Ollama Setup

.gitignore

@@ -18,3 +18,7 @@ Package.resolved
 .vscode
 _build
 docs/src
+
+env/
+envs/
+.envrc

llama_stack/providers/remote/inference/ollama/ollama.py

@@ -48,10 +48,10 @@ model_aliases = [
         "llama3.1:8b-instruct-fp16",
         CoreModelId.llama3_1_8b_instruct.value,
     ),
-    build_model_alias_with_just_provider_model_id(
-        "llama3.1:8b",
-        CoreModelId.llama3_1_8b_instruct.value,
-    ),
+    # build_model_alias_with_just_provider_model_id(
+    #     "llama3.1:8b",
+    #     CoreModelId.llama3_1_8b_instruct.value,
+    # ),
     build_model_alias(
         "llama3.1:70b-instruct-fp16",
         CoreModelId.llama3_1_70b_instruct.value,
@@ -347,7 +347,7 @@ class OllamaInferenceAdapter(Inference, ModelsProtocolPrivate):
                 )
             return model
         model = await self.register_helper.register_model(model)
-        models = await self.client.ps()
+        models = await self.client.list()
         available_models = [m["model"] for m in models["models"]]
         if model.provider_resource_id not in available_models:
             raise ValueError(

llama_stack/templates/ollama/run.yaml

@@ -79,6 +79,11 @@ models:
   provider_id: ollama
   provider_model_id: null
   model_type: llm
+- metadata: {}
+  model_id: Llama-3.2-3B-Instruct
+  provider_id: ollama
+  provider_model_id: null
+  model_type: llm
 - metadata:
     embedding_dimension: 384
   model_id: all-MiniLM-L6-v2