fix: use OLLAMA_URL to activate Ollama provider in starter (#2963)

We tried to always keep Ollama enabled. However doing so makes the
provider implementation half-assed -- should it error when it cannot
connect to Ollama or not? What happens during periodic model refresh?
Etc. Instead do the same thing we do for vLLM -- use the `OLLAMA_URL` to
conditionally enable the provider.

## Test Plan

Run `uv run llama stack build --template starter --image-type venv
--run` with and without `OLLAMA_URL` set. Verify using
`llama-stack-client provider list` that ollama is correctly enabled.

llama_stack/templates/ci-tests/run.yaml

@@ -19,7 +19,7 @@ providers:
     config:
       base_url: https://api.cerebras.ai
       api_key: ${env.CEREBRAS_API_KEY:=}
-  - provider_id: ollama
+  - provider_id: ${env.OLLAMA_URL:+ollama}
     provider_type: remote::ollama
     config:
       url: ${env.OLLAMA_URL:=http://localhost:11434}

llama_stack/templates/starter/run.yaml

@@ -19,7 +19,7 @@ providers:
     config:
       base_url: https://api.cerebras.ai
       api_key: ${env.CEREBRAS_API_KEY:=}
-  - provider_id: ollama
+  - provider_id: ${env.OLLAMA_URL:+ollama}
     provider_type: remote::ollama
     config:
       url: ${env.OLLAMA_URL:=http://localhost:11434}

llama_stack/templates/starter/starter.py

@@ -66,6 +66,7 @@ ENABLED_INFERENCE_PROVIDERS = [
 ]
 
 INFERENCE_PROVIDER_IDS = {
+    "ollama": "${env.OLLAMA_URL:+ollama}",
     "vllm": "${env.VLLM_URL:+vllm}",
     "tgi": "${env.TGI_URL:+tgi}",
     "cerebras": "${env.CEREBRAS_API_KEY:+cerebras}",