mirror of
https://github.com/BerriAI/litellm.git
synced 2026-11-02 07:13:02 +00:00
1f3f512e50
Reading the window newest first, the first row seen of each session is that session's newest row, so the per-row newer-row probe and the index it needed are unnecessary. The page walks the existing (startTime, request_id) index in batches sized from how quickly new sessions appear, keeps the first row of each session in SQL, and stops once page_size + 1 sessions are settled. Filters are evaluated on each batch rather than steering the index choice, and cursor pages drop sessions that already have a row at or above the cursor with one lookup per candidate session. A walk that would read more than 20000 rows falls back to the grouped page. Drops the session head index and its migration. |
||
|---|---|---|
| .. | ||
| dist | ||
| litellm_proxy_extras | ||
| tests | ||
| build_and_publish.md | ||
| LICENSE | ||
| migration_runbook.md | ||
| pyproject.toml | ||
| README.md | ||
Additional files for the proxy. Reduces the size of the main litellm package.
Currently, only stores the migration.sql files for litellm-proxy.
To install, run:
uv add litellm-proxy-extras
OR
uv tool install 'litellm[proxy]' # installs litellm-proxy-extras and other proxy dependencies
To use the migrations, run:
litellm --use_prisma_migrate