Merge 3ddf8c1b84 into 7cc4819e90

.github/workflows/integration-tests.yml

@@ -91,7 +91,7 @@ jobs:
 
       - name: Build Llama Stack
         run: |
-          uv run llama stack build --template ci-tests --image-type venv
+          uv run llama stack build --template starter --image-type venv
 
       - name: Check Storage and Memory Available Before Tests
         if: ${{ always() }}
@@ -101,6 +101,11 @@ jobs:
 
       - name: Run Integration Tests
         env:
+          OLLAMA_INFERENCE_MODEL: "llama3.2:3b-instruct-fp16" # for server tests
+          ENABLE_OLLAMA: "ollama" # for server tests
+          OLLAMA_URL: "http://0.0.0.0:11434"
+          SAFETY_MODEL: "llama-guard3:1b"
+          ENABLE_HUGGINGFACE: "huggingface"
           LLAMA_STACK_CLIENT_TIMEOUT: "300" # Increased timeout for eval operations
         # Use 'shell' to get pipefail behavior
         # https://docs.github.com/en/actions/reference/workflow-syntax-for-github-actions#exit-codes-and-error-action-preference
@@ -108,9 +113,9 @@ jobs:
         shell: bash
         run: |
           if [ "${{ matrix.client-type }}" == "library" ]; then
-            stack_config="ci-tests"
+            stack_config="starter"
           else
-            stack_config="server:ci-tests"
+            stack_config="server:starter"
           fi
 
           EXCLUDE_TESTS="builtin_tool or safety_with_image or code_interpreter or test_rag"

.github/workflows/integration-vector-io-tests.yml

@@ -93,7 +93,7 @@ jobs:
 
       - name: Build Llama Stack
         run: |
-          uv run llama stack build --template ci-tests --image-type venv
+          uv run llama stack build --template starter --image-type venv
 
       - name: Check Storage and Memory Available Before Tests
         if: ${{ always() }}

.github/workflows/providers-build.yml

@@ -97,9 +97,9 @@ jobs:
 
       - name: Build a single provider
         run: |
-          yq -i '.image_type = "container"' llama_stack/templates/ci-tests/build.yaml
-          yq -i '.image_name = "test"' llama_stack/templates/ci-tests/build.yaml
-          USE_COPY_NOT_MOUNT=true LLAMA_STACK_DIR=. uv run llama stack build --config llama_stack/templates/ci-tests/build.yaml
+          yq -i '.image_type = "container"' llama_stack/templates/starter/build.yaml
+          yq -i '.image_name = "test"' llama_stack/templates/starter/build.yaml
+          USE_COPY_NOT_MOUNT=true LLAMA_STACK_DIR=. uv run llama stack build --config llama_stack/templates/starter/build.yaml
 
       - name: Inspect the container image entrypoint
         run: |
@@ -126,14 +126,14 @@ jobs:
             .image_type    = "container" |
             .image_name    = "ubi9-test" |
             .distribution_spec.container_image = "registry.access.redhat.com/ubi9:latest"
-          ' llama_stack/templates/ci-tests/build.yaml
+          ' llama_stack/templates/starter/build.yaml
 
       - name: Build dev container (UBI9)
         env:
           USE_COPY_NOT_MOUNT: "true"
           LLAMA_STACK_DIR: "."
         run: |
-          uv run llama stack build --config llama_stack/templates/ci-tests/build.yaml
+          uv run llama stack build --config llama_stack/templates/starter/build.yaml
 
       - name: Inspect UBI9 image
         run: |

llama_stack/providers/inline/post_training/huggingface/post_training.py

@@ -22,9 +22,6 @@ from llama_stack.apis.post_training import (
 from llama_stack.providers.inline.post_training.huggingface.config import (
     HuggingFacePostTrainingConfig,
 )
-from llama_stack.providers.inline.post_training.huggingface.recipes.finetune_single_device import (
-    HFFinetuningSingleDevice,
-)
 from llama_stack.providers.utils.scheduler import JobArtifact, Scheduler
 from llama_stack.providers.utils.scheduler import JobStatus as SchedulerJobStatus
 from llama_stack.schema_utils import webmethod
@@ -83,6 +80,10 @@ class HuggingFacePostTrainingImpl:
         async def handler(on_log_message_cb, on_status_change_cb, on_artifact_collected_cb):
             on_log_message_cb("Starting HF finetuning")
 
+            from llama_stack.providers.inline.post_training.huggingface.recipes.finetune_single_device import (
+                HFFinetuningSingleDevice,
+            )
+
             recipe = HFFinetuningSingleDevice(
                 job_uuid=job_uuid,
                 datasetio_api=self.datasetio_api,

llama_stack/templates/ci-tests/__init__.py

@@ -1,7 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-from .ci_tests import get_distribution_template  # noqa: F401

llama_stack/templates/ci-tests/build.yaml

@@ -1,60 +0,0 @@
-version: 2
-distribution_spec:
-  description: CI tests for Llama Stack
-  providers:
-    inference:
-    - remote::cerebras
-    - remote::ollama
-    - remote::vllm
-    - remote::tgi
-    - remote::hf::serverless
-    - remote::hf::endpoint
-    - remote::fireworks
-    - remote::together
-    - remote::bedrock
-    - remote::databricks
-    - remote::nvidia
-    - remote::runpod
-    - remote::openai
-    - remote::anthropic
-    - remote::gemini
-    - remote::groq
-    - remote::llama-openai-compat
-    - remote::sambanova
-    - remote::passthrough
-    - inline::sentence-transformers
-    vector_io:
-    - inline::faiss
-    - inline::sqlite-vec
-    - inline::milvus
-    - remote::chromadb
-    - remote::pgvector
-    files:
-    - inline::localfs
-    safety:
-    - inline::llama-guard
-    agents:
-    - inline::meta-reference
-    telemetry:
-    - inline::meta-reference
-    post_training:
-    - inline::huggingface
-    eval:
-    - inline::meta-reference
-    datasetio:
-    - remote::huggingface
-    - inline::localfs
-    scoring:
-    - inline::basic
-    - inline::llm-as-judge
-    - inline::braintrust
-    tool_runtime:
-    - remote::brave-search
-    - remote::tavily-search
-    - inline::rag-runtime
-    - remote::model-context-protocol
-image_type: conda
-additional_pip_packages:
-- aiosqlite
-- asyncpg
-- sqlalchemy[asyncio]

llama_stack/templates/ci-tests/ci_tests.py

@@ -1,19 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-
-from llama_stack.templates.template import DistributionTemplate
-
-from ..starter.starter import get_distribution_template as get_starter_distribution_template
-
-
-def get_distribution_template() -> DistributionTemplate:
-    template = get_starter_distribution_template()
-    name = "ci-tests"
-    template.name = name
-    template.description = "CI tests for Llama Stack"
-
-    return template

llama_stack/templates/starter/run.yaml

@@ -179,7 +179,7 @@ providers:
       sqlite_db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/starter}/trace_store.db
       otel_exporter_otlp_endpoint: ${env.OTEL_EXPORTER_OTLP_ENDPOINT:=}
   post_training:
-  - provider_id: huggingface
+  - provider_id: ${env.ENABLE_HUGGINGFACE:=__disabled__}
     provider_type: inline::huggingface
     config:
       checkpoint_format: huggingface

llama_stack/templates/starter/starter.py

@@ -281,7 +281,7 @@ def get_distribution_template() -> DistributionTemplate:
         config=SentenceTransformersInferenceConfig.sample_run_config(),
     )
     post_training_provider = Provider(
-        provider_id="huggingface",
+        provider_id="${env.ENABLE_HUGGINGFACE:=__disabled__}",
         provider_type="inline::huggingface",
         config=HuggingFacePostTrainingConfig.sample_run_config(f"~/.llama/distributions/{name}"),
     )