Files
pascalkienast 82328b843d Retry LLM payloads with reasoning token params after 400/422
Some providers reject temperature/max_tokens. Remap once to
max_completion_tokens instead of matching model name strings.
2026-08-18 17:27:02 +00:00
..