Skip to content

Commit f192004

Browse files
are-cesare-ces
committed
Updating RHEL AI, RHAIIS, RHOAI configs to work with LLS 0.3.x
1 parent e93f5d9 commit f192004

8 files changed

Lines changed: 887 additions & 357 deletions

File tree

‎.github/workflows/e2e_tests_rhaiis.yaml‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -124,8 +124,8 @@ jobs:
124124
125125
- name: Docker Login for quay access
126126
env:
127-
QUAY_ROBOT_USERNAME: ${{ secrets.QUAY_ROBOT_USERNAME }}
128-
QUAY_ROBOT_TOKEN: ${{ secrets.QUAY_ROBOT_TOKEN }}
127+
QUAY_ROBOT_USERNAME: ${{ secrets.QUAY_DOWNSTREAM_USERNAME }}
128+
QUAY_ROBOT_TOKEN: ${{ secrets.QUAY_DOWNSTREAM_TOKEN }}
129129
run: |
130130
echo $QUAY_ROBOT_TOKEN | docker login quay.io -u=$QUAY_ROBOT_USERNAME --password-stdin
131131

‎.github/workflows/e2e_tests_rhelai.yaml‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -125,8 +125,8 @@ jobs:
125125
126126
- name: Docker Login for quay access
127127
env:
128-
QUAY_ROBOT_USERNAME: ${{ secrets.QUAY_ROBOT_USERNAME }}
129-
QUAY_ROBOT_TOKEN: ${{ secrets.QUAY_ROBOT_TOKEN }}
128+
QUAY_ROBOT_USERNAME: ${{ secrets.QUAY_DOWNSTREAM_USERNAME }}
129+
QUAY_ROBOT_TOKEN: ${{ secrets.QUAY_DOWNSTREAM_TOKEN }}
130130
run: |
131131
echo $QUAY_ROBOT_TOKEN | docker login quay.io -u=$QUAY_ROBOT_USERNAME --password-stdin
132132

‎examples/vllm-rhaiis.yaml‎

Lines changed: 157 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,157 @@
1+
version: '2'
2+
image_name: sample-notebook
3+
apis:
4+
- agents
5+
- batches
6+
- datasetio
7+
- eval
8+
- files
9+
- inference
10+
- safety
11+
- scoring
12+
- tool_runtime
13+
- vector_io
14+
15+
providers:
16+
inference:
17+
- provider_id: vllm
18+
provider_type: remote::vllm
19+
config:
20+
url: http://${env.RHAIIS_URL}:8000/v1/
21+
api_token: ${env.RHAIIS_API_KEY}
22+
tls_verify: false
23+
max_tokens: 2048
24+
- provider_id: sentence-transformers
25+
provider_type: inline::sentence-transformers
26+
config: {}
27+
vector_io:
28+
- provider_id: documentation_faiss
29+
provider_type: inline::faiss
30+
config:
31+
persistence:
32+
namespace: vector_io::faiss
33+
backend: kv_default
34+
files:
35+
- provider_id: meta-reference-files
36+
provider_type: inline::localfs
37+
config:
38+
storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/storage}
39+
metadata_store:
40+
table_name: files_metadata
41+
backend: sql_default
42+
safety:
43+
- provider_id: llama-guard
44+
provider_type: inline::llama-guard
45+
config:
46+
excluded_categories: []
47+
agents:
48+
- provider_id: meta-reference
49+
provider_type: inline::meta-reference
50+
config:
51+
persistence:
52+
agent_state:
53+
namespace: agents
54+
backend: kv_default
55+
responses:
56+
table_name: responses
57+
backend: sql_default
58+
max_write_queue_size: 10000
59+
num_writers: 4
60+
eval:
61+
- provider_id: meta-reference
62+
provider_type: inline::meta-reference
63+
config:
64+
kvstore:
65+
namespace: eval
66+
backend: kv_default
67+
datasetio:
68+
- provider_id: huggingface
69+
provider_type: remote::huggingface
70+
config:
71+
kvstore:
72+
namespace: datasetio::huggingface
73+
backend: kv_default
74+
- provider_id: localfs
75+
provider_type: inline::localfs
76+
config:
77+
kvstore:
78+
namespace: datasetio::localfs
79+
backend: kv_default
80+
scoring:
81+
- provider_id: basic
82+
provider_type: inline::basic
83+
config: {}
84+
tool_runtime:
85+
- provider_id: rag-runtime
86+
provider_type: inline::rag-runtime
87+
config: {}
88+
- provider_id: model-context-protocol
89+
provider_type: remote::model-context-protocol
90+
config: {}
91+
batches:
92+
- provider_id: reference
93+
provider_type: inline::reference
94+
config:
95+
kvstore:
96+
namespace: batches
97+
backend: kv_default
98+
storage:
99+
backends:
100+
kv_default:
101+
type: kv_sqlite
102+
db_path: ${env.SQLITE_STORE_DIR:=~/.llama/storage}/kv_store.db
103+
sql_default:
104+
type: sql_sqlite
105+
db_path: ${env.SQLITE_STORE_DIR:=~/.llama/storage}/sql_store.db
106+
stores:
107+
metadata:
108+
namespace: registry
109+
backend: kv_default
110+
inference:
111+
table_name: inference_store
112+
backend: sql_default
113+
max_write_queue_size: 10000
114+
num_writers: 4
115+
conversations:
116+
table_name: openai_conversations
117+
backend: sql_default
118+
prompts:
119+
namespace: prompts
120+
backend: kv_default
121+
registered_resources:
122+
models:
123+
- model_id: all-mpnet-base-v2
124+
provider_id: sentence-transformers
125+
provider_model_id: all-mpnet-base-v2
126+
model_type: embedding
127+
metadata:
128+
embedding_dimension: 768
129+
- model_id: ${env.RHAIIS_MODEL}
130+
provider_id: vllm
131+
model_type: llm
132+
provider_model_id: ${env.RHAIIS_MODEL}
133+
134+
shields:
135+
- shield_id: llama-guard
136+
provider_id: ${env.SAFETY_MODEL:+llama-guard}
137+
provider_shield_id: ${env.SAFETY_MODEL:=}
138+
- shield_id: code-scanner
139+
provider_id: ${env.CODE_SCANNER_MODEL:+code-scanner}
140+
provider_shield_id: ${env.CODE_SCANNER_MODEL:=}
141+
datasets: []
142+
scoring_fns: []
143+
benchmarks: []
144+
tool_groups:
145+
- toolgroup_id: builtin::rag
146+
provider_id: rag-runtime
147+
server:
148+
port: 8321
149+
telemetry:
150+
enabled: true
151+
vector_stores:
152+
default_provider_id: documentation_faiss
153+
default_embedding_model:
154+
provider_id: sentence-transformers
155+
model_id: all-mpnet-base-v2
156+
safety:
157+
default_shield_id: llama-guard

‎examples/vllm-rhelai.yaml‎

Lines changed: 157 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,157 @@
1+
version: '2'
2+
image_name: sample-notebook
3+
apis:
4+
- agents
5+
- batches
6+
- datasetio
7+
- eval
8+
- files
9+
- inference
10+
- safety
11+
- scoring
12+
- tool_runtime
13+
- vector_io
14+
15+
providers:
16+
inference:
17+
- provider_id: vllm
18+
provider_type: remote::vllm
19+
config:
20+
url: http://${env.RHEL_AI_URL}:${env.RHEL_AI_PORT}/v1/
21+
api_token: ${env.RHEL_AI_API_KEY}
22+
tls_verify: false
23+
max_tokens: 2048
24+
- provider_id: sentence-transformers
25+
provider_type: inline::sentence-transformers
26+
config: {}
27+
vector_io:
28+
- provider_id: documentation_faiss
29+
provider_type: inline::faiss
30+
config:
31+
persistence:
32+
namespace: vector_io::faiss
33+
backend: kv_default
34+
files:
35+
- provider_id: meta-reference-files
36+
provider_type: inline::localfs
37+
config:
38+
storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/storage}
39+
metadata_store:
40+
table_name: files_metadata
41+
backend: sql_default
42+
safety:
43+
- provider_id: llama-guard
44+
provider_type: inline::llama-guard
45+
config:
46+
excluded_categories: []
47+
agents:
48+
- provider_id: meta-reference
49+
provider_type: inline::meta-reference
50+
config:
51+
persistence:
52+
agent_state:
53+
namespace: agents
54+
backend: kv_default
55+
responses:
56+
table_name: responses
57+
backend: sql_default
58+
max_write_queue_size: 10000
59+
num_writers: 4
60+
eval:
61+
- provider_id: meta-reference
62+
provider_type: inline::meta-reference
63+
config:
64+
kvstore:
65+
namespace: eval
66+
backend: kv_default
67+
datasetio:
68+
- provider_id: huggingface
69+
provider_type: remote::huggingface
70+
config:
71+
kvstore:
72+
namespace: datasetio::huggingface
73+
backend: kv_default
74+
- provider_id: localfs
75+
provider_type: inline::localfs
76+
config:
77+
kvstore:
78+
namespace: datasetio::localfs
79+
backend: kv_default
80+
scoring:
81+
- provider_id: basic
82+
provider_type: inline::basic
83+
config: {}
84+
tool_runtime:
85+
- provider_id: rag-runtime
86+
provider_type: inline::rag-runtime
87+
config: {}
88+
- provider_id: model-context-protocol
89+
provider_type: remote::model-context-protocol
90+
config: {}
91+
batches:
92+
- provider_id: reference
93+
provider_type: inline::reference
94+
config:
95+
kvstore:
96+
namespace: batches
97+
backend: kv_default
98+
storage:
99+
backends:
100+
kv_default:
101+
type: kv_sqlite
102+
db_path: ${env.SQLITE_STORE_DIR:=~/.llama/storage}/kv_store.db
103+
sql_default:
104+
type: sql_sqlite
105+
db_path: ${env.SQLITE_STORE_DIR:=~/.llama/storage}/sql_store.db
106+
stores:
107+
metadata:
108+
namespace: registry
109+
backend: kv_default
110+
inference:
111+
table_name: inference_store
112+
backend: sql_default
113+
max_write_queue_size: 10000
114+
num_writers: 4
115+
conversations:
116+
table_name: openai_conversations
117+
backend: sql_default
118+
prompts:
119+
namespace: prompts
120+
backend: kv_default
121+
registered_resources:
122+
models:
123+
- model_id: all-mpnet-base-v2
124+
provider_id: sentence-transformers
125+
provider_model_id: all-mpnet-base-v2
126+
model_type: embedding
127+
metadata:
128+
embedding_dimension: 768
129+
- model_id: ${env.RHEL_AI_MODEL}
130+
provider_id: vllm
131+
model_type: llm
132+
provider_model_id: ${env.RHEL_AI_MODEL}
133+
134+
shields:
135+
- shield_id: llama-guard
136+
provider_id: ${env.SAFETY_MODEL:+llama-guard}
137+
provider_shield_id: ${env.SAFETY_MODEL:=}
138+
- shield_id: code-scanner
139+
provider_id: ${env.CODE_SCANNER_MODEL:+code-scanner}
140+
provider_shield_id: ${env.CODE_SCANNER_MODEL:=}
141+
datasets: []
142+
scoring_fns: []
143+
benchmarks: []
144+
tool_groups:
145+
- toolgroup_id: builtin::rag
146+
provider_id: rag-runtime
147+
server:
148+
port: 8321
149+
telemetry:
150+
enabled: true
151+
vector_stores:
152+
default_provider_id: documentation_faiss
153+
default_embedding_model:
154+
provider_id: sentence-transformers
155+
model_id: all-mpnet-base-v2
156+
safety:
157+
default_shield_id: llama-guard

0 commit comments

Comments
 (0)