fix: add missing files provider to NVIDIA distribution (#3479)

# What does this PR do?
The rag-runtime tool requires files API as a dependency, but the NVIDIA
distribution was missing the files provider configuration. Thus, when
running:

```
llama stack build --distro nvidia --image-type venv
```
And then:
```
llama stack run {path_to_distribution_config} --image-type venv
```
It would raise an error:
```
RuntimeError: Failed to resolve 'tool_runtime' provider 'rag-runtime' of type 'inline::rag-runtime': required dependency 'files' is not available. Please add a 'files' provider to your configuration or check if the provider is properly configured.
```

This PR fixes the issue by adding missing files provider to NVIDIA
distribution.

## Test Plan
N/A

docs/source/distributions/self_hosted_distro/nvidia.md

@@ -11,6 +11,7 @@ The `llamastack/distribution-nvidia` distribution consists of the following prov
 | agents | `inline::meta-reference` |
 | datasetio | `inline::localfs`, `remote::nvidia` |
 | eval | `remote::nvidia` |
+| files | `inline::localfs` |
 | inference | `remote::nvidia` |
 | post_training | `remote::nvidia` |
 | safety | `remote::nvidia` |

llama_stack/distributions/nvidia/build.yaml

@@ -23,6 +23,8 @@ distribution_spec:
     - provider_type: inline::basic
     tool_runtime:
     - provider_type: inline::rag-runtime
+    files:
+    - provider_type: inline::localfs
 image_type: venv
 additional_pip_packages:
 - aiosqlite

llama_stack/distributions/nvidia/nvidia.py

@@ -8,6 +8,7 @@ from pathlib import Path
 
 from llama_stack.core.datatypes import BuildProvider, ModelInput, Provider, ShieldInput, ToolGroupInput
 from llama_stack.distributions.template import DistributionTemplate, RunConfigSettings, get_model_registry
+from llama_stack.providers.inline.files.localfs.config import LocalfsFilesImplConfig
 from llama_stack.providers.remote.datasetio.nvidia import NvidiaDatasetIOConfig
 from llama_stack.providers.remote.eval.nvidia import NVIDIAEvalConfig
 from llama_stack.providers.remote.inference.nvidia import NVIDIAConfig
@@ -15,7 +16,7 @@ from llama_stack.providers.remote.inference.nvidia.models import MODEL_ENTRIES
 from llama_stack.providers.remote.safety.nvidia import NVIDIASafetyConfig
 
 
-def get_distribution_template() -> DistributionTemplate:
+def get_distribution_template(name: str = "nvidia") -> DistributionTemplate:
     providers = {
         "inference": [BuildProvider(provider_type="remote::nvidia")],
         "vector_io": [BuildProvider(provider_type="inline::faiss")],
@@ -30,6 +31,7 @@ def get_distribution_template() -> DistributionTemplate:
         ],
         "scoring": [BuildProvider(provider_type="inline::basic")],
         "tool_runtime": [BuildProvider(provider_type="inline::rag-runtime")],
+        "files": [BuildProvider(provider_type="inline::localfs")],
     }
 
     inference_provider = Provider(
@@ -52,6 +54,11 @@ def get_distribution_template() -> DistributionTemplate:
         provider_type="remote::nvidia",
         config=NVIDIAEvalConfig.sample_run_config(),
     )
+    files_provider = Provider(
+        provider_id="meta-reference-files",
+        provider_type="inline::localfs",
+        config=LocalfsFilesImplConfig.sample_run_config(f"~/.llama/distributions/{name}"),
+    )
     inference_model = ModelInput(
         model_id="${env.INFERENCE_MODEL}",
         provider_id="nvidia",
@@ -73,7 +80,7 @@ def get_distribution_template() -> DistributionTemplate:
 
     default_models, _ = get_model_registry(available_models)
     return DistributionTemplate(
-        name="nvidia",
+        name=name,
         distro_type="self_hosted",
         description="Use NVIDIA NIM for running LLM inference, evaluation and safety",
         container_image=None,
@@ -86,6 +93,7 @@ def get_distribution_template() -> DistributionTemplate:
                     "inference": [inference_provider],
                     "datasetio": [datasetio_provider],
                     "eval": [eval_provider],
+                    "files": [files_provider],
                 },
                 default_models=default_models,
                 default_tool_groups=default_tool_groups,
@@ -97,6 +105,7 @@ def get_distribution_template() -> DistributionTemplate:
                         safety_provider,
                     ],
                     "eval": [eval_provider],
+                    "files": [files_provider],
                 },
                 default_models=[inference_model, safety_model],
                 default_shields=[ShieldInput(shield_id="${env.SAFETY_MODEL}", provider_id="nvidia")],

llama_stack/distributions/nvidia/run-with-safety.yaml

@@ -4,6 +4,7 @@ apis:
 - agents
 - datasetio
 - eval
+- files
 - inference
 - post_training
 - safety
@@ -88,6 +89,14 @@ providers:
   tool_runtime:
   - provider_id: rag-runtime
     provider_type: inline::rag-runtime
+  files:
+  - provider_id: meta-reference-files
+    provider_type: inline::localfs
+    config:
+      storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/distributions/nvidia/files}
+      metadata_store:
+        type: sqlite
+        db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/files_metadata.db
 metadata_store:
   type: sqlite
   db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/registry.db

llama_stack/distributions/nvidia/run.yaml

@@ -4,6 +4,7 @@ apis:
 - agents
 - datasetio
 - eval
+- files
 - inference
 - post_training
 - safety
@@ -77,6 +78,14 @@ providers:
   tool_runtime:
   - provider_id: rag-runtime
     provider_type: inline::rag-runtime
+  files:
+  - provider_id: meta-reference-files
+    provider_type: inline::localfs
+    config:
+      storage_dir: ${env.FILES_STORAGE_DIR:=~/.llama/distributions/nvidia/files}
+      metadata_store:
+        type: sqlite
+        db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/files_metadata.db
 metadata_store:
   type: sqlite
   db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/registry.db