Exception thrown when calling the invoke() method on chat object of class ChatSnowflakeCortex

[sfc-gh-clakkad]

Checked other resources

• [X] I added a very descriptive title to this issue.
• [X] I searched the LangChain documentation with the integrated search.
• [X] I used the GitHub search to find a similar question and didn't find it.
• [X] I am sure that this is a bug in LangChain rather than my code.
• [X] The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

Issue reproduced using the Snowflake Cortex integration tutorial on the official documentation here : https://python.langchain.com/v0.2/docs/integrations/chat/snowflake/

during generation at the line chat.invoke(messages)

This is from the langchain-community package and not the core langchain.

Error Message and Stack Trace (if applicable)

` ... File ~/opt/anaconda3/envs/cortex_base/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py:624, in BaseChatModel.generate(self, messages, stop, callbacks, tags, metadata, run_name, run_id, kwargs) 621 for i, m in enumerate(messages): 622 try: 623 results.append( --> 624 self._generate_with_cache( 625 m, 626 stop=stop, 627 run_manager=run_managers[i] if run_managers else None, 628 kwargs, 629 ) 630 ) 631 except BaseException as e: 632 if run_managers:

File ~/opt/anaconda3/envs/cortex_base/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py:846, in BaseChatModel._generate_with_cache(self, messages, stop, run_manager, kwargs) 844 else: 845 if inspect.signature(self._generate).parameters.get("run_manager"): --> 846 result = self._generate( 847 messages, stop=stop, run_manager=run_manager, kwargs 848 ) 849 else: 850 result = self._generate(messages, stop=stop, **kwargs)

File ~/opt/anaconda3/envs/cortex_base/lib/python3.10/site-packages/langchain_community/chat_models/snowflake.py:220, in ChatSnowflakeCortex._generate(self, messages, stop, run_manager, **kwargs) 218 l_rows = self._sp_session.sql(sql_stmt).collect() 219 except Exception as e: --> 220 raise ChatSnowflakeCortexError( 221 f"Error while making request to Snowflake Cortex via Snowpark: {e}" 222 ) 224 response = json.loads(l_rows[0]["LLM_RESPONSE"]) 225 ai_message_content = response["choices"][0]["messages"]

ChatSnowflakeCortexError: Error while making request to Snowflake Cortex via Snowpark: 'NoneType' object has no attribute 'sql'`

Description

• Ideally the invoke() method should be able to return a completion as returned by Snowflake Cortex Complete function.
• The issue seems to be that "_sp_session" being used at invocation isnt set to the actual value grabbed during validation. Hence the sql() method supported on Snowpark session throws the above None error as by default _sp_session is set to None.

System Info

langchain==0.2.16 langchain-community==0.2.16 langchain-core==0.2.38 langchain-text-splitters==0.2.4

platform: mac python==3.10

[KuroAshi75]

facing similar issue

[NabilShabab]

facing similar issue..

[markns]

I'm not sure why yet, but making _sp_session a public field in the ChatSnowflakeCortex model makes this problem go away. Just replace _sp_session with sp_session everywhere in snowflake.py

[ritmar]

The public variable gets initialized on calling ChatSnowflakeCortex, but the internal variables initialization is not enforced (the one starting with _) , renaming _sp_session to sp_session enforces the initialization of the variable, i.e., snowflake session gets created in this case.