Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: use async lc tracer instead of run_sync #1529

Merged
merged 2 commits into from
Nov 19, 2024
Merged

feat: use async lc tracer instead of run_sync #1529

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
60cd178
feat: use async lc tracer instead of run_sync
willydouhard Nov 18, 2024
3cc6ace
Add fallback for pydantic v1.
dokterbob Nov 19, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
101 changes: 55 additions & 46 deletions backend/chainlit/langchain/callbacks.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,17 +3,19 @@
from typing import Any, Dict, List, Optional, Sequence, Tuple, TypedDict, Union
from uuid import UUID

from chainlit.context import context_var
from chainlit.message import Message
from chainlit.step import Step
from langchain.callbacks.tracers.base import BaseTracer
import pydantic
from langchain.callbacks.tracers.schemas import Run
from langchain.schema import BaseMessage
from langchain_core.outputs import ChatGenerationChunk, GenerationChunk
from langchain_core.tracers.base import AsyncBaseTracer
from literalai import ChatGeneration, CompletionGeneration, GenerationMessage
from literalai.helper import utc_now
from literalai.observability.step import TrueStepType

from chainlit.context import context_var
from chainlit.message import Message
from chainlit.step import Step

DEFAULT_ANSWER_PREFIX_TOKENS = ["Final", "Answer", ":"]


Expand Down Expand Up @@ -122,6 +124,14 @@ def ensure_values_serializable(self, data):
key: self.ensure_values_serializable(value)
for key, value in data.items()
}
elif isinstance(data, pydantic.BaseModel):
# Fallback to support pydantic v1
# https://docs.pydantic.dev/latest/migration/#changes-to-pydanticbasemodel
if pydantic.VERSION.startswith("1"):
return data.dict()

# pydantic v2
return data.model_dump() # pyright: ignore reportAttributeAccessIssue
elif isinstance(data, list):
return [self.ensure_values_serializable(item) for item in data]
elif isinstance(data, (str, int, float, bool, type(None))):
Expand Down Expand Up @@ -249,7 +259,7 @@ def process_content(content: Any) -> Tuple[Dict, Optional[str]]:
DEFAULT_TO_KEEP = ["retriever", "llm", "agent", "chain", "tool"]


class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
class LangchainTracer(AsyncBaseTracer, GenerationHelper, FinalStreamHelper):
steps: Dict[str, Step]
parent_id_map: Dict[str, str]
ignored_runs: set
Expand All @@ -268,7 +278,7 @@ def __init__(
to_keep: Optional[List[str]] = None,
**kwargs: Any,
) -> None:
BaseTracer.__init__(self, **kwargs)
AsyncBaseTracer.__init__(self, **kwargs)
GenerationHelper.__init__(self)
FinalStreamHelper.__init__(
self,
Expand Down Expand Up @@ -296,7 +306,7 @@ def __init__(
else:
self.to_keep = to_keep

def on_chat_model_start(
async def on_chat_model_start(
self,
serialized: Dict[str, Any],
messages: List[List[BaseMessage]],
Expand All @@ -305,8 +315,9 @@ def on_chat_model_start(
parent_run_id: Optional["UUID"] = None,
tags: Optional[List[str]] = None,
metadata: Optional[Dict[str, Any]] = None,
name: Optional[str] = None,
**kwargs: Any,
) -> Any:
) -> Run:
lc_messages = messages[0]
self.chat_generations[str(run_id)] = {
"input_messages": lc_messages,
Expand All @@ -315,54 +326,63 @@ def on_chat_model_start(
"tt_first_token": None,
}

return super().on_chat_model_start(
return await super().on_chat_model_start(
serialized,
messages,
run_id=run_id,
parent_run_id=parent_run_id,
tags=tags,
metadata=metadata,
name=name,
**kwargs,
)

def on_llm_start(
async def on_llm_start(
self,
serialized: Dict[str, Any],
prompts: List[str],
*,
run_id: "UUID",
parent_run_id: Optional[UUID] = None,
tags: Optional[List[str]] = None,
parent_run_id: Optional["UUID"] = None,
metadata: Optional[Dict[str, Any]] = None,
name: Optional[str] = None,
**kwargs: Any,
) -> Run:
self.completion_generations[str(run_id)] = {
"prompt": prompts[0],
"start": time.time(),
"token_count": 0,
"tt_first_token": None,
}
return super().on_llm_start(
) -> None:
await super().on_llm_start(
serialized,
prompts,
run_id=run_id,
parent_run_id=parent_run_id,
tags=tags,
metadata=metadata,
name=name,
**kwargs,
)

def on_llm_new_token(
self.completion_generations[str(run_id)] = {
"prompt": prompts[0],
"start": time.time(),
"token_count": 0,
"tt_first_token": None,
}

return None

async def on_llm_new_token(
self,
token: str,
*,
chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
run_id: "UUID",
parent_run_id: Optional["UUID"] = None,
**kwargs: Any,
) -> Run:
) -> None:
await super().on_llm_new_token(
token=token,
chunk=chunk,
run_id=run_id,
parent_run_id=parent_run_id,
**kwargs,
)
if isinstance(chunk, ChatGenerationChunk):
start = self.chat_generations[str(run_id)]
else:
Expand All @@ -377,24 +397,13 @@ def on_llm_new_token(
if self.answer_reached:
if not self.final_stream:
self.final_stream = Message(content="")
self._run_sync(self.final_stream.send())
self._run_sync(self.final_stream.stream_token(token))
await self.final_stream.send()
await self.final_stream.stream_token(token)
self.has_streamed_final_answer = True
else:
self.answer_reached = self._check_if_answer_reached()

return super().on_llm_new_token(
token,
chunk=chunk,
run_id=run_id,
parent_run_id=parent_run_id,
)

def _run_sync(self, co): # TODO: WHAT TO DO WITH THIS?
context_var.set(self.context)
self.context.loop.create_task(co)

def _persist_run(self, run: Run) -> None:
async def _persist_run(self, run: Run) -> None:
pass

def _get_run_parent_id(self, run: Run):
Expand Down Expand Up @@ -445,8 +454,8 @@ def _should_ignore_run(self, run: Run):
self.ignored_runs.add(str(run.id))
return ignore, parent_id

def _start_trace(self, run: Run) -> None:
super()._start_trace(run)
async def _start_trace(self, run: Run) -> None:
await super()._start_trace(run)
context_var.set(self.context)

ignore, parent_id = self._should_ignore_run(run)
Expand Down Expand Up @@ -489,9 +498,9 @@ def _start_trace(self, run: Run) -> None:

self.steps[str(run.id)] = step

self._run_sync(step.send())
await step.send()

def _on_run_update(self, run: Run) -> None:
async def _on_run_update(self, run: Run) -> None:
"""Process a run upon update."""
context_var.set(self.context)

Expand Down Expand Up @@ -576,10 +585,10 @@ def _on_run_update(self, run: Run) -> None:

if current_step:
current_step.end = utc_now()
self._run_sync(current_step.update())
await current_step.update()

if self.final_stream and self.has_streamed_final_answer:
self._run_sync(self.final_stream.update())
await self.final_stream.update()

return

Expand All @@ -599,16 +608,16 @@ def _on_run_update(self, run: Run) -> None:
else output
)
current_step.end = utc_now()
self._run_sync(current_step.update())
await current_step.update()

def _on_error(self, error: BaseException, *, run_id: UUID, **kwargs: Any):
async def _on_error(self, error: BaseException, *, run_id: UUID, **kwargs: Any):
context_var.set(self.context)

if current_step := self.steps.get(str(run_id), None):
current_step.is_error = True
current_step.output = str(error)
current_step.end = utc_now()
self._run_sync(current_step.update())
await current_step.update()

on_llm_error = _on_error
on_chain_error = _on_error
Expand Down