Merge branch 'meta-llama:main' into feat/litellm_sambanova_usage

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

@@ -89,7 +89,6 @@ class ChatAgent(ShieldRunnerMixin):
         self,
         agent_id: str,
         agent_config: AgentConfig,
-        tempdir: str,
         inference_api: Inference,
         safety_api: Safety,
         tool_runtime_api: ToolRuntime,
@@ -99,7 +98,6 @@ class ChatAgent(ShieldRunnerMixin):
     ):
         self.agent_id = agent_id
         self.agent_config = agent_config
-        self.tempdir = tempdir
         self.inference_api = inference_api
         self.safety_api = safety_api
         self.vector_io_api = vector_io_api

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

@@ -7,7 +7,6 @@
 import json
 import logging
 import shutil
-import tempfile
 import uuid
 from typing import AsyncGenerator, List, Optional, Union
 
@@ -64,7 +63,6 @@ class MetaReferenceAgentsImpl(Agents):
         self.tool_groups_api = tool_groups_api
 
         self.in_memory_store = InmemoryKVStoreImpl()
-        self.tempdir = tempfile.mkdtemp()
 
     async def initialize(self) -> None:
         self.persistence_store = await kvstore_impl(self.config.persistence_store)
@@ -107,7 +105,6 @@ class MetaReferenceAgentsImpl(Agents):
         return ChatAgent(
             agent_id=agent_id,
             agent_config=agent_config,
-            tempdir=self.tempdir,
             inference_api=self.inference_api,
             safety_api=self.safety_api,
             vector_io_api=self.vector_io_api,

llama_stack/providers/inline/inference/meta_reference/generators.py

@@ -259,7 +259,7 @@ class Llama3Generator:
 
         temperature, top_p = _infer_sampling_params(sampling_params)
         for result in self.inner_generator.generate(
-            llm_inputs=[self.formatter.encode_content(request.content)],
+            model_inputs=[self.formatter.encode_content(request.content)],
             max_gen_len=max_gen_len,
             temperature=temperature,
             top_p=top_p,
@@ -284,7 +284,7 @@ class Llama3Generator:
 
         temperature, top_p = _infer_sampling_params(sampling_params)
         for result in self.inner_generator.generate(
-            llm_inputs=[self.formatter.encode_dialog_prompt(request.messages, _infer_tool_prompt_format(request))],
+            model_inputs=[self.formatter.encode_dialog_prompt(request.messages, _infer_tool_prompt_format(request))],
             max_gen_len=max_gen_len,
             temperature=temperature,
             top_p=top_p,

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

@@ -307,9 +307,10 @@ class OllamaInferenceAdapter(Inference, ModelsProtocolPrivate):
         if model.model_type == ModelType.embedding:
             logger.info(f"Pulling embedding model `{model.provider_resource_id}` if necessary...")
             await self.client.pull(model.provider_resource_id)
-            response = await self.client.list()
-        else:
-            response = await self.client.ps()
+        # we use list() here instead of ps() -
+        #  - ps() only lists running models, not available models
+        #  - models not currently running are run by the ollama server as needed
+        response = await self.client.list()
         available_models = [m["model"] for m in response["models"]]
         if model.provider_resource_id not in available_models:
             raise ValueError(