Template rejig (#48)

* document-rag / graph-rag refactor of templates

* Tweaking the docs and categories

* Clarify triple store vs RAG

* Tweak knowledge graph linkage

* Doc embedding for Qdrant

* Fix document RAG on Qdrant

* Fix templates

* Bump version

* Updated templates
This commit is contained in:
cybermaggedon 2024-09-03 00:09:15 +01:00 committed by GitHub
parent 121f7bb776
commit 208c219962
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
47 changed files with 1407 additions and 454 deletions

View file

@ -7,10 +7,6 @@ local prompts = import "prompt-template.jsonnet";
"chunk-size":: 250,
"chunk-overlap":: 15,
"embeddings-model":: "all-MiniLM-L6-v2",
"graph-rag-entity-limit":: 50,
"graph-rag-triple-limit":: 30,
"graph-rag-max-subgraph-size":: 3000,
services +: {
@ -81,102 +77,6 @@ local prompts = import "prompt-template.jsonnet";
},
},
embeddings: base + {
image: images.trustgraph,
command: [
"embeddings-hf",
"-p",
url.pulsar,
"-m",
$["embeddings-model"],
],
deploy: {
resources: {
limits: {
cpus: '1.0',
memory: '256M'
},
reservations: {
cpus: '0.5',
memory: '256M'
}
}
},
},
"kg-extract-definitions": base + {
image: images.trustgraph,
command: [
"kg-extract-definitions",
"-p",
url.pulsar,
],
deploy: {
resources: {
limits: {
cpus: '0.5',
memory: '128M'
},
reservations: {
cpus: '0.1',
memory: '128M'
}
}
},
},
"kg-extract-relationships": base + {
image: images.trustgraph,
command: [
"kg-extract-relationships",
"-p",
url.pulsar,
],
deploy: {
resources: {
limits: {
cpus: '0.5',
memory: '128M'
},
reservations: {
cpus: '0.1',
memory: '128M'
}
}
},
},
"graph-rag": base + {
image: images.trustgraph,
command: [
"graph-rag",
"-p",
url.pulsar,
"--prompt-request-queue",
"non-persistent://tg/request/prompt-rag",
"--prompt-response-queue",
"non-persistent://tg/response/prompt-rag-response",
"--entity-limit",
std.toString($["graph-rag-entity-limit"]),
"--triple-limit",
std.toString($["graph-rag-triple-limit"]),
"--max-subgraph-size",
std.toString($["graph-rag-max-subgraph-size"]),
],
deploy: {
resources: {
limits: {
cpus: '0.5',
memory: '128M'
},
reservations: {
cpus: '0.1',
memory: '128M'
}
}
},
},
}
} + prompts