phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-03 09:53:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

update distro definition for starter wow

Browse source
This commit is contained in:
Ashwin Bharambe 2025-11-10 15:37:24 -08:00
parent 5fdedff37f
commit 7bd4c09aef
4 changed files with 56 additions and 49 deletions
Download patch file Download diff file Expand all files Collapse all files

25
src/llama_stack/distributions/ci-tests/run-with-postgres-store.yaml
View file

@ -165,20 +165,15 @@ providers:
- provider_id: meta-reference - provider_id: meta-reference
provider_type: inline::meta-reference provider_type: inline::meta-reference
config: config:
persistence_store: persistence:
type: sql_postgres agent_state:
host: ${env.POSTGRES_HOST:=localhost} namespace: agents
port: ${env.POSTGRES_PORT:=5432} backend: kv_default
db: ${env.POSTGRES_DB:=llamastack} responses:
user: ${env.POSTGRES_USER:=llamastack} table_name: agent_responses
password: ${env.POSTGRES_PASSWORD:=llamastack} backend: sql_default
responses_store: max_write_queue_size: 10000
type: sql_postgres num_writers: 4
host: ${env.POSTGRES_HOST:=localhost}
port: ${env.POSTGRES_PORT:=5432}
db: ${env.POSTGRES_DB:=llamastack}
user: ${env.POSTGRES_USER:=llamastack}
password: ${env.POSTGRES_PASSWORD:=llamastack}
post_training: post_training:
- provider_id: torchtune-cpu - provider_id: torchtune-cpu
provider_type: inline::torchtune-cpu provider_type: inline::torchtune-cpu
@ -235,6 +230,8 @@ providers:
kvstore: kvstore:
namespace: batches namespace: batches
backend: kv_default backend: kv_default
max_concurrent_batches: 1
max_concurrent_requests_per_batch: 10
storage: storage:
backends: backends:
kv_default: kv_default:

25
src/llama_stack/distributions/starter-gpu/run-with-postgres-store.yaml
View file

@ -165,20 +165,15 @@ providers:
- provider_id: meta-reference - provider_id: meta-reference
provider_type: inline::meta-reference provider_type: inline::meta-reference
config: config:
persistence_store: persistence:
type: sql_postgres agent_state:
host: ${env.POSTGRES_HOST:=localhost} namespace: agents
port: ${env.POSTGRES_PORT:=5432} backend: kv_default
db: ${env.POSTGRES_DB:=llamastack} responses:
user: ${env.POSTGRES_USER:=llamastack} table_name: agent_responses
password: ${env.POSTGRES_PASSWORD:=llamastack} backend: sql_default
responses_store: max_write_queue_size: 10000
type: sql_postgres num_writers: 4
host: ${env.POSTGRES_HOST:=localhost}
port: ${env.POSTGRES_PORT:=5432}
db: ${env.POSTGRES_DB:=llamastack}
user: ${env.POSTGRES_USER:=llamastack}
password: ${env.POSTGRES_PASSWORD:=llamastack}
post_training: post_training:
- provider_id: huggingface-gpu - provider_id: huggingface-gpu
provider_type: inline::huggingface-gpu provider_type: inline::huggingface-gpu
@ -238,6 +233,8 @@ providers:
kvstore: kvstore:
namespace: batches namespace: batches
backend: kv_default backend: kv_default
max_concurrent_batches: 1
max_concurrent_requests_per_batch: 10
storage: storage:
backends: backends:
kv_default: kv_default:

25
src/llama_stack/distributions/starter/run-with-postgres-store.yaml
View file

@ -165,20 +165,15 @@ providers:
- provider_id: meta-reference - provider_id: meta-reference
provider_type: inline::meta-reference provider_type: inline::meta-reference
config: config:
persistence_store: persistence:
type: sql_postgres agent_state:
host: ${env.POSTGRES_HOST:=localhost} namespace: agents
port: ${env.POSTGRES_PORT:=5432} backend: kv_default
db: ${env.POSTGRES_DB:=llamastack} responses:
user: ${env.POSTGRES_USER:=llamastack} table_name: agent_responses
password: ${env.POSTGRES_PASSWORD:=llamastack} backend: sql_default
responses_store: max_write_queue_size: 10000
type: sql_postgres num_writers: 4
host: ${env.POSTGRES_HOST:=localhost}
port: ${env.POSTGRES_PORT:=5432}
db: ${env.POSTGRES_DB:=llamastack}
user: ${env.POSTGRES_USER:=llamastack}
password: ${env.POSTGRES_PASSWORD:=llamastack}
post_training: post_training:
- provider_id: torchtune-cpu - provider_id: torchtune-cpu
provider_type: inline::torchtune-cpu provider_type: inline::torchtune-cpu
@ -235,6 +230,8 @@ providers:
kvstore: kvstore:
namespace: batches namespace: batches
backend: kv_default backend: kv_default
max_concurrent_batches: 1
max_concurrent_requests_per_batch: 10
storage: storage:
backends: backends:
kv_default: kv_default:

30
src/llama_stack/distributions/starter/starter.py
View file

@ -20,11 +20,19 @@ from llama_stack.core.datatypes import (
from llama_stack.core.storage.datatypes import ( from llama_stack.core.storage.datatypes import (
InferenceStoreReference, InferenceStoreReference,
KVStoreReference, KVStoreReference,
ResponsesStoreReference,
SqlStoreReference, SqlStoreReference,
) )
from llama_stack.core.utils.dynamic import instantiate_class_type from llama_stack.core.utils.dynamic import instantiate_class_type
from llama_stack.distributions.template import DistributionTemplate, RunConfigSettings from llama_stack.distributions.template import DistributionTemplate, RunConfigSettings
from llama_stack.providers.datatypes import RemoteProviderSpec from llama_stack.providers.datatypes import RemoteProviderSpec
from llama_stack.providers.inline.agents.meta_reference.config import (
AgentPersistenceConfig,
MetaReferenceAgentsImplConfig,
)
from llama_stack.providers.inline.batches.reference.config import (
ReferenceBatchesImplConfig,
)
from llama_stack.providers.inline.files.localfs.config import LocalfsFilesImplConfig from llama_stack.providers.inline.files.localfs.config import LocalfsFilesImplConfig
from llama_stack.providers.inline.inference.sentence_transformers import ( from llama_stack.providers.inline.inference.sentence_transformers import (
SentenceTransformersInferenceConfig, SentenceTransformersInferenceConfig,
@ -278,22 +286,30 @@ def get_distribution_template(name: str = "starter") -> DistributionTemplate:
Provider( Provider(
provider_id="meta-reference", provider_id="meta-reference",
provider_type="inline::meta-reference", provider_type="inline::meta-reference",
config=dict( config=MetaReferenceAgentsImplConfig(
persistence_store=postgres_sql_config, persistence=AgentPersistenceConfig(
responses_store=postgres_sql_config, agent_state=KVStoreReference(
), backend="kv_default",
namespace="agents",
),
responses=ResponsesStoreReference(
backend="sql_default",
table_name="agent_responses",
),
),
).model_dump(exclude_none=True),
) )
], ],
"batches": [ "batches": [
Provider( Provider(
provider_id="reference", provider_id="reference",
provider_type="inline::reference", provider_type="inline::reference",
config=dict( config=ReferenceBatchesImplConfig(
kvstore=KVStoreReference( kvstore=KVStoreReference(
backend="kv_default", backend="kv_default",
namespace="batches", namespace="batches",
).model_dump(exclude_none=True), ),
), ).model_dump(exclude_none=True),
) )
], ],
}, },