Add metering-rag for RAG token metrics

This commit is contained in:
Cyber MacGeddon 2024-09-29 20:02:01 +01:00
parent 6d57b7a6f3
commit db9c41d1c0

View file

@ -132,5 +132,37 @@ local prompt = import "prompt-template.jsonnet";
},
"metering-rag" +: {
create:: function(engine)
local container =
engine.container("metering-rag")
.with_image(images.trustgraph)
.with_command([
"metering",
"-p",
url.pulsar,
"-i",
"non-persistent://tg/request/text-completion-rag",
])
.with_limits("0.5", "128M")
.with_reservations("0.1", "128M");
local containerSet = engine.containers(
"metering-rag", [ container ]
);
local service =
engine.internalService(containerSet)
.with_port(8000, 8000, "metrics");
engine.resources([
containerSet,
service,
])
},
} + prompt