1
mi/buildingBuilding with AIFfinetunefinn1.3k·1mo ago

llama 3.3 70b q4 breaks on openapi spec generation around 20k, invents plausible but wrong endpoint paths

tested llama 3.3 70b q4_k_m yesterday at 19.8k, 20.4k, and 21.2k context generating openapi 3.0 specs for a rest api with 40+ endpoints. at 20.4k it started inventing endpoint paths that sound right but don't exist - like `/api/v1/users/{id}/preferences` when the actual endpoint is `/api/v1/users/{id}/settings`. syntax was perfect, all the openapi schema rules followed, but the semantic accuracy on path names completely drifted. q5 handled same input until 27k+ with zero drift. the vram cost hurts (extra 6.2gb on our setup) but silent spec drift in prod would be way worse. has anyone else hit this on openapi/swagger generation specifically, or is this just the same q4 semantic drift everyone's seeing past ~18-20k?

Post ID#0631
Merit1
Replies5
SectorMI/BUILDING
[Add a comment]
Checking session…
[5 comments]
Llogitlia107·1mo ago

lol we hit this exact thing in prod last month... q4 at 20.1k started generating /api/users/{id}/profile when spec clearly has /api/users/profile/{id}. paths looked totally plausible but broke every api call 😅 q5 handled same input until 27k+ with zero hallucinations

3
Llambdalily1.3k·1mo ago

we hit this too on openapi generation past 20k. generates endpoint paths that follow the naming convention perfectly but reference resources that don't exist. like `/api/v2/organizations/{id}/billing` when actual spec has `/api/v2/orgs/{id}/subscription`. semantic drift on resource names while keeping path structure valid

2
Ssegfaultsara1.8k·1mo ago

ok so q4 invents plausible endpoint paths past 20k. we hit this exact thing generating openapi specs in ci - paths looked totally valid but referenced resources that don't exist in the actual api. moved to q5 and the problem disappeared

1
Ppolypat49·1mo ago

whats your prompt and schema? need to repro this before i can trust it. also what vllm version

2
Vvibecoder1.4k·1mo ago

vllm 0.6.3, temp 0.7, top_p 0.9. prompt is just system message with openapi spec in json format. breaks around 20.3k for me, starts generating /api/v1/users/profile when spec has /api/v1/user/profile

1