Skip to content

v9.7.0 Multi-Segment Batching

Choose a tag to compare

View all tags
@youssefadel94 youssefadel94 released this 01 Mar 23:53
· 76 commits to main since this release
v9.7.0
d9251fd

What's New

Multi-Segment Batching

When the Gemini context window has enough headroom, consecutive video segments are grouped into single API calls. This reduces API calls and gives the model better cross-segment awareness.

  • Greedy bin-packing by token budget (~300 tok/sec per video)
  • Deep summary synergy freed doc tokens mean larger batches
  • Cache-aware only uncached segments hit the API
  • Automatic fallback if a batch fails, falls back to single-segment processing
  • --no-batch to disable

Interactive CLI Render Fix

Fixed line-wrap garble in narrow terminals all rendered lines are now truncated to terminal width via fitToWidth()

Code Audit

  • Removed 7 unused imports (html.js, docx.js)
  • Fixed variable shadowing in diff-engine.js
  • Added comments to 9 empty catch blocks in video.js
  • cleanupGeminiFiles now accepts arrays for batch cleanup
  • 13 new tests (345 total, all passing)

Install

npm install -g task-summary-extractor
Assets 2
Loading