mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-12 23:01:41 +00:00
4360c9fc43
A shadow eval job now covers a set of keys instead of exactly one, and each key carries its own max_turns budget, so one key exhausting its budget leaves its siblings sampling. Key, budget and stop state move to a new LiteLLM_ShadowEvalJobKey child table, taking the one-active-per-key partial unique index with them so the race-safe DB guarantee is unchanged. Attempts now record api_key_id, which is what makes per-key turn counting and the new by_key result slice possible The dashboard picker becomes a multi-select and the detail view gains a per key breakdown showing each key's own budget and status |
||
|---|---|---|
| .. | ||
| migrations | ||
| __init__.py | ||
| _logging.py | ||
| prisma_toolchain.py | ||
| py.typed | ||
| replica_identity.py | ||
| schema.prisma | ||
| utils.py | ||