v0.2.1 - Minimal Output

UX Improvements

Clean, minimal output:

Before:

                         DeepSeek OCR CLI                         
                  Version 0.2.0 (Ollama backend)                  

Connecting to Ollama...
2025-12-06 19:26:15 - deepseek_ocr.model - INFO - Initialized...
✓ Connected to Ollama (deepseek-ocr)
Processing: document.pdf
OCR pages: 0%|████...
✓ Done!

After:

deepseek-ocr v0.2.1
OCR pages: 0%|████...
→ output/document.md

Changes

• Minimal banner (one line)
• Hide logger output by default (use --verbose to see)
• Remove verbose status messages
• Clean completion: just show output path
• Simpler error messages

Install/upgrade:

pip install --upgrade deepseek-ocr-cli