[Bug]: KeyError: 'message'

[pavanpraneeth]

Current Behavior

When i do acheing with langchain for the first of a query say llm.predict('tell me a joke') it works and runs in 2 seconds , then when i again run the same llm.predict('tell me a joke') it throws this error. I am using AzureOpenai

57 @root_validator 58 def set_text(cls, values: Dict[str, Any]) -> Dict[str, Any]: 59 """Set the text attribute to be the contents of the message.""" ---> 60 values["text"] = values["message"].content 61 return values

KeyError: 'message'

Expected Behavior

It should typically run in in micro seconds or millis seconds

Steps To Reproduce

No response

Environment

No response

Anything else?

No response

[SimFG]

From the current problem description, I don't seem to know what caused the error.

[nikkoxgonzales]

I have this error as well, 21 days still no reply hmm.

[SimFG]

@nikkoxgonzales I tried the langchain example a few days ago and got no errors. And if you want me to help you, please give me more details.

[nikkoxgonzales]

I did the example on langchain's website for gptcache.

I am using AzureOpenAI/AzureChatOpenAI when I got this [KeyError: 'message'].

[SimFG]

@nikkoxgonzales i can't find the error. maybe it's caused the Incompatibility of Azure api. but I don't have the Azure account. you can check the difference of the openai and Azure. the following is all my demo code:

import gptcache
from gptcache import Cache
from gptcache.manager.factory import manager_factory
from gptcache.processor.pre import get_prompt
import hashlib

from langchain.cache import GPTCache
import langchain
from langchain.llms import OpenAIChat

def get_hashed_name(name):
    return hashlib.sha256(name.encode()).hexdigest()

def init_gptcache(cache_obj: Cache, llm: str):
    hashed_llm = get_hashed_name(llm)
    cache_obj.init(
        pre_embedding_func=get_prompt,
        data_manager=manager_factory(manager="map", data_dir=f"map_cache_{hashed_llm}"),
    )

langchain.llm_cache = GPTCache(init_gptcache)

print("langchain:", langchain.__version__)
print("gptcache:", gptcache.__version__)

llm = OpenAIChat()
now = time.time()
print(llm("Tell me a joke"))
print(time.time() - now)
now = time.time()
print(llm("Tell me a joke"))
print(time.time() - now)

output:

langchain: 0.0.288
gptcache: 0.1.41
xxxx/venv/lib/python3.8/site-packages/langchain/llms/openai.py:787: 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`
  warnings.warn(
Sure, here's a joke for you: 

Why don't scientists trust atoms? 

Because they make up everything!
2.1356260776519775
Sure, here's a joke for you: 

Why don't scientists trust atoms? 

Because they make up everything!
0.0004391670227050781

[SimFG]

I will close the issue. If you have other problem, please open a new issue.