Langchain Async explained. Make multiple OpenAI chatgpt API calls at the same time video files
Added 2023-03-06 01:07:30 +0000 UTCCode files are for video: https://youtu.be/eAikW9o1Ros
MAKING PARALLEL CALLS TO THE LLM WILL COST EXTRA TOKENS!!
Async is important because you can run multiple calls to the OpenAI API at the same time instead of one by one. Therefore it results in significant speed gains when you want to make multiple calls at the same time.
Relevant links:
langchain async blog post: https://blog.langchain.dev/async-api/
miniconda: https://docs.conda.io/en/latest/miniconda.html
Python Async walkthrough: https://realpython.com/async-io-python/#setting-up-your-environment
langchain async for llm calls: https://langchain.readthedocs.io/en/latest/modules/llms/async_llm.html
langchain async for chain calls: https://langchain.readthedocs.io/en/latest/modules/chains/async_chain.html
lanchain async for agent calls: https://langchain.readthedocs.io/en/latest/modules/agents/examples/async_agent.html
wikipedia article about Kaprekar's constant: https://en.wikipedia.org/wiki/6174_(number)
Comments
Thank you for letting me know. I was thinking of going over old projects when I find some time. Maybe I will make live streams while updating the code
Echo Hive
2023-05-30 10:30:56 +0000 UTCI've tried to replace llm = OpenAIChat(temperature=0.9) with llm = ChatOpenAI(temperature=0.9, model_name="gpt-3.5-turbo", openai_api_key=os.getenv('OPENAI_API_KEY')) After adding from langchain.chat_models import ChatOpenAI I also confirmed my API key is read correctly. But it seems it's not that simple. Now I'm getting File "/opt/homebrew/lib/python3.10/site-packages/langchain/chat_models/base.py", line 119, in agenerate results = await asyncio.gather( File "/opt/homebrew/lib/python3.10/site-packages/langchain/chat_models/openai.py", line 359, in _agenerate message_dicts, params = self._create_message_dicts(messages, stop) File "/opt/homebrew/lib/python3.10/site-packages/langchain/chat_models/openai.py", line 341, in _create_message_dicts message_dicts = [_convert_message_to_dict(m) for m in messages] File "/opt/homebrew/lib/python3.10/site-packages/langchain/chat_models/openai.py", line 341, in message_dicts = [_convert_message_to_dict(m) for m in messages] File "/opt/homebrew/lib/python3.10/site-packages/langchain/chat_models/openai.py", line 117, in _convert_message_to_dict raise ValueError(f"Got unknown type {message}") ValueError: Got unknown type T
vesper8
2023-05-30 09:12:47 +0000 UTCAlso getting this warning when running llm.py with instructions. Not sure if you'd want to update the scripts. It's still working though UserWarning: You are trying to use a chat model. This way of initializing it is no longer supported. Instead, please use: `from langchain.chat_models import ChatOpenAI`
vesper8
2023-05-30 09:01:26 +0000 UTCThanks for letting me know. I will try to find the new ones and update the links
Echo Hive
2023-05-30 08:09:31 +0000 UTCJust want to point out that the three readthedocs.io links are now 404s
vesper8
2023-05-30 08:08:45 +0000 UTC
