mirror of
https://github.com/BerriAI/litellm.git
synced 2026-10-15 04:19:08 +00:00
2c8a6e8a05
Address review feedback on the input cost formula rows. The cache-miss input cost subtracted only the cache-read portion, so when a request also had cache-write (creation) cost the miss rate was inflated; subtract both the cache-read and cache-write portions. Guard the derived rate against a negative cache-miss cost, which inconsistent backend costs could produce, so the formula hides instead of showing a negative rate. Show the per-token input formula only when cache_read_cost is present, since otherwise input_cost blends the miss and hit rates and cannot be split accurately, and render it directly under the Input Cost line so it attaches to the right row rather than floating below the whole section. |
||
|---|---|---|
| .. | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||