c795ef2203
Add helm deployment instructions for ChatQnA, AgentQnA, AudioQnA, CodeTrans, DocSum, FaqGen and VisualQnA Signed-off-by: Dolpher Du <dolpher.du@intel.com>
125 lines
3.9 KiB
YAML
125 lines
3.9 KiB
YAML
# Copyright (C) 2024 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
apiVersion: gmc.opea.io/v1alpha3
|
|
kind: GMConnector
|
|
metadata:
|
|
labels:
|
|
app.kubernetes.io/name: gmconnector
|
|
app.kubernetes.io/managed-by: kustomize
|
|
gmc/platform: gaudi
|
|
name: switch
|
|
namespace: switch
|
|
spec:
|
|
routerConfig:
|
|
name: router
|
|
serviceName: router-service
|
|
nodes:
|
|
root:
|
|
routerType: Sequence
|
|
steps:
|
|
- name: Embedding
|
|
nodeName: node1
|
|
- name: Reranking
|
|
data: $response
|
|
internalService:
|
|
serviceName: reranking-svc
|
|
config:
|
|
endpoint: /v1/reranking
|
|
TEI_RERANKING_ENDPOINT: tei-reranking-svc
|
|
- name: TeiReranking
|
|
internalService:
|
|
serviceName: tei-reranking-svc
|
|
config:
|
|
endpoint: /rerank
|
|
isDownstreamService: true
|
|
- name: Llm
|
|
data: $response
|
|
nodeName: node2
|
|
node1:
|
|
routerType: Switch
|
|
steps:
|
|
- name: Embedding
|
|
condition: embedding-model-id==large
|
|
internalService:
|
|
serviceName: embedding-svc-large
|
|
config:
|
|
endpoint: /v1/embeddings
|
|
TEI_EMBEDDING_ENDPOINT: tei-embedding-gaudi-svc-bge15
|
|
- name: Embedding
|
|
condition: embedding-model-id==small
|
|
internalService:
|
|
serviceName: embedding-svc-small
|
|
config:
|
|
endpoint: /v1/embeddings
|
|
TEI_EMBEDDING_ENDPOINT: tei-embedding-gaudi-svc-bge-small
|
|
- name: TeiEmbeddingGaudi
|
|
internalService:
|
|
serviceName: tei-embedding-gaudi-svc-bge15
|
|
config:
|
|
MODEL_ID: BAAI/bge-base-en-v1.5
|
|
isDownstreamService: true
|
|
- name: TeiEmbeddingGaudi
|
|
internalService:
|
|
serviceName: tei-embedding-gaudi-svc-bge-small
|
|
config:
|
|
MODEL_ID: BAAI/bge-base-en-v1.5
|
|
isDownstreamService: true
|
|
- name: Retriever
|
|
condition: embedding-model-id==large
|
|
data: $response
|
|
internalService:
|
|
serviceName: retriever-svc-large
|
|
config:
|
|
endpoint: /v1/retrieval
|
|
REDIS_URL: redis-vector-db-large
|
|
TEI_EMBEDDING_ENDPOINT: tei-embedding-gaudi-svc-bge15
|
|
- name: Retriever
|
|
condition: embedding-model-id==small
|
|
data: $response
|
|
internalService:
|
|
serviceName: retriever-svc-small
|
|
config:
|
|
endpoint: /v1/retrieval
|
|
REDIS_URL: redis-vector-db-small
|
|
TEI_EMBEDDING_ENDPOINT: tei-embedding-gaudi-svc-bge-small
|
|
- name: VectorDB
|
|
internalService:
|
|
serviceName: redis-vector-db-large
|
|
isDownstreamService: true
|
|
- name: VectorDB
|
|
internalService:
|
|
serviceName: redis-vector-db-small
|
|
isDownstreamService: true
|
|
node2:
|
|
routerType: Switch
|
|
steps:
|
|
- name: Llm
|
|
condition: model-id==intel
|
|
internalService:
|
|
serviceName: llm-svc-intel
|
|
config:
|
|
endpoint: /v1/chat/completions
|
|
TGI_LLM_ENDPOINT: tgi-gaudi-service-intel
|
|
- name: Llm
|
|
condition: model-id==llama
|
|
internalService:
|
|
serviceName: llm-svc-llama
|
|
config:
|
|
endpoint: /v1/chat/completions
|
|
TGI_LLM_ENDPOINT: tgi-gaudi-service-llama
|
|
- name: TgiGaudi
|
|
internalService:
|
|
serviceName: tgi-gaudi-service-intel
|
|
config:
|
|
endpoint: /generate
|
|
MODEL_ID: Intel/neural-chat-7b-v3-3
|
|
isDownstreamService: true
|
|
- name: TgiGaudi
|
|
internalService:
|
|
serviceName: tgi-gaudi-service-llama
|
|
config:
|
|
endpoint: /generate
|
|
MODEL_ID: openlm-research/open_llama_3b
|
|
isDownstreamService: true
|