This Python script automates the extraction of battery experimental data from PDF files using AI models. It employs a two-stage approach: first classifying whether a document is battery-related, then extracting detailed experimental conditions.

• Automated PDF Processing: Supports both PyPDF2 and PyMuPDF for robust PDF text extraction
• Two-Stage AI Pipeline:
  • Classification model to identify battery-related documents
  • Extraction model to extract detailed experimental conditions
• OpenAI-Compatible API: Works with DashScope and ModelScope APIs
• Comprehensive Statistics: Tracks processing time, token usage, and error rates
• JSON Output: Structured data output for easy integration
• Retry Mechanism: Automatic retry for failed API calls