phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-08-03 17:29:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

do not pass memory tools to inference

Browse source
This commit is contained in:
Dinesh Yeduguru 2025-01-08 18:53:32 -08:00
parent d0c8dced65
commit b46d94d87d
2 changed files with 4 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

13
llama_stack/providers/inline/agents/meta_reference/agent_instance.py
View file

@ -487,20 +487,13 @@ class ChatAgent(ShieldRunnerMixin):
stop_reason = None
with tracing.span("inference") as span:
def is_memory_group(tool):
memory_tool_group = tool_to_group.get(MEMORY_QUERY_TOOL, None)
has_memory_tool = MEMORY_QUERY_TOOL in tool_defs
return (
has_memory_tool
and tool_to_group.get(tool.tool_name, None) != memory_tool_group
)
async for chunk in await self.inference_api.chat_completion(
self.agent_config.model,
input_messages,
tools=[
tool for tool in tool_defs.values() if not is_memory_group(tool)
tool
for tool in tool_defs.values()
if tool_to_group.get(tool.tool_name, None) != MEMORY_GROUP
],
tool_prompt_format=self.agent_config.tool_prompt_format,
stream=True,

2
llama_stack/providers/inline/tool_runtime/memory/memory.py
View file

@ -60,7 +60,7 @@ class MemoryToolRuntimeImpl(ToolsProtocolPrivate, ToolRuntime):
description="Retrieve context from memory",
parameters=[
ToolParameter(
name="input_messages",
name="messages",
description="The input messages to search for",
parameter_type="array",
),