "ERROR" status

Slow Processing

Processing speed depends on your hardware and library size:

  • Embedding generation is the most compute-intensive step. A GPU significantly speeds this up.
  • OCR for scanned PDFs is slower than native text extraction. Consider increasing DPI rendering if quality is poor.
  • LLM operations (metadata fallback, Ask My Library) depend on your model size and hardware. A smaller model like llama3.2:3b is faster but less capable.
  • Adjust [jobs] max_concurrent to process more files in parallel (default: 2).

Metadata Extraction Issues

  • Wrong title extracted — the PDF may have misleading header text. Check the evidence panel to see what was parsed, then manually correct in the proposal review.
  • No metadata found — the paper may be a scanned image with poor OCR quality. Check the index status in the dashboard.
  • DOI not detected — some papers embed the DOI only in metadata (not visible text). FileMind extracts from the first 2 pages of text by default; increase [ingestion] metadata_pages if needed.

Search Returns No Results

  • Check that your library has been fully indexed (look for pending scan jobs in the dashboard)
  • For semantic search, embeddings must be generated first — this happens automatically during scanning
  • Try switching to "keyword" mode if semantic search returns nothing — the query may be too abstract
  • Check the [search] semantic_min_score threshold — lowering it returns more results

Database Issues

  • Database locked — another instance of FileMind may be running. Close it and try again. FileMind automatically retries locked operations up to 5 times.
  • Corrupt database — restore from a backup, or use Settings → Library → Clear Derived Data to rebuild the index while keeping your file registry.
  • Large database — the SQLite file grows with your library. Embeddings are the largest component. A 1,000-paper library typically uses 200-500 MB.

Health Check

FileMind's health endpoint reports the status of all subsystems. If something isn't working, check the dashboard status panel or visit http://localhost:{port}/health for details. Degraded subsystems are listed with a reason: 

{
  "ok": true,
  "degraded": ["llm"],
  "llm": {
    "available": false,
    "reason": "Ollama not reachable"
  }
}

FileMind degrades gracefully — if the LLM is unavailable, heuristic-only metadata extraction still works. If embeddings are unavailable, keyword search still works.