d79b3744cb
This change remove `BaseComponent`'s: - run_raw - run_batch_raw - run_document - run_batch_document - is_document - is_batch Each component is expected to support multiple types of inputs and a single type of output. Since we want the component to work out-of-the-box with both standardized and customized use cases, supporting multiple types of inputs are expected. At the same time, to reduce the complexity of understanding how to use a component, we restrict a component to only have a single output type. To accommodate these changes, we also refactor some components to remove their run_raw, run_batch_raw... methods, and to decide the common output interface for those components. Tests are updated accordingly. Commit changes: * Add kwargs to vector store's query * Simplify the BaseComponent * Update tests * Remove support for Python 3.8 and 3.9 * Bump version 0.3.0 * Fix github PR caching still use old environment after bumping version --------- Co-authored-by: ian <ian@cinnamon.is>
106 lines
3.3 KiB
Python
106 lines
3.3 KiB
Python
from __future__ import annotations
|
|
|
|
import logging
|
|
from typing import Type
|
|
|
|
from langchain.chat_models.base import BaseChatModel
|
|
from langchain.schema.messages import BaseMessage, HumanMessage
|
|
from theflow.base import Param
|
|
|
|
from ...base import BaseComponent
|
|
from ..base import LLMInterface
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class ChatLLM(BaseComponent):
|
|
def flow(self):
|
|
if self.inflow is None:
|
|
raise ValueError("No inflow provided.")
|
|
|
|
if not isinstance(self.inflow, BaseComponent):
|
|
raise ValueError(
|
|
f"inflow must be a BaseComponent, found {type(self.inflow)}"
|
|
)
|
|
|
|
text = self.inflow.flow().text
|
|
return self.__call__(text)
|
|
|
|
|
|
class LangchainChatLLM(ChatLLM):
|
|
_lc_class: Type[BaseChatModel]
|
|
|
|
def __init__(self, **params):
|
|
if self._lc_class is None:
|
|
raise AttributeError(
|
|
"Should set _lc_class attribute to the LLM class from Langchain "
|
|
"if using LLM from Langchain"
|
|
)
|
|
|
|
self._kwargs: dict = {}
|
|
for param in list(params.keys()):
|
|
if param in self._lc_class.__fields__:
|
|
self._kwargs[param] = params.pop(param)
|
|
super().__init__(**params)
|
|
|
|
@Param.auto(cache=False)
|
|
def agent(self) -> BaseChatModel:
|
|
return self._lc_class(**self._kwargs)
|
|
|
|
def run(
|
|
self, messages: str | BaseMessage | list[BaseMessage], **kwargs
|
|
) -> LLMInterface:
|
|
"""Generate response from messages
|
|
|
|
Args:
|
|
messages: history of messages to generate response from
|
|
**kwargs: additional arguments to pass to the langchain chat model
|
|
|
|
Returns:
|
|
LLMInterface: generated response
|
|
"""
|
|
input_: list[BaseMessage] = []
|
|
|
|
if isinstance(messages, str):
|
|
input_ = [HumanMessage(content=messages)]
|
|
elif isinstance(messages, BaseMessage):
|
|
input_ = [messages]
|
|
else:
|
|
input_ = messages
|
|
|
|
pred = self.agent.generate(messages=[input_], **kwargs)
|
|
all_text = [each.text for each in pred.generations[0]]
|
|
|
|
completion_tokens, total_tokens, prompt_tokens = 0, 0, 0
|
|
try:
|
|
if pred.llm_output is not None:
|
|
completion_tokens = pred.llm_output["token_usage"]["completion_tokens"]
|
|
total_tokens = pred.llm_output["token_usage"]["total_tokens"]
|
|
prompt_tokens = pred.llm_output["token_usage"]["prompt_tokens"]
|
|
except Exception:
|
|
logger.warning(
|
|
f"Cannot get token usage from LLM output for {self._lc_class.__name__}"
|
|
)
|
|
|
|
return LLMInterface(
|
|
text=all_text[0] if len(all_text) > 0 else "",
|
|
candidates=all_text,
|
|
completion_tokens=completion_tokens,
|
|
total_tokens=total_tokens,
|
|
prompt_tokens=prompt_tokens,
|
|
logits=[],
|
|
)
|
|
|
|
def __setattr__(self, name, value):
|
|
if name in self._lc_class.__fields__:
|
|
self._kwargs[name] = value
|
|
setattr(self.agent, name, value)
|
|
else:
|
|
super().__setattr__(name, value)
|
|
|
|
def __getattr__(self, name):
|
|
if name in self._lc_class.__fields__:
|
|
return getattr(self.agent, name)
|
|
|
|
return super().__getattr__(name) # type: ignore
|