Introduce Llama stack distributions (#22)

* Add distribution CLI scaffolding * More progress towards `llama distribution install` * getting closer to a distro definition, distro install + configure works * Distribution server now functioning * read existing configuration, save enums properly * Remove inference uvicorn server entrypoint and llama inference CLI command * updated dependency and client model name * Improved exception handling * local imports for faster cli * undo a typo, add a passthrough distribution * implement full-passthrough in the server * add safety adapters, configuration handling, server + clients * cleanup, moving stuff to common, nuke utils * Add a Path() wrapper at the earliest place * fixes * Bring agentic system api to toolchain Add adapter dependencies and resolve adapters using a topological sort * refactor to reduce size of `agentic_system` * move straggler files and fix some important existing bugs * ApiSurface -> Api * refactor a method out * Adapter -> Provider * Make each inference provider into its own subdirectory * installation fixes * Rename Distribution -> DistributionSpec, simplify RemoteProviders * dict key instead of attr * update inference config to take model and not model_dir * Fix passthrough streaming, send headers properly not part of body :facepalm * update safety to use model sku ids and not model dirs * Update cli_reference.md * minor fixes * add DistributionConfig, fix a bug in model download * Make install + start scripts do proper configuration automatically * Update CLI_reference * Nuke fp8_requirements, fold fbgemm into common requirements * Update README, add newline between API surface configurations * Refactor download functionality out of the Command so can be reused * Add `llama model download` alias for `llama download` * Show message about checksum file so users can check themselves * Simpler intro statements * get ollama working * Reduce a bunch of dependencies from toolchain Some improvements to the distribution install script * Avoid using `conda run` since it buffers everything * update dependencies and rely on LLAMA_TOOLCHAIN_DIR for dev purposes * add validation for configuration input * resort imports * make optional subclasses default to yes for configuration * Remove additional_pip_packages; move deps to providers * for inline make 8b model the default * Add scripts to MANIFEST * allow installing from test.pypi.org * Fix #2 to help with testing packages * Must install llama-models at that same version first * fix PIP_ARGS --------- Co-authored-by: Hardik Shah <hjshah@fb.com> Co-authored-by: Hardik Shah <hjshah@meta.com>
2024-08-08 13:38:41 -07:00 · 2024-08-08 13:38:41 -07:00 · e830814399
commit e830814399
parent da4645a27a
115 changed files with 5839 additions and 1120 deletions
--- a/llama_toolchain/agentic_system/tools/custom/execute.py
+++ b/llama_toolchain/agentic_system/tools/custom/execute.py
@ -0,0 +1,83 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, AsyncGenerator, List
+
+from llama_models.llama3_1.api.datatypes import StopReason, ToolResponseMessage
+
+from llama_toolchain.agentic_system.api import (
+    AgenticSystem,
+    AgenticSystemTurnCreateRequest,
+    AgenticSystemTurnResponseEventType as EventType,
+)
+
+from llama_toolchain.inference.api import Message
+
+
+async def execute_with_custom_tools(
+    system: AgenticSystem,
+    system_id: str,
+    session_id: str,
+    messages: List[Message],
+    custom_tools: List[Any],
+    max_iters: int = 5,
+    stream: bool = True,
+) -> AsyncGenerator:
+    # first create a session, or do you keep a persistent session?
+    tools_dict = {t.get_name(): t for t in custom_tools}
+
+    current_messages = messages.copy()
+    n_iter = 0
+    while n_iter < max_iters:
+        n_iter += 1
+
+        request = AgenticSystemTurnCreateRequest(
+            system_id=system_id,
+            session_id=session_id,
+            messages=current_messages,
+            stream=stream,
+        )
+
+        turn = None
+        async for chunk in system.create_agentic_system_turn(request):
+            if chunk.event.payload.event_type != EventType.turn_complete.value:
+                yield chunk
+            else:
+                turn = chunk.event.payload.turn
+
+        message = turn.output_message
+        if len(message.tool_calls) == 0:
+            yield chunk
+            return
+
+        if message.stop_reason == StopReason.out_of_tokens:
+            yield chunk
+            return
+
+        tool_call = message.tool_calls[0]
+        if tool_call.tool_name not in tools_dict:
+            m = ToolResponseMessage(
+                call_id=tool_call.call_id,
+                tool_name=tool_call.tool_name,
+                content=f"Unknown tool `{tool_call.tool_name}` was called. Try again with something else",
+            )
+            next_message = m
+        else:
+            tool = tools_dict[tool_call.tool_name]
+            result_messages = await execute_custom_tool(tool, message)
+            next_message = result_messages[0]
+
+        yield next_message
+        current_messages = [next_message]
+
+
+async def execute_custom_tool(tool: Any, message: Message) -> List[Message]:
+    result_messages = await tool.run([message])
+    assert (
+        len(result_messages) == 1
+    ), f"Expected single message, got {len(result_messages)}"
+
+    return result_messages