Merge-related changes.

2025-12-31 07:33:51 +00:00 · 2025-04-02 19:56:44 +02:00 · 2025-04-02 19:56:44 +02:00 · 60e9f46856
commit 60e9f46856
parent d38aea33c1 66d6c2580e
456 changed files with 38636 additions and 10892 deletions
--- a/llama_stack/templates/nvidia/build.yaml
+++ b/llama_stack/templates/nvidia/build.yaml
@ -1,32 +1,27 @@
 version: '2'
 distribution_spec:
-  description: Use NVIDIA NIM for running LLM inference
+  description: Use NVIDIA NIM for running LLM inference and safety
  providers:
    inference:
    - remote::nvidia
    vector_io:
    - inline::faiss
    safety:
-    - inline::llama-guard
+    - remote::nvidia
    agents:
    - inline::meta-reference
    telemetry:
    - inline::meta-reference
    eval:
    - inline::meta-reference
+    post_training:
+    - remote::nvidia
    datasetio:
-    - remote::huggingface
    - inline::localfs
    scoring:
    - inline::basic
-    - inline::llm-as-judge
-    - inline::braintrust
    tool_runtime:
-    - remote::brave-search
-    - remote::tavily-search
-    - inline::code-interpreter
    - inline::rag-runtime
-    - remote::model-context-protocol
    preprocessing:
    - inline::basic
    - inline::simple_chunking
--- a/llama_stack/templates/nvidia/doc_template.md
+++ b/llama_stack/templates/nvidia/doc_template.md
@ -39,9 +39,10 @@ You can do this via Conda (build code) or Docker which has a pre-built image.
 This method allows you to get started quickly without having to build the distribution code.

 ```bash
-LLAMA_STACK_PORT=5001
+LLAMA_STACK_PORT=8321
 docker run \
  -it \
+  --pull always \
  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
  -v ./run.yaml:/root/my-run.yaml \
  llamastack/distribution-{{ name }} \
@ -55,7 +56,7 @@ docker run \
 ```bash
 llama stack build --template nvidia --image-type conda
 llama stack run ./run.yaml \
-  --port 5001 \
+  --port 8321 \
  --env NVIDIA_API_KEY=$NVIDIA_API_KEY
  --env INFERENCE_MODEL=$INFERENCE_MODEL
 ```
--- a/llama_stack/templates/nvidia/nvidia.py
+++ b/llama_stack/templates/nvidia/nvidia.py
@ -7,9 +7,10 @@
 from pathlib import Path

 from llama_stack.apis.preprocessing.preprocessors import PreprocessorInput
-from llama_stack.distribution.datatypes import Provider, ToolGroupInput
+from llama_stack.distribution.datatypes import ModelInput, Provider, ShieldInput, ToolGroupInput
 from llama_stack.providers.remote.inference.nvidia import NVIDIAConfig
 from llama_stack.providers.remote.inference.nvidia.models import MODEL_ENTRIES
+from llama_stack.providers.remote.safety.nvidia import NVIDIASafetyConfig
 from llama_stack.templates.template import DistributionTemplate, RunConfigSettings, get_model_registry


@ -17,19 +18,14 @@ def get_distribution_template() -> DistributionTemplate:
    providers = {
        "inference": ["remote::nvidia"],
        "vector_io": ["inline::faiss"],
-        "safety": ["inline::llama-guard"],
+        "safety": ["remote::nvidia"],
        "agents": ["inline::meta-reference"],
        "telemetry": ["inline::meta-reference"],
        "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::code-interpreter",
-            "inline::rag-runtime",
-            "remote::model-context-protocol",
-        ],
+        "post_training": ["remote::nvidia"],
+        "datasetio": ["inline::localfs"],
+        "scoring": ["inline::basic"],
+        "tool_runtime": ["inline::rag-runtime"],
        "preprocessing": ["inline::basic", "inline::simple_chunking"],
    }

@ -38,23 +34,28 @@ def get_distribution_template() -> DistributionTemplate:
        provider_type="remote::nvidia",
        config=NVIDIAConfig.sample_run_config(),
    )
+    safety_provider = Provider(
+        provider_id="nvidia",
+        provider_type="remote::nvidia",
+        config=NVIDIASafetyConfig.sample_run_config(),
+    )
+    inference_model = ModelInput(
+        model_id="${env.INFERENCE_MODEL}",
+        provider_id="nvidia",
+    )
+    safety_model = ModelInput(
+        model_id="${env.SAFETY_MODEL}",
+        provider_id="nvidia",
+    )

    available_models = {
        "nvidia": MODEL_ENTRIES,
    }
    default_tool_groups = [
-        ToolGroupInput(
-            toolgroup_id="builtin::websearch",
-            provider_id="tavily-search",
-        ),
        ToolGroupInput(
            toolgroup_id="builtin::rag",
            provider_id="rag-runtime",
        ),
-        ToolGroupInput(
-            toolgroup_id="builtin::code_interpreter",
-            provider_id="code-interpreter",
-        ),
    ]
    default_preprocessors = [
        PreprocessorInput(
@ -71,7 +72,7 @@ def get_distribution_template() -> DistributionTemplate:
    return DistributionTemplate(
        name="nvidia",
        distro_type="remote_hosted",
-        description="Use NVIDIA NIM for running LLM inference",
+        description="Use NVIDIA NIM for running LLM inference and safety",
        container_image=None,
        template_path=Path(__file__).parent / "doc_template.md",
        providers=providers,
@ -85,15 +86,60 @@ def get_distribution_template() -> DistributionTemplate:
                default_tool_groups=default_tool_groups,
                default_preprocessors=default_preprocessors,
            ),
+            "run-with-safety.yaml": RunConfigSettings(
+                provider_overrides={
+                    "inference": [
+                        inference_provider,
+                        safety_provider,
+                    ]
+                },
+                default_models=[inference_model, safety_model],
+                default_shields=[ShieldInput(shield_id="${env.SAFETY_MODEL}", provider_id="nvidia")],
+                default_tool_groups=default_tool_groups,
+                default_preprocessors=default_preprocessors,
+            ),
        },
        run_config_env_vars={
-            "LLAMASTACK_PORT": (
-                "5001",
-                "Port for the Llama Stack distribution server",
-            ),
            "NVIDIA_API_KEY": (
                "",
                "NVIDIA API Key",
            ),
+            ## Nemo Customizer related variables
+            "NVIDIA_USER_ID": (
+                "llama-stack-user",
+                "NVIDIA User ID",
+            ),
+            "NVIDIA_DATASET_NAMESPACE": (
+                "default",
+                "NVIDIA Dataset Namespace",
+            ),
+            "NVIDIA_ACCESS_POLICIES": (
+                "{}",
+                "NVIDIA Access Policies",
+            ),
+            "NVIDIA_PROJECT_ID": (
+                "test-project",
+                "NVIDIA Project ID",
+            ),
+            "NVIDIA_CUSTOMIZER_URL": (
+                "https://customizer.api.nvidia.com",
+                "NVIDIA Customizer URL",
+            ),
+            "NVIDIA_OUTPUT_MODEL_DIR": (
+                "test-example-model@v1",
+                "NVIDIA Output Model Directory",
+            ),
+            "GUARDRAILS_SERVICE_URL": (
+                "http://0.0.0.0:7331",
+                "URL for the NeMo Guardrails Service",
+            ),
+            "INFERENCE_MODEL": (
+                "Llama3.1-8B-Instruct",
+                "Inference model",
+            ),
+            "SAFETY_MODEL": (
+                "meta/llama-3.1-8b-instruct",
+                "Name of the model to use for safety",
+            ),
        },
    )
--- a/llama_stack/templates/nvidia/run-with-safety.yaml
+++ b/llama_stack/templates/nvidia/run-with-safety.yaml
@ -0,0 +1,123 @@
+version: '2'
+image_name: nvidia
+apis:
+- agents
+- datasetio
+- eval
+- inference
+- post_training
+- preprocessing
+- safety
+- scoring
+- telemetry
+- tool_runtime
+- vector_io
+providers:
+  inference:
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      url: ${env.NVIDIA_BASE_URL:https://integrate.api.nvidia.com}
+      api_key: ${env.NVIDIA_API_KEY:}
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      guardrails_service_url: ${env.GUARDRAILS_SERVICE_URL:http://localhost:7331}
+      config_id: self-check
+  vector_io:
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/faiss_store.db
+  safety:
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      guardrails_service_url: ${env.GUARDRAILS_SERVICE_URL:http://localhost:7331}
+      config_id: self-check
+  agents:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      persistence_store:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/agents_store.db
+  telemetry:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      service_name: "${env.OTEL_SERVICE_NAME:\u200B}"
+      sinks: ${env.TELEMETRY_SINKS:console,sqlite}
+      sqlite_db_path: ${env.SQLITE_DB_PATH:~/.llama/distributions/nvidia/trace_store.db}
+  eval:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/meta_reference_eval.db
+  post_training:
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      api_key: ${env.NVIDIA_API_KEY:}
+      dataset_namespace: ${env.NVIDIA_DATASET_NAMESPACE:default}
+      project_id: ${env.NVIDIA_PROJECT_ID:test-project}
+      customizer_url: ${env.NVIDIA_CUSTOMIZER_URL:http://nemo.test}
+  datasetio:
+  - provider_id: localfs
+    provider_type: inline::localfs
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/localfs_datasetio.db
+  scoring:
+  - provider_id: basic
+    provider_type: inline::basic
+    config: {}
+  tool_runtime:
+  - provider_id: rag-runtime
+    provider_type: inline::rag-runtime
+    config: {}
+  preprocessing:
+  - provider_id: basic
+    provider_type: inline::basic
+    config: {}
+  - provider_id: simple_chunking
+    provider_type: inline::simple_chunking
+    config: {}
+metadata_store:
+  type: sqlite
+  db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/registry.db
+models:
+- metadata: {}
+  model_id: ${env.INFERENCE_MODEL}
+  provider_id: nvidia
+  model_type: llm
+- metadata: {}
+  model_id: ${env.SAFETY_MODEL}
+  provider_id: nvidia
+  model_type: llm
+shields:
+- shield_id: ${env.SAFETY_MODEL}
+  provider_id: nvidia
+vector_dbs: []
+datasets: []
+scoring_fns: []
+benchmarks: []
+tool_groups:
+- toolgroup_id: builtin::rag
+  provider_id: rag-runtime
+preprocessors:
+- preprocessor_id: builtin::basic
+  provider_id: basic
+- preprocessor_id: builtin::chunking
+  provider_id: simple_chunking
+server:
+  port: 8321
--- a/llama_stack/templates/nvidia/run.yaml
+++ b/llama_stack/templates/nvidia/run.yaml
@ -5,6 +5,7 @@ apis:
 - datasetio
 - eval
 - inference
+- post_training
 - preprocessing
 - safety
 - scoring
@ -27,9 +28,11 @@ providers:
        namespace: null
        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/faiss_store.db
  safety:
-  - provider_id: llama-guard
-    provider_type: inline::llama-guard
-    config: {}
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      guardrails_service_url: ${env.GUARDRAILS_SERVICE_URL:http://localhost:7331}
+      config_id: self-check
  agents:
  - provider_id: meta-reference
    provider_type: inline::meta-reference
@ -42,51 +45,41 @@ providers:
  - provider_id: meta-reference
    provider_type: inline::meta-reference
    config:
-      service_name: ${env.OTEL_SERVICE_NAME:llama-stack}
+      service_name: "${env.OTEL_SERVICE_NAME:\u200B}"
      sinks: ${env.TELEMETRY_SINKS:console,sqlite}
      sqlite_db_path: ${env.SQLITE_DB_PATH:~/.llama/distributions/nvidia/trace_store.db}
  eval:
  - provider_id: meta-reference
    provider_type: inline::meta-reference
-    config: {}
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/meta_reference_eval.db
+  post_training:
+  - provider_id: nvidia
+    provider_type: remote::nvidia
+    config:
+      api_key: ${env.NVIDIA_API_KEY:}
+      dataset_namespace: ${env.NVIDIA_DATASET_NAMESPACE:default}
+      project_id: ${env.NVIDIA_PROJECT_ID:test-project}
+      customizer_url: ${env.NVIDIA_CUSTOMIZER_URL:http://nemo.test}
  datasetio:
-  - provider_id: huggingface
-    provider_type: remote::huggingface
-    config: {}
  - provider_id: localfs
    provider_type: inline::localfs
-    config: {}
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nvidia}/localfs_datasetio.db
  scoring:
  - provider_id: basic
    provider_type: inline::basic
    config: {}
-  - provider_id: llm-as-judge
-    provider_type: inline::llm-as-judge
-    config: {}
-  - provider_id: braintrust
-    provider_type: inline::braintrust
-    config:
-      openai_api_key: ${env.OPENAI_API_KEY:}
  tool_runtime:
-  - provider_id: brave-search
-    provider_type: remote::brave-search
-    config:
-      api_key: ${env.BRAVE_SEARCH_API_KEY:}
-      max_results: 3
-  - provider_id: tavily-search
-    provider_type: remote::tavily-search
-    config:
-      api_key: ${env.TAVILY_SEARCH_API_KEY:}
-      max_results: 3
-  - provider_id: code-interpreter
-    provider_type: inline::code-interpreter
-    config: {}
  - provider_id: rag-runtime
    provider_type: inline::rag-runtime
    config: {}
-  - provider_id: model-context-protocol
-    provider_type: remote::model-context-protocol
-    config: {}
  preprocessing:
  - provider_id: basic
    provider_type: inline::basic
@ -222,12 +215,8 @@ datasets: []
 scoring_fns: []
 benchmarks: []
 tool_groups:
- toolgroup_id: builtin::websearch
-  provider_id: tavily-search
 - toolgroup_id: builtin::rag
  provider_id: rag-runtime
- toolgroup_id: builtin::code_interpreter
-  provider_id: code-interpreter
 preprocessors:
 - preprocessor_id: builtin::basic
  provider_id: basic