feat(lemonade): retry transient "llama-server failed to start" on model load#1876

Merged

kovtcharov-amd merged 9 commits into

mainfrom

feat/lemonade-load-retry

Jul 7, 2026

Commits on Jun 26, 2026

feat(lemonade): retry transient "llama-server failed to start" on model load
kovtcharov
committed
test(rag): load embedder via LemonadeClient (exercise the load retry, not raw REST)
kovtcharov
committed
Merge remote-tracking branch 'origin/main' into feat/lemonade-load-retry
kovtcharov
committed
test(ci): drive Lemonade model pull/load/verify through LemonadeClient, not raw REST
kovtcharov
committed
feat(lemonade): bump load retry to 3 with escalating backoff (windowed fault)
kovtcharov
committed
test(rag): tolerate non-fatal LLM/VLM pre-pull failures (match prior behavior)
kovtcharov
committed
test(ci): tolerate empty content from reasoning models in --chat check
kovtcharov
committed

Commits on Jul 7, 2026