Skip to content

CHANGELOG.md

Latest commit

 

History

History
100 lines (81 loc) · 3.66 KB

CHANGELOG.md

File metadata and controls

100 lines (81 loc) · 3.66 KB

SVECTOR Python SDK Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[1.7.6] - 2025-08-07

Added - Comprehensive Vision API Enhancement

  • Complete Vision API Parity: Full feature parity with Node.js/JSR SDK
  • Advanced Vision Methods:
    • analyze_with_confidence() - Analysis with confidence scoring
    • batch_analyze() - Batch processing for multiple images
    • create_response() - Advanced input format compatibility
    • compare_images() - Multi-image comparison
    • generate_caption() - Social media caption generation with style options
  • Enhanced Error Handling: Better timeout handling and endpoint fallback
  • Improved Retry Logic: Smarter retry with exponential backoff
  • Comprehensive Examples: Complete example files matching Node.js patterns
  • Updated Documentation: Extensive README with all vision capabilities
  • Utility Functions: encode_image() and create_data_url() helpers
  • Better Type Hints: Improved type annotations for vision API

Enhanced

  • Vision API Performance: Reduced default timeout to 60s for better responsiveness
  • Endpoint Management: Dynamic baseURL support with fallback endpoints
  • Logging: Added detailed request logging for debugging
  • Response Format: Consistent response format across all vision methods

Fixed

  • Timeout Issues: Better handling of gateway timeouts (524 errors)
  • Server Error Retries: Proper 5xx error retry logic
  • API Compatibility: Full compatibility with SVECTOR vision endpoints

[1.0.0] - 2025-06-17

Added

  • Initial release of SVECTOR Python SDK
  • Complete chat completions API with streaming support
  • File upload and RAG (Retrieval Augmented Generation) capabilities
  • Models API for listing available models
  • Comprehensive error handling with specific error types
  • Command-line interface (CLI) for easy interaction
  • Support for multiple file upload methods (path, bytes, file object)
  • Multi-turn conversation support
  • Automatic retry logic with exponential backoff
  • Type hints for better developer experience
  • Comprehensive documentation and examples

Features

  • Chat API: Complete chat completions with customizable parameters
  • Streaming: Real-time response streaming via Server-Sent Events
  • File Upload: Support for various file formats for RAG functionality
  • Error Handling: Specific error classes for different API scenarios
  • CLI Tool: Full-featured command-line interface
  • Python 3.8+: Support for modern Python versions
  • Production Ready: Robust error handling and retry mechanisms

CLI Commands

  • svector chat - Send chat messages
  • svector stream - Stream responses in real-time
  • svector models - List available models
  • svector config - Manage API key configuration
  • svector file upload - Upload files for RAG
  • svector ask - Ask questions about uploaded files

API Coverage

  • Chat completions with parameters (temperature, max_tokens, etc.)
  • Streaming responses
  • File uploads for RAG functionality
  • Model listing
  • Multi-file RAG queries
  • Response metadata access

Dependencies

  • requests>=2.25.0 - HTTP client
  • typing-extensions>=4.0.0 - Type hints for Python <3.10

Installation

pip install svector

Quick Start

from svector import SVECTOR

client = SVECTOR(api_key="your-api-key")
response = client.chat.create(
    model="spec-3-turbo",
    messages=[{"role": "user", "content": "Hello!"}]
)

For more information, visit: https://www.svector.co.in