v4.0.0b1

Overview

Major architectural enhancements, adding a distributed server/worker system powered by the pure Gemini Interactions API and a streamlined 3-stage processing pipeline.

---

Key Changes

1. Pure Gemini Interactions API Migration

• Removed dependency on pydantic-ai in favor of direct, low-level interactions with the Gemini API via src/ai_sub/gemini_executor.py.

2. Distributed Server & Worker Architecture

• Implemented a robust client-server task distribution model ([src/ai_sub/server/coordinator.py](src/ai_sub/server/coordinator.py), src/ai_sub/worker/runner.py).
• Added deployment configurations and deployment guides ([docs/SERVER_DEPLOYMENT.md](docs/SERVER_DEPLOYMENT.md), docs/worker.yaml.example, docs/server.yaml.example).

3. Streamlined Pipeline Stages

• Refactored pipeline execution into a strict 3-stage sequence:

1. Global Summary:
   src/ai_sub/jobs/summary.py generates initial overview and context.
2. Subtitle Generation:
   src/ai_sub/jobs/subtitles.py performs raw and refined subtitle segmentation and stitching.
3. Lyrics Refinement:
   src/ai_sub/jobs/lyrics.py handles lyrics synchronization and refinement (replacing legacy music-refine).

---

Breaking Changes

• Requires Python 3.11 or later; configuration options have been reorganized.

---

What's Changed

• V4 - Server/Worker Architecture & Pure Gemini API Migration by @FlippFuzz in #24
• Remove pydantic ai by @FlippFuzz in #23

Full Changelog: v3.6.0...v4.0.0b1