Run the script to produce vllm outputs

llama_stack/providers/inline/agents/meta_reference/config.py

@@ -21,5 +21,5 @@ class MetaReferenceAgentsImplConfig(BaseModel):
             "persistence_store": SqliteKVStoreConfig.sample_run_config(
                 __distro_dir__=__distro_dir__,
                 db_name="agents_store.db",
-            ).model_dump(),
+            )
         }

llama_stack/providers/inline/memory/faiss/config.py

@@ -25,5 +25,5 @@ class FaissImplConfig(BaseModel):
             "kvstore": SqliteKVStoreConfig.sample_run_config(
                 __distro_dir__=__distro_dir__,
                 db_name="faiss_store.db",
-            ).model_dump(),
+            )
         }

llama_stack/providers/utils/kvstore/config.py

@@ -61,8 +61,9 @@ class SqliteKVStoreConfig(CommonConfig):
             "type": "sqlite",
             "namespace": None,
             "db_path": "${env.SQLITE_STORE_DIR:~/.llama/"
-            + f"{__distro_dir__}/{db_name}"
-            + "}",
+            + __distro_dir__
+            + "}/"
+            + db_name,
         }
 
 

llama_stack/scripts/distro_codegen.py

@@ -50,6 +50,7 @@ def process_template(template_dir: Path, progress) -> None:
 
     except Exception as e:
         progress.print(f"[red]Error processing {template_dir.name}: {str(e)}")
+        raise e
 
 
 def main():

llama_stack/templates/remote-vllm/build.yaml

@@ -1,12 +1,19 @@
+version: '2'
 name: remote-vllm
 distribution_spec:
   description: Use (an external) vLLM server for running LLM inference
+  docker_image: llamastack/distribution-remote-vllm:test-0.0.52rc3
   providers:
-    inference: remote::vllm
+    inference:
+    - remote::vllm
     memory:
     - inline::faiss
     - remote::chromadb
     - remote::pgvector
-    safety: inline::llama-guard
-    agents: inline::meta-reference
-    telemetry: inline::meta-reference
+    safety:
+    - inline::llama-guard
+    agents:
+    - inline::meta-reference
+    telemetry:
+    - inline::meta-reference
+image_type: conda

llama_stack/templates/remote-vllm/vllm.py

@@ -41,6 +41,7 @@ def get_distribution_template() -> DistributionTemplate:
         name="remote-vllm",
         distro_type="self_hosted",
         description="Use (an external) vLLM server for running LLM inference",
+        docker_image="llamastack/distribution-remote-vllm:test-0.0.52rc3",
         template_path=Path(__file__).parent / "doc_template.md",
         providers=providers,
         default_models=[inference_model, safety_model],

llama_stack/templates/template.py

@@ -91,7 +91,7 @@ class RunConfigSettings(BaseModel):
             apis=list(apis),
             providers=provider_configs,
             metadata_store=SqliteKVStoreConfig.sample_run_config(
-                dir=f"distributions/{name}",
+                __distro_dir__=f"distributions/{name}",
                 db_name="registry.db",
             ),
             models=self.default_models,