issue_owner_repo listlengths 2 2 | issue_body stringlengths 0 261k ⌀ | issue_title stringlengths 1 925 | issue_comments_url stringlengths 56 81 | issue_comments_count int64 0 2.5k | issue_created_at stringlengths 20 20 | issue_updated_at stringlengths 20 20 | issue_html_url stringlengths 37 62 | issue_github_id int64 387k 2.46B | issue_number int64 1 127k |
|---|---|---|---|---|---|---|---|---|---|
[
"langchain-ai",
"langchain"
] | ### 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
If I use my fact checker:
```
import sys
import os
current_dir = os.path.dirname(os.path.abspath(__file__))
parent_dir = os.path.dirname(current_dir)
sys.path.insert(0, parent_dir)
sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
from llm_utils import get_prompt, AllowedModels, load_chat_model, get_chat_prompt, load_llm_model
from typing import List, Dict
from fact_check.checker import FactChecker
from langchain_core.output_parsers import StrOutputParser, PydanticOutputParser
from langchain_core.output_parsers.openai_tools import JsonOutputToolsParser
from models import QuestionSet, Question, FactCheckQuestion, StandardObject
import asyncio
from langchain.globals import set_debug
set_debug(False)
def parse_results(result):
fact_check = result[0][0]['args'] # Sometimes it doesnt use the freaking json parser
return {
'question': fact_check['question'],
'answer': fact_check['answer'],
'category': fact_check['category'],
'explanation': fact_check['explanation'],
'fact_check': fact_check['fact_check']
}
async def gather_tasks(tasks):
return await asyncio.gather(*tasks)
async def afact_checker(question: Question) -> FactCheckQuestion:
"""
Uses an OpenAI model to generate a list of questions for each category.
:param model: The model to use for question generation.
:param categories: A list of categories to generate questions for.
:return:
"""
fact_check = FactChecker(question.question, question.answer)
response = fact_check._get_answer()
model = AllowedModels('gpt-4')
prompt = get_chat_prompt('fact_checking')
llm = load_chat_model(model)
llm = llm.bind_tools([FactCheckQuestion])
parser = JsonOutputToolsParser()
chain = prompt['prompt'] | llm | parser # now lets the use the perplexity model to assert if the answer is correct
actively_grading = []
task = chain.ainvoke({
'question': question.question,
'answer': question.answer,
'category': question.category,
'findings': response,
})
actively_grading.append(task)
results = await asyncio.gather(*actively_grading)
parsed_results = parse_results(results)
return FactCheckQuestion(**parsed_results)
if __name__ == '__main__':
loop = asyncio.get_event_loop()
result = loop.run_until_complete(afact_checker(Question(question="What is the capital of Nigeria?",
answer="Abuja",
category="Geography",
difficulty="hard")))
loop.close()
print(result)
```
It returns an error 50% of the time due to the fact that the return from the ChatModel sometimes uses the tool and sometimes doesn't. I'm afraid I cant share my prompt, but its a pretty simply system and user prompt that makes no mention of how it should be structured as an output.
Here are two examples of returns from the same code:
```
# using the bound tool
{
"generations": [
[
{
"text": "",
"generation_info": {
"finish_reason": "tool_calls",
"logprobs": null
},
"type": "ChatGeneration",
"message": {
"lc": 1,
"type": "constructor",
"id": [
"langchain",
"schema",
"messages",
"AIMessage"
],
"kwargs": {
"content": "",
"additional_kwargs": {
"tool_calls": [
{
"id": "call_FetSvBCClds7wRDu7oEpfOD3",
"function": {
"arguments": "{\n\"question\": \"What is the capital of Nigeria?\",\n\"answer\": \"Abuja\",\n\"category\": \"Geography\",\n\"fact_check\": true,\n\"explanation\": \"correct\"\n}",
"name": "FactCheckQuestion"
},
"type": "function"
}
]
}
}
}
}
]
],
"llm_output": {
"token_usage": {
"completion_tokens": 48,
"prompt_tokens": 311,
"total_tokens": 359
},
"model_name": "gpt-4",
"system_fingerprint": null
},
"run": null
}
# forgoing the bound tool
{
"generations": [
[
{
"text": "{ \"question\": \"What is the capital of Nigeria?\", \"answer\": \"Abuja\", \"category\": \"Geography\", \"fact_check\": true, \"explanation\": \"correct\" }",
"generation_info": {
"finish_reason": "stop",
"logprobs": null
},
"type": "ChatGeneration",
"message": {
"lc": 1,
"type": "constructor",
"id": [
"langchain",
"schema",
"messages",
"AIMessage"
],
"kwargs": {
"content": "{ \"question\": \"What is the capital of Nigeria?\", \"answer\": \"Abuja\", \"category\": \"Geography\", \"fact_check\": true, \"explanation\": \"correct\" }",
"additional_kwargs": {}
}
}
}
]
],
"llm_output": {
"token_usage": {
"completion_tokens": 41,
"prompt_tokens": 311,
"total_tokens": 352
},
"model_name": "gpt-4",
"system_fingerprint": null
},
"run": null
}
```
There is no difference between these two runs. I simply called `chain.invoke{..}` twice.
Is there a way to __force__ the ChatModel to use the bound tool?
### Error Message and Stack Trace (if applicable)
_No response_
### Description
If I call the `invoke` function twice on a Pydantic tool bound `ChatModel` It alternates between using the tool to return a JSON object and returning raw text.
### System Info
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 23.2.0: Wed Nov 15 21:55:06 PST 2023; root:xnu-10002.61.3~2/RELEASE_ARM64_T6020
> Python Version: 3.11.7 (main, Dec 15 2023, 12:09:04) [Clang 14.0.6 ]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.25
> langchain_anthropic: 0.1.4
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
| Chat Agent doesn't always use bound tools for JsonOutputParser | https://api.github.com/repos/langchain-ai/langchain/issues/19474/comments | 1 | 2024-03-24T01:04:57Z | 2024-07-01T16:06:19Z | https://github.com/langchain-ai/langchain/issues/19474 | 2,204,122,965 | 19,474 |
[
"langchain-ai",
"langchain"
] | ### Example Code
1. Initialize a QdrantClient with `prefer_grpc=True`:
```python
class SparseVectorStore(ValidateQdrantClient):
...
self.client = QdrantClient(
url=os.getenv("QDRANT_URL"),
api_key=os.getenv("QDRANT_API_KEY"),
prefer_grpc=True,
)
...
```
2. Pass `self.client` to `QdrantSparseVectorRetriever`.
```python
def create_sparse_retriever(self):
...
return QdrantSparseVectorRetriever(
client=self.client,
collection_name=self.collection_name,
sparse_vector_name=self.vector_name,
sparse_encoder=self.sparse_encoder,
k=self.k,
)
```
### Error Message
A ValidationError is thrown with the message:
```
pydantic.v1.error_wrappers.ValidationError: 1 validation error for QdrantSparseVectorRetriever __root__ argument of type 'NoneType' is not iterable (type=type_error)
```
### Description
When initializing a `QdrantClient` with `prefer_grpc=True` and passing it to `QdrantSparseVectorRetriever`, a `ValidationError`is thrown. The error does not occur when `prefer_grpc=False`.
### Expected Behavior
`QdrantSparseVectorRetriever` should be initialized without any errors. I am also using the `Qdrant.from_documents()` to store the OpenAI text embedding (dense) for the hybrid search and it works fine.
```python
class DenseVectorStore(ValidateQdrantClient):
...
self._qdrant_db = Qdrant.from_documents(
self.documents,
embeddings,
url=os.getenv("QDRANT_URL"),
prefer_grpc=True,
api_key=os.getenv("QDRANT_API_KEY"),
collection_name=self.collection_name,
force_recreate=True,
)
...
```
### System Info
langchain==0.1.13
langchain-community==0.0.29
langchain-core==0.1.33
langchain-openai==0.1.1
langchain-text-splitters==0.0.1
qdrant-client==1.8.0 | `QdrantSparseVectorRetriever` throws `ValidationError` when `prefer_grpc=True` in `QdrantClient` | https://api.github.com/repos/langchain-ai/langchain/issues/19472/comments | 0 | 2024-03-23T23:19:14Z | 2024-06-29T16:09:17Z | https://github.com/langchain-ai/langchain/issues/19472 | 2,204,088,506 | 19,472 |
[
"langchain-ai",
"langchain"
] | ### 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
# Below code cannot work when attach "callbacks" to the Anthropic LLM:
## Code block-1:
```
## bug report
from langchain.chains.question_answering import load_qa_chain
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.messages.base import BaseMessage
from langchain_core.documents import Document
from langchain_openai import ChatOpenAI
from langchain_core.outputs.llm_result import LLMResult
from typing import Any, AsyncIterable, Awaitable, Dict, List
from langchain.callbacks import AsyncIteratorCallbackHandler
import asyncio
from langchain_anthropic import ChatAnthropic
from langchain_core.prompts import ChatPromptTemplate
callback_handler = AsyncIteratorCallbackHandler()
llm = ChatAnthropic(
temperature=0, model_name="claude-3-haiku-20240307", callbacks=[callback_handler]
)
chat_template = ChatPromptTemplate.from_template(
"""You are a document assistant expert, please reply any questions using below context text within 20 words."
Context: ```{context}```
Question: {question}
Answer:
"""
)
async def reply(question) -> AsyncIterable[str]: # type: ignore
chain = load_qa_chain(
llm=llm,
chain_type="stuff",
verbose=True,
prompt=chat_template,
)
# chain.callbacks = []
async def wrap_done(fn: Awaitable, event: asyncio.Event):
"""Wrap an awaitable with a event to signal when it's done or an exception is raised."""
try:
await fn
except Exception as e:
# TODO: handle exception
print(f"Caught exception: {e}")
finally:
# Signal the aiter to stop.
event.set()
# Begin a task that runs in the background.
task = asyncio.create_task(
wrap_done(
chain.ainvoke(
{
"question": question,
"chat_history": [],
"input_documents": [
Document(
page_content="StreamingStdOutCallbackHandler is a callback class in LangChain"
)
],
} # , config={"callbacks": [callback_handler]}
),
callback_handler.done,
),
)
# callback_handler.aiter
async for token in callback_handler.aiter():
yield token
await task # type: ignore
## test reply()
answer = reply("how StreamingStdOutCallbackHandler works? ")
res = ""
async for chunk in answer:
res += chunk
print(res)
### the output is EMPTY !
### (Should be not empty)
```
# The expected actions should be like below after fixing "**_agenerate**" in ChatAnthropic:
## Code block - 2:
```
class ChatAnthropic_(ChatAnthropic):
streaming = False
async def _agenerate(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
run_manager: Optional[AsyncCallbackManagerForLLMRun] = None,
stream: Optional[bool] = None,
**kwargs: Any,
) -> ChatResult:
should_stream = stream if stream is not None else self.streaming
if should_stream:
stream_iter = self._astream(
messages, stop=stop, run_manager=run_manager, **kwargs
)
return await agenerate_from_stream(stream_iter)
params = self._format_params(messages=messages, stop=stop, **kwargs)
data = await self._async_client.messages.create(**params)
return self._format_output(data, **kwargs)
llm = ChatAnthropic_(temperature=0, model_name="claude-3-haiku-20240307", callbacks=[callback_handler], streaming=True)
```
Replace code block-1 with the above llm in code block -2 and re-run the code block-1, the output is expected and not empty.
# Expected output:
```
'StreamingStdOutCallbackHandler is a callback class in LangChain that writes the output of a language model to the standard output in a streaming manner.'
```
### Error Message and Stack Trace (if applicable)
no error msg
### Description
I am trying to use Anthropic LLM which is the llm param of load_qa_chain() of with "callbacks" parameter attached.
I expected the output should be not empty after iterate the "callback" object(AsyncIteratorCallbackHandler)
Instead, it output nothing
### System Info
langchain==0.1.13
langchain-anthropic==0.1.3
langchain-cli==0.0.21
langchain-community==0.0.29
langchain-core==0.1.33
langchain-google-genai==0.0.9
langchain-openai==0.1.1
langchain-text-splitters==0.0.1 | ChatAnthropic cannot work as expected when callbacks set | https://api.github.com/repos/langchain-ai/langchain/issues/19466/comments | 0 | 2024-03-23T10:10:51Z | 2024-06-29T16:09:12Z | https://github.com/langchain-ai/langchain/issues/19466 | 2,203,806,587 | 19,466 |
[
"langchain-ai",
"langchain"
] | ### 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
see Description
### Error Message and Stack Trace (if applicable)
Input should be a subclass of BaseModel
### Description
The attribute `pydantic_object` in class `langchain_core.output_parsers.json.JsonOutputParser` still dosent support BaseModel -> V2 (just V1)
The line 197 should support ` Optional[Type[Union[BaseModelV1, BaseModelV2]]` instead just BaseModel from v1.
### System Info
dosen´t matter | Annotation for langchain_core.output_parsers.json.JsonOutputParser -> pydantic_object not compatible for v2 | https://api.github.com/repos/langchain-ai/langchain/issues/19441/comments | 0 | 2024-03-22T13:43:19Z | 2024-06-28T16:08:23Z | https://github.com/langchain-ai/langchain/issues/19441 | 2,202,528,487 | 19,441 |
[
"langchain-ai",
"langchain"
] | ### 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
import json
from langchain_community.agent_toolkits.openapi.spec import reduce_openapi_spec
from langchain_community.agent_toolkits.openapi import planner
from langchain_openai import OpenAI
from langchain.requests import RequestsWrapper
with open("leo_reduce_openapi.json") as f:
raw_mongodb_api_spec = json.load(f)
mongodb_api_spec = reduce_openapi_spec(raw_mongodb_api_spec)
# Get API credentials.
headers = construct_auth_headers(LEO_API_KEY)
requests_wrapper = RequestsWrapper(headers=headers)
llm = OpenAI(model_name="local_model", temperature=0.0,openai_api_base=AIURL,
openai_api_key="OPENAI_API_KEY",
max_tokens=10000)
mongodb_api_spec.servers[0]['url'] = f"http://{LEO_API_URL}:{LEO_API_PORT}" + mongodb_api_spec.servers[0]['url']
mongodb_api_agent = planner.create_openapi_agent(mongodb_api_spec, requests_wrapper, llm,
verbose=True,
agent_executor_kwargs={"handle_parsing_errors": "Check your output and make sure it conforms, use the Action/Action Input syntax"})
user_query = (
"""Do only one a request to get all namespaces, and return the list of namespaces
Use parameters:
database_name: telegraf
collection_name: 3ddlmlite
ATTENTION to keep parameters along the discussion
"""
)
result = mongodb_api_agent.invoke(user_query)
### Error Message and Stack Trace (if applicable)
File "C:\Program Files\JetBrains\PyCharm 2022.3.2\plugins\python\helpers\pydev\pydevconsole.py", line 364, in runcode
coro = func()
File "<input>", line 1, in <module>
File "C:\Program Files\JetBrains\PyCharm 2022.3.2\plugins\python\helpers\pydev\_pydev_bundle\pydev_umd.py", line 197, in runfile
pydev_imports.execfile(filename, global_vars, local_vars) # execute the script
File "C:\Program Files\JetBrains\PyCharm 2022.3.2\plugins\python\helpers\pydev\_pydev_imps\_pydev_execfile.py", line 18, in execfile
exec(compile(contents+"\n", file, 'exec'), glob, loc)
File "C:\Users\PYTHON\langChain\DLMDataRequest.py", line 84, in <module>
result = mongodb_api_agent.invoke(user_query)
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\chains\base.py", line 163, in invoke
raise e
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\chains\base.py", line 153, in invoke
self._call(inputs, run_manager=run_manager)
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\agents\agent.py", line 1432, in _call
next_step_output = self._take_next_step(
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\agents\agent.py", line 1138, in _take_next_step
[
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\agents\agent.py", line 1138, in <listcomp>
[
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\agents\agent.py", line 1223, in _iter_next_step
yield self._perform_agent_action(
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain\agents\agent.py", line 1245, in _perform_agent_action
observation = tool.run(
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_core\tools.py", line 422, in run
raise e
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_core\tools.py", line 381, in run
self._run(*tool_args, run_manager=run_manager, **tool_kwargs)
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_core\tools.py", line 587, in _run
else self.func(*args, **kwargs)
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_community\agent_toolkits\openapi\planner.py", line 321, in _create_and_run_api_controller_agent
agent = _create_api_controller_agent(base_url, docs_str, requests_wrapper, llm)
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_community\agent_toolkits\openapi\planner.py", line 263, in _create_api_controller_agent
RequestsGetToolWithParsing(
File "C:\Users\PYTHON\langChain\venv\lib\site-packages\langchain_community\tools\requests\tool.py", line 36, in __init__
raise ValueError(
ValueError: You must set allow_dangerous_requests to True to use this tool. Request scan be dangerous and can lead to security vulnerabilities. For example, users can ask a server to make a request to an internalserver. It's recommended to use requests through a proxy server and avoid accepting inputs from untrusted sources without proper sandboxing.Please see: https://python.langchain.com/docs/security for further security information.
### Description
Add allow_dangerous_requests has no effect. It's because kwargs is not add in create_openapi_agent() function and can't go to _create_api_controller_agent --> RequestsGetToolWithParsing
in awaiting, i've add:
in file venv/Lib/site-packages/langchain_community/agent_toolkits/openapi/planner.py
tools: List[BaseTool] = [
RequestsGetToolWithParsing(
requests_wrapper=requests_wrapper, llm_chain=get_llm_chain, allow_dangerous_requests=True
),
RequestsPostToolWithParsing(
requests_wrapper=requests_wrapper, llm_chain=post_llm_chain, allow_dangerous_requests=True
),
### System Info
langchain 0.1.13
langchain-community 0.0.29
langchain-core 0.1.33
langchain-openai 0.1.0
langchain-text-splitters 0.0.1
langchainplus-sdk 0.0.21
langsmith 0.1.31
windows 10 | [bug] [toolkit]: can't add allow_dangerous_requests in parameter | https://api.github.com/repos/langchain-ai/langchain/issues/19440/comments | 1 | 2024-03-22T12:42:14Z | 2024-07-01T16:06:14Z | https://github.com/langchain-ai/langchain/issues/19440 | 2,202,413,072 | 19,440 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from typing import Annotated
from langchain.tools import StructuredTool
from pydantic import BaseModel, Field
class ToolSchema(BaseModel):
question: Annotated[
str,
Field(
description="Question to be done to the search engine. Make it clear and complete. Should be formulated always in the language the user is using with you."
),
]
page: Annotated[int, Field(ge=0, le=1, description="Result page to be searched in")]
class SearchInternetTool(StructuredTool):
def __init__(self):
super(StructuredTool, self).__init__(
name="Search Internet",
description=f"""
Useful for getting recent & factual information.
If this tool is used, it is mandatory to include the sources in your response afterwards.
You can only use this tool 2 times.
""".replace(
" ", ""
).replace(
"\n", " "
),
args_schema=ToolSchema,
func=self.function,
)
def function(self, question: str, page: int) -> str:
return f"Question: {question}, Page: {page}"
tool = SearchInternetTool()
print(tool.run("""{"question": "How old is Snoop dogg", "page": 0}"""))
```
### Error Message and Stack Trace (if applicable)
```
Traceback (most recent call last):
File "<frozen runpy>", line 198, in _run_module_as_main
File "<frozen runpy>", line 88, in _run_code
File "/Users/alramalho/workspace/jarvis/jarvis-clean/backend/spike.py", line 42, in <module>
print(tool.run("""{"question": "How old is Snoop dogg", "page": 0}"""))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/alramalho/Library/Caches/pypoetry/virtualenvs/backend-jarvis-Xbe6mXa_-py3.11/lib/python3.11/site-packages/langchain_core/tools.py", line 388, in run
raise e
File "/Users/alramalho/Library/Caches/pypoetry/virtualenvs/backend-jarvis-Xbe6mXa_-py3.11/lib/python3.11/site-packages/langchain_core/tools.py", line 379, in run
parsed_input = self._parse_input(tool_input)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/alramalho/Library/Caches/pypoetry/virtualenvs/backend-jarvis-Xbe6mXa_-py3.11/lib/python3.11/site-packages/langchain_core/tools.py", line 279, in _parse_input
input_args.validate({key_: tool_input})
File "pydantic/main.py", line 711, in pydantic.main.BaseModel.validate
File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for ToolSchema
page
field required (type=value_error.missing)
```
### Description
i am trying to update my AgentExecutor to make use of `args_schema` on tool usage.
Nevertheless, it internally is calling the tool.run (`AgentExecutor._perform_agent_action`), which is failing to the error above, reproducible by the given code
### System Info
```
langchain==0.1.13
langchain-community==0.0.29
langchain-core==0.1.33
langchain-openai==0.1.0
langchain-text-splitters==0.0.1
``` | AgentExecutor fails to use StructuredTool | https://api.github.com/repos/langchain-ai/langchain/issues/19437/comments | 3 | 2024-03-22T11:23:33Z | 2024-07-21T16:19:57Z | https://github.com/langchain-ai/langchain/issues/19437 | 2,202,276,796 | 19,437 |
[
"langchain-ai",
"langchain"
] | ### 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
``` python
from langchain_text_splitters import MarkdownHeaderTextSplitter
markdown_document = """
# Heading 1
## Heading 2
This is a Markdown List with a nested List:
- Item 1
- Sub Item 1.1
- Item 2
- Sub Item 2.1
- Sub Item 2.2
- Item 3
"""
headers_to_split_on = [
("#", "Header 1"),
("##", "Header 2"),
("###", "Header 3"),
]
markdown_splitter = MarkdownHeaderTextSplitter(headers_to_split_on=headers_to_split_on)
md_header_splits = markdown_splitter.split_text(markdown_document)
print(md_header_splits[0].page_content)
```
**Expected output**
```markdown
This is a Markdown List with a nested List:
- Item 1
- Sub Item 1.1
- Item 2
- Sub Item 2.1
- Sub Item 2.2
- Item 3
```
Actual Output
```markdown
This is a Markdown List with a nested List:
- Item 1
- Sub Item 1.1
- Item 2
- Sub Item 2.1
- Sub Item 2.2
- Item 3
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
**MarkdownHeaderTextSplitter.split_text()** is removing the format from the nested lists / sublists.
The issue is caused in this part of the [code](https://github.com/langchain-ai/langchain/blob/53ac1ebbbccbc16ce57badf08522c6e59256fdfe/libs/text-splitters/langchain_text_splitters/markdown.py#L109C13-L109C41). Stripping the line removes the Markdown sublist / nested list indentation ( [Markdown Lists Docs](https://www.markdownguide.org/basic-syntax/#lists-1) ).
The same issue is also expected on [paragraphs ](https://www.markdownguide.org/basic-syntax/#paragraphs) [blockquotes ](https://www.markdownguide.org/basic-syntax/#blockquotes) and so on.
### System Info
Package Information
-------------------
> langchain_core: 0.1.28
> langchain: 0.0.350
> langchain_community: 0.0.3
> langsmith: 0.1.13
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | MarkdownHeaderTextSplitter removing format of nested lists / sublists | https://api.github.com/repos/langchain-ai/langchain/issues/19436/comments | 0 | 2024-03-22T10:42:43Z | 2024-06-28T16:08:13Z | https://github.com/langchain-ai/langchain/issues/19436 | 2,202,208,221 | 19,436 |
[
"langchain-ai",
"langchain"
] | ### 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
from typing import Optional
from langchain.chains import create_structured_output_runnable
from langchain_community.chat_models import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.pydantic_v1 import BaseModel, Field
class Dog(BaseModel):
'''Identifying information about a dog.'''
name: str = Field(..., description="The dog's name")
color: str = Field(..., description="The dog's color")
fav_food: Optional[str] = Field(None, description="The dog's favorite food")
llm = ChatOpenAI(model="gpt-4", temperature=0)
structured_llm = create_structured_output_runnable(
Dog,
llm,
mode="openai-json",
enforce_function_usage=False,
return_single=False
)
system = '''You are a world class assistant for extracting information in structured JSON formats
Extract a valid JSON blob from the user input that matches the following JSON Schema:
{output_schema}'''
prompt = ChatPromptTemplate.from_messages(
[("system", system), ("human", "{input}"),]
)
llm_chain = prompt | structured_llm
rsp2 = llm_chain.invoke({"input": "There are three dogs here. You need to return all the information of the three dogs。dog1：Yellow Lili likes to eat meat Black；dog2： Hutch loves to eat hot dogs ；dog3：White flesh likes to eat bones",
"output_schema":Dog})
print(rsp2)
### Error Message and Stack Trace (if applicable)
/Users/anker/cy/code/python/claud_api_test/env/bin/python /Users/anker/cy/code/python/claud_api_test/3.py
/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/_api/deprecation.py:117: LangChainDeprecationWarning: The class `langchain_community.chat_models.openai.ChatOpenAI` was deprecated in langchain-community 0.0.10 and will be removed in 0.2.0. An updated version of the class exists in the langchain-openai package and should be used instead. To use it run `pip install -U langchain-openai` and import as `from langchain_openai import ChatOpenAI`.
warn_deprecated(
Traceback (most recent call last):
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/output_parsers/pydantic.py", line 27, in parse_result
return self.pydantic_object.parse_obj(json_object)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/pydantic/v1/main.py", line 526, in parse_obj
return cls(**obj)
^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/pydantic/v1/main.py", line 341, in __init__
raise validation_error
pydantic.v1.error_wrappers.ValidationError: 2 validation errors for Dog
name
field required (type=value_error.missing)
color
field required (type=value_error.missing)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/Users/anker/cy/code/python/claud_api_test/3.py", line 39, in <module>
rsp2 = llm_chain.invoke({"input": "There are three dogs here. You need to return all the information of the three dogs。dog1：Yellow Lili likes to eat meat Black；dog2： Hutch loves to eat hot dogs ；dog3：White flesh likes to eat bones",
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2309, in invoke
input = step.invoke(
^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/output_parsers/base.py", line 169, in invoke
return self._call_with_config(
^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1488, in _call_with_config
context.run(
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/runnables/config.py", line 347, in call_func_with_variable_args
return func(input, **kwargs) # type: ignore[call-arg]
^^^^^^^^^^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/output_parsers/base.py", line 170, in <lambda>
lambda inner_input: self.parse_result(
^^^^^^^^^^^^^^^^^^
File "/Users/anker/cy/code/python/claud_api_test/env/lib/python3.11/site-packages/langchain_core/output_parsers/pydantic.py", line 31, in parse_result
raise OutputParserException(msg, llm_output=json_object)
langchain_core.exceptions.OutputParserException: Failed to parse Dog from completion {'dogs': [{'name': 'Lili', 'color': 'Yellow', 'favorite_food': 'meat'}, {'name': 'Hutch', 'color': 'Black', 'favorite_food': 'hot dogs'}, {'name': 'Flesh', 'color': 'White', 'favorite_food': 'bones'}]}. Got: 2 validation errors for Dog
name
field required (type=value_error.missing)
color
field required (type=value_error.missing)
### Description
### System Info
open ai返回数据正常，但是langchain解析报错 | field required (type=value_error.missing) | https://api.github.com/repos/langchain-ai/langchain/issues/19431/comments | 2 | 2024-03-22T09:16:12Z | 2024-07-09T09:45:31Z | https://github.com/langchain-ai/langchain/issues/19431 | 2,202,044,183 | 19,431 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
**I was trying langserve recently, and I found that all the client examples were using python, which confused me very much. If they were all using python, then why not just call chain directly? Now I need to use python to write the chain. Then use llangserve to encapsulate it into a rest API, and use RemoteRunnable in python to call the deployed chain. Isn't this unnecessary?**
For example：
### Idea or request for content:
**The problem I have now is:**
I create server
But I want to call it in JS, the page to upload files, rather than in another python code, which is really strange.
| DOC: Why use python SDK in Client？ | https://api.github.com/repos/langchain-ai/langchain/issues/19428/comments | 0 | 2024-03-22T08:17:06Z | 2024-06-28T16:08:03Z | https://github.com/langchain-ai/langchain/issues/19428 | 2,201,942,590 | 19,428 |
[
"langchain-ai",
"langchain"
] | ### 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
```
import os
from pinecone import Pinecone
from dotenv import load_dotenv
load_dotenv()
# Create empty index
PINECONE_KEY, PINECONE_INDEX_NAME = os.getenv(
'PINECONE_API_KEY'), os.getenv('PINECONE_INDEX_NAME')
pc = Pinecone(api_key=PINECONE_KEY)
from langchain_core.documents import Document
from langchain_openai import OpenAIEmbeddings
from langchain_pinecone import PineconeVectorStore
embeddings = OpenAIEmbeddings()
# create new index
# pc.create_index(
# name="film-bot-index",
# dimension=1536,
# metric="cosine",
# spec=PodSpec(
# environment="gcp-starter"
# )
# )
# Target index and check status
index_name = "film-bot-index"
pc_index = pc.Index(index_name)
docs = [
Document(
page_content="A bunch of scientists bring back dinosaurs and mayhem breaks loose",
metadata={"year": 1993, "rating": 7.7,
"genre": ["action", "science fiction"]},
),
Document(
page_content="Leo DiCaprio gets lost in a dream within a dream within a dream within a ...",
metadata={"year": 2010, "director": "Christopher Nolan", "rating": 8.2},
),
Document(
page_content="A psychologist / detective gets lost in a series of dreams within dreams within dreams and Inception reused the idea",
metadata={"year": 2006, "director": "Satoshi Kon", "rating": 8.6},
),
Document(
page_content="A bunch of normal-sized women are supremely wholesome and some men pine after them",
metadata={"year": 2019, "director": "Greta Gerwig", "rating": 8.3},
),
Document(
page_content="Toys come alive and have a blast doing so",
metadata={"year": 1995, "genre": "animated"},
),
Document(
page_content="Three men walk into the Zone, three men walk out of the Zone",
metadata={
"year": 1979,
"director": "Andrei Tarkovsky",
"genre": ["science fiction", "thriller"],
"rating": 9.9,
},
),
]
vectorstore = PineconeVectorStore.from_documents(
docs, embeddings, index_name=index_name
)
from langchain.chains.query_constructor.base import AttributeInfo
from langchain.retrievers.self_query.base import SelfQueryRetriever
from langchain_openai import OpenAI
metadata_field_info = [
AttributeInfo(
name="genre",
description="The genre of the movie",
type="string or list[string]",
),
AttributeInfo(
name="year",
description="The year the movie was released",
type="integer",
),
AttributeInfo(
name="director",
description="The name of the movie director",
type="string",
),
AttributeInfo(
name="rating", description="A 1-10 rating for the movie", type="float"
),
]
document_content_description = "Brief summary of a movie"
llm = OpenAI(temperature=0)
retriever = SelfQueryRetriever.from_llm(
llm, vectorstore, document_content_description, metadata_field_info, verbose=True
)
```
### Error Message and Stack Trace (if applicable)
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[3], [line 27](vscode-notebook-cell:?execution_count=3&line=27)
[25](vscode-notebook-cell:?execution_count=3&line=25) document_content_description = "Brief summary of a movie"
[26](vscode-notebook-cell:?execution_count=3&line=26) llm = OpenAI(temperature=0)
---> [27](vscode-notebook-cell:?execution_count=3&line=27) retriever = SelfQueryRetriever.from_llm(
[28](vscode-notebook-cell:?execution_count=3&line=28) llm, vectorstore, document_content_description, metadata_field_info, verbose=True
[29](vscode-notebook-cell:?execution_count=3&line=29) )
File [~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:227](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:227), in SelfQueryRetriever.from_llm(cls, llm, vectorstore, document_contents, metadata_field_info, structured_query_translator, chain_kwargs, enable_limit, use_original_query, **kwargs)
[213](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:213) @classmethod
[214](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:214) def from_llm(
[215](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:215) cls,
(...)
[224](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:224) **kwargs: Any,
[225](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:225) ) -> "SelfQueryRetriever":
[226](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:226) if structured_query_translator is None:
--> [227](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:227) structured_query_translator = _get_builtin_translator(vectorstore)
[228](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:228) chain_kwargs = chain_kwargs or {}
[230](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:230) if (
[231](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:231) "allowed_comparators" not in chain_kwargs
[232](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:232) and structured_query_translator.allowed_comparators is not None
[233](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:233) ):
File [~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:101](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:101), in _get_builtin_translator(vectorstore)
[98](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:98) except ImportError:
[99](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:99) pass
--> [101](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:101) raise ValueError(
[102](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:102) f"Self query retriever with Vector Store type {vectorstore.__class__}"
[103](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:103) f" not supported."
[104](https://file+.vscode-resource.vscode-cdn.net/Users/ed/Developer/FilmBot/~/miniconda3/envs/FilmBot/lib/python3.12/site-packages/langchain/retrievers/self_query/base.py:104) )
ValueError: Self query retriever with Vector Store type <class 'langchain_pinecone.vectorstores.PineconeVectorStore'> not supported.
### Description
I am trying to create a self-querying retriever using the Pinecone database. The documentation makes it appear as though Pinecone is supported, but sadly it appears as though it is not. Fingers crossed support hasn't been pulled for Chroma DB as well. The code provided above is lightly modified from the documentation ([see here](https://python.langchain.com/docs/integrations/retrievers/self_query/pinecone)).
### System Info
langchain==0.1.13
langchain-community==0.0.29
langchain-core==0.1.33
langchain-experimental==0.0.54
langchain-openai==0.0.8
langchain-pinecone==0.0.3
langchain-text-splitters==0.0.1
Mac
Python Version 3.12.2 | ValueError: Self query retriever with Vector Store type <class 'langchain_pinecone.vectorstores.PineconeVectorStore'> not supported. | https://api.github.com/repos/langchain-ai/langchain/issues/19418/comments | 6 | 2024-03-21T22:27:51Z | 2024-05-28T11:28:32Z | https://github.com/langchain-ai/langchain/issues/19418 | 2,201,301,599 | 19,418 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
import logging
import os
from langchain.indexes import SQLRecordManager, index
from langchain.vectorstores.qdrant import Qdrant
from langchain_community.embeddings import CohereEmbeddings
from langchain_core.documents import Document
from qdrant_client import QdrantClient
from qdrant_client.models import Distance, VectorParams
DOCUMENT_COUNT = 100
COLLECTION_NAME = "test_index"
COHERE_EMBED_MODEL = os.getenv("COHERE_EMBED_MODEL")
COHERE_API_KEY = os.getenv("COHERE_API_KEY")
QDRANT_API_KEY = os.getenv("QDRANT_API_KEY")
# Setup embeddings and vector store
embeddings = CohereEmbeddings(model=COHERE_EMBED_MODEL, cohere_api_key=COHERE_API_KEY)
vectorstore = Qdrant(
client=QdrantClient(url="http://localhost:6333", api_key=QDRANT_API_KEY),
collection_name=COLLECTION_NAME,
embeddings=embeddings,
)
# Init Qdrant collection for vectors
vectorstore.client.create_collection(
collection_name=COLLECTION_NAME,
vectors_config=VectorParams(size=1024, distance=Distance.COSINE),
)
# Init the record manager using SQLite
namespace = f"qdrant/{COLLECTION_NAME}"
record_manager = SQLRecordManager(
namespace, db_url="sqlite:///record_manager_cache.sql"
)
record_manager.create_schema()
# Init 100 example documents
documents = [Document(page_content=f"example{i}", metadata={"source": f"example{i}.txt"}) for i in range(DOCUMENT_COUNT)]
# Log at the INFO level so we can see output from httpx
logging.basicConfig(level=logging.INFO)
# Index 100 documents with a batch size of 100.
# EXPECTED: 1 call to Qdrant with 100 documents per call
# ACTUAL : 2 calls to Qdrant with 64 and 36 documents per call, respectively
result = index(
documents,
record_manager,
vectorstore,
batch_size=100,
cleanup="incremental",
source_id_key="source",
)
print(result)
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
* I'm trying to index documents to a vector store (Qdrant) using the `index()` API to support a record manager. I specify a `batch_size` that is larger than the vector store's default `batch_size` on my `index()` call.
* I expect to see my calls to Qdrant respect the `batch_size`
* LangChain indexes using the vector store implementation's default `batch_size` parameter (Qdrant uses 64)
Running the example code with `DOCUMENT_COUNT` set to 100, you would see two PUTs to Qdrant:
```shell
INFO:httpx:HTTP Request: PUT http://localhost:6333/collections/test_index/points?wait=true "HTTP/1.1 200 OK"
INFO:httpx:HTTP Request: PUT http://localhost:6333/collections/test_index/points?wait=true "HTTP/1.1 200 OK"
{'num_added': 100, 'num_updated': 0, 'num_skipped': 0, 'num_deleted': 0}
```
Running the example code with `DOCUMENT_COUNT` set to 64, you would see one PUT to Qdrant:
```shell
INFO:httpx:HTTP Request: PUT http://localhost:6333/collections/test_index/points?wait=true "HTTP/1.1 200 OK"
{'num_added': 64, 'num_updated': 0, 'num_skipped': 0, 'num_deleted': 0}
```
This is because the `batch_size` is not passed on calls to `vector_store.add_documents()`, which itself calls `add_texts()`:
```python
if docs_to_index:
vector_store.add_documents(docs_to_index, ids=uids)
```
([link](https://github.com/langchain-ai/langchain/blob/v0.1.13/libs/langchain/langchain/indexes/_api.py#L333))
As a result, the vector store implementation's default `batch_size` parameter is used instead:
```python
def add_texts(
self,
texts: Iterable[str],
metadatas: Optional[List[dict]] = None,
ids: Optional[Sequence[str]] = None,
batch_size: int = 64, # Here's the parameter
```
([link](https://github.com/langchain-ai/langchain/blob/v0.1.13/libs/community/langchain_community/vectorstores/qdrant.py#L168))
### Suggested Fix
Update the the `vector_store.add_documents()` call in `index()` to include `batch_size=batch_size`:
https://github.com/langchain-ai/langchain/blob/v0.1.13/libs/langchain/langchain/indexes/_api.py#L333
```python
if docs_to_index:
vector_store.add_documents(docs_to_index, ids=uids, batch_size=batch_size)
```
In doing so, the parameter is passed onward through `kwargs` to the final `add_texts` calls.
If you folks are good with this as a fix, I'm happy to open a PR (since this is my first issue on LangChain, I wanted to make sure I'm not barking up the wrong tree).
### System Info
```shell
System Information
------------------
> OS: Linux
> OS Version: #1 SMP Wed Mar 2 00:30:59 UTC 2022
> Python Version: 3.10.13 (main, Aug 25 2023, 13:20:03) [GCC 9.4.0]
Package Information
-------------------
> langchain_core: 0.1.30
> langchain: 0.1.11
> langchain_community: 0.0.27
> langsmith: 0.1.23
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
``` | index() API does not respect batch_size on vector_store.add_documents() | https://api.github.com/repos/langchain-ai/langchain/issues/19415/comments | 0 | 2024-03-21T20:31:36Z | 2024-07-01T16:06:09Z | https://github.com/langchain-ai/langchain/issues/19415 | 2,201,124,108 | 19,415 |
[
"langchain-ai",
"langchain"
] | I confirmed that WebBaseLoader(\<url\>, session=session) works fine.
WebBaseLoader uses the requests.Session and the defined session headers to make the request.
However, SitemapLoader(\<url\>, session=session) is not working.
SitemapLoader on the same URL and session returns and empty response.
The SitemapLoader() __init__ method has argument **kwargs, which are passed to the WebBaseLoader base class.
However, something is missing in the SitemapLoader implementation, as the session is not correctly used in its logic.
_Originally posted by @GuillermoGarciaF in https://github.com/langchain-ai/langchain/discussions/12844#discussioncomment-8870006_ | SitemapLoader not using requests.Session headers even if base class WebBaseLoader implements it | https://api.github.com/repos/langchain-ai/langchain/issues/19412/comments | 0 | 2024-03-21T19:06:36Z | 2024-06-27T16:09:04Z | https://github.com/langchain-ai/langchain/issues/19412 | 2,200,953,763 | 19,412 |
[
"langchain-ai",
"langchain"
] | ### 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
```python from langchain_community.chat_message_histories import SQLChatMessageHistory
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.runnables.history import RunnableWithMessageHistory
from langchain_openai import ChatOpenAI
import os
from dotenv import load_dotenv
load_dotenv(r'/<PATH>/<TO>/<.ENV_FILE>/.env',override=True)
user = os.environ["SNOWFLAKE_USER"]
password = os.environ["SNOWFLAKE_PASSWORD"]
account = os.environ["SNOWFLAKE_ACCOUNT"]
database = os.environ["SNOWFLAKE_DB"]
schema = os.environ["SNOWFLAKE_SCHEMA"]
warehouse = os.environ["SNOWFLAKE_WAREHOUSE"]
role = os.environ["SNOWFLAKE_ROLE"]
prompt = ChatPromptTemplate.from_messages(
[
("system", "You are a helpful assistant."),
MessagesPlaceholder(variable_name="history"),
("human", "{question}"),
]
)
chain = prompt | ChatOpenAI()
snowflake_uri = f"snowflake://{user}:{password}@{account}/{database}/{schema}?warehouse={warehouse}&role={role}"
session_id = "test_user"
chain_with_history = RunnableWithMessageHistory(
chain,
lambda session_id: SQLChatMessageHistory(
session_id=session_id, connection_string=snowflake_uri
),
input_messages_key="question",
history_messages_key="history",
)
config = {"configurable": {"session_id": session_id}}
chain_with_history.invoke({"question": "Hi! I'm bob"}, config=config)
### Error Message and Stack Trace (if applicable)
Error in RootListenersTracer.on_chain_end callback: FlushError('Instance <Message at 0x1285435b0> has a NULL identity key. If this is an auto-generated value, check that the database table allows generation of new primary key values, and that the mapped Column object is configured to expect these generated values. Ensure also that this flush() is not occurring at an inappropriate time, such as within a load() event.')
AIMessage(content='Hello Bob! How can I assist you today?')
### Description
The SQLChatMessageHistory class errors out when trying to connect to Snowflake as a chat history database. There are three things to do to make this work with Snowflake:
1. Make sure the Snowflake user/role has the privileges to create Sequences and create Tables.
2. Create the Sequence in Snowflake first. This can be done in the Snowflake UI or creating a function utilizing sqlalchemy that creates the Sequence before anything else.
3. Use the Sequence in the `Message` class within the `create_message_model()` function, for the "id" column. Should look like: `id = Column(Integer, Sequence("NAME_OF_SEQUENCE"), primary_key=True,autoincrement=True)`
### System Info
python 3.10
| SQLChatMessageHistory does not support Snowflake integration - for storing and retrieving chat history from Snowflake database. | https://api.github.com/repos/langchain-ai/langchain/issues/19411/comments | 0 | 2024-03-21T18:47:32Z | 2024-06-27T16:08:59Z | https://github.com/langchain-ai/langchain/issues/19411 | 2,200,917,365 | 19,411 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
memory = ConversationBufferMemory(return_messages=True)
mem_vars = memory.load_memory_variables({})
pretty_print("Memory Variables init", mem_vars)
pretty_print("Memory Variables in str list (buffer_as_str) init", memory.buffer_as_str)
memory.buffer.append(AIMessage(content="This is a Gaming Place"))
mem_vars = memory.load_memory_variables({})
pretty_print("Memory Variables seeded", mem_vars)
pretty_print(
"Memory Variables in str list (buffer_as_str), seeded", memory.buffer_as_str
)
memory.buffer.append(HumanMessage(content="Hello dudes", id="user-1"))
memory.buffer.append(HumanMessage(content="hi", id="user-2"))
memory.buffer.append(HumanMessage(content="yo yo", id="user-3"))
memory.buffer.append(HumanMessage(content="nice to see you", id="user-4"))
memory.buffer.append(HumanMessage(content="hoho dude", id="user-5"))
memory.buffer.append(HumanMessage(content="o lalala", id="user-L"))
memory.buffer.append(HumanMessage(content="guten tag", id="user-XXXXL"))
memory.buffer.append(HumanMessage(content="Let's get started, ok?", id="user-1"))
memory.buffer.append(HumanMessage(content="YES", id="user-2"))
memory.buffer.append(HumanMessage(content="YEAH....", id="user-3"))
memory.buffer.append(HumanMessage(content="Cool..", id="user-4"))
memory.buffer.append(HumanMessage(content="yup.", id="user-5"))
memory.buffer.append(HumanMessage(content="Great.....", id="user-L"))
memory.buffer.append(HumanMessage(content="alles klar", id="user-XXXXL"))
memory.buffer.append(HumanMessage(content="Opppsssssss.", id="user-5"))
mem_vars = memory.load_memory_variables({})
pretty_print("Memory Variables", mem_vars)
pretty_print("Memory Variables in str list (buffer_as_str)", memory.buffer_as_str)
def convert_memory_to_dict(memory: ConversationBufferMemory) -> List[Dict[str, str]]:
"""Convert the memory to the dict, role is id, content is the message content."""
res = [
"""The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context.
If the AI does not know the answer to a question, it truthfully says it does not know.
Notice: The 'uid' is user-id, 'role' is user role for human or ai, 'content' is the message content.
"""
]
history = memory.load_memory_variables({})["history"]
for hist_item in history:
role = "human" if isinstance(hist_item, HumanMessage) else "ai"
res.append(
{
"role": role,
"content": hist_item.content,
"uid": hist_item.id if role == "human" else "",
}
)
return res
cxt_dict = convert_memory_to_dict(memory)
pretty_print("cxt_dict", cxt_dict)
def build_chain_without_parsing(
model: BaseChatModel,
) -> RunnableSerializable[Dict, str]:
prompt = ChatPromptTemplate.from_messages(
[
SystemMessage(
content=("You are an AI assistant." "You can handle the query of user.")
),
MessagesPlaceholder(variable_name="history"),
HumanMessagePromptTemplate.from_template("{query}"),
]
)
return (
prompt | model
) # comment model, you can see the filled template after invoking the chain.
model = llm
human_query = HumanMessage(
"""Count the number of 'uid'.""",
id="user-X",
)
res = build_chain_without_parsing(model).invoke(
{
"history": cxt_dict,
"query": human_query,
}
)
pretty_print("Result", res)
```
### Error Message and Stack Trace (if applicable)
The LLM returns me:
```python
AIMessage(
│ content="It seems like you're asking for a count of unique 'uid' values based on the previous conversation structure you've outlined. However, in the conversation snippets you've provided, there are no explicit 'uid' values or a structured format that includes 'uid', 'role', and 'content' fields as you initially described. The conversation appears to be a series of greetings and affirmations without any structured data or identifiers that would allow for counting unique user IDs ('uid').\n\nIf you have a specific dataset or a list of entries that include 'uid', 'role', and 'content' fields, please provide that data. Then, I could help you determine the number of unique 'uid' values within that context."
)
```
### Description
Hello,
I am trying to assign meaningful identifiers to the `id` of `HumanMessage` for downstream tasks of user-id or item-id. I have two approaches to do this:
1. Set the identifiers to the `id` of `HumanMessage`, but I checked on LangSmith and found that all `id`s are not visible by the LLM.
2. Set the identifiers to be a part of `additional_kwargs` (ie. `uid`) as shown in the code I pasted. While I can see them in LangSmith, the LLM cannot see them and gives me a negative response.
Could you please confirm if my understanding is correct?
### System Info
```
langchain==0.1.11
langchain-anthropic==0.1.3
langchain-community==0.0.27
langchain-core==0.1.30
langchain-groq==0.0.1
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
langchainhub==0.1.14
``` | It turns out that the LLM cannot see the information in `additional_kwargs` or `id` of HumanMessage. | https://api.github.com/repos/langchain-ai/langchain/issues/19401/comments | 0 | 2024-03-21T14:04:17Z | 2024-06-27T16:08:54Z | https://github.com/langchain-ai/langchain/issues/19401 | 2,200,287,868 | 19,401 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_core import runnables
class MyRunnable(runnables.RunnableSerializable[str, str]):
def invoke(self,
input: str,
config: runnables.RunnableConfig | None = None) -> str:
if config:
md = config.get('metadata', {})
md['len'] = len(input)
return input[::-1]
# Normal Runnable properly preserves config
mr = MyRunnable()
rc = runnables.RunnableConfig(metadata={'starting_text': '123'})
mr.invoke('hello', config=rc)
print(rc)
# Outputs: {'metadata': {'starting_text': '123', 'len': 5}}
# RetryRunnable's metadata changes do not get preserved
retry_mr = MyRunnable().with_retry(stop_after_attempt=3)
rc = runnables.RunnableConfig(metadata={'starting_text': '123'})
retry_mr.invoke('hello', config=rc)
print(rc)
# Outputs: {'metadata': {'starting_text': '123'}}
# (should be the same as above)
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
I noticed that none of the metadata added by any runnable wrapped in retry is being preserved outside of the retry.
I realize `RunnableConfig`'s metadata probably isn't heavily used. But we've been using it as a side-channel to collect a lot of info during our chain runs, which has been incredibly useful.
Hoping this isn't intended behavior.
If there are multiple retry attempts, at the least we would want the metadata from any successful invocation to make it back up.
### System Info
Standard Google Colab (but also seen in other environments)
```
!pip freeze | grep langchain
langchain-core==0.1.33
```
| RunnableRetry does not preserve metadata in RunnableConfig | https://api.github.com/repos/langchain-ai/langchain/issues/19397/comments | 0 | 2024-03-21T12:06:15Z | 2024-06-27T16:08:50Z | https://github.com/langchain-ai/langchain/issues/19397 | 2,200,011,408 | 19,397 |
[
"langchain-ai",
"langchain"
] | ### 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.
- [ ] 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
I edited the LangChain code to print the `function` variable in both `create_tagging_chain_pydantic()` and `create_extraction_chain_pydantic()` and then ran this script:
```python
from langchain.chains import create_extraction_chain_pydantic, create_tagging_chain_pydantic
from langchain_openai import ChatOpenAI
from pydantic.v1 import BaseModel, Field
class NameAndAge(BaseModel):
name: str = Field(description="The name of the person.")
age: int = Field(description="The age of the person.")
class NamesAndAges(BaseModel):
names_and_ages: list[NameAndAge] = Field(description="The names and ages of the people.")
llm = ChatOpenAI(api_key="sk-XXX")
tagging_chain = create_tagging_chain_pydantic(pydantic_schema=NamesAndAges, llm=llm)
extraction_chain = create_extraction_chain_pydantic(pydantic_schema=NamesAndAges, llm=llm)
```
Which printed:
```
TAGGING:
{
"name": "information_extraction",
"description": "Extracts the relevant information from the passage.",
"parameters": {
"type": "object",
"properties": {
"names_and_ages": {
"title": "Names And Ages",
"description": "The names and ages of the people.",
"type": "array",
"items": {
"$ref": "#/definitions/NameAndAge"
}
}
},
"required": [
"names_and_ages"
]
}
}
EXTRACTION:
{
"name": "information_extraction",
"description": "Extracts the relevant information from the passage.",
"parameters": {
"type": "object",
"properties": {
"info": {
"type": "array",
"items": {
"type": "object",
"properties": {
"names_and_ages": {
"title": "Names And Ages",
"description": "The names and ages of the people.",
"type": "array",
"items": {
"title": "NameAndAge",
"type": "object",
"properties": {
"name": {
"title": "Name",
"description": "The name of the person.",
"type": "string"
},
"age": {
"title": "Age",
"description": "The age of the person.",
"type": "integer"
}
},
"required": [
"name",
"age"
]
}
}
},
"required": [
"names_and_ages"
]
}
}
},
"required": [
"info"
]
}
}
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
`langchain.chains.openai_functions.extraction.py` has these 3 lines in the function `create_extraction_chain_pydantic()`:
```python
openai_schema = pydantic_schema.schema()
openai_schema = _resolve_schema_references(
openai_schema, openai_schema.get("definitions", {})
)
function = _get_extraction_function(openai_schema)
```
However, in `langchain.chains.openai_functions.tagging.py`, the `create_tagging_chain_pydantic()` function is implemented as:
```python
openai_schema = pydantic_schema.schema()
function = _get_tagging_function(openai_schema)
```
This means that any nested objects in the schema are not passed as references in the JSON schema to the LLMChain.
Is this intentional or is it a bug?
### System Info
```
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-openai==0.0.3
langchain-text-splitters==0.0.1
``` | create_tagging_chain_pydantic() doesn't call _resolve_schema_references() like create_extraction_chain_pydantic() does | https://api.github.com/repos/langchain-ai/langchain/issues/19394/comments | 0 | 2024-03-21T10:24:55Z | 2024-06-27T16:08:44Z | https://github.com/langchain-ai/langchain/issues/19394 | 2,199,776,457 | 19,394 |
[
"langchain-ai",
"langchain"
] | ### 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
call `ainvoke` with `**kwargs` has no effect
### Error Message and Stack Trace (if applicable)
_No response_
### Description
https://github.com/langchain-ai/langchain/blob/b20c2640dac79551685b8aba095ebc6125df928c/libs/core/langchain_core/runnables/base.py#L2984-2995
```
results = await asyncio.gather(
*(
step.ainvoke(
input,
# mark each step as a child run
patch_config(
config, callbacks=run_manager.get_child(f"map:key:{key}")
),
)
for key, step in steps.items()
)
)
```
does not correctly pass through `**kwargs` to child `Runnable`s
### System Info
langchian 0.1.13 | RunnableParallel does not correctly pass through **kwargs to child Runnables | https://api.github.com/repos/langchain-ai/langchain/issues/19386/comments | 0 | 2024-03-21T08:34:18Z | 2024-06-27T16:08:39Z | https://github.com/langchain-ai/langchain/issues/19386 | 2,199,532,724 | 19,386 |
[
"langchain-ai",
"langchain"
] | ### 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
```
response = chain.invoke(inputs, extra_headers={"x-request-id": "each call with each id"})
```
I don't want to re-create llm with `default_headers` since that would cost too much time.
The ability to pass extra_headers which is accepted by `openai` client is really useful
### Error Message and Stack Trace (if applicable)
_No response_
### Description
https://github.com/langchain-ai/langchain/blob/master/libs/community/langchain_community/chat_models/openai.py#L421-L440
How to pass in these `**kwargs` in `generate`, both `invoke` or `ainvoke` just ignore the passed in `**kwargs`
### System Info
langchain 0.1.2 | Unable to pass openai extra headers or `**kwargs` from `invoke` or `ainvoke` | https://api.github.com/repos/langchain-ai/langchain/issues/19383/comments | 0 | 2024-03-21T08:08:25Z | 2024-06-27T16:08:34Z | https://github.com/langchain-ai/langchain/issues/19383 | 2,199,475,134 | 19,383 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain.schema import Document
from langchain.vectorstores.mongodb_atlas import MongoDBAtlasVectorSearch
from langchain_openai.embeddings import AzureOpenAIEmbeddings
from pymongo.mongo_client import MongoClient
from pymongo.server_api import ServerApi
os.environ["OPENAI_API_KEY"] = "asd"
os.environ["OPENAI_API_TYPE"] = "azure"
os.environ["OPENAI_API_VERSION"] = "2023-05-15"
ATLAS_CONNECTION_STRING = "asd"
COLLECTION_NAME = "documents"
DB_NAME = "FraDev"
embeddings = AzureOpenAIEmbeddings(
deployment="text-embedding-ada-002",
chunk_size=1, # we need to use one because azure is poop
azure_endpoint="asd",
)
# Create a new client and connect to the server
client = MongoClient(ATLAS_CONNECTION_STRING, server_api=ServerApi("1"))
collection = client["FraDev"][COLLECTION_NAME]
print(collection)
def create_vector_search():
"""
Creates a MongoDBAtlasVectorSearch object using the connection string, database, and collection names, along with the OpenAI embeddings and index configuration.
:return: MongoDBAtlasVectorSearch object
"""
vector_search = MongoDBAtlasVectorSearch.from_connection_string(
ATLAS_CONNECTION_STRING,
f"{DB_NAME}.{COLLECTION_NAME}",
embeddings,
index_name="default",
)
return vector_search
docs = [Document(page_content="foo", metadata={"id": 123})]
vector_search = MongoDBAtlasVectorSearch.from_documents(
documents=docs,
embedding=embeddings,
collection=collection,
index_name="default", # Use a predefined index name
)
```
### Error Message and Stack Trace (if applicable)
The `__init__` from `MongoDBAtlasVectorSearch` defines the keys to be stored inside the db
```python
def __init__(
self,
collection: Collection[MongoDBDocumentType],
embedding: Embeddings,
*,
index_name: str = "default",
text_key: str = "text",
embedding_key: str = "embedding",
relevance_score_fn: str = "cosine",
):
```
See an example from mongo compass
The `Document` structure is destroyed, there is no `page_content` no `metadata` object inside - what is going on?
### Description
See above, thanks a lot
### System Info
```
langchain==0.1.9
langchain-community==0.0.24
langchain-core==0.1.26
langchain-openai==0.0.7
``` | MongoAtlas DB destroys Document structure | https://api.github.com/repos/langchain-ai/langchain/issues/19379/comments | 0 | 2024-03-21T07:50:49Z | 2024-06-27T16:08:29Z | https://github.com/langchain-ai/langchain/issues/19379 | 2,199,433,206 | 19,379 |
[
"langchain-ai",
"langchain"
] | ### 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
```from langchain.agents import AgentExecutor
from langchain.agents.format_scratchpad.openai_tools import format_to_openai_tool_messages
from langchain.agents.output_parsers.openai_tools import OpenAIToolsAgentOutputParser
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.tools import tool
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
openai_api_base=f"http://192.168.1.201:18000/v1",
openai_api_key="EMPTY",
model="qwen",
# temperature=0.0,
# top_p=0.2,
# max_tokens=settings.INFERENCE_MAX_TOKENS,
verbose=True,
)
@tool
def get_word_length(word: str) -> int:
"""返回一个单词的长度。"""
print("-----")
return len(word)
print(get_word_length.invoke("flower"))
print(get_word_length.name)
print(get_word_length.description)
# 6
# get_word_length
# get_word_length(word: str) -> int - 返回一个单词的长度。
tools = [get_word_length]
prompt = ChatPromptTemplate.from_messages(
[
("system", "You are very powerful assistant, but don't know current events"),
("user", "{input}"),
MessagesPlaceholder(variable_name="agent_scratchpad"),
]
)
llm_with_tools = llm.bind_tools(tools=tools)
agent = (
{
"input": lambda x: x["input"],
"agent_scratchpad": lambda x: format_to_openai_tool_messages(x["intermediate_steps"]),
}
| prompt
| llm_with_tools
| OpenAIToolsAgentOutputParser()
)
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True)
agent_executor.invoke({"input": "How many letters in the word flower"})
# > Entering new AgentExecutor chain...
# The word "flower" has 5 letters.
#
# > Finished chain.```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
* i am trying to use langchain to bind tools to llm , then join to agent. but when I invoke the agent, I found the tool not using, could you help to figure the issue out?
* I am using local LLM based on qwen host on private PC
### System Info
langchain: 0.1.12
langchain-core: 0.1.32
langchain-openai: 0.0.8
openai: 1.14.2
OS: ubuntu22.04 (docker)
CUDA: 12.4
| Tool cannot using after llm.bind_tools | https://api.github.com/repos/langchain-ai/langchain/issues/19368/comments | 0 | 2024-03-21T01:41:29Z | 2024-06-27T16:08:24Z | https://github.com/langchain-ai/langchain/issues/19368 | 2,198,920,597 | 19,368 |
[
"langchain-ai",
"langchain"
] | ### Privileged issue
- [X] I am a LangChain maintainer, or was asked directly by a LangChain maintainer to create an issue here.
### Issue Content
Callbacks are a bit out of date: https://python.langchain.com/docs/modules/callbacks/
We need to document updates in the supported callbacks
| Document on_retriever_x callbacks | https://api.github.com/repos/langchain-ai/langchain/issues/19361/comments | 2 | 2024-03-20T21:35:28Z | 2024-07-04T16:08:43Z | https://github.com/langchain-ai/langchain/issues/19361 | 2,198,612,174 | 19,361 |
[
"langchain-ai",
"langchain"
] | I've encountered an issue with `ChatLiteLLMRouter` where it ignores the model I specify. Instead, it defaults to using the first model in its list (which can be even with a wrong type). This behavior seems tied to the invocation of:
```python
self._set_model_for_completion()
```
Here's what's happening: I set up the router with my chosen model like this:
```python
chat = ChatLiteLLMRouter(model="gpt-4-0613", router=LiteLLMRouterFactory().router())
```
And then, when I attempt to process messages:
```python
message = HumanMessage(content="Hello")
response = await chat.agenerate([[message], [message]])
```
It ignores my model choice. The culprit appears to be the line that sets the model to the first item in the model list within `_set_model_for_completion`:
https://github.com/langchain-ai/langchain/blob/5d220975fc563a92f41aeb0907e8c3819da073f5/libs/community/langchain_community/chat_models/litellm_router.py#L176
because of:
```python
def _set_model_for_completion(self) -> None:
# use first model name (aka: model group),
# since we can only pass one to the router completion functions
self.model = self.router.model_list[0]["model_name"]
```
Removing the mentioned line corrects the issue, and the router then correctly uses the model I initially specified.
Is this intended behavior, or a bug that we can fix?
If in some cases that setting is still needed, it can be fixed like this:
```python
if self.model is None:
self.model = self.router.model_list[0]["model_name"]
```
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-experimental==0.0.49
langchain-openai==0.0.8
langchain-text-splitters==0.0.1 | ChatLiteLLMRouter ignores specified model selection (overrides it by taking the 1st) | https://api.github.com/repos/langchain-ai/langchain/issues/19356/comments | 2 | 2024-03-20T19:07:36Z | 2024-07-31T16:06:55Z | https://github.com/langchain-ai/langchain/issues/19356 | 2,198,344,884 | 19,356 |
[
"langchain-ai",
"langchain"
] | ### 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
This discussion is not related to any specific python code; this is more like a promotion or idea.
### Error Message and Stack Trace (if applicable)
_No response_
### Description
### Intro
I am a software engineer at MediaTek, and my project involves using LangChain to address some of our challenges and to conduct research on topics related to LangChain. I believe a member of our team has already initiated contact with the vendor regarding the purchase of a [LangSmith](https://smith.langchain.com/) License.
### Motivation
Today, I delved into the source code and discovered that this package heavily relies on Pydantic, specifically version 1. However, the OpenAI API is currently utilizing `Pydantic==2.4.2` [Ref](https://github.com/openai/openai-python/blob/main/requirements.lock#L40), there is no reason we don't upgrade it as a developer.
### Observation of current repository and needs
Here are some observations and understandings I have gathered:
1. In [langchain_core](https://github.com/langchain-ai/langchain/tree/master/libs/core/langchain_core), `langchain.pydantic_v1` is used solely for invoking `pydantic.v1`.
2. There are significant differences between Pydantic v1 and v2, such as:
- `root_validator` has been replaced by `model_validator`.
- `validator` has been replaced by `field_validator`.
- etc.
### Question
Should we consider updating this module?
If so, it would be my honor to undertake this task.
### Workflow
If I am to proceed, my approach would include:
1. Replacing all instances of `from langchain_core.pydantic_v1 import XXX` with `from pydantic import XXX` within the `langchain` codebase.
2. Making the necessary updates for Pydantic, including changes to `model_validator`, `field_validator`, etc.
3. Keeping `langchain_core.pydantic_v1` unchanged to avoid conflicts with other repositories, but issuing a deprecation warning to inform users and developers.
After this task has been done, I can keep upgrading other related repositories from `langgraph` to more.
### System Info
None | [Chore] upgrading pydantic from v1 to v2 with solution | https://api.github.com/repos/langchain-ai/langchain/issues/19355/comments | 2 | 2024-03-20T18:57:57Z | 2024-03-20T19:09:21Z | https://github.com/langchain-ai/langchain/issues/19355 | 2,198,320,708 | 19,355 |
[
"langchain-ai",
"langchain"
] | ### 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
**Example 1**
```py
from langchain_openai import ChatOpenAI
import httpx
http_client = httpx.Client()
llm = ChatOpenAI(
model_name="gpt-4-1106-preview",
openai_api_key="foo",
http_client=http_client,
)
```
**Example 2**
```py
from langchain_openai import ChatOpenAI
import httpx
http_async_client = httpx.AsyncClient()
llm = ChatOpenAI(
model_name="gpt-4-1106-preview",
openai_api_key="foo",
http_client=http_async_client,
)
```
### Error Message and Stack Trace (if applicable)
**Example 1**
```
Traceback (most recent call last):
File "/home/justin/example.py", line 7, in <module>
llm = ChatOpenAI(
^^^^^^^^^^^
File "/home/justin/venv/lib/python3.11/site-packages/langchain_core/load/serializable.py", line 120, in __init__
super().__init__(**kwargs)
File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for ChatOpenAI
__root__
Invalid `http_client` argument; Expected an instance of `httpx.AsyncClient` but got <class 'httpx.Client'> (type=type_error)
```
**Example 2**
```
Traceback (most recent call last):
File "/home/justin/example.py", line 7, in <module>
llm = ChatOpenAI(
^^^^^^^^^^^
File "/home/justin/venv/lib/python3.11/site-packages/langchain_core/load/serializable.py", line 120, in __init__
super().__init__(**kwargs)
File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__
pydantic.error_wrappers.ValidationError: 1 validation error for ChatOpenAI
__root__
Invalid `http_client` argument; Expected an instance of `httpx.Client` but got <class 'httpx.AsyncClient'> (type=type_error)
```
### Description
When attempting to instantiate the `ChatOpenAI` model with a custom `httpx.Client`, I realized that I receive an error stating that the `http_client` needs to be of type `htttpx.AsyncClient`. This is also true when I try using a custom `htttpx.AsyncClient`, I get an error stating the type needs to be `httpx.Client`.
I noticed this error only occured when I updated my `openai` package to `1.14.2`, before this, the error did not occur.
I have found a similar issue here: https://github.com/langchain-ai/langchain/issues/19116. However, the bug fix was merged, and it did not fix my issue.
### System Info
platform: Ubuntu 22.04.4 LTS
python version: 3.11.6
Error occurs with these package versions
```
langchain==0.1.12
langchain-community==0.0.29
langchain-core==0.1.33
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
openinference-instrumentation-langchain==0.1.12
openai==1.14.2
openinference-instrumentation-openai==0.1.4
```
Note that with `openai` version 1.13.3, this error does not occur | ChatOpenAI http_client cannot be specified due to client being checked for httpx.SyncClient and httpx.AsyncClient simultaneously with openai 1.14.2 | https://api.github.com/repos/langchain-ai/langchain/issues/19354/comments | 9 | 2024-03-20T18:41:46Z | 2024-06-06T01:58:01Z | https://github.com/langchain-ai/langchain/issues/19354 | 2,198,291,029 | 19,354 |
[
"langchain-ai",
"langchain"
] | ### Privileged issue
- [X] I am a LangChain maintainer, or was asked directly by a LangChain maintainer to create an issue here.
### Issue Content
Conversational RAG (e.g., LCEL analogues `ConversationalRetrievalChain`) is implemented across the docs to varying degrees of consistency. Users seeking to quickly get set up with conversational RAG might need to contend with these varying implementations.
We propose to update these implementations to use the abstractions used in [get_started/quickstart#conversation-retrieval-chain](https://python.langchain.com/docs/get_started/quickstart#conversation-retrieval-chain).
Known implementations:
- [ ] [get_started/quickstart#conversation-retrieval-chain](https://python.langchain.com/docs/get_started/quickstart#conversation-retrieval-chain)
- [x] [use_cases/question_answering/chat_history](https://python.langchain.com/docs/use_cases/question_answering/chat_history) ([PR](https://github.com/langchain-ai/langchain/pull/19349))
- [x] [expression_language/cookbook/retrieval#conversational-retrieval-chain](https://python.langchain.com/docs/expression_language/cookbook/retrieval#conversational-retrieval-chain)
- [ ] [use_cases/chatbots/quickstart](https://python.langchain.com/docs/use_cases/chatbots/quickstart)
- [ ] [use_cases/chatbots/retrieval](https://python.langchain.com/docs/use_cases/chatbots/retrieval) | [docs] Consolidate logic for conversational RAG | https://api.github.com/repos/langchain-ai/langchain/issues/19344/comments | 0 | 2024-03-20T17:07:11Z | 2024-07-08T16:05:55Z | https://github.com/langchain-ai/langchain/issues/19344 | 2,198,074,722 | 19,344 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
class CustomRunnable(RunnableSerializable):
def transform_meaning(self, input):
if input["input"].find("meaning"):
input["input"] = input["input"].replace("meaning", "purpose")
return input
def invoke(
self,
input: Any,
config: RunnableConfig = None,
**kwargs: Any,
) -> Any:
# Implement the custom logic here
return self.transform_meaning(input)
llm = ChatOpenAI()
question = 'What is the meaning of life?'
prompt = ChatPromptTemplate.from_messages([
("system", "You are a helpful assistant."),
("user", "{input}")
])
output_parser = StrOutputParser()
original_chain = CustomRunnable() | prompt | llm | output_parser
serialized_chain = langchain_core.load.dumps(original_chain.to_json())
deserialized_chain = langchain_core.load.loads(serialized_chain)
deserialized_chain.invoke({
"input": question
})
```
### Error Message and Stack Trace (if applicable)
```
---------------------------------------------------------------------------
NotImplementedError Traceback (most recent call last)
Cell In[93], line 46
41
42
43
45 serialized_chain = langchain_core.load.dumps(chain.to_json())
---> 46 deserialized_chain = langchain_core.load.loads(serialized_chain, valid_namespaces=["langchain", "__main__"])
48 deserialized_chain.invoke({
49 "input": input
50 })
File ./site-packages/langchain_core/_api/beta_decorator.py:109, in beta.<locals>.beta.<locals>.warning_emitting_wrapper(*args, **kwargs)
107 warned = True
108 emit_warning()
--> 109 return wrapped(*args, **kwargs)
File ./site-packages/langchain_core/load/load.py:132, in loads(text, secrets_map, valid_namespaces)
113 @beta()
114 def loads(
115 text: str,
(...)
118 valid_namespaces: Optional[List[str]] = None,
119 ) -> Any:
120 """Revive a LangChain class from a JSON string.
121 Equivalent to `load(json.loads(text))`.
122
(...)
130 Revived LangChain objects.
131 """
--> 132 return json.loads(text, object_hook=Reviver(secrets_map, valid_namespaces))
File /usr/lib/python3.10/json/__init__.py:359, in loads(s, cls, object_hook, parse_float, parse_int, parse_constant, object_pairs_hook, **kw)
357 if parse_constant is not None:
358 kw['parse_constant'] = parse_constant
--> 359 return cls(**kw).decode(s)
File /usr/lib/python3.10/json/decoder.py:337, in JSONDecoder.decode(self, s, _w)
332 def decode(self, s, _w=WHITESPACE.match):
333 """Return the Python representation of ``s`` (a ``str`` instance
334 containing a JSON document).
335
336 """
--> 337 obj, end = self.raw_decode(s, idx=_w(s, 0).end())
338 end = _w(s, end).end()
339 if end != len(s):
File /usr/lib/python3.10/json/decoder.py:353, in JSONDecoder.raw_decode(self, s, idx)
344 """Decode a JSON document from ``s`` (a ``str`` beginning with
345 a JSON document) and return a 2-tuple of the Python
346 representation and the index in ``s`` where the document ended.
(...)
350
351 """
352 try:
--> 353 obj, end = self.scan_once(s, idx)
354 except StopIteration as err:
355 raise JSONDecodeError("Expecting value", s, err.value) from None
File ./site-packages/langchain_core/load/load.py:60, in Reviver.__call__(self, value)
53 raise KeyError(f'Missing key "{key}" in load(secrets_map)')
55 if (
56 value.get("lc", None) == 1
57 and value.get("type", None) == "not_implemented"
58 and value.get("id", None) is not None
59 ):
---> 60 raise NotImplementedError(
61 "Trying to load an object that doesn't implement "
62 f"serialization: {value}"
63 )
65 if (
66 value.get("lc", None) == 1
67 and value.get("type", None) == "constructor"
68 and value.get("id", None) is not None
69 ):
70 [*namespace, name] = value["id"]
NotImplementedError: Trying to load an object that doesn't implement serialization: {'lc': 1, 'type': 'not_implemented', 'id': ['__main__', 'CustomRunnable'], 'repr': 'CustomRunnable()', 'name': 'CustomRunnable', 'graph': {'nodes': [{'id': 0, 'type': 'schema', 'data': {'title': 'CustomRunnableInput'}}, {'id': 1, 'type': 'runnable', 'data': {'id': ['__main__', 'CustomRunnable'], 'name': 'CustomRunnable'}}, {'id': 2, 'type': 'schema', 'data': {'title': 'CustomRunnableOutput'}}], 'edges': [{'source': 0, 'target': 1}, {'source': 1, 'target': 2}]}}
```
### Description
After testing many suggestions from the kaga-ai and dosu bots, it seems that classes that extends `RunnableSerializable` are still not considered to be serializable.
What was expected:
- To be able to create a custom logic to transform the input using `RunnableSerializable`..
- Pipe this to a chain..
- Serialized it so it can be stored somewhere..
- And then be able to deserialize it before running `invoke`
What happened:
- When I add the class based on `RunnableSerializable` to the chain, it breaks the deserialization because the graph node has the `'type': 'not_implemented'` property
Original discussion: https://github.com/langchain-ai/langchain/discussions/19307
### System Info
```
langchain==0.1.9
langchain-community==0.0.24
langchain-core==0.1.27
langchain-openai==0.0.8
Python 3.10.12
``` | Class that extends RunnableSerializable makes the Chain not serializable | https://api.github.com/repos/langchain-ai/langchain/issues/19338/comments | 2 | 2024-03-20T14:16:26Z | 2024-03-20T15:00:11Z | https://github.com/langchain-ai/langchain/issues/19338 | 2,197,653,861 | 19,338 |
[
"langchain-ai",
"langchain"
] | ### 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
```
import logging
from chromadb import PersistentClient
from langchain.vectorstores.chroma import Chroma
from langchain.indexes import SQLRecordManager, index
from langchain_openai import OpenAIEmbeddings
from matextract_langchain_prototype.text_splitter import MatSplitter
TEST_FILE_NAME = "nlo_test.txt"
# logging.basicConfig(level=10)
# logger = logging.getLogger(__name__)
def create_vector_db():
"""simply create a vector database for a paper"""
chroma_client = PersistentClient()
embedding = OpenAIEmbeddings()
chroma_db = Chroma(client=chroma_client, collection_name="vector_database", embedding_function=embedding)
record_manager = SQLRecordManager(namespace="chroma/vector_database", db_url="sqlite:///record_manager.db")
record_manager.create_schema()
text_splitter = MatSplitter(chunk_size=100, chunk_overlap=0)
with open(TEST_FILE_NAME, encoding='utf-8') as file:
content = file.read()
documents = text_splitter.create_documents([content], [{"source": TEST_FILE_NAME}])
info = index(
docs_source=documents,
record_manager=record_manager,
vector_store=chroma_db,
cleanup="incremental",
source_id_key="source",
batch_size=100
)
print(info)
```
### Error Message and Stack Trace (if applicable)
I run the function twice, below is the second returned info.
{'num_added': 42, 'num_updated': 0, 'num_skipped': 100, 'num_deleted': 42}
### Description
This is not as I expected, It should return message like num_skipped 142 instead of 100. I think there is something wrong with the record manager. Hope the developer of langchain can fix it.
### System Info
langchain = 0.1.12
windows 11
python = 3.10 | langchain index incremental mode failed to detect existed documents once exceed the default batch_size | https://api.github.com/repos/langchain-ai/langchain/issues/19335/comments | 11 | 2024-03-20T13:25:47Z | 2024-08-10T16:07:30Z | https://github.com/langchain-ai/langchain/issues/19335 | 2,197,531,580 | 19,335 |
[
"langchain-ai",
"langchain"
] | ### 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
```
from langchain.cache import SQLiteCache
from langchain.globals import set_llm_cache
from langchain_anthropic import ChatAnthropic
from langchain_core.messages import HumanMessage
set_llm_cache(SQLiteCache(database_path=".langchain_cache.db"))
chat_model = ChatAnthropic(
model="claude-3-sonnet-20240229",
temperature=1.0,
max_tokens=2048,
)
message = HumanMessage(content="Hello World!")
print(response)
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
The caching is only dependent on the messages and not on the parameters given to the `ChatAnthropic` class.
This results in langchain hitting the cache instead of sending a new requests to the API even so parameters like `temperature`, `max_tokens` or even the `model` have been changed.
I.e. when the first request containg just the message `"Hello World"` was send to ` "claude-3-sonnet-20240229"` and one changes the model to ` "claude-3-opus-20240229"` afterward langchain will still fetch the response for from the first request.
### System Info
System Information
------------------
> OS: Linux
> OS Version: #25~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Feb 20 16:09:15 UTC 2
> Python Version: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
Package Information
-------------------
> langchain_core: 0.1.32
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.29
> langchain_anthropic: 0.1.4
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | Caching for ChatAnthropic is not working as expected | https://api.github.com/repos/langchain-ai/langchain/issues/19328/comments | 4 | 2024-03-20T10:54:52Z | 2024-06-26T15:11:40Z | https://github.com/langchain-ai/langchain/issues/19328 | 2,197,235,323 | 19,328 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from common_base.llm_base import embedding_model
redis_url = "redis://:mypassword@127.0.0.1:6379"
from langchain_community.vectorstores.redis import Redis
from langchain_community.vectorstores.redis.filters import RedisFilter
rds: Redis = Redis.from_existing_index(
embedding_model,
redis_url=redis_url,
index_name="teacher_report",
schema="teacher_report/teacher_report.yaml",
)
email_filter = RedisFilter.text('teacher_email') == 'asdf@path.how'
asdf = rds.similarity_search(query='asdf',k=3, filter=email_filter)
print(asdf)
```
```yaml
text:
- name: teacher_email
no_index: false
no_stem: false
sortable: false
weight: 1
withsuffixtrie: false
- name: clinic_title
no_index: false
no_stem: false
sortable: false
weight: 1
withsuffixtrie: false
- name: content
no_index: false
no_stem: false
sortable: false
weight: 1
withsuffixtrie: false
vector:
- algorithm: FLAT
block_size: 1000
datatype: FLOAT32
dims: 1536
distance_metric: COSINE
initial_cap: 20000
name: content_vector
```
### Error Message and Stack Trace (if applicable)
```
Traceback (most recent call last):
File "/opt/homebrew/lib/python3.11/site-packages/langchain_community/vectorstores/redis/base.py", line 946, in similarity_search_by_vector
results = self.client.ft(self.index_name).search(redis_query, params_dict) # type: ignore # noqa: E501
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/commands/search/commands.py", line 501, in search
res = self.execute_command(SEARCH_CMD, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/client.py", line 543, in execute_command
return conn.retry.call_with_retry(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/retry.py", line 46, in call_with_retry
return do()
^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/client.py", line 544, in <lambda>
lambda: self._send_command_parse_response(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/client.py", line 520, in _send_command_parse_response
return self.parse_response(conn, command_name, **options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/client.py", line 560, in parse_response
response = connection.read_response()
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/redis/connection.py", line 536, in read_response
raise response
redis.exceptions.ResponseError: Syntax error at offset 22 near path
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/Users/jimmykim/workspace/deus-path-machina/test2.py", line 15, in <module>
asdf = rds.similarity_search(query='asdf',k=3, fetch_k=10, filter=email_filter)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/langchain_community/vectorstores/redis/base.py", line 882, in similarity_search
return self.similarity_search_by_vector(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/langchain_community/vectorstores/redis/base.py", line 950, in similarity_search_by_vector
raise ValueError(
ValueError: Query failed with syntax error. This is likely due to malformation of filter, vector, or query argument
```
### Description
I have written a simple sample code to filter and search data using Redis VectorStore. Despite creating a filter with RedisFilterExpression and passing it as a parameter as described in the documentation, I encounter a syntax error. When I do not pass the filter, the search works correctly across all vector data. I suspect this might be a bug, what do you think?
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-experimental==0.0.45
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
MAC OS M1
Python 3.11.8 | Why redis vectorstore filter parameter make syntax error | https://api.github.com/repos/langchain-ai/langchain/issues/19323/comments | 1 | 2024-03-20T10:02:16Z | 2024-06-27T16:08:19Z | https://github.com/langchain-ai/langchain/issues/19323 | 2,197,130,983 | 19,323 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
# postion: langchain_openai/chat_models/base.py
def _convert_delta_to_message_chunk(
_dict: Mapping[str, Any], default_class: Type[BaseMessageChunk]
) -> BaseMessageChunk:
role = cast(str, _dict.get("role"))
content = cast(str, _dict.get("content") or "")
additional_kwargs: Dict = {}
if _dict.get("function_call"):
function_call = dict(_dict["function_call"])
if "name" in function_call and function_call["name"] is None:
function_call["name"] = ""
additional_kwargs["function_call"] = function_call
if _dict.get("tool_calls"):
additional_kwargs["tool_calls"] = _dict["tool_calls"]
if role == "user" or default_class == HumanMessageChunk:
return HumanMessageChunk(content=content)
elif role == "assistant" or default_class == AIMessageChunk:
return AIMessageChunk(content=content, additional_kwargs=additional_kwargs)
elif role == "system" or default_class == SystemMessageChunk:
return SystemMessageChunk(content=content)
elif role == "function" or default_class == FunctionMessageChunk:
return FunctionMessageChunk(content=content, name=_dict["name"])
elif role == "tool" or default_class == ToolMessageChunk:
return ToolMessageChunk(content=content, tool_call_id=_dict["tool_call_id"])
elif role or default_class == ChatMessageChunk:
return ChatMessageChunk(content=content, role=role)
else:
return default_class(content=content) # type: ignore
```
### Error Message and Stack Trace (if applicable)
Traceback (most recent call last):
File "/Users/sunny/Documents/Codes/ai/themis/api/utils.py", line 15, in wrap_done
await fn
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_core/_api/deprecation.py", line 154, in awarning_emitting_wrapper
return await wrapped(*args, **kwargs)
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain/chains/base.py", line 428, in acall
return await self.ainvoke(
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain/chains/base.py", line 212, in ainvoke
raise e
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain/chains/base.py", line 203, in ainvoke
await self._acall(inputs, run_manager=run_manager)
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain/chains/llm.py", line 275, in _acall
response = await self.agenerate([inputs], run_manager=run_manager)
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain/chains/llm.py", line 142, in agenerate
return await self.llm.agenerate_prompt(
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 581, in agenerate_prompt
return await self.agenerate(
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 541, in agenerate
raise exceptions[0]
File "/Users/sunny/opt/anaconda3/envs/py3.10/lib/python3.10/asyncio/tasks.py", line 232, in __step
result = coro.send(None)
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 645, in _agenerate_with_cache
result = await self._agenerate(
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_openai/chat_models/base.py", line 553, in _agenerate
return await agenerate_from_stream(stream_iter)
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_core/language_models/chat_models.py", line 84, in agenerate_from_stream
async for chunk in stream:
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_openai/chat_models/base.py", line 521, in _astream
chunk = _convert_delta_to_message_chunk(
File "/Users/sunny/.local/share/virtualenvs/themis-l6RndCmc/lib/python3.10/site-packages/langchain_openai/chat_models/base.py", line 176, in _convert_delta_to_message_chunk
role = cast(str, _dict.get("role"))
AttributeError: 'NoneType' object has no attribute 'get'
### Description
I'm currently using langchain_openai to execute Azure's Streaming API, but it encountered the above error. From the Azure API result, I noticed:
{'delta': None, 'finish_reason': None, 'index': 0, 'logprobs': None, 'content_filter_results': {'hate': {'filtered': False, 'severity': 'safe'}, 'self_harm': {'filtered': False, 'severity': 'safe'}, 'sexual': {'filtered': False, 'severity': 'safe'}, 'violence': {'filtered': False, 'severity': 'safe'}}, 'content_filter_offsets': {'check_offset': 252, 'start_offset': 252, 'end_offset': 263}}
The delta is None. However, in the function langchain_openai/chat_models/base.py/_convert_delta_to_message_chunk, the _dict parameter is None, which leads to the error.
### System Info
langchain==0.1.10
langchain-community==0.0.25
langchain-core==0.1.32
langchain-experimental==0.0.48
langchain-google-genai==0.0.11
langchain-openai==0.0.7
langchain-text-splitters==0.0.1
| langchain_openai - bug: 'NoneType' object has no attribute 'get' | https://api.github.com/repos/langchain-ai/langchain/issues/19318/comments | 0 | 2024-03-20T07:41:16Z | 2024-03-20T07:43:02Z | https://github.com/langchain-ai/langchain/issues/19318 | 2,196,869,387 | 19,318 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
On the web page https://python.langchain.com/docs/use_cases/question_answering/quickstart when user clicks on open in colab the notebook is not present at
https://colab.research.google.com/github/langchain-ai/langchain/blob/master/docs/docs/use_cases/question_answering/quickstart.ipynb
### Idea or request for content:
please fix the link | colab link is not working to quickstart.ipynb | https://api.github.com/repos/langchain-ai/langchain/issues/19304/comments | 0 | 2024-03-20T03:25:08Z | 2024-06-26T16:07:55Z | https://github.com/langchain-ai/langchain/issues/19304 | 2,196,594,832 | 19,304 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_core.pydantic_v1 import BaseModel, Field
from langchain_core.runnables import ConfigurableField
from langchain_openai import ChatOpenAI
class Joke(BaseModel):
setup: str = Field(description="The setup of the joke")
punchline: str = Field(description="The punchline to the joke")
model = ChatOpenAI(model="gpt-3.5-turbo",temperature=0).configurable_alternatives(
ConfigurableField(id="chat_model"),
default_key="gpt-3.5-turbo",
gpt_4=ChatOpenAI(model="gpt-4-0125-preview"),
).with_structured_output(Joke)
```
### Error Message and Stack Trace (if applicable)
AttributeError: 'RunnableConfigurableAlternatives' object has no attribute 'with_structured_output'
### Description
Need to think through how to support cases like this without overwriting methods like `with_structured_output`.
### System Info
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 22.4.0: Mon Mar 6 20:59:58 PST 2023; root:xnu-8796.101.5~3/RELEASE_ARM64_T6020
> Python Version: 3.11.7 (main, Feb 12 2024, 12:44:48) [Clang 14.0.3 (clang-1403.0.22.14.1)]
Package Information
-------------------
> langchain_core: 0.1.32
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.26
> langchain_fireworks: 0.1.1
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
> langserve: 0.0.45
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph | core: binding sometimes overrides methods | https://api.github.com/repos/langchain-ai/langchain/issues/19279/comments | 0 | 2024-03-19T16:04:35Z | 2024-06-25T16:41:42Z | https://github.com/langchain-ai/langchain/issues/19279 | 2,195,379,111 | 19,279 |
[
"langchain-ai",
"langchain"
] | ### 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
from langchain_community.llms import Ollama
llm = Ollama(base_url = )
llm.invoke("Tell me a joke")
### Error Message and Stack Trace (if applicable)
\python\python39\lib\site-packages (from dataclasses-json<0.7,>=0.5.7->langchain_community) (0.9.0)
Requirement already satisfied: pydantic<3,>=1 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from langchain-core<0.2.0,>=0.1.31->langchain_community) (1.10.14)
Requirement already satisfied: anyio<5,>=3 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from langchain-core<0.2.0,>=0.1.31->langchain_community) (4.3.0)
Requirement already satisfied: jsonpatch<2.0,>=1.33 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from langchain-core<0.2.0,>=0.1.31->langchain_community) (1.33)
Requirement already satisfied: packaging<24.0,>=23.2 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from langchain-core<0.2.0,>=0.1.31->langchain_community) (23.2)Requirement already satisfied: urllib3<3,>=1.21.1 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from requests<3,>=2->langchain_community) (2.2.1)
Requirement already satisfied: idna<4,>=2.5 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from requests<3,>=2->langchain_community) (3.6)
Requirement already satisfied: certifi>=2017.4.17 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from requests<3,>=2->langchain_community) (2024.2.2)
Requirement already satisfied: charset-normalizer<4,>=2 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from requests<3,>=2->langchain_community) (3.3.2)
Requirement already satisfied: aiosignal>=1.1.2 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (1.3.1)
Requirement already satisfied: async-timeout<5.0,>=4.0; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (4.0.3)
Requirement already satisfied: yarl<2.0,>=1.0 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (1.9.4)
Requirement already satisfied: frozenlist>=1.1.1 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (1.4.1)
Requirement already satisfied: multidict<7.0,>=4.5 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (6.0.5)
Requirement already satisfied: attrs>=17.3.0 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from aiohttp<4.0.0,>=3.8.3->langchain_community) (23.2.0)
Requirement already satisfied: orjson<4.0.0,>=3.9.14 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from langsmith<0.2.0,>=0.1.0->langchain_community) (3.9.15)
Requirement already satisfied: greenlet!=0.4.17; python_version >= "3" and (platform_machine ==
"aarch64" or (platform_machine == "ppc64le" or (platform_machine == "x86_64" or (platform_machine == "amd64" or (platform_machine == "AMD64" or (platform_machine == "win32" or platform_machine == "WIN32")))))) in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from SQLAlchemy<3,>=1.4->langchain_community) (3.0.3)
Requirement already satisfied: typing-extensions>=3.7.4 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from typing-inspect<1,>=0.4.0->dataclasses-json<0.7,>=0.5.7->langchain_community) (4.10.0)
Requirement already satisfied: mypy-extensions>=0.3.0 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from typing-inspect<1,>=0.4.0->dataclasses-json<0.7,>=0.5.7->langchain_community) (1.0.0)
Requirement already satisfied: sniffio>=1.1 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio<5,>=3->langchain-core<0.2.0,>=0.1.31->langchain_community) (1.3.1)
Requirement already satisfied: exceptiongroup>=1.0.2; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio<5,>=3->langchain-core<0.2.0Requirement already satisfied: jsonpointer>=1.9 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from jsonpatch<2.0,>=1.33->langchain-core<0.2.0,>=0.1.31->langchain_community) (2.4)
WARNING: You are using pip version 20.2.3; however, version 24.0 is available.
You should consider upgrading via the 'c:\users\maste\appdata\local\programs\python\python39\python.exe -m pip install --upgrade pip' command.
PS C:\Users\maste> pip install Ollama
Collecting Ollama
Downloading ollama-0.1.7-py3-none-any.whl (9.4 kB)
Collecting httpx<0.26.0,>=0.25.2
Downloading httpx-0.25.2-py3-none-any.whl (74 kB)
|████████████████████████████████| 74 kB 2.6 MB/s
Requirement already satisfied: anyio in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->Ollama) (4.3.0)
Requirement already satisfied: certifi in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->Ollama) (2024.2.2)
Requirement already satisfied: sniffio in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->Ollama) (1.3.1)
Collecting httpcore==1.*
Downloading httpcore-1.0.4-py3-none-any.whl (77 kB)
|████████████████████████████████| 77 kB ...
Requirement already satisfied: idna in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->Ollama) (3.6)
Requirement already satisfied: exceptiongroup>=1.0.2; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio->httpx<0.26.0,>=0.25.2->Ollama) (1.2.0)
Requirement already satisfied: typing-extensions>=4.1; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio->httpx<0.26.0,>=0.25.2->Ollama) (4.10.0)
Collecting h11<0.15,>=0.13
|████████████████████████████████| 58 kB ...
Installing collected packages: h11, httpcore, httpx, Ollama
Successfully installed Ollama-0.1.7 h11-0.14.0 httpcore-1.0.4 httpx-0.25.2
WARNING: You are using pip version 20.2.3; however, version 24.0 is available.
You should consider upgrading via the 'c:\users\maste\appdata\local\programs\python\python39\python.exe -m pip install --upgrade pip' command.
PS C:\Users\maste> pip install ollama
Requirement already satisfied: ollama in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (0.1.7)
Requirement already satisfied: httpx<0.26.0,>=0.25.2 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from ollama) (0.25.2)
Requirement already satisfied: httpcore==1.* in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->ollama) (1.0.4)
Requirement already satisfied: sniffio in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->ollama) (1.3.1)
Requirement already satisfied: anyio in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->ollama) (4.3.0)
Requirement already satisfied: certifi in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->ollama) (2024.2.2)
Requirement already satisfied: idna in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpx<0.26.0,>=0.25.2->ollama) (3.6)
Requirement already satisfied: h11<0.15,>=0.13 in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from httpcore==1.*->httpx<0.26.0,>=0.25.2->ollama) (0.14.0)
Requirement already satisfied: exceptiongroup>=1.0.2; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio->httpx<0.26.0,>=0.25.2->ollRequirement already satisfied: typing-extensions>=4.1; python_version < "3.11" in c:\users\maste\appdata\local\programs\python\python39\lib\site-packages (from anyio->httpx<0.26.0,>=0.25.2->ollama) (4.10.0)
WARNING: You are using pip version 20.2.3; however, version 24.0 is available.
You should consider upgrading via the 'c:\users\maste\appdata\local\programs\python\python39\python.exe -m pip install --upgrade pip' command.
PS C:\Users\maste> python hello.py
Traceback (most recent call last):
File "C:\Users\maste\hello.py", line 5, in <module>
llm.invoke("Tell me a joke")
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\language_models\llms.py", line 246, in invoke
self.generate_prompt(
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\language_models\llms.py", line 541, in generate_prompt
return self.generate(prompt_strings, stop=stop, callbacks=callbacks, **kwargs)
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\language_models\llms.py", line 671, in generate
CallbackManager.configure(
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\callbacks\manager.py", line 1443, in configure
return _configure(
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\callbacks\manager.py", line 1950, in _configure
debug = _get_debug()
PS C:\Users\maste> python hello.py
Traceback (most recent call last):
File "C:\Users\maste\hello.py", line 19, in <module>
print(chain.invoke({"topic": "Space travel"}))
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\runnables\base.py", line 2209, in invoke
callback_manager = get_callback_manager_for_config(config)
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\runnables\config.py", line 381, in get_callback_manager_for_config
return CallbackManager.configure(
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\callbacks\manager.py", line 1443, in configure
return _configure(
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\callbacks\manager.py", line 1950, in _configure
debug = _get_debug()
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\callbacks\manager.py", line 57, in _get_debug
return get_debug()
File "C:\Users\maste\AppData\Local\Programs\Python\Python39\lib\site-packages\langchain_core\globals\__init__.py", line 129, in get_debug
old_debug = langchain.debug
AttributeError: module '
### Description
All I am trying to do is use Ollama using langchain for my web app. The invoke function always says debug does not exists and gives me an attribute error
### System Info
from langchain_community.llms import Ollama
llm = Ollama(base_url = )
llm.invoke("Tell me a joke") | AttributeError: module 'langchain' has no attribute 'debug' | https://api.github.com/repos/langchain-ai/langchain/issues/19278/comments | 4 | 2024-03-19T16:04:33Z | 2024-08-02T06:25:55Z | https://github.com/langchain-ai/langchain/issues/19278 | 2,195,379,015 | 19,278 |
[
"langchain-ai",
"langchain"
] | ### Privileged issue
- [X] I am a LangChain maintainer, or was asked directly by a LangChain maintainer to create an issue here.
### Issue Content
# Goal
Allow instantiating language models with specific caches provided as an init parameter. This will bring language models on feature parity with chat models w/ respect to caching behavior.
This is the `cache` parameter: https://github.com/langchain-ai/langchain/blob/50f93d86ec56a92e1d0f5b390514d9a67a95d083/libs/core/langchain_core/language_models/base.py#L82-L82
Implementation is required in BaseLLM for both sync and async paths: https://github.com/langchain-ai/langchain/blob/50f93d86ec56a92e1d0f5b390514d9a67a95d083/libs/core/langchain_core/language_models/llms.py#L737-L737
Here's a reference implementation for chat models: https://github.com/langchain-ai/langchain/pull/17386
## Acceptance criteria
* The PR must include unit tests that provide coverage of the various caching configurations. You can look at the reference PR for Chat Models which covers the relevant scenarios. | langchain-core: Allow passing local cache to language models | https://api.github.com/repos/langchain-ai/langchain/issues/19276/comments | 3 | 2024-03-19T15:36:18Z | 2024-04-05T15:19:56Z | https://github.com/langchain-ai/langchain/issues/19276 | 2,195,311,866 | 19,276 |
[
"langchain-ai",
"langchain"
] | I get an error when following the example notebook
https://python.langchain.com/docs/use_cases/extraction/quickstart
The script below runs fine if the schema is set to `Person`
```
runnable = prompt | llm.with_structured_output(schema=Person)
```
However, it failed when the schema is set to `Data`
```python
runnable = prompt | llm.with_structured_output(schema=Data)
```
### Version
```python
import langchain
from google.cloud import aiplatform
print(f"LangChain version: {langchain.__version__}")
print(f"Vertex AI SDK version: {aiplatform.__version__}")
```
LangChain version: 0.1.11
Vertex AI SDK version: 1.43.0
### Script
```python
from typing import List, Optional
from langchain_core.pydantic_v1 import BaseModel, Field
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.pydantic_v1 import BaseModel, Field
from langchain_google_vertexai import ChatVertexAI, HarmCategory, HarmBlockThreshold
class Person(BaseModel):
"""Information about a person."""
name: Optional[str] = Field(..., description="The name of the person")
hair_color: Optional[str] = Field(
..., description="The color of the person's hair if known"
)
height_in_meters: Optional[str] = Field(
..., description="Height measured in meters"
)
class Data(BaseModel):
"""Extracted data about people."""
people: List[Person]
prompt = ChatPromptTemplate.from_messages(
[
(
"system",
"You are an expert extraction algorithm. "
"Only extract relevant information from the text. "
"If you do not know the value of an attribute asked to extract, "
"return null for the attribute's value.",
),
# Please see the how-to about improving performance with
# reference examples.
# MessagesPlaceholder('examples'),
("human", "{text}"),
]
)
llm = ChatVertexAI(
model_name="gemini-pro",
temperature=0,
safety_settings={
HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT: HarmBlockThreshold.BLOCK_NONE,
},
convert_system_message_to_human=True,
)
text = "My name is Jeff, my hair is black and i am 6 feet tall. Anna has the same color hair as me."
runnable = prompt | llm.with_structured_output(schema=Data)
runnable.invoke({"text": text})
```
### Error
```
File ~/scratch/conda/envs/langchain-vertexai/lib/python3.10/site-packages/langchain_google_vertexai/functions_utils.py:160, in PydanticFunctionsOutputParser.parse_result(self, result, partial)
ValidationError: 2 validation errors for Data
people -> 0 -> hair_color
field required (type=value_error.missing)
people -> 1 -> hair_color
```
### Other questions
Is there an analog function to `convert_to_openai_tool` called `convert_to_vertexai_tool`?
```python
import json
from langchain_core.utils.function_calling import convert_to_openai_tool
print(json.dumps(convert_to_openai_tool(Data), indent=2))
```
Output:
```json
{
"type": "function",
"function": {
"name": "Data",
"description": "Extracted data about people.",
"parameters": {
"type": "object",
"properties": {
"people": {
"type": "array",
"items": {
"description": "Information about a person.",
"type": "object",
"properties": {
"name": {
"description": "The name of the person",
"type": "string"
},
"hair_color": {
"description": "The color of the person's hair if known",
"type": "string"
},
"height_in_meters": {
"description": "Height measured in meters",
"type": "string"
}
},
"required": [
"name",
"hair_color",
"height_in_meters"
]
}
}
},
"required": [
"people"
]
}
}
}
```
The `Person` class parameters were set to optional.
Why are the `Person` parameters set to required in the OpenAI tool?
How to see the generated Schema used in the call of the tool?
https://ai.google.dev/api/python/google/ai/generativelanguage/Schema
_Originally posted by @schinto in https://github.com/langchain-ai/langchain/discussions/18975#discussioncomment-8840417_ | Multiple entities extraction in quickstart demo fails with ChatVertexAI | https://api.github.com/repos/langchain-ai/langchain/issues/19272/comments | 0 | 2024-03-19T14:20:02Z | 2024-06-25T16:13:27Z | https://github.com/langchain-ai/langchain/issues/19272 | 2,195,111,331 | 19,272 |
[
"langchain-ai",
"langchain"
] | ### 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
import os
from langchain.llms import OpenAI
from langchain.utilities import SQLDatabase
# ... Your database credentials ...
os.environ["OPENAI_API_KEY"] = "sk-myapikey"
db_uri = f"postgresql://{user}:{password}@{host}/{database}"
db = SQLDatabase(db_uri) # Error occurs here
### Error Message and Stack Trace (if applicable)
NoInspectionAvailable: No inspection system is available for object of type <class 'str'>
Traceback:
File "C:\Python311\Lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 542, in _run_script
exec(code, module.__dict__)
File "C:\opensql\gemSQL.py", line 19, in <module>
db = SQLDatabase(db_uri)
^^^^^^^^^^^^^^^^^^^
File "C:\Python311\Lib\site-packages\langchain_community\utilities\sql_database.py", line 69, in __init__
self._inspector = inspect(self._engine)
^^^^^^^^^^^^^^^^^^^^^
File "C:\Python311\Lib\site-packages\sqlalchemy\inspection.py", line 71, in inspect
raise exc.NoInspectionAvailable(
### Description
I'm encountering a NoInspectionAvailable error when using the SQLDatabase class in LangChain. However, basic SQLAlchemy connections and introspection work correctly.
### System Info
OS: windows 10
Python: 3.19
LangChain: 0.1.12
SQLAlchemy: 1.4.47 | SQLDatabase introspection fails with NoInspectionAvailable | https://api.github.com/repos/langchain-ai/langchain/issues/19264/comments | 2 | 2024-03-19T12:02:31Z | 2024-06-25T16:13:24Z | https://github.com/langchain-ai/langchain/issues/19264 | 2,194,779,898 | 19,264 |
[
"langchain-ai",
"langchain"
] | ### 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
Impacted code is https://github.com/langchain-ai/langchain/blob/514fe807784f520449914a64ffc983538fa6743d/libs/community/langchain_community/utilities/sql_database.py#L426
```
elif self.dialect == "trino":
connection.exec_driver_sql(
"USE ?",
(self._schema,),
execution_options=execution_options,
```
Trino cannot use parameterized syntax for USE
Generated SQL statement to Trino should be
`EXECUTE IMMEDIATE 'USE mycatalog';`
instead of
`EXECUTE IMMEDIATE 'USE ?' USING 'mycatalog';`
### Error Message and Stack Trace (if applicable)
On Trino side, we got
```
io.trino.sql.parser.ParsingException: line 1:24: mismatched input '?'. Expecting: <identifier>
at io.trino.sql.parser.SqlParser.lambda$invokeParser$1(SqlParser.java:183)
at java.base/java.util.Optional.ifPresent(Optional.java:178)
```
### Description
Without patching Trino dialect for SQL prompts, Trino cannot be used as a SQL source for LangChain
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.31
langchain-experimental==0.0.54
langchain-openai==0.0.8
langchain-text-splitters==0.0.1 | SQL_database Trino dialect - wrong usage of USE to set catalog, as Trino does not use parameters for identifiers | https://api.github.com/repos/langchain-ai/langchain/issues/19261/comments | 1 | 2024-03-19T10:57:49Z | 2024-07-04T16:08:33Z | https://github.com/langchain-ai/langchain/issues/19261 | 2,194,647,431 | 19,261 |
[
"langchain-ai",
"langchain"
] | ### 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
```
import `json`
import redis
from fastapi import APIRouter, status
from fastapi.encoders import jsonable_encoder
from fastapi.responses import JSONResponse
from langchain.chains import ConversationalRetrievalChain, ConversationChain
from langchain.callbacks.base import AsyncCallbackHandler
from langchain.callbacks.manager import AsyncCallbackManager
from fastapi.responses import StreamingResponse
from typing import Any, Awaitable, Callable, Iterator, Optional, Union
from langchain.chains.conversational_retrieval.prompts import (
CONDENSE_QUESTION_PROMPT,
QA_PROMPT,
)
from langchain.chains.llm import LLMChain
from langchain.chains.question_answering import load_qa_chain
from langchain_community.chat_models import AzureChatOpenAI
from langchain_openai import AzureOpenAIEmbeddings
from langchain.memory import ConversationBufferMemory, ConversationBufferWindowMemory
from langchain_community.chat_message_histories import RedisChatMessageHistory
from langchain.prompts import (
ChatPromptTemplate,
HumanMessagePromptTemplate,
MessagesPlaceholder,
SystemMessagePromptTemplate,
)
from langchain.prompts.prompt import PromptTemplate
from langchain_community.vectorstores import Milvus
from pydantic import BaseModel, validator
from typing import Optional
from starlette.types import Send
from genai_openwork_backend.services.coroutine.loop import get_loop
from genai_openwork_backend.app.api.chat_history.views import (
get_recent_chats_rag,
cache_chat
)
from genai_openwork_backend.db.connection import get_connection, aget_connection,release_connection
from genai_openwork_backend.config import config
from datetime import datetime
import re
from enum import Enum as PyEnum
router = APIRouter()
redis_host = config.redis['REDIS_HOST']
redis_port = config.redis['REDIS_PORT']
openai_api_version = config.llm["OPENAI_API_VERSION"]
deployment_name=config.llm['DEPLOYMENT_NAME']
model_name=config.llm['LLM_MODEL']
openai_api_base=config.llm['AZURE_OPENAI_ENDPOINT']
deployment=config.llm['EMBEDDING']
model=config.llm['MODEL']
openai_api_type=config.llm['OPENAI_API_TYPE']
milvus_host = config.vectordb['MILIVS_HOST']
milvus_port = config.vectordb['MILIVUS_PORT']
# No change here - Using the default version
CONDENSE_QUESTION_PROMPT = PromptTemplate(
input_variables=[
"chat_history",
"question",
],
template="Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question, in its original language.\n\nChat History:\n{chat_history}\nFollow Up Input: {question}\nStandalone question:",
)
# QA prompt updated
_template = """You are a helpful AI assisstant. The following is a friendly conversation between a human and you.
Use the following documents provided as context to answer the question at the end . If you don't know the answer, just say you don't know. DO NOT try to make up an answer.
If the question is not related to the context, politely respond that you are tuned to only answer questions that are related to the context.
Also, generate three brief follow-up questions that the user would likely ask next. Try not to repeat questions that have already been asked. Only generate questions in next line with a tag 'Next Questions' as a Markdown list.
{question}
=========
{context}
=========
Answer:"""
variables = ["context", "question"]
QA_PROMPT = PromptTemplate(
template=_template,
input_variables=variables,
)
class AllowedChatModels(str, PyEnum):
value1 = "gpt-3.5"
value2 = "gpt-4"
class ConversationStyles(str, PyEnum):
value1 = "precise"
value2 = "balanced"
value3 = "creative"
class RagChatRequest(BaseModel):
"""Request model for chat requests.
Includes the conversation ID and the message from the user.
"""
user_id: str
conversation_id: str
question: str
collection: Optional[list] = ["all"]
vectordb_collection : Optional[str] = "openwork"
chatModel : Optional[AllowedChatModels] = "gpt-3.5"
conversationStyle : Optional[ConversationStyles] = "precise"
Sender = Callable[[Union[str, bytes]], Awaitable[None]]
class EmptyIterator(Iterator[Union[str, bytes]]):
def __iter__(self):
return self
def __next__(self):
raise StopIteration
class AsyncStreamCallbackHandler(AsyncCallbackHandler):
"""Callback handler for streaming, inheritance from AsyncCallbackHandler."""
def __init__(self, send: Sender):
super().__init__()
self.send = send
async def on_llm_new_token(self, token: str, **kwargs: Any) -> None:
"""Rewrite on_llm_new_token to send token to client."""
await self.send(f"{token}")
class ChatOpenAIStreamingResponse(StreamingResponse):
"""Streaming response for openai chat model, inheritance from StreamingResponse."""
def __init__(
self,
generate: Callable[[Sender], Awaitable[None]],
request,
status_code: int = 200,
media_type: Optional[str] = None,
) -> None:
super().__init__(
content=EmptyIterator(),
status_code=status_code,
media_type=media_type,
)
self.generate = generate
self.request = request
self.answer = b''
async def stream_response(self, send: Send) -> None:
"""Rewrite stream_response to send response to client."""
await send(
{
"type": "http.response.start",
"status": self.status_code,
"headers": self.raw_headers,
},
)
async def send_chunk(chunk: Union[str, bytes]):
if not isinstance(chunk, bytes):
chunk = chunk.encode(self.charset)
self.answer += chunk
await send({"type": "http.response.body", "body": chunk, "more_body": True})
# send body to client
await self.generate(send_chunk)
# send empty body to client to close connection
await send({"type": "http.response.body", "body": b"", "more_body": False})
def transformConversationStyleToTemperature(style : str):
style_temp_obj = {
"precise" : 1,
"balanced" : 0.5,
"creative" : 0.1
}
return style_temp_obj.get(style)
def send_message_tredence_llm(
query: RagChatRequest,
) -> Callable[[Sender], Awaitable[None]]:
async def generate(send: Sender):
temperature = transformConversationStyleToTemperature(query.conversationStyle)
chat_model = AzureChatOpenAI(
streaming=True,
azure_endpoint=openai_api_base,
deployment_name=deployment_name,
model_name=model_name,
openai_api_version=openai_api_version,
verbose=True,
)
chat_model2 = AzureChatOpenAI(
streaming=False,
azure_endpoint=openai_api_base,
deployment_name=deployment_name,
model_name=model_name,
openai_api_version=openai_api_version,
verbose=True,
)
chat_model.temperature = temperature
chat_model2.temperature = temperature
embeddings = AzureOpenAIEmbeddings(
deployment=deployment,
model=str(model),
azure_endpoint=openai_api_base,
openai_api_type=openai_api_type,
openai_api_version=openai_api_version,
)
vectorstore = Milvus(
embeddings,
collection_name=query.vectordb_collection,
connection_args={"host": milvus_host, "port": milvus_port},
)
chain_input = query.question
memory = ConversationBufferWindowMemory( k=10, return_messages=True,memory_key="chat_history")
chat_list = await get_recent_chats_rag(query.conversation_id)
if(len(chat_list)):
for c in chat_list:
memory.save_context({"input": c["input"]}, {"output": c["output"]})
# Set up the chain
question_generator = LLMChain(
llm=chat_model2,
prompt=CONDENSE_QUESTION_PROMPT,
)
doc_chain = load_qa_chain(
llm=chat_model,
chain_type="stuff",
prompt=QA_PROMPT,
# callback_manager=AsyncCallbackManager(
# [AsyncStreamCallbackHandler(send)],
# ),
)
if len(query.collection) == 1 and query.collection[0] == "all":
expression = ""
else:
expression = f'group in ["{query.collection[0]}"'
for i in range(1,len(query.collection)):
expression += f',"{query.collection[i]}"'
expression += "]"
print("expression", expression)
chain = ConversationalRetrievalChain(
memory=memory,
combine_docs_chain=doc_chain,
question_generator=question_generator,
retriever=vectorstore.as_retriever(
search_type="similarity", search_kwargs={"k": 4, "expr": f"{expression}"}
),
verbose=True,
return_source_documents=True
)
history = memory.chat_memory.messages
print(history)
await chain.acall(chain_input, callbacks=[AsyncStreamCallbackHandler(send)])
return generate
@router.post("/rag/stream")
async def stream(request: RagChatRequest):
return ChatOpenAIStreamingResponse(
send_message_tredence_llm(request),
request,
media_type="text/event-stream",
)
```
### Error Message and Stack Trace (if applicable)
> Entering new ConversationalRetrievalChain chain...
> Finished chain.
2024-03-19 09:40:13.368 | ERROR | trace_id=0 | span_id=0 | uvicorn.protocols.http.httptools_impl:run_asgi:424 - Exception in ASGI application
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/spawn.py", line 116, in spawn_main
exitcode = _main(fd, parent_sentinel)
│ │ └ 4
│ └ 10
└ <function _main at 0x7f0ec497bd00>
File "/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/spawn.py", line 129, in _main
return self._bootstrap(parent_sentinel)
│ │ └ 4
│ └ <function BaseProcess._bootstrap at 0x7f0ec4b627a0>
└ <SpawnProcess name='SpawnProcess-7' parent=2267916 started>
File "/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/process.py", line 314, in _bootstrap
self.run()
│ └ <function BaseProcess.run at 0x7f0ec4b61e10>
└ <SpawnProcess name='SpawnProcess-7' parent=2267916 started>
File "/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
│ │ │ │ │ └ {'config': <uvicorn.config.Config object at 0x7f0ec4ca2950>, 'target': <bound method Server.run of <uvicorn.server.Server obj...
│ │ │ │ └ <SpawnProcess name='SpawnProcess-7' parent=2267916 started>
│ │ │ └ ()
│ │ └ <SpawnProcess name='SpawnProcess-7' parent=2267916 started>
│ └ <function subprocess_started at 0x7f0ec3fc9ea0>
└ <SpawnProcess name='SpawnProcess-7' parent=2267916 started>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/uvicorn/_subprocess.py", line 76, in subprocess_started
target(sockets=sockets)
│ └ [<socket.socket fd=3, family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0, laddr=('0.0.0.0', 1785)>]
└ <bound method Server.run of <uvicorn.server.Server object at 0x7f0ec4ca28f0>>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/uvicorn/server.py", line 60, in run
return asyncio.run(self.serve(sockets=sockets))
│ │ │ │ └ [<socket.socket fd=3, family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0, laddr=('0.0.0.0', 1785)>]
│ │ │ └ <function Server.serve at 0x7f0ec3fc9360>
│ │ └ <uvicorn.server.Server object at 0x7f0ec4ca28f0>
│ └ <function run at 0x7f0ec4991360>
└ <module 'asyncio' from '/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/asyncio/__init__.py'>
File "/home/azureuser/.pyenv/versions/3.10.13/lib/python3.10/asyncio/runners.py", line 44, in run
return loop.run_until_complete(main)
│ │ └ <coroutine object Server.serve at 0x7f0ec3eb21f0>
│ └ <method 'run_until_complete' of 'uvloop.loop.Loop' objects>
└ <uvloop.Loop running=True closed=False debug=False>
> File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/uvicorn/protocols/http/httptools_impl.py", line 419, in run_asgi
result = await app( # type: ignore[func-returns-value]
└ <uvicorn.middleware.proxy_headers.ProxyHeadersMiddleware object at 0x7f0e8dc1cd00>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/uvicorn/middleware/proxy_headers.py", line 78, in __call__
return await self.app(scope, receive, send)
│ │ │ │ └ <bound method RequestResponseCycle.send of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc707...
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <fastapi.applications.FastAPI object at 0x7f0ea55efb80>
└ <uvicorn.middleware.proxy_headers.ProxyHeadersMiddleware object at 0x7f0e8dc1cd00>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/fastapi/applications.py", line 270, in __call__
await super().__call__(scope, receive, send)
│ │ └ <bound method RequestResponseCycle.send of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc707...
│ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
└ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/applications.py", line 124, in __call__
await self.middleware_stack(scope, receive, send)
│ │ │ │ └ <bound method RequestResponseCycle.send of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc707...
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <starlette.middleware.errors.ServerErrorMiddleware object at 0x7f0e8dc1df30>
└ <fastapi.applications.FastAPI object at 0x7f0ea55efb80>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/middleware/errors.py", line 184, in __call__
raise exc
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/middleware/errors.py", line 162, in __call__
await self.app(scope, receive, _send)
│ │ │ │ └ <function ServerErrorMiddleware.__call__.<locals>._send at 0x7f0e8dc743a0>
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <starlette.middleware.exceptions.ExceptionMiddleware object at 0x7f0e8dc1c850>
└ <starlette.middleware.errors.ServerErrorMiddleware object at 0x7f0e8dc1df30>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/middleware/exceptions.py", line 79, in __call__
raise exc
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/middleware/exceptions.py", line 68, in __call__
await self.app(scope, receive, sender)
│ │ │ │ └ <function ExceptionMiddleware.__call__.<locals>.sender at 0x7f0e8dc74430>
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <fastapi.middleware.asyncexitstack.AsyncExitStackMiddleware object at 0x7f0ec1d63bb0>
└ <starlette.middleware.exceptions.ExceptionMiddleware object at 0x7f0e8dc1c850>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/fastapi/middleware/asyncexitstack.py", line 21, in __call__
raise e
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/fastapi/middleware/asyncexitstack.py", line 18, in __call__
await self.app(scope, receive, send)
│ │ │ │ └ <function ExceptionMiddleware.__call__.<locals>.sender at 0x7f0e8dc74430>
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <fastapi.routing.APIRouter object at 0x7f0ec1dc0be0>
└ <fastapi.middleware.asyncexitstack.AsyncExitStackMiddleware object at 0x7f0ec1d63bb0>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/routing.py", line 706, in __call__
await route.handle(scope, receive, send)
│ │ │ │ └ <function ExceptionMiddleware.__call__.<locals>.sender at 0x7f0e8dc74430>
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <function Route.handle at 0x7f0ec31b5000>
└ APIRoute(path='/api/openwork/rag/stream', name='stream', methods=['POST'])
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/routing.py", line 276, in handle
await self.app(scope, receive, send)
│ │ │ │ └ <function ExceptionMiddleware.__call__.<locals>.sender at 0x7f0e8dc74430>
│ │ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ │ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
│ └ <function request_response.<locals>.app at 0x7f0e8dc23d00>
└ APIRoute(path='/api/openwork/rag/stream', name='stream', methods=['POST'])
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/routing.py", line 69, in app
await response(scope, receive, send)
│ │ │ └ <function ExceptionMiddleware.__call__.<locals>.sender at 0x7f0e8dc74430>
│ │ └ <bound method RequestResponseCycle.receive of <uvicorn.protocols.http.httptools_impl.RequestResponseCycle object at 0x7f0e8dc...
│ └ {'type': 'http', 'asgi': {'version': '3.0', 'spec_version': '2.3'}, 'http_version': '1.1', 'server': ('127.0.0.1', 1785), 'cl...
└ <genai_openwork_backend.app.api.openwork.views.ChatOpenAIStreamingResponse object at 0x7f0e8dc70c40>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/responses.py", line 266, in __call__
async with anyio.create_task_group() as task_group:
│ │ └ <anyio._backends._asyncio.TaskGroup object at 0x7f0e8dc70b20>
│ └ <function create_task_group at 0x7f0ec3e409d0>
└ <module 'anyio' from '/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/an...
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 597, in __aexit__
raise exceptions[0]
└ [ValueError("One output key expected, got dict_keys(['answer', 'source_documents'])")]
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/starlette/responses.py", line 269, in wrap
await func()
└ functools.partial(<bound method ChatOpenAIStreamingResponse.stream_response of <genai_openwork_backend.app.api.openwork.views...
File "/home/azureuser/anindya/genai_openwork_backend/genai_openwork_backend/app/api/openwork/views.py", line 233, in stream_response
await self.generate(send_chunk)
│ │ └ <function ChatOpenAIStreamingResponse.stream_response.<locals>.send_chunk at 0x7f0e8dc74a60>
│ └ <function send_message_tredence_llm.<locals>.generate at 0x7f0e8dc74550>
└ <genai_openwork_backend.app.api.openwork.views.ChatOpenAIStreamingResponse object at 0x7f0e8dc70c40>
File "/home/azureuser/anindya/genai_openwork_backend/genai_openwork_backend/app/api/openwork/views.py", line 392, in generate
await chain.acall(chain_input, callbacks=[AsyncStreamCallbackHandler(send)])
│ │ │ │ └ <function ChatOpenAIStreamingResponse.stream_response.<locals>.send_chunk at 0x7f0e8dc74a60>
│ │ │ └ <class 'genai_openwork_backend.app.api.openwork.views.AsyncStreamCallbackHandler'>
│ │ └ 'explain about supply chain tower'
│ └ <function Chain.acall at 0x7f0eb5e0c940>
└ ConversationalRetrievalChain(memory=ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10), ve...
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain_core/_api/deprecation.py", line 154, in awarning_emitting_wrapper
return await wrapped(*args, **kwargs)
│ │ └ {'callbacks': [<genai_openwork_backend.app.api.openwork.views.AsyncStreamCallbackHandler object at 0x7f0e8db997b0>]}
│ └ (ConversationalRetrievalChain(memory=ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10), v...
└ <function Chain.acall at 0x7f0eb5e0c430>
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain/chains/base.py", line 413, in acall
return await self.ainvoke(
│ └ <function Chain.ainvoke at 0x7f0eb5ddbb50>
└ ConversationalRetrievalChain(memory=ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10), ve...
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain/chains/base.py", line 211, in ainvoke
final_outputs: Dict[str, Any] = self.prep_outputs(
│ │ │ └ <function Chain.prep_outputs at 0x7f0eb5e0c160>
│ │ └ ConversationalRetrievalChain(memory=ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10), ve...
│ └ typing.Any
└ typing.Dict
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain/chains/base.py", line 440, in prep_outputs
self.memory.save_context(inputs, outputs)
│ │ │ │ └ {'answer': 'The article explains that a Supply Chain Control Tower (SCCT) is a cross-departmental, system-integrated “informa...
│ │ │ └ {'question': 'explain about supply chain tower', 'chat_history': []}
│ │ └ <function BaseChatMemory.save_context at 0x7f0eb5c085e0>
│ └ ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10)
└ ConversationalRetrievalChain(memory=ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10), ve...
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain/memory/chat_memory.py", line 37, in save_context
input_str, output_str = self._get_input_output(inputs, outputs)
│ │ │ └ {'answer': 'The article explains that a Supply Chain Control Tower (SCCT) is a cross-departmental, system-integrated “informa...
│ │ └ {'question': 'explain about supply chain tower', 'chat_history': []}
│ └ <function BaseChatMemory._get_input_output at 0x7f0eb5c08550>
└ ConversationBufferWindowMemory(return_messages=True, memory_key='chat_history', k=10)
File "/home/azureuser/.pyenv/versions/3.10.13/envs/genai_openwork_backend_ani/lib/python3.10/site-packages/langchain/memory/chat_memory.py", line 29, in _get_input_output
raise ValueError(f"One output key expected, got {outputs.keys()}")
ValueError: One output key expected, got dict_keys(['answer', 'source_documents'])
### Description
I am trying to return the source documents along with the answer as a part of streamed response. I have put return_source_documents=True in ConversationalRetrievalChain parameters, however error is coming. If I comment it, then the answer gets streamed without any error. How to return the source documents in the stream?
### System Info
langchain==0.1.0
langchain-community==0.0.20
langchain-core==0.1.23
langchain-openai==0.0.5
openinference-instrumentation-langchain==0.1.12
python --> 3.10.13 | Streaming of Source Documents not working in ConversationalRetrievalChain | https://api.github.com/repos/langchain-ai/langchain/issues/19259/comments | 2 | 2024-03-19T09:44:53Z | 2024-07-04T16:08:28Z | https://github.com/langchain-ai/langchain/issues/19259 | 2,194,484,797 | 19,259 |
[
"langchain-ai",
"langchain"
] | ### 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
```
from langchain_openai import AzureChatOpenAI
from langchain_core.output_parsers import StrOutputParser
from langchain_core.runnables import RunnableLambda, RunnableSequence
from langchain_core.messages import HumanMessage
def langchain_model(prompt_func: callable) -> RunnableSequence:
model = AzureChatOpenAI(azure_deployment="gpt-35-16k", temperature=0)
return RunnableLambda(prompt_func) | model | StrOutputParser()
def prompt_func(
_dict: dict,
) -> list:
question = _dict.get("question")
texts = _dict.get("texts")
text_message = {
"type": "text",
"text": (
"You are a classification system for Procurement Documents. Answer the question solely on the provided Reference texts.\n"
"If you cant find a answer reply exactly like this: 'Sorry i dont have an answer for youre question'\n"
"Return the answer as a string in the language the question is written in.\n\n "
f"User-provided question: \n"
f"{question} \n\n"
"Reference texts:\n"
f"{texts}"
),
}
return [HumanMessage(content=[text_message])]
model = langchain_model(prompt_func=prompt_func)
for steps, runnable in model:
try:
print(runnable[0].dict())
except:
print(runnable)
```
### Error Message and Stack Trace (if applicable)
for AzureChatOpenAI as a component you will just get this output as dict:
{'model': 'gpt-3.5-turbo', 'stream': False, 'n': 1, 'temperature': 0.0, '_type': 'azure-openai-chat'}.
### Description
This is not enough if you want to log the model with MLFlow, the deployment_name is definitely needed and needs to be gotten from dict. See related issue: https://github.com/mlflow/mlflow/issues/11439. Expected output would have more details at least a deployment name.
### System Info
System Information
------------------
> OS: Darwin
> Python Version: 3.10.11 (v3.10.11:7d4cc5aa85, Apr 4 2023, 19:05:19) [Clang 13.0.0]
Package Information
-------------------
> langchain_core: 0.1.32
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.26
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | Getting from runnables.dict() in RunnableSequence not all expected variables. | https://api.github.com/repos/langchain-ai/langchain/issues/19255/comments | 3 | 2024-03-19T08:50:44Z | 2024-03-27T05:31:37Z | https://github.com/langchain-ai/langchain/issues/19255 | 2,194,374,025 | 19,255 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain.agents import AgentExecutor, create_openai_tools_agent
from langchain_core.prompts import (
ChatPromptTemplate,
HumanMessagePromptTemplate,
MessagesPlaceholder,
PromptTemplate,
SystemMessagePromptTemplate,
)
prompt = ChatPromptTemplate.from_messages(
[
_system_message,
MessagesPlaceholder(variable_name=memory_key, optional=True),
HumanMessagePromptTemplate.from_template("{input}"),
MessagesPlaceholder(variable_name="agent_scratchpad"),
]
)
agent = create_openai_tools_agent(llm, [], prompt)
agent_executor = AgentExecutor(
agent=agent,
tools=[], # type: ignore
memory=memory,
verbose=True,
return_intermediate_steps=True,
handle_parsing_errors=True,
)
print(agent_executor.input_keys) # Return empty list
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
Since `create_openai_tools_agent` returns a `RunnableSequence` and not a `BaseSingleActionAgent`, the property `input_keys` of the `AgentExecutor` doesn't work for this agent anymore
```python
@property
def input_keys(self) -> List[str]:
"""Return the input keys.
:meta private:
"""
return self.agent.input_keys
```
### System Info
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 23.4.0: Wed Feb 21 21:44:54 PST 2024; root:xnu-10063.101.15~2/RELEASE_ARM64_T6030
> Python Version: 3.9.18 (main, Sep 11 2023, 08:25:10)
[Clang 14.0.6 ]
Package Information
-------------------
> langchain_core: 0.1.27
> langchain: 0.1.9
> langchain_community: 0.0.24
> langsmith: 0.1.10
> langchain_experimental: 0.0.52
> langchain_google_genai: 0.0.6
> langchain_openai: 0.0.6
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | `create_openai_tools_agent` (RunnableSequence) doesn't return input_keys, breaks `AgentExecutor.input_keys` | https://api.github.com/repos/langchain-ai/langchain/issues/19251/comments | 2 | 2024-03-19T07:12:22Z | 2024-06-28T16:07:38Z | https://github.com/langchain-ai/langchain/issues/19251 | 2,194,205,736 | 19,251 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
There is a spelling mistake in line 32.
### Idea or request for content:
I can fix this mistake.
Please assign this issue to me. | DOC: typo error in https://python.langchain.com/docs/modules/model_io/chat/index.mdx, line 32. | https://api.github.com/repos/langchain-ai/langchain/issues/19247/comments | 0 | 2024-03-19T01:44:49Z | 2024-06-25T16:13:28Z | https://github.com/langchain-ai/langchain/issues/19247 | 2,193,829,820 | 19,247 |
[
"langchain-ai",
"langchain"
] | ### 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
Bellow is the code where i am getting "HTTPError('403 Client Error: Forbidden for url: https://storage.googleapis.com/flash-rank/ms-marco-MultiBERT-L-12.zip')"
compressor = FlashrankRerank()
compressor_retriever = ContextualCompressionRetriever(base_compressor=compressor,
base_retriever= base_retriever
)
compressed_docs = compressor_retriever.get_relevant_documents(query)
return compressed_docs
### Error Message and Stack Trace (if applicable)
Tool run errored with error:
HTTPError('403 Client Error: Forbidden for url: https://storage.googleapis.com/flash-rank/ms-marco-MultiBERT-L-12.zip')Traceback (most recent call last):
### Description
1. I have developed an LLM app with chromadb and its working
2. I am integrating FlashRank Retriever using langchain, but i am getting the error
### System Info
python==3.10.4
langchain==0.1.9
FlashRank==0.1.69 | Not able to utilize flashrank reranker in langchain and getting error "HTTPError('403 Client Error: Forbidden for url: https://storage.googleapis.com/flash-rank/ms-marco-MultiBERT-L-12.zip')" | https://api.github.com/repos/langchain-ai/langchain/issues/19241/comments | 1 | 2024-03-18T19:47:20Z | 2024-06-24T16:14:32Z | https://github.com/langchain-ai/langchain/issues/19241 | 2,193,132,295 | 19,241 |
[
"langchain-ai",
"langchain"
] | ### 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
import sqlite3
from langchain_openai import ChatOpenAI
from langchain_community.utilities import SQLDatabase
from langchain_community.agent_toolkits import create_sql_agent
# Make new test database, create a timestamp table and column and insert value
database_path = 'testdb.sqlite'
timestamp_value = 672457193.7343056
create_table_sql = """
CREATE TABLE IF NOT EXISTS timestamps (
timestamp TIMESTAMP
);
"""
insert_sql = """
INSERT INTO timestamps (timestamp) VALUES (?);
"""
conn = sqlite3.connect(database_path)
cursor = conn.cursor()
cursor.execute(create_table_sql)
cursor.execute(insert_sql, (timestamp_value,))
conn.commit()
conn.close()
print("The table has been created and the value has been inserted successfully.")
db = SQLDatabase.from_uri(f"sqlite:///{database_path}")
prefix = """" You are a SQL expert that answers questions about a database. Use the tools below.
"""
agent_executor = create_sql_agent(
llm=ChatOpenAI(model="gpt-3.5-turbo", temperature=0, openai_api_key=openai_api_key, ),
db=db,
agent_type="openai-tools",
verbose=True,
prefix=prefix
)
response = agent_executor.invoke("What is this database used for?")
print(response)
### Error Message and Stack Trace (if applicable)
> Entering new SQL Agent Executor chain...
Invoking: `sql_db_list_tables` with ``
timestamps
Invoking: `sql_db_schema` with `{'table_names': 'timestamps'}`
responded: The database contains a table named "timestamps." Let me query the schema of this table to understand its structure.
Traceback (most recent call last):
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/main.py", line 305, in <module>
response = agent_executor.invoke("What is this databse used for?")
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/chains/base.py", line 163, in invoke
raise e
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/chains/base.py", line 153, in invoke
self._call(inputs, run_manager=run_manager)
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/agents/agent.py", line 1432, in _call
next_step_output = self._take_next_step(
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/agents/agent.py", line 1138, in _take_next_step
[
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/agents/agent.py", line 1138, in <listcomp>
[
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/agents/agent.py", line 1223, in _iter_next_step
yield self._perform_agent_action(
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain/agents/agent.py", line 1245, in _perform_agent_action
observation = tool.run(
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_core/tools.py", line 417, in run
raise e
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_core/tools.py", line 376, in run
self._run(*tool_args, run_manager=run_manager, **tool_kwargs)
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_community/tools/sql_database/tool.py", line 75, in _run
return self.db.get_table_info_no_throw(
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_community/utilities/sql_database.py", line 532, in get_table_info_no_throw
return self.get_table_info(table_names)
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_community/utilities/sql_database.py", line 352, in get_table_info
table_info += f"\n{self._get_sample_rows(table)}\n"
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/langchain_community/utilities/sql_database.py", line 377, in _get_sample_rows
sample_rows = list(
File "/Users/noneofyourbusiness/FICT/AFSTUDEER/testopenaidb/venv/lib/python3.10/site-packages/sqlalchemy/engine/result.py", line 529, in iterrows
make_row(raw_row) if make_row else raw_row
File "lib/sqlalchemy/cyextension/resultproxy.pyx", line 22, in sqlalchemy.cyextension.resultproxy.BaseRow.__init__
File "lib/sqlalchemy/cyextension/resultproxy.pyx", line 79, in sqlalchemy.cyextension.resultproxy._apply_processors
File "lib/sqlalchemy/cyextension/processors.pyx", line 40, in sqlalchemy.cyextension.processors.str_to_datetime
TypeError: fromisoformat: argument must be str
### Description
I'm using Langchain SQL-agent to connect a LLM to a existing database. This database has multiple tables, columns and rows. One of these columns is a timestamp column with the value 672457193.7343056
Note i didnt design this database(s), but want to work with them. Ive encountered multiple databases that use this formatting.
I have posted this on the github of SQLAlchemy aswel, and they have answerd in the following reply https://github.com/sqlalchemy/sqlalchemy/discussions/9990#discussioncomment-8828699
I have no idea how to fix this issue. I hope i can get some help here.
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-experimental==0.0.32
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
Mac OS
Python 3.10.4
| SQL agent - SQLAlchemy - timestamp formatting issues that make the SQL-agent crash | https://api.github.com/repos/langchain-ai/langchain/issues/19234/comments | 3 | 2024-03-18T15:14:43Z | 2024-06-25T16:13:23Z | https://github.com/langchain-ai/langchain/issues/19234 | 2,192,497,427 | 19,234 |
[
"langchain-ai",
"langchain"
] | ### 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
While using the load_and_split function in AzureAIDocumentIntelligenceLoader with mode="object". I'm getting Validation error with the page_content when the content_type is a table.
### My code
loader = AzureAIDocumentIntelligenceLoader(
api_endpoint=endpoint,
api_key=key,
file_path=file_path,
api_model="prebuilt-layout",
mode="object"
)
docs = loader.load_and_split()
On looking into the classes here's what I found.
### langchain_community.document_loaders.parsers.doc_intelligence.py
#### AzureAIDocumentIntelligenceParser > line 75 in _generate_docs_object
for table in result.tables:
yield Document(
page_content=table.cells, # json object
metadata={
"footnote": table.footnotes,
"caption": table.caption,
"page": para.bounding_regions[0].page_number,
"bounding_box": para.bounding_regions[0].polygon,
"row_count": table.row_count,
"column_count": table.column_count,
"type": "table",
},
)
A json object is getting passed in the page_content of the Document class. Due to which pydantic throws a validation error.
### langchain_core.documents.base.py
#### Document > line 12 in page_content
class Document(Serializable):
"""Class for storing a piece of text and associated metadata."""
page_content: str
"""String text."""
metadata: dict = Field(default_factory=dict)
"""Arbitrary metadata about the page content (e.g., source, relationships to other
documents, etc.).
"""
This code shows that only string type is accepted in the page_content field.
### Error Message and Stack Trace (if applicable)
---------------------------------------------------------------------------
ValidationError Traceback (most recent call last)
Cell In[22], line 1
----> 1 azure_documents = loader.load_and_split(text_splitter=RecursiveCharacterTextSplitter(chunk_size=800, chunk_overlap=200))
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_core\document_loaders\base.py:59, in BaseLoader.load_and_split(self, text_splitter)
57 else:
58 _text_splitter = text_splitter
---> 59 docs = self.load()
60 return _text_splitter.split_documents(docs)
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_core\document_loaders\base.py:29, in BaseLoader.load(self)
27 def load(self) -> List[Document]:
28 """Load data into Document objects."""
---> 29 return list(self.lazy_load())
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_community\document_loaders\doc_intelligence.py:86, in AzureAIDocumentIntelligenceLoader.lazy_load(self)
84 if self.file_path is not None:
85 blob = Blob.from_path(self.file_path)
---> 86 yield from self.parser.parse(blob)
87 else:
88 yield from self.parser.parse_url(self.url_path)
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_core\document_loaders\base.py:121, in BaseBlobParser.parse(self, blob)
106 def parse(self, blob: Blob) -> List[Document]:
107 """Eagerly parse the blob into a document or documents.
108
109 This is a convenience method for interactive development environment.
(...)
119 List of documents
120 """
--> 121 return list(self.lazy_parse(blob))
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_community\document_loaders\parsers\doc_intelligence.py:104, in AzureAIDocumentIntelligenceParser.lazy_parse(self, blob)
102 yield from self._generate_docs_page(result)
103 else:
--> 104 yield from self._generate_docs_object(result)
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_community\document_loaders\parsers\doc_intelligence.py:74, in AzureAIDocumentIntelligenceParser._generate_docs_object(self, result)
72 # table
73 for table in result.tables:
---> 74 yield Document(
75 page_content=table.cells, # json object
76 metadata={
77 "footnote": table.footnotes,
78 "caption": table.caption,
79 "page": para.bounding_regions[0].page_number,
80 "bounding_box": para.bounding_regions[0].polygon,
81 "row_count": table.row_count,
82 "column_count": table.column_count,
83 "type": "table",
84 },
85 )
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_core\documents\base.py:22, in Document.__init__(self, page_content, **kwargs)
20 def __init__(self, page_content: str, **kwargs: Any) -> None:
21 """Pass page_content in as positional or named arg."""
---> 22 super().__init__(page_content=page_content, **kwargs)
File d:\Users\RajatKumar.Roy\Anaconda3\envs\llm_exp\lib\site-packages\langchain_core\load\serializable.py:120, in Serializable.__init__(self, **kwargs)
119 def __init__(self, **kwargs: Any) -> None:
--> 120 super().__init__(**kwargs)
121 self._lc_kwargs = kwargs
File ~\AppData\Roaming\Python\Python310\site-packages\pydantic\main.py:341, in pydantic.main.BaseModel.__init__()
ValidationError: 1 validation error for Document
page_content
str type expected (type=type_error.str)
### Description
I'm trying to use the load_and_split function with the AzureAIDocumentIntelligenceLoader class which would extract paragraphs and tables from a pdf and perform chunking on the content. For this, I've passed mode="object" in the class argument. During execution, I'm getting an error which is mentioned as a validation error from the pydantic "Document" class.
### System Info
## Langchain version
langchain==0.1.12
langchain-community==0.0.28
## Python version
python=3.10.12 | ValidationError: Pydantic validation error on "page_content" with "object" mode in AzureAIDocumentIntelligenceLoader | https://api.github.com/repos/langchain-ai/langchain/issues/19229/comments | 0 | 2024-03-18T11:19:00Z | 2024-06-24T16:13:49Z | https://github.com/langchain-ai/langchain/issues/19229 | 2,191,926,207 | 19,229 |
[
"langchain-ai",
"langchain"
] | ### 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
```Python
from llama_index.core import VectorStoreIndex
from llama_index.core import (
StorageContext,
load_index_from_storage,
)
from llama_index.core.node_parser import SentenceSplitter
from loader import get_documents
import os
from langchain.text_splitter import RecursiveCharacterTextSplitter
from llama_index.core.node_parser import LangchainNodeParser
from llama_index.core.node_parser import HierarchicalNodeParser
def get_index(source_dir, persist_dir, split_type="sentence", chunk_size=1024):
if not os.path.exists(persist_dir):
# load the documents and create the index
documents = get_documents(source_dir)
if split_type == "sentence":
index = VectorStoreIndex.from_documents(documents=documents,transformations=[SentenceSplitter(chunk_size=chunk_size, chunk_overlap=20)], show_progress=False)
```
### Error Message and Stack Trace (if applicable)
File "D:\RAG_benchmark\RAG-benchmark\main.py", line 26, in <module>
index = get_index("D:\RAG_benchmark\data", cfg.persist_dir, split_type=cfg.split_type, chunk_size=cfg.chunk_size)
File "D:\RAG_benchmark\RAG-benchmark\index.py", line 18, in get_index
index = VectorStoreIndex.from_documents(documents=documents,transformations=[SentenceSplitter(chunk_size=chunk_size, chunk_overlap=20)], show_progress=False)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\base.py", line 145, in from_documents
return cls(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\vector_store\base.py", line 75, in __init__
super().__init__(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\base.py", line 94, in __init__
index_struct = self.build_index_from_nodes(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\vector_store\base.py", line 308, in build_index_from_nodes
return self._build_index_from_nodes(nodes, **insert_kwargs)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\vector_store\base.py", line 280, in _build_index_from_nodes
self._add_nodes_to_index(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\vector_store\base.py", line 233, in _add_nodes_to_index
nodes_batch = self._get_node_with_embedding(nodes_batch, show_progress)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\vector_store\base.py", line 141, in _get_node_with_embedding
id_to_embed_map = embed_nodes(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\indices\utils.py", line 138, in embed_nodes
new_embeddings = embed_model.get_text_embedding_batch(
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\instrumentation\dispatcher.py", line 102, in wrapper
self.span_drop(id=id, err=e)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\instrumentation\dispatcher.py", line 77, in span_drop
h.span_drop(id, err, **kwargs)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\instrumentation\span_handlers\base.py", line 45, in span_drop
self.prepare_to_drop_span(id, err, **kwargs)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\instrumentation\span_handlers\null.py", line 33, in prepare_to_drop_span
raise err
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\instrumentation\dispatcher.py", line 100, in wrapper
result = func(*args, **kwargs)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\core\base\embeddings\base.py", line 280, in get_text_embedding_batch
embeddings = self._get_text_embeddings(cur_batch)
File "C:\Users\hhw\miniconda3\lib\site-packages\llama_index\embeddings\langchain\base.py", line 87, in _get_text_embeddings
return self._langchain_embedding.embed_documents(texts)
File "C:\Users\hhw\miniconda3\lib\site-packages\langchain_community\embeddings\huggingface.py", line 93, in embed_documents
embeddings = self.client.encode(
TypeError: sentence_transformers.SentenceTransformer.SentenceTransformer.encode() got multiple values for keyword argument 'show_progress_bar'
### Description
It should work corrently, but it got a problems that can't be fixed
### System Info
Windows11 The latest version of langchain | .langchain_community\embeddings\huggingface.py embeddings = self.client.encode( texts, show_progress_bar=self.show_progress, **self.encode_kwargs ) | https://api.github.com/repos/langchain-ai/langchain/issues/19228/comments | 0 | 2024-03-18T11:01:30Z | 2024-06-24T16:13:48Z | https://github.com/langchain-ai/langchain/issues/19228 | 2,191,886,887 | 19,228 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
def list_pull_request_files(self, pr_number: int) -> List[Dict[str, Any]]:
"""Fetches the full text of all files in a PR. Truncates after first 3k tokens.
# TODO: Enhancement to summarize files with ctags if they're getting long.
Args:
pr_number(int): The number of the pull request on Github
Returns:
dict: A dictionary containing the issue's title,
body, and comments as a string
"""
tiktoken = _import_tiktoken()
MAX_TOKENS_FOR_FILES = 3_000
pr_files = []
pr = self.github_repo_instance.get_pull(number=int(pr_number))
total_tokens = 0
page = 0
while True: # or while (total_tokens + tiktoken()) < MAX_TOKENS_FOR_FILES:
files_page = pr.get_files().get_page(page)
if len(files_page) == 0:
break
for file in files_page:
try:
file_metadata_response = requests.get(file.contents_url)
if file_metadata_response.status_code == 200:
download_url = json.loads(file_metadata_response.text)[
"download_url"
]
else:
print(f"Failed to download file: {file.contents_url}, skipping") # noqa: T201
continue
file_content_response = requests.get(download_url)
if file_content_response.status_code == 200:
# Save the content as a UTF-8 string
file_content = file_content_response.text
else:
print( # noqa: T201
"Failed downloading file content "
f"(Error {file_content_response.status_code}). Skipping"
)
continue
file_tokens = len(
tiktoken.get_encoding("cl100k_base").encode(
file_content + file.filename + "file_name file_contents"
)
)
if (total_tokens + file_tokens) < MAX_TOKENS_FOR_FILES:
pr_files.append(
{
"filename": file.filename,
"contents": file_content,
"additions": file.additions,
"deletions": file.deletions,
}
)
total_tokens += file_tokens
except Exception as e:
print(f"Error when reading files from a PR on github. {e}") # noqa: T201
page += 1
return pr_files
```
### Error Message and Stack Trace (if applicable)
`Failed to download file <file.contents_url>, skipping`
### Description
I have a LangChain GitHub Agent that has to retrieve pull requests files to analyze their content from a repository. The GitHub App has all the necessary permissions to do so (which are the ones in the [documentation](https://python.langchain.com/docs/integrations/toolkits/github#create-a-github-app)).
However, when the `GET /repos/{owner}/{repo}/contents/{path}` function is called for a pull request file with a GitHub App (for each file inside the PR), the following error appears:
`Failed to download file <file.contents_url>, skipping`.
What I noticed is that this is the only function that needs to call a GitHub API explicitely (through the `requests.get` instruction).
This code is inside the `langchain_community\utilities\github.py` file
### System Info
Windows 10 Home 22H2
Python 3.10.6
PyGithub == 2.2.0
langchain == 0.1.11
langchain-core == 0.1.30
langchain-community == 0.0.27
langchain-openai == 0.0.8 | LangChain GitHub: list_pull_request_files function doesn't work correctly when using a GitHub App | https://api.github.com/repos/langchain-ai/langchain/issues/19222/comments | 0 | 2024-03-18T09:39:35Z | 2024-06-24T16:13:51Z | https://github.com/langchain-ai/langchain/issues/19222 | 2,191,693,323 | 19,222 |
[
"langchain-ai",
"langchain"
] | ### 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
def get_language_model(self, streaming: bool = False, temperature: int = 0, callbacks: Callbacks = None) -> BaseLanguageModel:
"""
Get the language model.
Args:
streaming (bool): Use streaming or not.
temperature (int): Temperature for language model generation.
callbacks (Callbacks): Callbacks for language model.
Returns:
BaseLanguageModel: The language model.
"""
llm_token_usage_callback = llm_token_usage_callback_var.get()
callbacks = callbacks or []
if llm_token_usage_callback and self.include_token_usage_cb:
callbacks.append(llm_token_usage_callback)
if logger.is_debug:
callbacks.append(LLMLogCallbackHandlerAsync())
if self._chat_model.openai_api_key:
return ChatOpenAI(openai_api_key=self._chat_model.openai_api_key, temperature=temperature, model=self._chat_model.model, streaming=streaming, callbacks=callbacks)
if self._chat_model.google_api_key:
safety_settings = {
HarmCategory.HARM_CATEGORY_HARASSMENT: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_HATE_SPEECH: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT: HarmBlockThreshold.BLOCK_NONE,
HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT: HarmBlockThreshold.BLOCK_NONE
}
return ChatGoogleGenerativeAI(
google_api_key=self._chat_model.google_api_key, temperature=temperature, model=self._chat_model.model,
streaming=streaming, callbacks=callbacks, convert_system_message_to_human=True, safety_settings=safety_settings
)
client = boto3.client(
'bedrock-runtime',
region_name=self._chat_model.bedrock_aws_region_name,
aws_access_key_id=self._chat_model.bedrock_aws_access_key_id,
aws_secret_access_key=self._chat_model.bedrock_aws_secret_access_key
)
return BedrockChat(client=client, model_id=self._chat_model.model, streaming=streaming, callbacks=callbacks, model_kwargs={"temperature": temperature})
async def _get_agent_executor_async(self) -> AgentExecutor:
"""
Prepare agent using tools and prompt
Returns:
AgentExecutor: AgentExecutor to invoke the Agent.
"""
self.tools = await self._get_tools()
llm = LLMSelector(self.model).get_language_model().with_config(RunnableConfig(run_name="Agent"))
prompt = ChatPromptTemplate.from_messages([
(EnumChatMessageType.SYSTEM, REACT_AGENT_SYSTEM_TEMPLATE),
(EnumChatMessageType.HUMAN, REACT_AGENT_USER_TEMPLATE),
]).partial(system_context=self.model.system_context or '', human_context=self.model.human_context or '')
agent = create_react_agent(llm=llm, tools=self.tools, prompt=prompt) # noqa
return AgentExecutor(agent=agent, tools=self.tools, handle_parsing_errors=self._handle_parser_exception, max_iterations=3)
### Error Message and Stack Trace (if applicable)
def _prepare_input_and_invoke_stream(
self,
prompt: Optional[str] = None,
system: Optional[str] = None,
messages: Optional[List[Dict]] = None,
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> Iterator[GenerationChunk]:
_model_kwargs = self.model_kwargs or {}
provider = self._get_provider()
if stop:
if provider not in self.provider_stop_sequence_key_name_map:
raise ValueError(
f"Stop sequence key name for {provider} is not supported."
)
# stop sequence from _generate() overrides
# stop sequences in the class attribute
_model_kwargs[self.provider_stop_sequence_key_name_map.get(provider)] = stop
if provider == "cohere":
_model_kwargs["stream"] = True
params = {**_model_kwargs, **kwargs}
if self._guardrails_enabled:
params.update(self._get_guardrails_canonical())
input_body = LLMInputOutputAdapter.prepare_input(
provider=provider,
prompt=prompt,
system=system,
messages=messages,
model_kwargs=params,
)
body = json.dumps(input_body)
request_options = {
"body": body,
"modelId": self.model_id,
"accept": "application/json",
"contentType": "application/json",
}
if self._guardrails_enabled:
request_options["guardrail"] = "ENABLED"
if self.guardrails.get("trace"): # type: ignore[union-attr]
request_options["trace"] = "ENABLED"
try:
response = self.client.invoke_model_with_response_stream(**request_options)
except Exception as e:
raise ValueError(f"Error raised by bedrock service: {e}")
for chunk in LLMInputOutputAdapter.prepare_output_stream(
provider, response, stop, True if messages else False
):
yield chunk
# verify and raise callback error if any middleware intervened
self._get_bedrock_services_signal(chunk.generation_info) # type: ignore[arg-type]
if run_manager is not None:
run_manager.on_llm_new_token(chunk.text, chunk=chunk)
### Description
When invoking BedrockChat with the meta-model, an error occurs stating that the "Stop sequence" key name is not supported. This issue arises because the create_react_agent function includes a parameter llm_with_stop which is bound to the stop sequence ["\nObservation"], but the meta-model does not support the use of a stop sequence.
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-google-genai==0.0.11
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
langchainhub==0.1.15
Platform: Mac
Python 3.11.6
| Stop sequence key name for meta is not supported, For meta-model (exp: meta.llama2-13b-chat-v1) in BedrockChat | https://api.github.com/repos/langchain-ai/langchain/issues/19220/comments | 3 | 2024-03-18T08:33:52Z | 2024-07-19T14:21:53Z | https://github.com/langchain-ai/langchain/issues/19220 | 2,191,540,570 | 19,220 |
[
"langchain-ai",
"langchain"
] | ### 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
```
#1 import the OS, Bedrock, ConversationChain, ConversationBufferMemory Langchain Modules
import os
from langchain.llms.bedrock import Bedrock
from langchain.memory import ConversationBufferMemory
from langchain.chains import ConversationChain
#2a Write a function for invoking model- client connection with Bedrock with profile, model_id & Inference params- model_kwargs
# def demo_chatbot():
def demo_chatbot(input_text):
demo_llm = Bedrock(
credentials_profile_name='default',
model_id='mistral.mixtral-8x7b-instruct-v0:1',
model_kwargs= {
"temperature": 0.9,
"top_p": 0.5,
"max_gen_len": 512})
# return demo_llm
#2b Test out the LLM with Predict method
return demo_llm.predict(input_text)
response = demo_chatbot('what is the temprature in london like ?')
print(response)
```
### Error Message and Stack Trace (if applicable)
```
[C:\Users\leo_c\AppData\Roaming\Python\Python310\site-packages\langchain_core\_api\deprecation.py:117](file:///C:/Users/leo_c/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:117): LangChainDeprecationWarning: The function `predict` was deprecated in LangChain 0.1.7 and will be removed in 0.2.0. Use invoke instead.
warn_deprecated(
```
```
---------------------------------------------------------------------------
ValidationException Traceback (most recent call last)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_community\llms\bedrock.py:536](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:536), in BedrockBase._prepare_input_and_invoke(self, prompt, system, messages, stop, run_manager, **kwargs)
[535](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:535) try:
--> [536](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:536) response = self.client.invoke_model(**request_options)
[538](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:538) text, body = LLMInputOutputAdapter.prepare_output(
[539](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:539) provider, response
[540](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:540) ).values()
File [c:\Users\leo_c\anaconda3\lib\site-packages\botocore\client.py:535](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:535), in ClientCreator._create_api_method.<locals>._api_call(self, *args, **kwargs)
[534](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:534) # The "self" in this scope is referring to the BaseClient.
--> [535](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:535) return self._make_api_call(operation_name, kwargs)
File [c:\Users\leo_c\anaconda3\lib\site-packages\botocore\client.py:983](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:983), in BaseClient._make_api_call(self, operation_name, api_params)
[982](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:982) error_class = self.exceptions.from_code(error_code)
--> [983](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:983) raise error_class(parsed_response, operation_name)
[984](file:///C:/Users/leo_c/anaconda3/lib/site-packages/botocore/client.py:984) else:
ValidationException: An error occurred (ValidationException) when calling the InvokeModel operation: Operation not allowed
During handling of the above exception, another exception occurred:
ValueError Traceback (most recent call last)
Cell In[1], [line 22](vscode-notebook-cell:?execution_count=1&line=22)
[18](vscode-notebook-cell:?execution_count=1&line=18) # return demo_llm
[19](vscode-notebook-cell:?execution_count=1&line=19)
[20](vscode-notebook-cell:?execution_count=1&line=20) #2b Test out the LLM with Predict method
[21](vscode-notebook-cell:?execution_count=1&line=21) return demo_llm.predict(input_text)
---> [22](vscode-notebook-cell:?execution_count=1&line=22) response = demo_chatbot('what is the temprature in london like ?')
[23](vscode-notebook-cell:?execution_count=1&line=23) print(response)
Cell In[1], [line 21](vscode-notebook-cell:?execution_count=1&line=21)
[10](vscode-notebook-cell:?execution_count=1&line=10) demo_llm = Bedrock(
[11](vscode-notebook-cell:?execution_count=1&line=11) credentials_profile_name='default',
[12](vscode-notebook-cell:?execution_count=1&line=12) model_id='meta.llama2-70b-chat-v1',
(...)
[16](vscode-notebook-cell:?execution_count=1&line=16) "top_p": 0.5,
[17](vscode-notebook-cell:?execution_count=1&line=17) "max_gen_len": 512})
[18](vscode-notebook-cell:?execution_count=1&line=18) # return demo_llm
[19](vscode-notebook-cell:?execution_count=1&line=19)
[20](vscode-notebook-cell:?execution_count=1&line=20) #2b Test out the LLM with Predict method
---> [21](vscode-notebook-cell:?execution_count=1&line=21) return demo_llm.predict(input_text)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\_api\deprecation.py:145](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:145), in deprecated.<locals>.deprecate.<locals>.warning_emitting_wrapper(*args, **kwargs)
[143](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:143) warned = True
[144](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:144) emit_warning()
--> [145](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:145) return wrapped(*args, **kwargs)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:1013](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1013), in BaseLLM.predict(self, text, stop, **kwargs)
[1011](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1011) else:
[1012](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1012) _stop = list(stop)
-> [1013](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1013) return self(text, stop=_stop, **kwargs)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\_api\deprecation.py:145](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:145), in deprecated.<locals>.deprecate.<locals>.warning_emitting_wrapper(*args, **kwargs)
[143](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:143) warned = True
[144](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:144) emit_warning()
--> [145](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/_api/deprecation.py:145) return wrapped(*args, **kwargs)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:972](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:972), in BaseLLM.__call__(self, prompt, stop, callbacks, tags, metadata, **kwargs)
[965](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:965) if not isinstance(prompt, str):
[966](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:966) raise ValueError(
[967](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:967) "Argument `prompt` is expected to be a string. Instead found "
[968](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:968) f"{type(prompt)}. If you want to run the LLM on multiple prompts, use "
[969](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:969) "`generate` instead."
[970](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:970) )
[971](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:971) return (
--> [972](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:972) self.generate(
[973](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:973) [prompt],
[974](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:974) stop=stop,
[975](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:975) callbacks=callbacks,
[976](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:976) tags=tags,
[977](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:977) metadata=metadata,
[978](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:978) **kwargs,
[979](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:979) )
[980](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:980) .generations[0][0]
[981](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:981) .text
[982](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:982) )
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:714](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:714), in BaseLLM.generate(self, prompts, stop, callbacks, tags, metadata, run_name, **kwargs)
[698](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:698) raise ValueError(
[699](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:699) "Asked to cache, but no cache found at `langchain.cache`."
[700](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:700) )
[701](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:701) run_managers = [
[702](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:702) callback_manager.on_llm_start(
[703](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:703) dumpd(self),
(...)
[712](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:712) )
[713](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:713) ]
--> [714](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:714) output = self._generate_helper(
[715](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:715) prompts, stop, run_managers, bool(new_arg_supported), **kwargs
[716](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:716) )
[717](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:717) return output
[718](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:718) if len(missing_prompts) > 0:
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:578](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:578), in BaseLLM._generate_helper(self, prompts, stop, run_managers, new_arg_supported, **kwargs)
[576](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:576) for run_manager in run_managers:
[577](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:577) run_manager.on_llm_error(e, response=LLMResult(generations=[]))
--> [578](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:578) raise e
[579](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:579) flattened_outputs = output.flatten()
[580](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:580) for manager, flattened_output in zip(run_managers, flattened_outputs):
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:565](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:565), in BaseLLM._generate_helper(self, prompts, stop, run_managers, new_arg_supported, **kwargs)
[555](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:555) def _generate_helper(
[556](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:556) self,
[557](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:557) prompts: List[str],
(...)
[561](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:561) **kwargs: Any,
[562](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:562) ) -> LLMResult:
[563](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:563) try:
[564](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:564) output = (
--> [565](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:565) self._generate(
[566](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:566) prompts,
[567](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:567) stop=stop,
[568](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:568) # TODO: support multiple run managers
[569](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:569) run_manager=run_managers[0] if run_managers else None,
[570](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:570) **kwargs,
[571](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:571) )
[572](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:572) if new_arg_supported
[573](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:573) else self._generate(prompts, stop=stop)
[574](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:574) )
[575](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:575) except BaseException as e:
[576](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:576) for run_manager in run_managers:
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_core\language_models\llms.py:1153](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1153), in LLM._generate(self, prompts, stop, run_manager, **kwargs)
[1150](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1150) new_arg_supported = inspect.signature(self._call).parameters.get("run_manager")
[1151](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1151) for prompt in prompts:
[1152](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1152) text = (
-> [1153](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1153) self._call(prompt, stop=stop, run_manager=run_manager, **kwargs)
[1154](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1154) if new_arg_supported
[1155](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1155) else self._call(prompt, stop=stop, **kwargs)
[1156](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1156) )
[1157](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1157) generations.append([Generation(text=text)])
[1158](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_core/language_models/llms.py:1158) return LLMResult(generations=generations)
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_community\llms\bedrock.py:831](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:831), in Bedrock._call(self, prompt, stop, run_manager, **kwargs)
[828](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:828) completion += chunk.text
[829](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:829) return completion
--> [831](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:831) return self._prepare_input_and_invoke(
[832](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:832) prompt=prompt, stop=stop, run_manager=run_manager, **kwargs
[833](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:833) )
File [~\AppData\Roaming\Python\Python310\site-packages\langchain_community\llms\bedrock.py:543](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:543), in BedrockBase._prepare_input_and_invoke(self, prompt, system, messages, stop, run_manager, **kwargs)
[538](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:538) text, body = LLMInputOutputAdapter.prepare_output(
[539](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:539) provider, response
[540](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:540) ).values()
[542](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:542) except Exception as e:
--> [543](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:543) raise ValueError(f"Error raised by bedrock service: {e}")
[545](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:545) if stop is not None:
[546](https://file+.vscode-resource.vscode-cdn.net/c%3A/Users/leo_c/OneDrive/Desktop/Openmesh/Pythia/Python%20scripts/~/AppData/Roaming/Python/Python310/site-packages/langchain_community/llms/bedrock.py:546) text = enforce_stop_tokens(text, stop)
ValueError: Error raised by bedrock service: An error occurred (ValidationException) when calling the InvokeModel operation: Operation not allowed
```
### Description
My script above is fairly basic but I am still getting the error above.
### System Info
pip freeze langchain:
```
WARNING: Ignoring invalid distribution -orch (c:\users\leo_c\anaconda3\lib\site-packages)
accelerate==0.23.0
aiohttp==3.8.4
aiosignal==1.3.1
alabaster @ file:///home/ktietz/src/ci/alabaster_1611921544520/work
altair==5.1.2
anaconda-client==1.11.2
anaconda-navigator==2.4.1
anaconda-project @ file:///C:/Windows/TEMP/abs_91fu4tfkih/croots/recipe/anaconda-project_1660339890874/work
ansible==9.1.0
ansible-core==2.16.2
anthropic==0.2.10
anyio==3.7.1
appdirs==1.4.4
argon2-cffi==23.1.0
argon2-cffi-bindings @ file:///C:/ci/argon2-cffi-bindings_1644569876605/work
arrow @ file:///C:/b/abs_cal7u12ktb/croot/arrow_1676588147908/work
ascii-magic==2.3.0
astroid @ file:///C:/b/abs_d4lg3_taxn/croot/astroid_1676904351456/work
astropy @ file:///C:/ci/astropy_1657719642921/work
asttokens @ file:///opt/conda/conda-bld/asttokens_1646925590279/work
async-timeout==4.0.2
atomicwrites==1.4.0
attrs @ file:///C:/b/abs_09s3y775ra/croot/attrs_1668696195628/work
Authlib==1.2.1
auto-gptq==0.4.2+cu118
Automat @ file:///tmp/build/80754af9/automat_1600298431173/work
autopep8 @ file:///opt/conda/conda-bld/autopep8_1650463822033/work
azure-cognitiveservices-speech==1.32.1
Babel @ file:///C:/b/abs_a2shv_3tqi/croot/babel_1671782804377/work
backcall @ file:///home/ktietz/src/ci/backcall_1611930011877/work
backports.functools-lru-cache @ file:///tmp/build/80754af9/backports.functools_lru_cache_1618170165463/work
backports.tempfile @ file:///home/linux1/recipes/ci/backports.tempfile_1610991236607/work
backports.weakref==1.0.post1
bcrypt==4.0.1
beautifulsoup4 @ file:///home/conda/feedstock_root/build_artifacts/beautifulsoup4_1680888073205/work
binaryornot @ file:///tmp/build/80754af9/binaryornot_1617751525010/work
black @ file:///C:/ci/black_1660221726201/work
bleach @ file:///opt/conda/conda-bld/bleach_1641577558959/work
blinker==1.6.3
blis==0.7.9
bokeh @ file:///C:/Windows/TEMP/abs_4a259bc2-ed05-4a1f-808e-ac712cc0900cddqp8sp7/croots/recipe/bokeh_1658136660686/work
boltons @ file:///C:/b/abs_707eo7c09t/croot/boltons_1677628723117/work
boto3==1.28.65
botocore==1.31.85
Bottleneck @ file:///C:/Windows/Temp/abs_3198ca53-903d-42fd-87b4-03e6d03a8381yfwsuve8/croots/recipe/bottleneck_1657175565403/work
brotlipy==0.7.0
bs4==0.0.1
cachelib==0.12.0
cachetools==5.3.1
catalogue==2.0.8
certifi==2023.7.22
cffi @ file:///C:/b/abs_49n3v2hyhr/croot/cffi_1670423218144/work
chardet @ file:///C:/ci_310/chardet_1642114080098/work
charset-normalizer @ file:///tmp/build/80754af9/charset-normalizer_1630003229654/work
click==8.1.7
cloudpickle @ file:///tmp/build/80754af9/cloudpickle_1632508026186/work
clyent==1.2.2
colorama @ file:///C:/b/abs_a9ozq0l032/croot/colorama_1672387194846/work
colorcet @ file:///C:/b/abs_46vyu0rpdl/croot/colorcet_1668084513237/work
coloredlogs==15.0.1
comm @ file:///C:/b/abs_1419earm7u/croot/comm_1671231131638/work
conda==23.3.1
conda-build==3.24.0
conda-content-trust @ file:///C:/Windows/TEMP/abs_4589313d-fc62-4ccc-81c0-b801b4449e833j1ajrwu/croots/recipe/conda-content-trust_1658126379362/work
conda-pack @ file:///tmp/build/80754af9/conda-pack_1611163042455/work
conda-package-handling @ file:///C:/b/abs_fcga8w0uem/croot/conda-package-handling_1672865024290/work
conda-repo-cli==1.0.41
conda-token @ file:///Users/paulyim/miniconda3/envs/c3i/conda-bld/conda-token_1662660369760/work
conda-verify==3.4.2
conda_package_streaming @ file:///C:/b/abs_0e5n5hdal3/croot/conda-package-streaming_1670508162902/work
confection==0.1.0
constantly==15.1.0
contourpy @ file:///C:/b/abs_d5rpy288vc/croots/recipe/contourpy_1663827418189/work
cookiecutter @ file:///opt/conda/conda-bld/cookiecutter_1649151442564/work
cryptography==41.0.7
cssselect @ file:///home/conda/feedstock_root/build_artifacts/cssselect_1666980406338/work
cycler @ file:///tmp/build/80754af9/cycler_1637851556182/work
cymem==2.0.7
cytoolz @ file:///C:/b/abs_61m9vzb4qh/croot/cytoolz_1667465938275/work
daal4py==2023.0.2
dask @ file:///C:/ci/dask-core_1658497112560/work
dataclasses-json==0.5.9
datasets==2.14.5
datashader @ file:///C:/b/abs_e80f3d7ac0/croot/datashader_1676023254070/work
datashape==0.5.4
dateparser==1.1.8
debugpy @ file:///C:/ci_310/debugpy_1642079916595/work
decorator @ file:///opt/conda/conda-bld/decorator_1643638310831/work
defusedxml @ file:///tmp/build/80754af9/defusedxml_1615228127516/work
Deprecated==1.2.14
diff-match-patch @ file:///Users/ktietz/demo/mc3/conda-bld/diff-match-patch_1630511840874/work
dill==0.3.7
distlib==0.3.8
distributed @ file:///C:/ci/distributed_1658523963030/work
dnspython==2.3.0
docker==6.1.3
docstring-to-markdown @ file:///C:/b/abs_cf10j8nr4q/croot/docstring-to-markdown_1673447652942/work
docutils @ file:///C:/Windows/TEMP/abs_24e5e278-4d1c-47eb-97b9-f761d871f482dy2vg450/croots/recipe/docutils_1657175444608/work
elastic-transport==8.4.0
elasticsearch==8.8.2
email-validator==2.1.0.post1
entrypoints @ file:///C:/ci/entrypoints_1649926676279/work
et-xmlfile==1.1.0
exceptiongroup==1.1.2
executing @ file:///opt/conda/conda-bld/executing_1646925071911/work
faiss-cpu==1.7.4
fake-useragent==1.1.3
fastapi==0.103.2
fastcore==1.5.29
fastjsonschema @ file:///C:/Users/BUILDE~1/AppData/Local/Temp/abs_ebruxzvd08/croots/recipe/python-fastjsonschema_1661376484940/work
ffmpeg-python==0.2.0
filelock==3.13.1
flake8 @ file:///C:/b/abs_9f6_n1jlpc/croot/flake8_1674581816810/work
Flask @ file:///C:/b/abs_ef16l83sif/croot/flask_1671217367534/work
Flask-Session==0.6.0
flit_core @ file:///opt/conda/conda-bld/flit-core_1644941570762/work/source/flit_core
fonttools==4.25.0
forbiddenfruit==0.1.4
frozenlist==1.3.3
fsspec==2023.6.0
future @ file:///C:/b/abs_3dcibf18zi/croot/future_1677599891380/work
gensim @ file:///C:/b/abs_a5vat69tv8/croot/gensim_1674853640591/work
gevent==23.9.1
gitdb==4.0.10
GitPython==3.1.40
glob2 @ file:///home/linux1/recipes/ci/glob2_1610991677669/work
google-api-core==2.11.1
google-api-python-client==2.70.0
google-auth==2.21.0
google-auth-httplib2==0.1.0
googleapis-common-protos==1.59.1
greenlet==2.0.2
grpcio==1.56.0
grpcio-tools==1.56.0
h11==0.14.0
h2==4.1.0
h5py @ file:///C:/ci/h5py_1659089830381/work
hagrid==0.3.97
HeapDict @ file:///Users/ktietz/demo/mc3/conda-bld/heapdict_1630598515714/work
holoviews @ file:///C:/b/abs_bbf97_0kcd/croot/holoviews_1676372911083/work
hpack==4.0.0
httpcore==0.17.3
httplib2==0.22.0
httptools==0.6.1
httpx==0.24.1
huggingface-hub==0.20.3
humanfriendly==10.0
hvplot @ file:///C:/b/abs_13un17_4x_/croot/hvplot_1670508919193/work
hyperframe==6.0.1
hyperlink @ file:///tmp/build/80754af9/hyperlink_1610130746837/work
idna @ file:///C:/b/abs_bdhbebrioa/croot/idna_1666125572046/work
imagecodecs @ file:///C:/b/abs_f0cr12h73p/croot/imagecodecs_1677576746499/work
imageio @ file:///C:/b/abs_27kq2gy1us/croot/imageio_1677879918708/work
imagesize @ file:///C:/Windows/TEMP/abs_3cecd249-3fc4-4bfc-b80b-bb227b0d701en12vqzot/croots/recipe/imagesize_1657179501304/work
imbalanced-learn @ file:///C:/b/abs_1911ryuksz/croot/imbalanced-learn_1677191585237/work
importlib-metadata==6.8.0
incremental @ file:///tmp/build/80754af9/incremental_1636629750599/work
inflection==0.5.1
iniconfig @ file:///home/linux1/recipes/ci/iniconfig_1610983019677/work
intake @ file:///C:/b/abs_42yyb2lhwx/croot/intake_1676619887779/work
intervaltree @ file:///Users/ktietz/demo/mc3/conda-bld/intervaltree_1630511889664/work
ipykernel @ file:///C:/b/abs_b4f07tbsyd/croot/ipykernel_1672767104060/work
ipython @ file:///C:/b/abs_d3h279dv3h/croot/ipython_1676582236558/work
ipython-genutils @ file:///tmp/build/80754af9/ipython_genutils_1606773439826/work
ipywidgets==7.7.2
isort @ file:///tmp/build/80754af9/isort_1628603791788/work
itables==1.6.2
itemadapter @ file:///tmp/build/80754af9/itemadapter_1626442940632/work
itemloaders @ file:///opt/conda/conda-bld/itemloaders_1646805235997/work
itsdangerous @ file:///tmp/build/80754af9/itsdangerous_1621432558163/work
janus==1.0.0
jaraco.context==4.3.0
jax==0.4.20
jaxlib==0.4.20
jedi @ file:///C:/ci/jedi_1644315428305/work
jellyfish @ file:///C:/ci/jellyfish_1647962737334/work
Jinja2 @ file:///C:/b/abs_7cdis66kl9/croot/jinja2_1666908141852/work
jinja2-time @ file:///opt/conda/conda-bld/jinja2-time_1649251842261/work
jmespath @ file:///Users/ktietz/demo/mc3/conda-bld/jmespath_1630583964805/work
joblib @ file:///C:/b/abs_e60_bwl1v6/croot/joblib_1666298845728/work
json5 @ file:///tmp/build/80754af9/json5_1624432770122/work
jsonpatch==1.33
jsonpointer==2.1
jsonschema @ file:///C:/b/abs_6ccs97j_l8/croot/jsonschema_1676558690963/work
jupyter @ file:///C:/Windows/TEMP/abs_56xfdi__li/croots/recipe/jupyter_1659349053177/work
jupyter-console @ file:///C:/b/abs_68ttzd5p9c/croot/jupyter_console_1677674667636/work
jupyter-server @ file:///C:/b/abs_1cfi3__jl8/croot/jupyter_server_1671707636383/work
jupyter_client @ file:///C:/ci/jupyter_client_1661834530766/work
jupyter_core @ file:///C:/b/abs_bd7elvu3w2/croot/jupyter_core_1676538600510/work
jupyterlab @ file:///C:/b/abs_513jt6yy74/croot/jupyterlab_1675354138043/work
jupyterlab-pygments @ file:///tmp/build/80754af9/jupyterlab_pygments_1601490720602/work
jupyterlab-widgets @ file:///tmp/build/80754af9/jupyterlab_widgets_1609884341231/work
jupyterlab_server @ file:///C:/b/abs_d1z_g1swc8/croot/jupyterlab_server_1677153204814/work
jupytext==1.15.2
keyring @ file:///C:/ci_310/keyring_1642165564669/work
kiwisolver @ file:///C:/b/abs_88mdhvtahm/croot/kiwisolver_1672387921783/work
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-text-splitters==0.0.1
langchainplus-sdk==0.0.20
langcodes==3.3.0
langsmith==0.1.27
lazy-object-proxy @ file:///C:/ci_310/lazy-object-proxy_1642083437654/work
libarchive-c @ file:///tmp/build/80754af9/python-libarchive-c_1617780486945/work
llvmlite==0.39.1
locket @ file:///C:/ci/locket_1652904090946/work
loguru==0.7.2
lxml @ file:///C:/ci/lxml_1657527492694/work
lz4 @ file:///C:/ci_310/lz4_1643300078932/work
manifest-ml==0.0.1
Markdown @ file:///C:/b/abs_98lv_ucina/croot/markdown_1671541919225/work
markdown-it-py==3.0.0
MarkupSafe @ file:///C:/ci/markupsafe_1654508036328/work
marshmallow==3.19.0
marshmallow-enum==1.5.1
matplotlib==3.8.0
matplotlib-inline @ file:///C:/ci/matplotlib-inline_1661934094726/work
mccabe @ file:///opt/conda/conda-bld/mccabe_1644221741721/work
mdit-py-plugins==0.4.0
mdurl==0.1.2
menuinst @ file:///C:/Users/BUILDE~1/AppData/Local/Temp/abs_455sf5o0ct/croots/recipe/menuinst_1661805970842/work
miniaudio==1.59
mistune @ file:///C:/ci_310/mistune_1642084168466/work
mkl-fft==1.3.1
mkl-random @ file:///C:/ci_310/mkl_random_1643050563308/work
mkl-service==2.4.0
ml-dtypes==0.3.1
mock @ file:///tmp/build/80754af9/mock_1607622725907/work
more-itertools==9.1.0
mpmath==1.2.1
msgpack @ file:///C:/ci/msgpack-python_1652348582618/work
multidict==6.0.4
multipledispatch @ file:///C:/ci_310/multipledispatch_1642084438481/work
multiprocess==0.70.15
munkres==1.1.4
murmurhash==1.0.9
mypy-extensions==0.4.3
names==0.3.0
navigator-updater==0.3.0
nbclassic @ file:///C:/b/abs_d0_ze5q0j2/croot/nbclassic_1676902914817/work
nbclient @ file:///C:/ci/nbclient_1650308592199/work
nbconvert @ file:///C:/b/abs_4av3q4okro/croot/nbconvert_1668450658054/work
nbformat @ file:///C:/b/abs_85_3g7dkt4/croot/nbformat_1670352343720/work
nest-asyncio @ file:///C:/b/abs_3a_4jsjlqu/croot/nest-asyncio_1672387322800/work
networkx==2.8
nltk==3.8.1
notebook @ file:///C:/b/abs_ca13hqvuzw/croot/notebook_1668179888546/work
notebook_shim @ file:///C:/b/abs_ebfczttg6x/croot/notebook-shim_1668160590914/work
numba @ file:///C:/b/abs_e53pp2e4k7/croot/numba_1670258349527/work
numexpr @ file:///C:/b/abs_a7kbak88hk/croot/numexpr_1668713882979/work
numpy @ file:///C:/b/abs_datssh7cer/croot/numpy_and_numpy_base_1672336199388/work
numpydoc @ file:///C:/b/abs_cfdd4zxbga/croot/numpydoc_1668085912100/work
opacus==1.4.0
openai==0.27.10
openapi-schema-pydantic==1.2.4
openpyxl==3.0.10
opentelemetry-api==1.20.0
opentelemetry-sdk==1.20.0
opentelemetry-semantic-conventions==0.41b0
opt-einsum==3.3.0
optimum==1.13.2
orjson==3.9.15
outcome==1.2.0
packaging==23.2
pandas==2.1.4
pandocfilters @ file:///opt/conda/conda-bld/pandocfilters_1643405455980/work
panel @ file:///C:/b/abs_55ujq2fpyh/croot/panel_1676379705003/work
param @ file:///C:/b/abs_d799n8xz_7/croot/param_1671697759755/work
paramiko @ file:///opt/conda/conda-bld/paramiko_1640109032755/work
parse==1.19.1
parsel @ file:///C:/ci/parsel_1646722035970/work
parso @ file:///opt/conda/conda-bld/parso_1641458642106/work
partd @ file:///opt/conda/conda-bld/partd_1647245470509/work
pathlib @ file:///Users/ktietz/demo/mc3/conda-bld/pathlib_1629713961906/work
pathspec @ file:///C:/b/abs_9cu5_2yb3i/croot/pathspec_1674681579249/work
pathy==0.10.2
patsy==0.5.3
peft==0.5.0
pep8==1.7.1
pexpect @ file:///tmp/build/80754af9/pexpect_1605563209008/work
pickleshare @ file:///tmp/build/80754af9/pickleshare_1606932040724/work
Pillow==9.4.0
pinecone-client==2.2.2
pkginfo @ file:///C:/b/abs_d18srtr68x/croot/pkginfo_1679431192239/work
platformdirs==4.1.0
plotly @ file:///C:/ci/plotly_1658160673416/work
pluggy @ file:///C:/ci/pluggy_1648042746254/work
ply==3.11
pooch @ file:///tmp/build/80754af9/pooch_1623324770023/work
-e git+https://github.com/alessandriniluca/postget.git@da51db8edbfe065062899b0bfee577d66be0c1e2#egg=postget
poyo @ file:///tmp/build/80754af9/poyo_1617751526755/work
praw==7.7.1
prawcore==2.3.0
preshed==3.0.8
prometheus-client @ file:///C:/Windows/TEMP/abs_ab9nx8qb08/croots/recipe/prometheus_client_1659455104602/work
prompt-toolkit @ file:///C:/b/abs_6coz5_9f2s/croot/prompt-toolkit_1672387908312/work
Protego @ file:///tmp/build/80754af9/protego_1598657180827/work
protobuf==4.23.3
psutil==5.9.6
psycopg2-binary==2.9.9
ptyprocess @ file:///tmp/build/80754af9/ptyprocess_1609355006118/work/dist/ptyprocess-0.7.0-py2.py3-none-any.whl
pure-eval @ file:///opt/conda/conda-bld/pure_eval_1646925070566/work
py @ file:///opt/conda/conda-bld/py_1644396412707/work
pyaes==1.6.1
pyarrow==14.0.1
pyasn1 @ file:///Users/ktietz/demo/mc3/conda-bld/pyasn1_1629708007385/work
pyasn1-modules==0.2.8
pycapnp==1.3.0
pycodestyle @ file:///C:/b/abs_d77nxvklcq/croot/pycodestyle_1674267231034/work
pycosat @ file:///C:/b/abs_4b1rrw8pn9/croot/pycosat_1666807711599/work
pycparser @ file:///tmp/build/80754af9/pycparser_1636541352034/work
pycryptodome==3.18.0
pyct @ file:///C:/b/abs_92z17k7ig2/croot/pyct_1675450330889/work
pycurl==7.45.1
pydantic==1.10.13
pydeck==0.8.1b0
PyDispatcher==2.0.5
pydocstyle @ file:///C:/b/abs_6dz687_5i3/croot/pydocstyle_1675221688656/work
pydub==0.25.1
pyee==8.2.2
pyerfa @ file:///C:/ci_310/pyerfa_1642088497201/work
pyflakes @ file:///C:/b/abs_6dve6e13zh/croot/pyflakes_1674165143327/work
Pygments==2.16.1
PyHamcrest @ file:///tmp/build/80754af9/pyhamcrest_1615748656804/work
PyJWT @ file:///C:/ci/pyjwt_1657529477795/work
pylint @ file:///C:/b/abs_83sq99jc8i/croot/pylint_1676919922167/work
pylint-venv @ file:///C:/b/abs_bf0lepsbij/croot/pylint-venv_1673990138593/work
pyls-spyder==0.4.0
pymongo==4.6.1
PyNaCl @ file:///C:/Windows/Temp/abs_d5c3ajcm87/croots/recipe/pynacl_1659620667490/work
pyodbc @ file:///C:/Windows/Temp/abs_61e3jz3u05/croots/recipe/pyodbc_1659513801402/work
pyOpenSSL==23.3.0
pyparsing @ file:///C:/Users/BUILDE~1/AppData/Local/Temp/abs_7f_7lba6rl/croots/recipe/pyparsing_1661452540662/work
pyppeteer==1.0.2
PyQt5==5.15.7
PyQt5-sip @ file:///C:/Windows/Temp/abs_d7gmd2jg8i/croots/recipe/pyqt-split_1659273064801/work/pyqt_sip
PyQtWebEngine==5.15.4
pyquery==2.0.0
pyreadline3==3.4.1
pyrsistent @ file:///C:/ci_310/pyrsistent_1642117077485/work
PySocks @ file:///C:/ci_310/pysocks_1642089375450/work
pytest==7.1.2
pytest-base-url==2.0.0
python-binance==1.0.17
python-dateutil @ file:///tmp/build/80754af9/python-dateutil_1626374649649/work
python-dotenv==1.0.0
python-lsp-black @ file:///C:/Users/BUILDE~1/AppData/Local/Temp/abs_dddk9lhpp1/croots/recipe/python-lsp-black_1661852041405/work
python-lsp-jsonrpc==1.0.0
python-lsp-server @ file:///C:/b/abs_e44khh1wya/croot/python-lsp-server_1677296772730/work
python-multipart==0.0.6
python-slugify @ file:///home/conda/feedstock_root/build_artifacts/python-slugify-split_1694282063120/work
python-snappy @ file:///C:/b/abs_61b1fmzxcn/croot/python-snappy_1670943932513/work
pytoolconfig @ file:///C:/b/abs_18sf9z_iwl/croot/pytoolconfig_1676315065270/work
pytz @ file:///C:/b/abs_22fofvpn1x/croot/pytz_1671698059864/work
pyviz-comms @ file:///tmp/build/80754af9/pyviz_comms_1623747165329/work
PyWavelets @ file:///C:/b/abs_a8r4b1511a/croot/pywavelets_1670425185881/work
pywin32==305.1
pywin32-ctypes @ file:///C:/ci_310/pywin32-ctypes_1642657835512/work
pywinpty @ file:///C:/b/abs_73vshmevwq/croot/pywinpty_1677609966356/work/target/wheels/pywinpty-2.0.10-cp310-none-win_amd64.whl
PyYAML==6.0.1
pyzmq==25.1.1
QDarkStyle @ file:///tmp/build/80754af9/qdarkstyle_1617386714626/work
qdrant-client==0.11.10
qstylizer @ file:///C:/b/abs_ef86cgllby/croot/qstylizer_1674008538857/work/dist/qstylizer-0.2.2-py2.py3-none-any.whl
QtAwesome @ file:///C:/b/abs_c5evilj98g/croot/qtawesome_1674008690220/work
qtconsole @ file:///C:/b/abs_5bap7f8n0t/croot/qtconsole_1674008444833/work
QtPy @ file:///C:/ci/qtpy_1662015130233/work
queuelib==1.5.0
redis==4.6.0
regex @ file:///C:/ci/regex_1658258299320/work
replicate==0.22.0
requests==2.31.0
requests-file @ file:///Users/ktietz/demo/mc3/conda-bld/requests-file_1629455781986/work
requests-html==0.10.0
requests-toolbelt @ file:///Users/ktietz/demo/mc3/conda-bld/requests-toolbelt_1629456163440/work
resolvelib==1.0.1
RestrictedPython==7.0
result==0.10.0
rich==13.6.0
river==0.21.0
rope @ file:///C:/b/abs_55g_tm_6ff/croot/rope_1676675029164/work
rouge==1.0.1
rsa==4.9
Rtree @ file:///C:/b/abs_e116ltblik/croot/rtree_1675157871717/work
ruamel-yaml-conda @ file:///C:/b/abs_6ejaexx82s/croot/ruamel_yaml_1667489767827/work
ruamel.yaml @ file:///C:/b/abs_30ee5qbthd/croot/ruamel.yaml_1666304562000/work
ruamel.yaml.clib @ file:///C:/b/abs_aarblxbilo/croot/ruamel.yaml.clib_1666302270884/work
s3transfer==0.7.0
safetensors==0.4.1
scikit-image @ file:///C:/b/abs_63r0vmx78u/croot/scikit-image_1669241746873/work
scikit-learn @ file:///C:/b/abs_7ck_bnw91r/croot/scikit-learn_1676911676133/work
scikit-learn-intelex==20230228.214818
scipy==1.11.3
Scrapy @ file:///C:/b/abs_9fn69i_d86/croot/scrapy_1677738199744/work
seaborn @ file:///C:/b/abs_68ltdkoyoo/croot/seaborn_1673479199997/work
selenium==4.9.0
Send2Trash @ file:///tmp/build/80754af9/send2trash_1632406701022/work
sentencepiece==0.1.99
service-identity @ file:///Users/ktietz/demo/mc3/conda-bld/service_identity_1629460757137/work
sherlock==0.4.1
sip @ file:///C:/Windows/Temp/abs_b8fxd17m2u/croots/recipe/sip_1659012372737/work
six @ file:///tmp/build/80754af9/six_1644875935023/work
smart-open @ file:///C:/ci/smart_open_1651235038100/work
smmap==5.0.1
sniffio @ file:///C:/ci_310/sniffio_1642092172680/work
snowballstemmer @ file:///tmp/build/80754af9/snowballstemmer_1637937080595/work
snscrape==0.7.0.20230622
sortedcontainers @ file:///tmp/build/80754af9/sortedcontainers_1623949099177/work
sounddevice==0.4.6
soupsieve @ file:///C:/b/abs_fasraqxhlv/croot/soupsieve_1666296394662/work
spacy==3.5.4
spacy-legacy==3.0.12
spacy-loggers==1.0.4
SpeechRecognition==3.10.0
Sphinx @ file:///C:/ci/sphinx_1657617157451/work
sphinxcontrib-applehelp @ file:///home/ktietz/src/ci/sphinxcontrib-applehelp_1611920841464/work
sphinxcontrib-devhelp @ file:///home/ktietz/src/ci/sphinxcontrib-devhelp_1611920923094/work
sphinxcontrib-htmlhelp @ file:///tmp/build/80754af9/sphinxcontrib-htmlhelp_1623945626792/work
sphinxcontrib-jsmath @ file:///home/ktietz/src/ci/sphinxcontrib-jsmath_1611920942228/work
sphinxcontrib-qthelp @ file:///home/ktietz/src/ci/sphinxcontrib-qthelp_1611921055322/work
sphinxcontrib-serializinghtml @ file:///tmp/build/80754af9/sphinxcontrib-serializinghtml_1624451540180/work
spyder @ file:///C:/b/abs_93s9xkw3pn/croot/spyder_1677776163871/work
spyder-kernels @ file:///C:/b/abs_feh4xo1mrn/croot/spyder-kernels_1673292245176/work
SQLAlchemy @ file:///C:/Windows/Temp/abs_f8661157-660b-49bb-a790-69ab9f3b8f7c8a8s2psb/croots/recipe/sqlalchemy_1657867864564/work
sqlitedict==2.1.0
srsly==2.4.6
stack-data @ file:///opt/conda/conda-bld/stack_data_1646927590127/work
starlette==0.27.0
statsmodels @ file:///C:/b/abs_bdqo3zaryj/croot/statsmodels_1676646249859/work
stqdm==0.0.5
streamlit==1.27.2
streamlit-jupyter==0.2.1
syft==0.8.3
sympy @ file:///C:/b/abs_95fbf1z7n6/croot/sympy_1668202411612/work
tables==3.7.0
tabulate @ file:///C:/ci/tabulate_1657600805799/work
TBB==0.2
tblib @ file:///Users/ktietz/demo/mc3/conda-bld/tblib_1629402031467/work
Telethon==1.29.2
tenacity @ file:///home/conda/feedstock_root/build_artifacts/tenacity_1692026804430/work
terminado @ file:///C:/b/abs_25nakickad/croot/terminado_1671751845491/work
text-unidecode @ file:///Users/ktietz/demo/mc3/conda-bld/text-unidecode_1629401354553/work
textdistance @ file:///tmp/build/80754af9/textdistance_1612461398012/work
thinc==8.1.10
threadpoolctl @ file:///Users/ktietz/demo/mc3/conda-bld/threadpoolctl_1629802263681/work
three-merge @ file:///tmp/build/80754af9/three-merge_1607553261110/work
tifffile @ file:///tmp/build/80754af9/tifffile_1627275862826/work
tiktoken==0.4.0
tinycss2 @ file:///C:/b/abs_52w5vfuaax/croot/tinycss2_1668168823131/work
tldextract @ file:///opt/conda/conda-bld/tldextract_1646638314385/work
tokenizers==0.15.2
toml @ file:///tmp/build/80754af9/toml_1616166611790/work
tomli @ file:///C:/Windows/TEMP/abs_ac109f85-a7b3-4b4d-bcfd-52622eceddf0hy332ojo/croots/recipe/tomli_1657175513137/work
tomlkit @ file:///C:/Windows/TEMP/abs_3296qo9v6b/croots/recipe/tomlkit_1658946894808/work
toolz @ file:///C:/b/abs_cfvk6rc40d/croot/toolz_1667464080130/work
torch==2.1.2
torchaudio==2.0.2+cu118
torchdata==0.7.1
torchtext==0.16.2
torchvision==0.15.2+cu118
tornado @ file:///C:/ci_310/tornado_1642093111997/work
tqdm==4.66.1
traitlets @ file:///C:/b/abs_e5m_xjjl94/croot/traitlets_1671143896266/work
transformers==4.38.1
trio==0.22.1
trio-websocket==0.10.3
Twisted @ file:///C:/Windows/Temp/abs_ccblv2rzfa/croots/recipe/twisted_1659592764512/work
twisted-iocpsupport @ file:///C:/ci/twisted-iocpsupport_1646817083730/work
typeguard==2.13.3
typer==0.9.0
typing-inspect==0.9.0
typing_extensions==4.8.0
tzdata==2023.3
tzlocal==5.0.1
ujson @ file:///C:/ci/ujson_1657525893897/work
Unidecode @ file:///tmp/build/80754af9/unidecode_1614712377438/work
update-checker==0.18.0
uritemplate==4.1.1
urllib3 @ file:///C:/b/abs_9bcwxczrvm/croot/urllib3_1673575521331/work
uvicorn==0.24.0.post1
validators==0.20.0
virtualenv==20.25.0
virtualenv-api==2.1.18
vocode==0.1.111
w3lib @ file:///Users/ktietz/demo/mc3/conda-bld/w3lib_1629359764703/work
wasabi==1.1.2
watchdog @ file:///C:/ci_310/watchdog_1642113443984/work
watchfiles==0.21.0
wcwidth @ file:///Users/ktietz/demo/mc3/conda-bld/wcwidth_1629357192024/work
weaviate-client==3.22.0
webencodings==0.5.1
websocket-client @ file:///C:/ci_310/websocket-client_1642093970919/work
websockets==11.0.3
Werkzeug @ file:///C:/b/abs_17q5kgb8bo/croot/werkzeug_1671216014857/work
whatthepatch @ file:///C:/Users/BUILDE~1/AppData/Local/Temp/abs_e7bihs8grh/croots/recipe/whatthepatch_1661796085215/work
widgetsnbextension==3.6.6
wikipedia==1.4.0
win-inet-pton @ file:///C:/ci_310/win_inet_pton_1642658466512/work
win32-setctime==1.1.0
wincertstore==0.2
wolframalpha==5.0.0
wrapt @ file:///C:/Windows/Temp/abs_7c3dd407-1390-477a-b542-fd15df6a24085_diwiza/croots/recipe/wrapt_1657814452175/work
wsproto==1.2.0
xarray @ file:///C:/b/abs_2fi_umrauo/croot/xarray_1668776806973/work
xlwings @ file:///C:/b/abs_1ejhh6s00l/croot/xlwings_1677024180629/work
xmltodict==0.13.0
xxhash==3.3.0
yapf @ file:///tmp/build/80754af9/yapf_1615749224965/work
yarl==1.9.2
zict==2.1.0
zipp @ file:///C:/b/abs_b9jfdr908q/croot/zipp_1672387552360/work
zope.event==5.0
zope.interface @ file:///C:/ci_310/zope.interface_1642113633904/work
zstandard==0.19.0
```
Platform:
Windows 11
Python version:
Python 3.10.9 | ValueError: Error raised by bedrock service: An error occurred (ValidationException) when calling the InvokeModel operation: Operation not allowed | https://api.github.com/repos/langchain-ai/langchain/issues/19215/comments | 1 | 2024-03-18T04:52:21Z | 2024-03-19T06:55:07Z | https://github.com/langchain-ai/langchain/issues/19215 | 2,191,228,767 | 19,215 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
def add_texts(
self,
texts: Iterable[str],
metadatas: Optional[List[dict]] = None,
timeout: Optional[int] = None,
batch_size: int = 1000,
**kwargs: Any,
) -> List[str]:
"""Insert text data into Milvus.
Inserting data when the collection has not be made yet will result
in creating a new Collection. The data of the first entity decides
the schema of the new collection, the dim is extracted from the first
embedding and the columns are decided by the first metadata dict.
Metadata keys will need to be present for all inserted values. At
the moment there is no None equivalent in Milvus.
Args:
texts (Iterable[str]): The texts to embed, it is assumed
that they all fit in memory.
metadatas (Optional[List[dict]]): Metadata dicts attached to each of
the texts. Defaults to None.
timeout (Optional[int]): Timeout for each batch insert. Defaults
to None.
batch_size (int, optional): Batch size to use for insertion.
Defaults to 1000.
Raises:
MilvusException: Failure to add texts
Returns:
List[str]: The resulting keys for each inserted element.
"""
from pymilvus import Collection, MilvusException
texts = list(texts)
try:
embeddings = self.embedding_func.embed_documents(texts)
except NotImplementedError:
embeddings = [self.embedding_func.embed_query(x) for x in texts]
if len(embeddings) == 0:
logger.debug("Nothing to insert, skipping.")
return []
# If the collection hasn't been initialized yet, perform all steps to do so
if not isinstance(self.col, Collection):
kwargs = {"embeddings": embeddings, "metadatas": metadatas}
if self.partition_names:
kwargs["partition_names"] = self.partition_names
if self.replica_number:
kwargs["replica_number"] = self.replica_number
if self.timeout:
kwargs["timeout"] = self.timeout
self._init(**kwargs)
# Dict to hold all insert columns
insert_dict: dict[str, list] = {
self._text_field: texts,
self._vector_field: embeddings,
}
if self._metadata_field is not None:
for d in metadatas:
insert_dict.setdefault(self._metadata_field, []).append(d)
else:
# Collect the metadata into the insert dict.
if metadatas is not None:
for d in metadatas:
for key, value in d.items():
if key in self.fields:
insert_dict.setdefault(key, []).append(value)
# Total insert count
vectors: list = insert_dict[self._vector_field]
total_count = len(vectors)
pks: list[str] = []
assert isinstance(self.col, Collection)
for i in range(0, total_count, batch_size):
# Grab end index
end = min(i + batch_size, total_count)
# Convert dict to list of lists batch for insertion
insert_list = [insert_dict[x][i:end] for x in self.fields]
# Insert into the collection.
try:
res: Collection
res = self.col.insert(insert_list, timeout=timeout, **kwargs)
pks.extend(res.primary_keys)
except MilvusException as e:
logger.error(
"Failed to insert batch starting at entity: %s/%s", i, total_count
)
raise e
self.col.flush()
return pks
```
self.col.flush() very slowly .
maybe we can use milvus its auto flush
https://milvus.io/docs/configure_quota_limits.md#quotaAndLimitsflushRateenabled
### Error Message and Stack Trace (if applicable)
_No response_
### Description
I'm trying to use langchain 0.0.332 and 0.1.12, but col.insert is still slowly.
maybe we can use milvus its auto flush
https://milvus.io/docs/configure_quota_limits.md#quotaAndLimitsflushRateenabled
### System Info
langchain version 0.0.332
linux ubuntu 20.0 | milvus col.flush() slowly | https://api.github.com/repos/langchain-ai/langchain/issues/19213/comments | 0 | 2024-03-18T02:26:40Z | 2024-06-24T16:13:52Z | https://github.com/langchain-ai/langchain/issues/19213 | 2,191,092,437 | 19,213 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from typing import Annotated, List, Tuple, Union
from langchain_community.tools.tavily_search import TavilySearchResults
from langchain_core.tools import tool
from langchain_experimental.tools import PythonREPLTool
tavily_tool = TavilySearchResults(max_results=5)
# This executes code locally, which can be unsafe
python_repl_tool = PythonREPLTool()
from langchain.agents import AgentExecutor, create_openai_tools_agent
from langchain_core.messages import BaseMessage, HumanMessage
from langchain_openai import ChatOpenAI
def create_agent(
llm: ChatOpenAI, tools: list, system_prompt: str
):
# Each worker node will be given a name and some tools.
prompt = ChatPromptTemplate.from_messages(
[
(
"system",
system_prompt,
),
MessagesPlaceholder(variable_name="messages"),
MessagesPlaceholder(variable_name="agent_scratchpad"),
]
)
agent = create_openai_tools_agent(llm, tools, prompt)
executor = AgentExecutor(agent=agent, tools=tools)
return executor
########
# Async agent nodes
########
async def agent_node(state, agent, name):
result = await agent.ainvoke(state)
return {"messages": [HumanMessage(content=result["output"], name=name)]}
from langchain.output_parsers.openai_functions import JsonOutputFunctionsParser
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
members = ["Researcher", "Coder"]
system_prompt = (
"You are a supervisor tasked with managing a conversation between the"
" following workers: {members}. Given the following user request,"
" respond with the worker to act next. Each worker will perform a"
" task and respond with their results and status. When finished,"
" respond with FINISH."
)
# Our team supervisor is an LLM node. It just picks the next agent to process
# and decides when the work is completed
options = ["FINISH"] + members
# Using openai function calling can make output parsing easier for us
function_def = {
"name": "route",
"description": "Select the next role.",
"parameters": {
"title": "routeSchema",
"type": "object",
"properties": {
"next": {
"title": "Next",
"anyOf": [
{"enum": options},
],
}
},
"required": ["next"],
},
}
prompt = ChatPromptTemplate.from_messages(
[
("system", system_prompt),
MessagesPlaceholder(variable_name="messages"),
(
"system",
"Given the conversation above, who should act next?"
" Or should we FINISH? Select one of: {options}",
),
]
).partial(options=str(options), members=", ".join(members))
llm = ChatOpenAI(model="gpt-3.5-turbo", streaming=True) # allow streaming
supervisor_chain = (
prompt
| llm.bind_functions(functions=[function_def], function_call="route")
| JsonOutputFunctionsParser()
)
import operator
from typing import Annotated, Any, Dict, List, Optional, Sequence, TypedDict
import functools
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langgraph.graph import StateGraph, END
# The agent state is the input to each node in the graph
class AgentState(TypedDict):
# The annotation tells the graph that new messages will always
# be added to the current states
messages: Annotated[Sequence[BaseMessage], operator.add]
# The 'next' field indicates where to route to next
next: str
research_agent = create_agent(llm, [tavily_tool], "You are a web researcher.")
research_node = functools.partial(agent_node, agent=research_agent, name="Researcher")
# NOTE: THIS PERFORMS ARBITRARY CODE EXECUTION. PROCEED WITH CAUTION
code_agent = create_agent(llm, [python_repl_tool], "You may generate safe python code to analyze data and generate charts using matplotlib.")
code_node = functools.partial(agent_node, agent=code_agent, name="Coder")
workflow = StateGraph(AgentState)
workflow.add_node("Researcher", research_node)
workflow.add_node("Coder", code_node)
workflow.add_node("supervisor", supervisor_chain)
for member in members:
# We want our workers to ALWAYS "report back" to the supervisor when done
workflow.add_edge(member, "supervisor")
# The supervisor populates the "next" field in the graph state
# which routes to a node or finishes
conditional_map = {k: k for k in members}
conditional_map["FINISH"] = END
workflow.add_conditional_edges("supervisor", lambda x: x["next"], conditional_map)
# Finally, add entrypoint
workflow.set_entry_point("supervisor")
graph = workflow.compile()
#######
# streaming events
#######
inputs = {'messages':[HumanMessage(content="write a research report on pikas.")]}
#inputs = {"messages": [HumanMessage(content="What is 1+1?")]}
async for event in graph.astream_events(inputs, version="v1"):
kind = event["event"]
if kind == "on_chat_model_stream":
content = event["data"]["chunk"].content
if content:
# Empty content in the context of OpenAI means
# that the model is asking for a tool to be invoked.
# So we only print non-empty content
print(content, end="|")
elif kind == "on_tool_start":
print("--")
print(
f"Starting tool: {event['name']} with inputs: {event['data'].get('input')}"
)
elif kind == "on_tool_end":
print(f"Done tool: {event['name']}")
print(f"Tool output was: {event['data'].get('output')}")
print("--")
```
### Error Message and Stack Trace (if applicable)
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
Cell In[14], line 4
1 inputs = {'messages':[HumanMessage(content="write a research report on pikas.")]}
2 #inputs = {"messages": [HumanMessage(content="What is 1+1?")]}
----> 4 async for event in graph.astream_events(inputs, version="v1"):
5 kind = event["event"]
6 if kind == "on_chat_model_stream":
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:1063, in Runnable.astream_events(self, input, config, version, include_names, include_types, include_tags, exclude_names, exclude_types, exclude_tags, **kwargs)
1059 root_name = config.get("run_name", self.get_name())
1061 # Ignoring mypy complaint about too many different union combinations
1062 # This arises because many of the argument types are unions
-> 1063 async for log in _astream_log_implementation( # type: ignore[misc]
1064 self,
1065 input,
1066 config=config,
1067 stream=stream,
1068 diff=True,
1069 with_streamed_output_list=True,
1070 **kwargs,
1071 ):
1072 run_log = run_log + log
1074 if not encountered_start_event:
1075 # Yield the start event for the root runnable.
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\tracers\log_stream.py:616, in _astream_log_implementation(runnable, input, config, stream, diff, with_streamed_output_list, **kwargs)
613 finally:
614 # Wait for the runnable to finish, if not cancelled (eg. by break)
615 try:
--> 616 await task
617 except asyncio.CancelledError:
618 pass
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\tracers\log_stream.py:570, in _astream_log_implementation.<locals>.consume_astream()
567 prev_final_output: Optional[Output] = None
568 final_output: Optional[Output] = None
--> 570 async for chunk in runnable.astream(input, config, **kwargs):
571 prev_final_output = final_output
572 if final_output is None:
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\pregel\__init__.py:872, in Pregel.astream(self, input, config, output_keys, input_keys, interrupt_before_nodes, interrupt_after_nodes, debug, **kwargs)
869 async def input_stream() -> AsyncIterator[Union[dict[str, Any], Any]]:
870 yield input
--> 872 async for chunk in self.atransform(
873 input_stream(),
874 config,
875 output_keys=output_keys,
876 input_keys=input_keys,
877 interrupt_before_nodes=interrupt_before_nodes,
878 interrupt_after_nodes=interrupt_after_nodes,
879 debug=debug,
880 **kwargs,
881 ):
882 yield chunk
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\pregel\__init__.py:896, in Pregel.atransform(self, input, config, output_keys, input_keys, interrupt_before_nodes, interrupt_after_nodes, debug, **kwargs)
884 async def atransform(
885 self,
886 input: AsyncIterator[Union[dict[str, Any], Any]],
(...)
894 **kwargs: Any,
895 ) -> AsyncIterator[Union[dict[str, Any], Any]]:
--> 896 async for chunk in self._atransform_stream_with_config(
897 input,
898 self._atransform,
899 config,
900 output_keys=output_keys,
901 input_keys=input_keys,
902 interrupt_before_nodes=interrupt_before_nodes,
903 interrupt_after_nodes=interrupt_after_nodes,
904 debug=debug,
905 **kwargs,
906 ):
907 yield chunk
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:1783, in Runnable._atransform_stream_with_config(self, input, transformer, config, run_type, **kwargs)
1781 while True:
1782 if accepts_context(asyncio.create_task):
-> 1783 chunk: Output = await asyncio.create_task( # type: ignore[call-arg]
1784 py_anext(iterator), # type: ignore[arg-type]
1785 context=context,
1786 )
1787 else:
1788 chunk = cast(Output, await py_anext(iterator))
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\tracers\log_stream.py:237, in LogStreamCallbackHandler.tap_output_aiter(self, run_id, output)
233 async def tap_output_aiter(
234 self, run_id: UUID, output: AsyncIterator[T]
235 ) -> AsyncIterator[T]:
236 """Tap an output async iterator to stream its values to the log."""
--> 237 async for chunk in output:
238 # root run is handled in .astream_log()
239 if run_id != self.root_id:
240 # if we can't find the run silently ignore
241 # eg. because this run wasn't included in the log
242 if key := self._key_map_by_run_id.get(run_id):
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\pregel\__init__.py:697, in Pregel._atransform(self, input, run_manager, config, **kwargs)
690 done, inflight = await asyncio.wait(
691 futures,
692 return_when=asyncio.FIRST_EXCEPTION,
693 timeout=self.step_timeout,
694 )
696 # panic on failure or timeout
--> 697 _panic_or_proceed(done, inflight, step)
699 # apply writes to channels
700 _apply_writes(
701 checkpoint, channels, pending_writes, config, step + 1
702 )
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\pregel\__init__.py:922, in _panic_or_proceed(done, inflight, step)
920 inflight.pop().cancel()
921 # raise the exception
--> 922 raise exc
923 # TODO this is where retry of an entire step would happen
925 if inflight:
926 # if we got here means we timed out
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\pregel\__init__.py:1071, in _aconsume(iterator)
1069 async def _aconsume(iterator: AsyncIterator[Any]) -> None:
1070 """Consume an async iterator."""
-> 1071 async for _ in iterator:
1072 pass
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:4435, in RunnableBindingBase.astream(self, input, config, **kwargs)
4429 async def astream(
4430 self,
4431 input: Input,
4432 config: Optional[RunnableConfig] = None,
4433 **kwargs: Optional[Any],
4434 ) -> AsyncIterator[Output]:
-> 4435 async for item in self.bound.astream(
4436 input,
4437 self._merge_configs(config),
4438 **{**self.kwargs, **kwargs},
4439 ):
4440 yield item
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:3920, in RunnableLambda.astream(self, input, config, **kwargs)
3917 async def input_aiter() -> AsyncIterator[Input]:
3918 yield input
-> 3920 async for chunk in self.atransform(input_aiter(), config, **kwargs):
3921 yield chunk
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:3903, in RunnableLambda.atransform(self, input, config, **kwargs)
3897 async def atransform(
3898 self,
3899 input: AsyncIterator[Input],
3900 config: Optional[RunnableConfig] = None,
3901 **kwargs: Optional[Any],
3902 ) -> AsyncIterator[Output]:
-> 3903 async for output in self._atransform_stream_with_config(
3904 input,
3905 self._atransform,
3906 self._config(config, self.afunc if hasattr(self, "afunc") else self.func),
3907 **kwargs,
3908 ):
3909 yield output
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:1783, in Runnable._atransform_stream_with_config(self, input, transformer, config, run_type, **kwargs)
1781 while True:
1782 if accepts_context(asyncio.create_task):
-> 1783 chunk: Output = await asyncio.create_task( # type: ignore[call-arg]
1784 py_anext(iterator), # type: ignore[arg-type]
1785 context=context,
1786 )
1787 else:
1788 chunk = cast(Output, await py_anext(iterator))
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\tracers\log_stream.py:237, in LogStreamCallbackHandler.tap_output_aiter(self, run_id, output)
233 async def tap_output_aiter(
234 self, run_id: UUID, output: AsyncIterator[T]
235 ) -> AsyncIterator[T]:
236 """Tap an output async iterator to stream its values to the log."""
--> 237 async for chunk in output:
238 # root run is handled in .astream_log()
239 if run_id != self.root_id:
240 # if we can't find the run silently ignore
241 # eg. because this run wasn't included in the log
242 if key := self._key_map_by_run_id.get(run_id):
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:3872, in RunnableLambda._atransform(self, input, run_manager, config, **kwargs)
3870 output = chunk
3871 else:
-> 3872 output = await acall_func_with_variable_args(
3873 cast(Callable, afunc), cast(Input, final), config, run_manager, **kwargs
3874 )
3876 # If the output is a runnable, use its astream output
3877 if isinstance(output, Runnable):
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:3847, in RunnableLambda._atransform.<locals>.f(*args, **kwargs)
3845 @wraps(func)
3846 async def f(*args, **kwargs): # type: ignore[no-untyped-def]
-> 3847 return await run_in_executor(config, func, *args, **kwargs)
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\config.py:493, in run_in_executor(executor_or_config, func, *args, **kwargs)
480 """Run a function in an executor.
481
482 Args:
(...)
489 Output: The output of the function.
490 """
491 if executor_or_config is None or isinstance(executor_or_config, dict):
492 # Use default executor with context copied from current context
--> 493 return await asyncio.get_running_loop().run_in_executor(
494 None,
495 cast(Callable[..., T], partial(copy_context().run, func, *args, **kwargs)),
496 )
498 return await asyncio.get_running_loop().run_in_executor(
499 executor_or_config, partial(func, **kwargs), *args
500 )
File c:\ProgramData\Anaconda3\envs\llm\Lib\concurrent\futures\thread.py:58, in _WorkItem.run(self)
55 return
57 try:
---> 58 result = self.fn(*self.args, **self.kwargs)
59 except BaseException as exc:
60 self.future.set_exception(exc)
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\base.py:3841, in RunnableLambda._atransform.<locals>.func(input, run_manager, config, **kwargs)
3835 def func(
3836 input: Input,
3837 run_manager: AsyncCallbackManagerForChainRun,
3838 config: RunnableConfig,
3839 **kwargs: Any,
3840 ) -> Output:
-> 3841 return call_func_with_variable_args(
3842 self.func, input, config, run_manager.get_sync(), **kwargs
3843 )
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langchain_core\runnables\config.py:326, in call_func_with_variable_args(func, input, config, run_manager, **kwargs)
324 if run_manager is not None and accepts_run_manager(func):
325 kwargs["run_manager"] = run_manager
--> 326 return func(input, **kwargs)
File c:\ProgramData\Anaconda3\envs\llm\Lib\site-packages\langgraph\graph\graph.py:37, in Branch.runnable(self, input)
35 result = self.condition(input)
36 if self.ends:
---> 37 destination = self.ends[result]
38 else:
39 destination = result
KeyError: 'ResearchResearcher'
### Description
This is a follow up to the recent updates to astream_events in #18743 and #19051. The discussion originally started in [langchain-ai/langgraph#136](https://github.com/langchain-ai/langgraph/issues/136).
I'm testing `astream_event` method on [this langgraph example notebook](https://github.com/langchain-ai/langgraph/blob/main/examples/multi_agent/agent_supervisor.ipynb). I modified several lines of code to allow LLM streaming and convert the nodes to async functions. Without `astream_events`, the supervisor node will output "next: Researcher", however it returns "next: ResearchResearcher", which breaks graph streaming since the name is not recognized as a graph node. Another example is "next: FINFINISH". The only scenario it's working is when router outputs "next: Coder" (one token).
My suspicion is the recently updated AddableDict logic in stream_events has a problem with `JSONOutputFunctionParser` in the supervisor chain. Any help from @eyurtsev or other team members would be greatly appreciated!
Here's the full Langsmith Trace: https://smith.langchain.com/public/e09b671b-325d-477e-bd29-e017d30c6741/r
### System Info
langchain=0.1.12
langchain-core=0.1.33rc1
langgraph=0.0.28 | astream_event produces redundant tokens and breaks graph streams | https://api.github.com/repos/langchain-ai/langchain/issues/19211/comments | 3 | 2024-03-18T01:11:51Z | 2024-08-04T16:07:16Z | https://github.com/langchain-ai/langchain/issues/19211 | 2,191,027,572 | 19,211 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
Documentation Page: [Evaluation](https://python.langchain.com/docs/guides/evaluation/)
### Idea or request for content:
Links to LangSmith documentation are broken. They should point to:
- **LangSmith Evaluation**: https://docs.smith.langchain.com/evaluation
- **cookbooks**: https://docs.smith.langchain.com/evaluation/faq | DOC: Broken links on Evaluation page | https://api.github.com/repos/langchain-ai/langchain/issues/19210/comments | 0 | 2024-03-18T00:42:48Z | 2024-03-19T02:13:11Z | https://github.com/langchain-ai/langchain/issues/19210 | 2,191,005,514 | 19,210 |
[
"langchain-ai",
"langchain"
] | ### 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
```
chain = ConversationalRetrievalChain.from_llm(
llm = llm,
memory = memory,
chain_type = "stuff",
retriever = retriever,
combine_docs_chain_kwargs = ...,
condense_question_prompt = ...,
return_source_documents = True,
return_generated_question = True,
verbose = True, debug = True
)
# Adding none of these will bypass the condensing phase and have the chain complete:
chain.rephrase_question = False
chain.question_generator = None
chain.condense_question_llm = None
```
### Error Message and Stack Trace (if applicable)
See #6879 for a description of the issue.
### Description
The goal is: if requested, the `ConversationRetreivalChain` should skip the question condensing phase.
(The work-around presented no longer seems to work, but ISTM shouldn't be necessary - I tried adding all the default Runnable methods, but was still getting the error I described - but maybe I was doing it wrong.)
The reason not to use the `RetrievalQAChain` is that the RQAC and CRC are inconsistent in their input/output parameters for some reason - this makes it a PITA to switch between them dynamically at run-time, including changing the keys used for memory, dealing with extra or invalid chain dicts or kwargs, etc.
**Desired behavior**
The RQAC is a simple sub-set of the CRC, so the CRC should behave just like the RQAC if you set either:
(a) `question_generator = None`
(b) `condense_question_llm = None`
(c) `rephrase_question = False`
(or perhaps a combination of (c) and either (a) or (b) if you want to handle (a) or (b) is None as an error condition without (c) == True. (This would make the RQAC redundant, but that's a fixture now.)
Setting `rephrase_question = False` currently is curious because I believe it uses the user's question verbatim, _but still calls the condensing llm_ - I'm not sure what the rationale is here.
If there is currently an LCEL equivalent of the CRC that behaves with or without the `condense_question_llm` in the chain, that would be another option, but simply honoring `rephrase_question = False` fully seems straightforward here.
### System Info
```
langchain==0.1.11
langchain-community==0.0.28
langchain-core==0.1.31
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
```
Python 3.10 | Unable to defeat question condensing in ConversationRetrievalChain (see #6879) | https://api.github.com/repos/langchain-ai/langchain/issues/19200/comments | 0 | 2024-03-17T16:04:46Z | 2024-06-23T16:09:30Z | https://github.com/langchain-ai/langchain/issues/19200 | 2,190,741,952 | 19,200 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
[cookbook/multiple_chains](https://python.langchain.com/docs/expression_language/cookbook/multiple_chains)
[cookbook/sql_db](https://python.langchain.com/docs/expression_language/cookbook/sql_db)
### Idea or request for content:
_No response_ | DOC: Updating `pip install` format in cookbook | https://api.github.com/repos/langchain-ai/langchain/issues/19197/comments | 0 | 2024-03-17T12:12:52Z | 2024-06-23T16:09:25Z | https://github.com/langchain-ai/langchain/issues/19197 | 2,190,634,544 | 19,197 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
Valid OCI authentication types include `INSTANCE_PRINCIPAL` and `RESOURCE_PRINCIPAL`. However, comments in the code for [llms](https://github.com/langchain-ai/langchain/blob/master/libs/community/langchain_community/llms/oci_generative_ai.py) and [embeddings](libs/community/langchain_community/embeddings/oci_generative_ai.py) incorrectly described these as `INSTANCE_PRINCIPLE` and `RESOURCE_PRINCIPLE` respectively. The incorrect information is also presented in an error message.
### Idea or request for content:
The code comments that show up in the API documentation, and error messages should be corrected to reflect the correct values. | DOC: Incorrect description of valid OCI authentication types for OCI Generative AI LLM and embeddings | https://api.github.com/repos/langchain-ai/langchain/issues/19194/comments | 0 | 2024-03-17T05:31:38Z | 2024-06-23T16:09:29Z | https://github.com/langchain-ai/langchain/issues/19194 | 2,190,488,898 | 19,194 |
[
"langchain-ai",
"langchain"
] | ### 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
Current implementation:
```python
def _stream(
self,
prompt: str,
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> Iterator[GenerationChunk]:
params = {**self._invocation_params, **kwargs, "stream": True}
self.get_sub_prompts(params, [prompt], stop) # this mutates params
for stream_resp in self.client.create(prompt=prompt, **params):
if not isinstance(stream_resp, dict):
stream_resp = stream_resp.model_dump()
chunk = _stream_response_to_generation_chunk(stream_resp)
yield chunk
if run_manager:
run_manager.on_llm_new_token(
chunk.text,
chunk=chunk,
verbose=self.verbose,
logprobs=(
chunk.generation_info["logprobs"]
if chunk.generation_info
else None
),
)
```
I believe this would correct and produce the intended behavior:
```python
def _stream(
self,
prompt: str,
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> Iterator[GenerationChunk]:
params = {**self._invocation_params, **kwargs, "stream": True}
self.get_sub_prompts(params, [prompt], stop) # this mutates params
for stream_resp in self.client.create(prompt=prompt, **params):
if not isinstance(stream_resp, dict):
stream_resp = stream_resp.model_dump()
chunk = _stream_response_to_generation_chunk(stream_resp)
if run_manager:
run_manager.on_llm_new_token(
chunk.text,
chunk=chunk,
verbose=self.verbose,
logprobs=(
chunk.generation_info["logprobs"]
if chunk.generation_info
else None
),
)
yield chunk
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
When streaming via ``langchain_openai.llms.base.BaseOpenAI._stream`` the yield appears before triggering the run manager event. This makes it impossible to invoke ``on_llm_new_token`` methods in a callback until the full response is received.
### System Info
```text
System Information
------------------
> OS: Linux
> OS Version: #21~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Feb 9 13:32:52 UTC 2
> Python Version: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
Package Information
-------------------
> langchain_core: 0.1.30
> langchain: 0.1.11
> langchain_community: 0.0.27
> langsmith: 0.1.23
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
``` | on_llm_new_token event broken in langchain_openai when streaming | https://api.github.com/repos/langchain-ai/langchain/issues/19185/comments | 2 | 2024-03-16T12:03:38Z | 2024-06-29T16:08:37Z | https://github.com/langchain-ai/langchain/issues/19185 | 2,189,935,758 | 19,185 |
[
"langchain-ai",
"langchain"
] | ### 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
from langchain.document_loaders import CSVLoader
loader = CSVLoader("./institution_all.csv")
data = loader.load()
### Error Message and Stack Trace (if applicable)
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
File /opt/conda/lib/python3.10/site-packages/langchain_community/document_loaders/csv_loader.py:67, in CSVLoader.lazy_load(self)
66 with open(self.file_path, newline="", encoding=self.encoding) as csvfile:
---> 67 yield from self.__read_file(csvfile)
68 except UnicodeDecodeError as e:
File /opt/conda/lib/python3.10/site-packages/langchain_community/document_loaders/csv_loader.py:98, in CSVLoader.__read_file(self, csvfile)
95 raise ValueError(
96 f"Source column '{self.source_column}' not found in CSV file."
97 )
---> 98 content = "\n".join(
99 f"{k.strip()}: {v.strip() if v is not None else v}"
100 for k, v in row.items()
101 if k not in self.metadata_columns
102 )
103 metadata = {"source": source, "row": i}
File /opt/conda/lib/python3.10/site-packages/langchain_community/document_loaders/csv_loader.py:99, in <genexpr>(.0)
95 raise ValueError(
96 f"Source column '{self.source_column}' not found in CSV file."
97 )
98 content = "\n".join(
---> 99 f"{k.strip()}: {v.strip() if v is not None else v}"
100 for k, v in row.items()
101 if k not in self.metadata_columns
102 )
103 metadata = {"source": source, "row": i}
AttributeError: 'NoneType' object has no attribute 'strip'
The above exception was the direct cause of the following exception:
RuntimeError Traceback (most recent call last)
Cell In[7], line 5
3 from langchain.document_loaders import CSVLoader
4 loader = CSVLoader("./institution_all.csv")
----> 5 data = loader.load()
File /opt/conda/lib/python3.10/site-packages/langchain_core/document_loaders/base.py:29, in BaseLoader.load(self)
27 def load(self) -> List[Document]:
28 """Load data into Document objects."""
---> 29 return list(self.lazy_load())
File /opt/conda/lib/python3.10/site-packages/langchain_community/document_loaders/csv_loader.py:83, in CSVLoader.lazy_load(self)
81 raise RuntimeError(f"Error loading {self.file_path}") from e
82 except Exception as e:
---> 83 raise RuntimeError(f"Error loading {self.file_path}") from e
RuntimeError: Error loading ./institution_all.csv
### Description
我正在尝试导入csv文档
### System Info
langchain-0.1.12 | CSVloader RuntimeError: Error loading ./institution_all.csv | https://api.github.com/repos/langchain-ai/langchain/issues/19174/comments | 3 | 2024-03-16T05:48:21Z | 2024-06-04T14:30:06Z | https://github.com/langchain-ai/langchain/issues/19174 | 2,189,795,094 | 19,174 |
[
"langchain-ai",
"langchain"
] | ### 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
tools = load_tools(["ddg-search", "wikipedia", "llm-math"], llm=llm)
retriever = db.as_retriever()
repl_tool = Tool(
name="python_repl",
description="A Python shell. Use this to execute python commands. Input should be a valid python command. If you want to see the output of a value, you should print it out with `print(...)`.",
func=PythonREPL().run,
)
retrieve_tool = create_retriever_tool(
retriever,
"SearchDocuments",
"Searches and returns results from documents to answer queries.",
)
tools += [repl_tool, retrieve_tool]
return tools
### Error Message and Stack Trace (if applicable)
> Entering new AgentExecutor chain...
Invoking: `SearchDocuments` with `list all documents`
2024-03-15 23:40:14.938 Uncaught app exception
Traceback (most recent call last):
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 535, in _run_script
exec(code, module.__dict__)
File "C:\Users\Rafael\OneDrive - Artificial Intelligence Expert\Documents\BioChat_v2\BioChat\src\app.py", line 196, in <module>
main(constants_path=f"./config/CONSTANTS.json")
File "C:\Users\Rafael\OneDrive - Artificial Intelligence Expert\Documents\BioChat_v2\BioChat\src\app.py", line 165, in main
result = st.session_state["agent"].invoke(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\chains\base.py", line 163, in invoke
raise e
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\chains\base.py", line 153, in invoke
self._call(inputs, run_manager=run_manager)
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\agents\agent.py", line 1432, in _call
next_step_output = self._take_next_step(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\agents\agent.py", line 1138, in _take_next_step
[
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\agents\agent.py", line 1138, in <listcomp>
[
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\agents\agent.py", line 1223, in _iter_next_step
yield self._perform_agent_action(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain\agents\agent.py", line 1245, in _perform_agent_action
observation = tool.run(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\tools.py", line 417, in run
raise e
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\tools.py", line 376, in run
self._run(*tool_args, run_manager=run_manager, **tool_kwargs)
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\tools.py", line 580, in _run
else self.func(*args, **kwargs)
File "C:\Users\Rafael\OneDrive - Artificial Intelligence Expert\Documents\BioChat_v2\BioChat/src\agent.py", line 165, in <lambda>
func=lambda query: retriever_tool(query),
File "C:\Users\Rafael\OneDrive - Artificial Intelligence Expert\Documents\BioChat_v2\BioChat/src\agent.py", line 148, in retriever_tool
docs = retriever.get_relevant_documents(query)
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\retrievers.py", line 244, in get_relevant_documents
raise e
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\retrievers.py", line 237, in get_relevant_documents
result = self._get_relevant_documents(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_core\vectorstores.py", line 674, in _get_relevant_documents
docs = self.vectorstore.similarity_search(query, **self.search_kwargs)
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_community\vectorstores\deeplake.py", line 541, in similarity_search
return self._search(
File "C:\Users\Rafael\anaconda3\envs\kaggle\lib\site-packages\langchain_community\vectorstores\deeplake.py", line 421, in _search
_embedding_function = self._embedding_function.embed_query
AttributeError: 'function' object has no attribute 'embed_query'
### Description
I'm trying to add a custom tool so that an agent can retrieve information from data lake activeloop however I keep getting the error
`AttributeError: 'function' object has no attribute 'embed_query'`
How can I fix this?
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-experimental==0.0.54
langchain-google-vertexai==0.1.0
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
langchainhub==0.1.15 | AttributeError: 'function' object has no attribute 'embed_query' with OpenAI llm and custom tool for Data Lake with Activeloop | https://api.github.com/repos/langchain-ai/langchain/issues/19171/comments | 1 | 2024-03-15T22:45:08Z | 2024-06-24T16:08:13Z | https://github.com/langchain-ai/langchain/issues/19171 | 2,189,579,272 | 19,171 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
The [link](https://api.python.langchain.com/en/latest/tracing.html?ref=blog.langchain.dev) to the tracing documentation, which is referenced [here](https://blog.langchain.dev/tracing/) is broken.
### Idea or request for content:
Please have documentation how to enable tracing of the tool.
Other [pages](https://js.langchain.com/docs/modules/agents/how_to/logging_and_tracing) imply that tracing and verbose are not the same.
If there is a way to time a tool, please add documentation on that too. | DOC: Broken link to Langchain Tracing | https://api.github.com/repos/langchain-ai/langchain/issues/19165/comments | 0 | 2024-03-15T20:06:34Z | 2024-06-21T16:37:05Z | https://github.com/langchain-ai/langchain/issues/19165 | 2,189,389,814 | 19,165 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
import boto3
import json
from langchain_community.chat_models import BedrockChat
from langchain.agents import tool
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
boto3_bedrock = boto3.client('bedrock-runtime')
modelId = "anthropic.claude-3-sonnet-20240229-v1:0"
llm = BedrockChat(
model_id=modelId,
client = boto3_bedrock
)
@tool
def get_word_length(word: str) -> int:
"""Returns the length of a word."""
return len(word)
#get_word_length.invoke("abc")
tools = [get_word_length]
prompt = ChatPromptTemplate.from_messages(
[
(
"system",
"You are very powerful assistant, but don't know current events",
),
("user", "{input}"),
MessagesPlaceholder(variable_name="agent_scratchpad"),
]
)
llm_with_tools = llm.bind_tools(tools)
```
### Error Message and Stack Trace (if applicable)
```python
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Cell In[7], line 1
----> 1 llm_with_tools = llm.bind_tools(tools)
AttributeError: 'BedrockChat' object has no attribute 'bind_tools'
```
### Description
I am trying to follow [this](https://python.langchain.com/docs/modules/agents/how_to/custom_agent) guide with Bedrock and it's throwing this error.
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
langchain-text-splitters==0.0.1 | 'BedrockChat' object has no attribute 'bind_tools' | https://api.github.com/repos/langchain-ai/langchain/issues/19162/comments | 5 | 2024-03-15T18:44:12Z | 2024-07-04T15:30:45Z | https://github.com/langchain-ai/langchain/issues/19162 | 2,189,262,902 | 19,162 |
[
"langchain-ai",
"langchain"
] | ### 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
This is the code for my server.py
```python
from fastapi import FastAPI
from fastapi.responses import RedirectResponse
from langserve import add_routes
from rag_conversation import chain as rag_conversation_chain
app = FastAPI()
@app.get("/")
async def redirect_root_to_docs():
return RedirectResponse("/docs")
add_routes(app, rag_conversation_chain, path="/rag-conversation")
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=8000)
```
### Error Message and Stack Trace (if applicable)
```
INFO: Will watch for changes in these directories:
INFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO: Started reloader process [10212] using WatchFiles
ERROR: Error loading ASGI app. Could not import module "app.server".
```
### Description
I am trying to use the tool lanchain and langserver, and create a api from a template. Steps I took:
1. create server template
```
langchain app new my-app --package rag-conversation
```
2. copy in the code provided in the cmd after the installation is ready
```
@app.get("/")
async def redirect_root_to_docs():
return RedirectResponse("/docs")
add_routes(app, rag_conversation_chain, path="/rag-conversation")
```
3. cd into my-app folder and run langchain serve in the cmd.
After which the server cant seem to start and throws this error
```
INFO: Will watch for changes in these directories:
INFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO: Started reloader process [10212] using WatchFiles
ERROR: Error loading ASGI app. Could not import module "app.server".
```
Does anyone know how to aproach this issue. There is no -v command option so this is all the information I have to go by.
### System Info
python -m langchain_core.sys_info:
```
System Information
------------------
> OS: Windows
> OS Version: 10.0.22631
> Python Version: 3.12.2 (tags/v3.12.2:6abddd9, Feb 6 2024, 21:26:36) [MSC v.1937 64 bit (AMD64)]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.25
> langchain_cli: 0.0.21
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
> langserve: 0.0.51
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
```
pip freeze:
```
aiohttp==3.9.3
aiosignal==1.3.1
annotated-types==0.6.0
anyio==4.3.0
attrs==23.2.0
beautifulsoup4==4.12.3
build==1.1.1
CacheControl==0.14.0
certifi==2024.2.2
charset-normalizer==3.3.2
cleo==2.1.0
click==8.1.7
colorama==0.4.6
crashtest==0.4.1
dataclasses-json==0.6.4
distlib==0.3.8
distro==1.9.0
dulwich==0.21.7
fastapi==0.110.0
fastjsonschema==2.19.1
filelock==3.13.1
frozenlist==1.4.1
gitdb==4.0.11
GitPython==3.1.42
greenlet==3.0.3
h11==0.14.0
httpcore==1.0.4
httptools==0.6.1
httpx==0.27.0
httpx-sse==0.4.0
idna==3.6
installer==0.7.0
jaraco.classes==3.3.1
jsonpatch==1.33
jsonpointer==2.4
keyring==24.3.1
langchain==0.1.12
langchain-cli==0.0.21
langchain-community==0.0.28
langchain-core==0.1.31
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
langserve==0.0.51
langsmith==0.1.25
markdown-it-py==3.0.0
marshmallow==3.21.1
mdurl==0.1.2
more-itertools==10.2.0
msgpack==1.0.8
multidict==6.0.5
mypy-extensions==1.0.0
numpy==1.26.4
openai==1.14.0
orjson==3.9.15
packaging==23.2
pexpect==4.9.0
pinecone-client==3.1.0
pkginfo==1.10.0
platformdirs==4.2.0
poetry==1.8.2
poetry-core==1.9.0
poetry-dotenv-plugin==0.2.0
poetry-plugin-export==1.6.0
ptyprocess==0.7.0
pydantic==2.6.4
pydantic_core==2.16.3
Pygments==2.17.2
pyproject_hooks==1.0.0
python-dotenv==1.0.1
pywin32-ctypes==0.2.2
PyYAML==6.0.1
rapidfuzz==3.6.2
regex==2023.12.25
requests==2.31.0
requests-toolbelt==1.0.0
rich==13.7.1
shellingham==1.5.4
smmap==5.0.1
sniffio==1.3.1
soupsieve==2.5
SQLAlchemy==2.0.28
sse-starlette==1.8.2
starlette==0.36.3
tenacity==8.2.3
tiktoken==0.6.0
tomlkit==0.12.4
tqdm==4.66.2
trove-classifiers==2024.3.3
typer==0.9.0
typing-inspect==0.9.0
typing_extensions==4.10.0
urllib3==2.2.1
uvicorn==0.23.2
virtualenv==20.25.1
watchfiles==0.21.0
websockets==12.0
yarl==1.9.4
``` | Langserver could not import module app.server | https://api.github.com/repos/langchain-ai/langchain/issues/19150/comments | 1 | 2024-03-15T16:38:47Z | 2024-06-26T16:07:40Z | https://github.com/langchain-ai/langchain/issues/19150 | 2,189,059,636 | 19,150 |
[
"langchain-ai",
"langchain"
] | ### 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
```
tools = [CorrelationTool(), DataTool()]
llm = Llm().azure_openai
prompt = hub.pull("hwchase17/react")
agent = create_react_agent(llm, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True, handle_parsing_errors=True, max_iterations = 5, early_stopping_method="generate")
with tracing_v2_enabled(project_name="default"):
ans = agent_executor.invoke({"input": "Present me some conclusions on data of criminality"})
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
The following code only executes once, presenting me with the data from the first action but not going after it, not writing a final answer. I feel like it's missing some loop in langchain code.
### System Info
```
> Entering new AgentExecutor chain...
I should use the Correlation Calculator to find some correlations
Action: Correlation Calculator
Action Input: "criminality in Brazil"
Roubo Furto Homicídio
Roubo 1.000000 0.984111 0.936390
Furto 0.984111 1.000000 0.983135
Homicídio 0.936390 0.983135 1.000000
> Finished chain.
```
| React Agent stops at first observation | https://api.github.com/repos/langchain-ai/langchain/issues/19149/comments | 2 | 2024-03-15T16:33:09Z | 2024-03-18T09:59:51Z | https://github.com/langchain-ai/langchain/issues/19149 | 2,189,049,809 | 19,149 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain.output_parsers import RetryOutputParser
from langchain_core.output_parsers.pydantic import PydanticOutputParser
from langchain_core.pydantic_v1 import BaseModel
from langchain_openai import OpenAI
class TestModel(BaseModel):
a: int
b: str
data_pydantic = TestModel(a=1, b="2")
data_json = data_pydantic.json()
parser = PydanticOutputParser(pydantic_object=TestModel)
retry_parser = RetryOutputParser.from_llm(parser=parser, llm=OpenAI(temperature=0))
retry_parser.parse_with_prompt(completion=data_json, prompt_value="Test prompt")
retry_parser.parse_with_prompt(completion=data_pydantic, prompt_value="Test prompt") # Error
```
### Error Message and Stack Trace (if applicable)
```
ValidationError Traceback (most recent call last)
Cell In[3], [line 20](vscode-notebook-cell:?execution_count=3&line=20)
[16](vscode-notebook-cell:?execution_count=3&line=16) retry_parser = RetryOutputParser.from_llm(parser=parser, llm=OpenAI(temperature=0))
[18](vscode-notebook-cell:?execution_count=3&line=18) retry_parser.parse_with_prompt(completion=data_json, prompt_value="Test prompt")
---> [20](vscode-notebook-cell:?execution_count=3&line=20) retry_parser.parse_with_prompt(completion=data_pydantic, prompt_value="Test prompt")
File [c:\Users\Asus\anaconda3\envs\dev\Lib\site-packages\langchain\output_parsers\retry.py:89](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:89), in RetryOutputParser.parse_with_prompt(self, completion, prompt_value)
[87](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:87) while retries <= self.max_retries:
[88](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:88) try:
---> [89](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:89) return self.parser.parse(completion)
[90](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:90) except OutputParserException as e:
[91](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain/output_parsers/retry.py:91) if retries == self.max_retries:
File [c:\Users\Asus\anaconda3\envs\dev\Lib\site-packages\langchain_core\output_parsers\json.py:218](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/output_parsers/json.py:218), in JsonOutputParser.parse(self, text)
[217](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/output_parsers/json.py:217) def parse(self, text: str) -> Any:
--> [218](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/output_parsers/json.py:218) return self.parse_result([Generation(text=text)])
File [c:\Users\Asus\anaconda3\envs\dev\Lib\site-packages\langchain_core\load\serializable.py:120](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/load/serializable.py:120), in Serializable.__init__(self, **kwargs)
[119](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/load/serializable.py:119) def __init__(self, **kwargs: Any) -> None:
--> [120](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/load/serializable.py:120) super().__init__(**kwargs)
[121](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/langchain_core/load/serializable.py:121) self._lc_kwargs = kwargs
File [c:\Users\Asus\anaconda3\envs\dev\Lib\site-packages\pydantic\main.py:341](file:///C:/Users/Asus/anaconda3/envs/dev/Lib/site-packages/pydantic/main.py:341), in pydantic.main.BaseModel.__init__()
ValidationError: 1 validation error for Generation
text
str type expected (type=type_error.str)
```
### Description
The `RetryOutputParser` does not seem to work correctly when used with `PydanticOutputParser`. I guess it won't work correctly whenever used with a parser that does not output a string.
In the code above, it works when receiving a string, but when receiving anything else, it throws:
```
ValidationError: 1 validation error for Generation
text
str type expected (type=type_error.str)
```
In the context of a chain with a `PydanticOutputParser`, when the llm call returns a correct result as the pydantic model, the retry parser throws an error.
I see no mention about it (the `RetryOutputParser` only accepting a string) in the docs: https://python.langchain.com/docs/modules/model_io/output_parsers/types/retry
I was able to avoid this issue by converting the `completion` value to a json string (shown below), if the type is the same as the expected pydantic model.
```python
def parse_with_prompt(args):
completion = args['completion']
if (type(completion) is TestModel):
args = args.copy()
del args['completion']
completion = completion.json(ensure_ascii=False)
args['completion'] = completion
return retry_parser.parse_with_prompt(**args)
chain = RunnableParallel(
completion=completion_chain, prompt_value=prompt
) | RunnableLambda(parse_with_prompt)
```
The problem is that this seems hackish, and I don't know if this will be portable in new versions of the parser (at least, in the example in the docs, I see no reference to the params that should be passed to `parse_with_prompt`, although I can see in the source code that they are `completion: str` and `prompt_value: PromptValue`, but I'm not sure if this should be considered an implementation detail, considering that there is no mention in the docs). Furthermore, if this issue is fixed in new versions, I may end up converting the model to json when I shouldn't.
For now I'm not using the `RetryOutputParser`, because it seems to not be production ready yet (at least with a parser that does not output a string).
### System Info
```
System Information
------------------
> OS: Windows
> OS Version: 10.0.22621
> Python Version: 3.11.5 | packaged by conda-forge | (main, Aug 27 2023, 03:23:48) [MSC v.1936 64 bit (AMD64)]
Package Information
-------------------
> langchain_core: 0.1.28
> langchain: 0.1.6
> langchain_community: 0.0.19
> langsmith: 0.1.14
> langchain_openai: 0.0.8
> langchainhub: 0.1.14
> langgraph: 0.0.28
> langserve: 0.0.46
``` | `RetryOutputParser` error when used with `PydanticOutputParser` | https://api.github.com/repos/langchain-ai/langchain/issues/19145/comments | 2 | 2024-03-15T16:21:21Z | 2024-07-22T16:08:16Z | https://github.com/langchain-ai/langchain/issues/19145 | 2,189,019,290 | 19,145 |
[
"langchain-ai",
"langchain"
] | ### 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
```
response = llm_chain.invoke(
input={"query": query}, config={"callbacks": [token_counter]}
)
```
### Error Message and Stack Trace (if applicable)
```
Traceback (most recent call last):
File "/app-root/ols/app/endpoints/ols.py", line 216, in validate_question
return question_validator.validate_question(conversation_id, llm_request.query)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app-root/ols/src/query_helpers/question_validator.py", line 78, in validate_question
response = llm_chain.invoke(
^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain/chains/base.py", line 163, in invoke
raise e
File "/usr/local/lib/python3.11/site-packages/langchain/chains/base.py", line 153, in invoke
self._call(inputs, run_manager=run_manager)
File "/usr/local/lib/python3.11/site-packages/langchain/chains/llm.py", line 103, in _call
response = self.generate([inputs], run_manager=run_manager)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain/chains/llm.py", line 115, in generate
return self.llm.generate_prompt(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 544, in generate_prompt
return self.generate(prompt_messages, stop=stop, callbacks=callbacks, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 408, in generate
raise e
File "/usr/local/lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 398, in generate
self._generate_with_cache(
File "/usr/local/lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 577, in _generate_with_cache
return self._generate(
^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain_openai/chat_models/base.py", line 444, in _generate
return generate_from_stream(stream_iter)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 65, in generate_from_stream
for chunk in stream:
File "/usr/local/lib/python3.11/site-packages/langchain_openai/chat_models/base.py", line 408, in _stream
for chunk in self.client.create(messages=message_dicts, **params):
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_utils/_utils.py", line 271, in wrapper
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/resources/chat/completions.py", line 659, in create
return self._post(
^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 1180, in post
return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 869, in request
return self._request(
^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 890, in _request
request = self._build_request(options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 452, in _build_request
headers = self._build_headers(options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/openai/_base_client.py", line 413, in _build_headers
headers = httpx.Headers(headers_dict)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/httpx/_models.py", line 70, in __init__
self._list = [
^
File "/usr/local/lib/python3.11/site-packages/httpx/_models.py", line 74, in <listcomp>
normalize_header_value(v, encoding),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/httpx/_utils.py", line 53, in normalize_header_value
return value.encode(encoding or "ascii")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
UnicodeEncodeError: 'ascii' codec can't encode character '\u23da' in position 58: ordinal not in range(128)
```
### Description
* A unicode character was accidentally pasted into the API key environment variable
* Langchain throws a difficult to understand unicode error because it does not appear to have "sanity checked" the value it was passing to httpx
### System Info
```
langchain==0.1.11
langchain-community==0.0.26
langchain-core==0.1.29
langchain-openai==0.0.5
langchain-text-splitters==0.0.1
```
Fedora release 39 (Thirty Nine)
Python 3.11.5 | When unicode character is in API key, non-specific error is returned (instead of invalid API key) | https://api.github.com/repos/langchain-ai/langchain/issues/19144/comments | 0 | 2024-03-15T15:54:24Z | 2024-06-21T16:37:07Z | https://github.com/langchain-ai/langchain/issues/19144 | 2,188,945,438 | 19,144 |
[
"langchain-ai",
"langchain"
] | ### 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
auth_client_secret = weaviate.AuthApiKey(WEAVIATE_ADMIN_APIKEY)
wv_conn = weaviate.Client(url=WEAVIATE_URL,
auth_client_secret=auth_client_secret))
wvdb = Weaviate(client=self.wv_conn, index_name=index_name,
text_key="text", embedding=embeddings,
attributes=["a1", "a2", "a3", "a4"])
wvdb.add_texts(texts=texts,
metadatas=metadatas,
ids=ids)
wvdb.similarity_search_with_score(query=query, k=top_k)
### Error Message and Stack Trace (if applicable)
ValueError: Error during query: [{'locations': [{'column': 34, 'line': 1}], 'message': 'Unknown argument "nearText" on field "OAAIndexOpenAITest" of type "GetObjectsObj". Did you mean "nearObject" or "nearVector"?', 'path': None}]
### Description
I'm having issues with Langchain creating index and vectorizing text in Weaviate. I do not want to set up a vectorizer in Weaviate but want Langchain to directly do that for me. Vectorizing ElasticSearch using Langchain seems to be working fine but not in Weaviate
### System Info
langchain==0.0.308
langchain-community==0.0.20
langchain-core==0.1.23
langchain-text-splitters==0.0.1
weaviate-client==3.24.2
weaviate== 1.21.2 | Langchain not able to create a new index and/or generate vectors to store in weaviate? | https://api.github.com/repos/langchain-ai/langchain/issues/19143/comments | 1 | 2024-03-15T15:49:11Z | 2024-03-28T18:33:51Z | https://github.com/langchain-ai/langchain/issues/19143 | 2,188,928,384 | 19,143 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
# Helper function for printing docs
def pretty_print_docs(docs):
print(
f"\n{'-' * 100}\n".join(
[f"Document {i+1}:\n\n" + d.page_content for i, d in enumerate(docs)]
)
)
import getpass
import os
os.environ["OPENAI_API_KEY"] = getpass.getpass()
from langchain_community.document_loaders import TextLoader
from langchain_community.vectorstores import FAISS
from langchain_openai import OpenAIEmbeddings
from langchain_text_splitters import RecursiveCharacterTextSplitter
documents = TextLoader(
"../../modules/state_of_the_union.txt",
).load()
text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)
texts = text_splitter.split_documents(documents)
embedding = OpenAIEmbeddings(model="text-embedding-ada-002")
retriever = FAISS.from_documents(texts, embedding).as_retriever(search_kwargs={"k": 20})
from langchain.retrievers import ContextualCompressionRetriever, FlashrankRerank
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(temperature=0)
compressor = FlashrankRerank()
compression_retriever = ContextualCompressionRetriever(
base_compressor=compressor, base_retriever=retriever
)
compressed_docs = compression_retriever.get_relevant_documents(
"What did the president say about Ketanji Jackson Brown"
)
print([doc.metadata["source"] for doc in compressed_docs])
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
When using `FlashrankRerank`, the original documents' metadata is overwritten with just an `id` and the `relevance_score`. This leads to complications if the documents were retrieved with important metadata.
`id` is also a commonly used metadata when documents are stored in a database and should not be overwritten.
As it is implemented, the `id` assigned by `FlashrankRerank` is trivial, because it is simply the index in the list of documents.
What I would like to happen is that the list of input documents is returned as is, just reordered and filtered, with the metadata intact. The relevance score should be added to the documents' existing metadata. No additional id is necessary to avoid clashing with existing ids.
### System Info
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.31
langchain-openai==0.0.8
langchain-text-splitters==0.0.1 | `FlashrankRerank` drops document metadata | https://api.github.com/repos/langchain-ai/langchain/issues/19142/comments | 0 | 2024-03-15T14:56:27Z | 2024-03-19T10:43:39Z | https://github.com/langchain-ai/langchain/issues/19142 | 2,188,745,224 | 19,142 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain.retrievers.document_compressors import FlashrankRerank
from langchain.retrievers import ContextualCompressionRetriever
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
The import path in the example notebook for the `FlashrankReranker` is wrong. It is listed as `from langchain.retrievers ...` but should be `from langchain.retrievers.document_compressors ...`.
### System Info
```
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.31
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
``` | Wrong import path in `docs/integrations/retrievers/flashrank-reranker.ipynb` | https://api.github.com/repos/langchain-ai/langchain/issues/19139/comments | 0 | 2024-03-15T14:25:20Z | 2024-06-21T16:37:05Z | https://github.com/langchain-ai/langchain/issues/19139 | 2,188,642,452 | 19,139 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
def get_message_history(collection_id: str) -> SQLChatMessageHistoryExtended:
return SQLChatMessageHistoryExtended(
session_id=collection_id,
connection=db_conn,
custom_message_converter=CustomMessageConverter(),
k_latest_interactions=self.k_memory_interactions
)
def format_docs(docs):
return "\n\n".join(doc.page_content for doc in docs)
db = PGVectorExtended.from_existing_index(
connection=db_conn,
embedding=self.embeddings,
collection_name=str(collection_id),
distance_strategy="cosine",
)
retriever = db.as_retriever(
search_type="similarity", search_kwargs={"k": self.k_similar_chunks}
)
chat_template = ChatPromptTemplate.from_messages(
[
("system", ("foo."),),
MessagesPlaceholder(variable_name="history"),
("human",("{question} \n\n ...foo... {context} \n\n")),
]
)
llm_with_tool = self.llm.bind_tools(
[quoted_answer],
tool_choice="quoted_answer",
)
output_parser = JsonOutputKeyToolsParser(key_name="quoted_answer", return_single=True)
chain = (
{
"context": itemgetter("question") | retriever | format_docs,
"question": itemgetter("question"),
"history": itemgetter("history"),
}
| chat_template
| llm_with_tool
| output_parser
# with these the history works
# | self.llm
# | StrOutputParser()
)
chain_with_message_history = RunnableWithMessageHistory(
chain,
get_message_history,
input_messages_key="question",
history_messages_key="history",
output_messages_key="quoted_answer",
)
response = chain_with_message_history.invoke(
{"question": question},
config={"configurable": {"session_id": str(collection_id)}}
)
# I have to add the history manually when using `llm_with_tool` and `output_parser`
# with `self.llm` and `StrOutputParser()` it is added automatically
get_message_history(str(collection_id)).add_ai_message(response['answer'])
get_message_history(str(collection_id)).add_user_message(question)
return response['answer'], response['citations']
```
### Error Message and Stack Trace (if applicable)
Attached is output from the chain
[chain_logs.txt](https://github.com/langchain-ai/langchain/files/14615664/chain_logs.txt)
### Description
## My goal
* I have a RAG chat which answers questions based on a document.
* I want to have history (which is stored in postgres) so when the users says "translate the previous answer to french" it will do so, for example.
* I also want to return citations for the user to see where the answer comes from.
* _(I have also [posted this on the Q&A](https://github.com/langchain-ai/langchain/discussions/19118) but based on the bot's output, seems more like a bug)_
## What's wrong
### 1) History is not inserted
- if I do not include the code below, nothing is inserted in the message store
```python
get_message_history(str(collection_id)).add_ai_message(response['answer'])
get_message_history(str(collection_id)).add_user_message(question)
```
- note that if I do not use the chain with OpenAI tools for citations, _the history is inserted correctly_
```python
chain = (
{
"context": itemgetter("question") | retriever | format_docs,
"question": itemgetter("question"),
"history": itemgetter("history"),
}
| chat_template
# with these the history works
| self.llm
| StrOutputParser()
)
```
### 2) History is ignored
- if I add the messages manually to the history, the chain cannot answer basic questions, like "translate the previous answer to french" and instead returns the previous answer
```
+---------------------------------------------------------------------------------+------+
|message |author|
+---------------------------------------------------------------------------------+------+
|The document is a draft report overview of the Civil UAV Capability ......... |ai |
|Summarise the document and write two main takeaways |human |
|Based on the definition provided in the 'Lexicon of UAV/ROA Terminology', .... |ai |
|Can it also carry animals? |human |
|Based on the provided context, UAVs are not intended to carry animals. |ai |
|Can it fly to space? |human |
|Based on the provided context, UAVs are not intended to carry animals. |ai |
|translate the previous answer to german |human |
+---------------------------------------------------------------------------------+------+
```
- Resolution is the same as above ☝️ : this works if the chain with OpenAI tools for citations is not used
## Solution
- Based on the logs from the chain, I suspect that the key `quoted_answer` is somehow not handled well
- I also see this suspicious warning in the logs
```
Error in RootListenersTracer.on_chain_end callback: KeyError('quoted_answer')
```
### System Info
```
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 23.4.0: Wed Feb 21 21:44:31 PST 2024; root:xnu-10063.101.15~2/RELEASE_X86_64
> Python Version: 3.11.1 (main, Aug 9 2023, 13:06:45) [Clang 14.0.3 (clang-1403.0.22.14.1)]
Package Information
-------------------
> langchain_core: 0.1.32
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.26
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
``` | When using citations from OpenAI's tools with `RunnableWithMessageHistory`, history is ignored and not inserted | https://api.github.com/repos/langchain-ai/langchain/issues/19136/comments | 1 | 2024-03-15T13:02:27Z | 2024-07-16T16:06:33Z | https://github.com/langchain-ai/langchain/issues/19136 | 2,188,468,815 | 19,136 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
There is a minor mistake in neo4j documentation. In [Seeding the database](https://python.langchain.com/docs/use_cases/graph/integrations/graph_cypher_qa#seeding-the-database), movie titles are inserted as 'name', and then later in [Add examples in the Cypher generation prompt](https://python.langchain.com/docs/use_cases/graph/integrations/graph_cypher_qa#add-examples-in-the-cypher-generation-prompt) section, example has 'title' instead of 'name': <code># How many people played in Top Gun? MATCH (m:Movie {{title:"Top Gun"}})<-[:ACTED_IN]-() RETURN count(*) AS numberOfActors</code>. I have also done cypher querying with LLaMA 2, and if you are interested, I can provide you code for template...
### Idea or request for content:
Change <code># How many people played in Top Gun? MATCH (m:Movie {{title:"Top Gun"}})<-[:ACTED_IN]-() RETURN count(*) AS numberOfActors</code> to <code># How many people played in Top Gun?
MATCH (m:Movie {{name:"Top Gun"}})<-[:ACTED_IN]-()
RETURN count(*) AS numberOfActors</code>. | Neo4j documentation mistake | https://api.github.com/repos/langchain-ai/langchain/issues/19134/comments | 1 | 2024-03-15T12:30:26Z | 2024-05-22T20:48:10Z | https://github.com/langchain-ai/langchain/issues/19134 | 2,188,411,236 | 19,134 |
[
"langchain-ai",
"langchain"
] | ### 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
1. Fetched documents similar to my query using below code using
> staff_knowledge_base.similarity_search(user_question, k=10)
```
staff_knowledge_base = PGVector(
embedding_function=embeddings,
connection_string=conn,
collection_name=collection)
```
**NOTE:** here default distance_strategy is used, i.e., COSINE (as per documentation)
2. Fetched documents similar to my query using below code using
> staff_knowledge_base.similarity_search(user_question, k=10)
```
from langchain_community.vectorstores.pgvector import DistanceStrategy
staff_knowledge_base = PGVector(
embedding_function=embeddings,
connection_string=conn,
collection_name=collection,
distance_strategy= DistanceStrategy.MAX_INNER_PRODUCT
)
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
**Observation:** There is no change in list of fetched documents
**Expected Result:** Documents fetched should have been changed with change in distance_strategy
### System Info
System Information
------------------
> OS: Windows
> OS Version: 10.0.22621
> Python Version: 3.10.9 (tags/v3.10.9:1dd9be6, Dec 6 2022, 20:01:21) [MSC v.1934 64 bit (AMD64)]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.25
> langchain_openai: 0.0.6
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | PGVector distance_strategy methods seems to be not working | https://api.github.com/repos/langchain-ai/langchain/issues/19129/comments | 1 | 2024-03-15T11:38:41Z | 2024-06-24T16:13:29Z | https://github.com/langchain-ai/langchain/issues/19129 | 2,188,319,485 | 19,129 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
1) Can you confirm which language model is being used in **create_sql_query_chain** method for text to sql conversion? How can I **configure it**, if I want to try with some other language model?
https://python.langchain.com/docs/use_cases/sql/quickstart#convert-question-to-sql-query
2) Can you confirm which language model is being used in **create_sql_agent** method for text to sql conversion? How can I **configure it**, if I want to try with some other language model?
https://python.langchain.com/docs/use_cases/sql/agents#agent
### Idea or request for content:
_No response_ | DOC: [Question]: How to change the text to sql model in create_sql_query_chain , create_sql_agent ? | https://api.github.com/repos/langchain-ai/langchain/issues/19124/comments | 1 | 2024-03-15T09:37:14Z | 2024-07-05T16:06:13Z | https://github.com/langchain-ai/langchain/issues/19124 | 2,188,098,957 | 19,124 |
[
"langchain-ai",
"langchain"
] | ### 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
with get_openai_callback() as cb:
output_parser = StrOutputParser()
llm_chain = prompt_main | llm | output_parser
all_text = str(prompt) + str(topics)
threshold = (llm.get_num_tokens(text=all_text) + tokens)
chatgpt_output = llm_chain.invoke({"prompt": prompt, "topics": topics})
chatgpt_output = chatgpt_output.replace("```", "").strip()
# Parse the string as a dictionary
data_dict = parse_json_output(chatgpt_output)
# Extract the categories list
subcat_list = data_dict.get('Sub-Categories', [])
total_cost = round(cb.total_cost, 5)
total_tokens = cb.total_tokens
print(f"Topics Tokens: {llm.get_num_tokens(text=''.join(topics))}")
print(f"Sub-Categories Tokens: {llm.get_num_tokens(text=''.join(subcat_list))}")
print(f"Estimated Total Tokens: {threshold}")
print(f"Total Tokens: {cb.total_tokens}")
print(f"Prompt Tokens: {cb.prompt_tokens}")
print(f"Completion Tokens: {cb.completion_tokens}")
print(f"Total Cost (USD): ${cb.total_cost}")
### Error Message and Stack Trace (if applicable)
_No response_
### Description
I've successfully calculated the Input Tokens as well as individual tokens for a list called Topics (contains 4000 topics).
According to the OpenAI Tokenizer Web Tool and my calculated tokens using llm.get_num_tokens the Tokens for the Topics should be at least ~8K meanwhile all the Total tokens (including Input and Output) should be around 13K but LangChain shows it simply as 815 Tokens.
Also the Cost should be different too but it shows Total Cost to be 0.00134 USD. I'm using gpt-3.5-turbo. Please look at the code and image attached and you'd understand something's wrong here.
### System Info
System Information
------------------
> OS: Linux
> OS Version: #1 SMP Tue Jan 30 20:59:52 UTC 2024
> Python Version: 3.10.13 | packaged by conda-forge | (main, Oct 26 2023, 18:07:37) [GCC 12.3.0]
Package Information
-------------------
> langchain_core: 0.1.32
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.26
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | OpenAI Tokens and Cost are inaccurate | https://api.github.com/repos/langchain-ai/langchain/issues/19120/comments | 1 | 2024-03-15T08:45:37Z | 2024-07-04T16:08:18Z | https://github.com/langchain-ai/langchain/issues/19120 | 2,188,002,825 | 19,120 |
[
"langchain-ai",
"langchain"
] | ### 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
Minimum working example:
```python
from langchain_openai import AzureChatOpenAI
import httpx
http_client = httpx.Client()
model = AzureChatOpenAI(
http_client=http_client,
api_key="foo",
api_version="2023-07-01-preview",
azure_endpoint="https://example.com",
)
```
### Error Message and Stack Trace (if applicable)
```
Traceback (most recent call last):
File "/home/ec2-user/environment/test/test-openai.py", line 7, in <module>
model = AzureChatOpenAI(http_client=http_client, api_key="foo", api_version="2023-07-01-preview", azure_endpoint="https://example.com")
File "/home/ec2-user/.local/share/virtualenvs/test/lib/python3.10/site-packages/langchain_core/load/serializable.py", line 120, in __init__
super().__init__(**kwargs)
File "/home/ec2-user/.local/share/virtualenvs/test/lib/python3.10/site-packages/pydantic/v1/main.py", line 341, in __init__
raise validation_error
pydantic.v1.error_wrappers.ValidationError: 1 validation error for AzureChatOpenAI
__root__
Invalid `http_client` argument; Expected an instance of `httpx.AsyncClient` but got <class 'httpx.Client'> (type=type_error)
```
### Description
Due to recent changes in openai-python the way langchain passes custom instances of `http_client` does not work anymore. The example code shows this behavior for `AzureChatOpenAI` but it holds for all instances where OpenAI clients are used.
See [here](https://github.com/langchain-ai/langchain/blob/9e569d85a45fd9e89f85f5c93e61940e36176076/libs/partners/openai/langchain_openai/chat_models/azure.py#L190-L191) for one example of how the custom `http_client` gets passed to openai-python. LangChain would need to ensure to properly pass an instance of `httpx.AsyncClient` or `httpx.Client`.
The reason for the exception are some new type checks in openai-python v1.13.4 (see [here](https://github.com/openai/openai-python/compare/v1.13.3...v1.13.4#diff-aca4f4354075e3c75151a8de08daeb25d4db0af2564381c26aba33a49c9dc829R783-R1334)).
PS: As a temporary solution, I pinned openai-python to an older version. It is not pinned in langchain_openai (see [here](https://github.com/langchain-ai/langchain/blob/9e569d85a45fd9e89f85f5c93e61940e36176076/libs/partners/openai/pyproject.toml#L16)).
### System Info
```
System Information
------------------
> OS: Linux
> OS Version: #1 SMP Sat Feb 24 09:50:35 UTC 2024
> Python Version: 3.10.12 (main, Nov 10 2023, 12:43:56) [GCC 7.3.1 20180712 (Red Hat 7.3.1-17)]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.11
> langchain_community: 0.0.28
> langsmith: 0.1.25
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
``` | openai-python 1.13.4 with custom http_client breaks OpenAI clients in langchain | https://api.github.com/repos/langchain-ai/langchain/issues/19116/comments | 6 | 2024-03-15T07:59:15Z | 2024-05-22T13:55:39Z | https://github.com/langchain-ai/langchain/issues/19116 | 2,187,931,701 | 19,116 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_openai import ChatOpenAI
from langchain.output_parsers import OutputFixingParser, XMLOutputParser
llm = ChatOpenAI()
parser = XMLOutputParser()
output_fixing_parser = OutputFixingParser.from_llm(llm=llm, parser=parser)
# OK: nicely parsed
llm_output = "<movies>\n <actor>Tom Hanks</actor></movies>"
parsed_output = output_fixing_parser.parse(llm_output)
# ERROR 1: mismatched brace
llm_output = "<moviesss>\n <actor>Tom Hanks</actor></movies>"
parsed_output = output_fixing_parser.parse(llm_output)
# xml.etree.ElementTree.ParseError: mismatched tag: line 2, column 30
# ERROR 2: unexpected string
llm_output = "movie actor: Tom Hanks"
parsed_output = output_fixing_parser.parse(llm_output)
# ValueError: Could not parse output: movie actor: Tom Hanks
```
### Error Message and Stack Trace (if applicable)
```python
# ERROR 1: mismatched tag
llm_output = "<moviesss>\n <actor>Tom Hanks</actor></movies>"
parsed_output = output_fixing_parser.parse(llm_output)
# xml.etree.ElementTree.ParseError: mismatched tag: line 2, column 30
```
```python
# ERROR 2: unexpected string
llm_output = "movie actor: Tom Hanks"
parsed_output = output_fixing_parser.parse(llm_output)
# ValueError: Could not parse output: movie actor: Tom Hanks
```
### Description
I'm trying to use the `OutputFixingParser` to wrap `XMLOutputParser`.
But, `OutputFixingParser` couldn't handle some kinds of llm-generated output.
Instead, it raises `xml.etree.ElementTree.ParseError` and `ValueError`.
### System Info
```
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.32
``` | `OutputFixingParser` raises unhandled exception while wrapping `XMLOutputParser` | https://api.github.com/repos/langchain-ai/langchain/issues/19107/comments | 0 | 2024-03-15T05:46:44Z | 2024-03-19T04:49:05Z | https://github.com/langchain-ai/langchain/issues/19107 | 2,187,771,203 | 19,107 |
[
"langchain-ai",
"langchain"
] | ### 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
```py
user_input = "你好"
embedding_model_name = fr"/usr/local/models/BAAI/bge-m3"
hf = HuggingFaceBgeEmbeddings(
model_name=embedding_model_name,
model_kwargs={"device": "cpu"},
encode_kwargs={
"normalize_embeddings": True,
"batch_size": 32
}
)
milvus_store = Milvus(
embedding_function=hf,
collection_name="qa",
drop_old=False,
text_field="content",
primary_field= "id",
vector_field = "vector",
# search_params={
# "metric_type": "IP",
# "index_type": "IVF_FLAT",
# "params": {"nprobe": 10, "nlist": 128}
# },
connection_args={
"host": "10.3.1.187",
"port": "19530",
"user": "",
"password": "",
"db_name": "qa"
},
)
retriever=milvus_store.as_retriever(
search_type="similarity_score_threshold",
search_kwargs={"score_threshold": 0.8}
)
docs= retriever.get_relevant_documents(query=user_input)
```
### Error Message and Stack Trace (if applicable)
```
Traceback (most recent call last):
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/retrievers.py", line 244, in get_relevant_documents
raise e
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/retrievers.py", line 237, in get_relevant_documents
result = self._get_relevant_documents(
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/vectorstores.py", line 677, in _get_relevant_documents
self.vectorstore.similarity_search_with_relevance_scores(
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/vectorstores.py", line 324, in similarity_search_with_relevance_scores
docs_and_similarities = self._similarity_search_with_relevance_scores(
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/vectorstores.py", line 271, in _similarity_search_with_relevance_scores
relevance_score_fn = self._select_relevance_score_fn()
File "/usr/local/churchill/lib/python3.10/site-packages/langchain_core/vectorstores.py", line 228, in _select_relevance_score_fn
raise NotImplementedError
NotImplementedError
```
### Description
I want to get `get_relevant_documents` using retriever ,but it has bug.
### System Info
langchain 0.1.9
python version 3.10
centos 8 | milvus retriever.get_relevant_documents has bug? | https://api.github.com/repos/langchain-ai/langchain/issues/19106/comments | 5 | 2024-03-15T05:38:46Z | 2024-07-11T06:57:54Z | https://github.com/langchain-ai/langchain/issues/19106 | 2,187,763,655 | 19,106 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
model_name = "sentence-transformers/all-mpnet-base-v2"
model_kwargs = {'device': 'cpu'}
encode_kwargs = {'normalize_embeddings': False}
embeddings = HuggingFaceEmbeddings(
model_name=model_name,
model_kwargs=model_kwargs,
encode_kwargs=encode_kwargs
)
```
### Error Message and Stack Trace (if applicable)
File "/app/llm/lfqa_langchain.py", line 29, in <module>
2024-03-14 18:45:10 embeddings = HuggingFaceEmbeddings(
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/langchain_community/embeddings/huggingface.py", line 59, in __init__
2024-03-14 18:45:10 import sentence_transformers
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/sentence_transformers/__init__.py", line 3, in <module>
2024-03-14 18:45:10 from .datasets import SentencesDataset, ParallelSentencesDataset
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/sentence_transformers/datasets/__init__.py", line 1, in <module>
2024-03-14 18:45:10 from .DenoisingAutoEncoderDataset import DenoisingAutoEncoderDataset
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/sentence_transformers/datasets/DenoisingAutoEncoderDataset.py", line 1, in <module>
2024-03-14 18:45:10 from torch.utils.data import Dataset
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/torch/__init__.py", line 236, in <module>
2024-03-14 18:45:10 _load_global_deps()
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/torch/__init__.py", line 197, in _load_global_deps
2024-03-14 18:45:10 _preload_cuda_deps(lib_folder, lib_name)
2024-03-14 18:45:10 File "/.venv/lib/python3.9/site-packages/torch/__init__.py", line 162, in _preload_cuda_deps
2024-03-14 18:45:10 raise ValueError(f"{lib_name} not found in the system path {sys.path}")
2024-03-14 18:45:10 ValueError: libcublas.so.*[0-9] not found in the system path ['', '/.venv/bin', '/usr/local/lib/python39.zip', '/usr/local/lib/python3.9', '/usr/local/lib/python3.9/lib-dynload', '/.venv/lib/python3.9/site-packages']
### Description
After generating Pipfile.lock and docker image, it shuts down automatically after 2-3 seconds with the given error. Using langchain = "==0.0.351".
### System Info
Platform linux (Docker)
Pip version: 23.0.1 | Docker ValueError: libcublas.so.*[0-9] not found in the system path | https://api.github.com/repos/langchain-ai/langchain/issues/19078/comments | 1 | 2024-03-14T15:50:12Z | 2024-03-14T18:30:48Z | https://github.com/langchain-ai/langchain/issues/19078 | 2,186,711,784 | 19,078 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
AI_URL = env["AI_URL"] or "http://localhost:11434"
OPENSEARCH_URL = env["OPENSEARCH_URL"] or "https://0.0.0.0:9200"
OPENSEARCH_USERNAME = env["OPENSEARCH_USERNAME"] or "admin"
OPENSEARCH_PASSWORD = env["OPENSEARCH_PASSWORD"] or "Admin123_"
OPENSEARCH_INDEX_NAME = env["OPENSEARCH_INDEX_NAME"] or "index-name"
embedding_function = OllamaEmbeddings(model="mistral", base_url=AI_URL)
os_client = OpenSearch(
hosts=[OPENSEARCH_URL],
http_auth=(OPENSEARCH_USERNAME, OPENSEARCH_PASSWORD),
use_ssl=False,
verify_certs=False,
)
try:
res = embedding_function.embed_query(thematic)
text = {
"thematic": thematic,
"extensions": thematics[thematic],
}
document = {
"vector_field": res,
"text": str(text),
"thematic": thematic,
"extensions": thematics[thematic],
}
os_client.index(index=OPENSEARCH_INDEX_NAME, body=document, refresh=True)
except Exception as e:
logger.error("Error pushing data", exc_info=e)
return JSONResponse({"message": "Error pushing data"}, status_code=500)
```
Here is the docker compose that I use:
```yml
version: '3'
services:
server:
container_name: server
build:
context: .
dockerfile: Dockerfile
ports:
- 8080:8080
env_file:
- path: .env
required: true
server-opensearch:
container_name: server-opensearch
image: opensearchproject/opensearch:2.12.0
env_file:
- path: .env
required: true
environment:
- discovery.type=single-node
- plugins.security.disabled=true
ulimits:
memlock:
soft: -1
hard: -1
ports:
- 9200:9200
- 9600:9600
```
The Dockerfile just package my app into a tar file and launch it with uvicorn like so:
`CMD ["uvicorn", "app.__init__:app", "--host", "0.0.0.0", "--port", "8080"]`
### Error Message and Stack Trace (if applicable)
(I'm sorry for the formatting of the stacktrace, that's the fault of the logger lib...)
Traceback (most recent call last):\n File \"/usr/lib/python3.11/site-packages/urllib3/connection.py\", line 198, in _new_conn\n sock = connection.create_connection(\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/urllib3/util/connection.py\", line 85, in create_connection\n raise err\n File \"/usr/lib/python3.11/site-packages/urllib3/util/connection.py\", line 73, in create_connection\n sock.connect(sa)\nConnectionRefusedError: [Errno 111] Connection refused\n\nThe above exception was the direct cause of the following exception:\n\nTraceback (most recent call last):\n File \"/usr/lib/python3.11/site-packages/urllib3/connectionpool.py\", line 793, in urlopen\n response = self._make_request(\n ^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/urllib3/connectionpool.py\", line 496, in _make_request\n conn.request(\n File \"/usr/lib/python3.11/site-packages/urllib3/connection.py\", line 400, in request\n self.endheaders()\n File \"/usr/lib/python3.11/http/client.py\", line 1293, in endheaders\n self._send_output(message_body, encode_chunked=encode_chunked)\n File \"/usr/lib/python3.11/http/client.py\", line 1052, in _send_output\n self.send(msg)\n File \"/usr/lib/python3.11/http/client.py\", line 990, in send\n self.connect()\n File \"/usr/lib/python3.11/site-packages/urllib3/connection.py\", line 238, in connect\n self.sock = self._new_conn()\n ^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/urllib3/connection.py\", line 213, in _new_conn\n raise NewConnectionError(\nurllib3.exceptions.NewConnectionError: <urllib3.connection.HTTPConnection object at 0xffff80db6590>: Failed to establish a new connection: [Errno 111] Connection refused\n\nThe above exception was the direct cause of the following exception:\n\nTraceback (most recent call last):\n File \"/usr/lib/python3.11/site-packages/requests/adapters.py\", line 486, in send\n resp = conn.urlopen(\n ^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/urllib3/connectionpool.py\", line 847, in urlopen\n retries = retries.increment(\n ^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/urllib3/util/retry.py\", line 515, in increment\n raise MaxRetryError(_pool, url, reason) from reason # type: ignore[arg-type]\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\nurllib3.exceptions.MaxRetryError: HTTPConnectionPool(host='0.0.0.0', port=11434): Max retries exceeded with url: /api/embeddings (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0xffff80db6590>: Failed to establish a new connection: [Errno 111] Connection refused'))\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n File \"/usr/lib/python3.11/site-packages/langchain_community/embeddings/ollama.py\", line 157, in _process_emb_response\n res = requests.post(\n ^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/requests/api.py\", line 115, in post\n return request(\"post\", url, data=data, json=json, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/requests/api.py\", line 59, in request\n return session.request(method=method, url=url, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/requests/sessions.py\", line 589, in request\n resp = self.send(prep, **send_kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/requests/sessions.py\", line 703, in send\n r = adapter.send(request, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/requests/adapters.py\", line 519, in send\n raise ConnectionError(e, request=request)\nrequests.exceptions.ConnectionError: HTTPConnectionPool(host='0.0.0.0', port=11434): Max retries exceeded with url: /api/embeddings (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0xffff80db6590>: Failed to establish a new connection: [Errno 111] Connection refused'))\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n File \"/opt/app/main/rag.py\", line 64, in recreate\n res = embedding_function.embed_query(thematic)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/langchain_community/embeddings/ollama.py\", line 217, in embed_query\n embedding = self._embed([instruction_pair])[0]\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/langchain_community/embeddings/ollama.py\", line 192, in _embed\n return [self._process_emb_response(prompt) for prompt in iter_]\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/langchain_community/embeddings/ollama.py\", line 192, in <listcomp>\n return [self._process_emb_response(prompt) for prompt in iter_]\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/usr/lib/python3.11/site-packages/langchain_community/embeddings/ollama.py\", line 163, in _process_emb_response\n raise ValueError(f\"Error raised by inference endpoint: {e}\")\nValueError: Error raised by inference endpoint: HTTPConnectionPool(host='0.0.0.0', port=11434): Max retries exceeded with url: /api/embeddings (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0xffff80db6590>: Failed to establish a new connection: [Errno 111] Connection refused'))
### Description
I'm trying to use the OllamaEmbeddings class to generate embeddings when my server is hit on a certain endpoint. When running locally, there are no problems, but if I'm running the server inside a docker container, I get this error. The ollama server is running on my machine locally, not inside a container.
### System Info
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 23.3.0: Wed Dec 20 21:30:27 PST 2023; root:xnu-10002.81.5~7/RELEASE_ARM64_T8103
> Python Version: 3.12.2 (main, Feb 6 2024, 20:19:44) [Clang 15.0.0 (clang-1500.1.0.2.5)]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.11
> langchain_community: 0.0.27
> langsmith: 0.1.24
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | "Max retries exceeded with url: /api/embeddings" when using OllamaEmbeddings inside a container | https://api.github.com/repos/langchain-ai/langchain/issues/19074/comments | 1 | 2024-03-14T13:31:19Z | 2024-03-14T18:34:04Z | https://github.com/langchain-ai/langchain/issues/19074 | 2,186,395,518 | 19,074 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
def generate_prompt(system, human_question):
return ChatPromptTemplate.from_messages(
[
("system", system),
MessagesPlaceholder(variable_name="chat_history"),
("human", human_question),
]
)
if __name__ == '__main__':
system="A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions."
human_question="""
{'EARLY_PHASE1': [],
'PHASE1': [{'nctId': 'NCT06277609', 'briefTitle': 'A Trial Investiga', 'officialTitle': 'Interventional,', 'url': '', 'phases': 'PHASE1', 'studyStatus': 'RECRUITING', 'conditions': ['Healthy Participants'],'NA': []}]
}
"""
print(generate_prompt(system, human_question))
```
### Error Message and Stack Trace (if applicable)
The input_variables variable should be chat_history instead of containing 'EARLY-PHASE1'
### Description
no
### System Info
langchain 0.1.11
langchain-community 0.0.27
langchain-core 0.1.30
langchain-openai 0.0.8
langchain-text-splitters 0.0.1
langchainhub 0.1.15
langsmith 0.1.23
| Text with {will be parsed as input_variables, causing an error | https://api.github.com/repos/langchain-ai/langchain/issues/19067/comments | 1 | 2024-03-14T08:14:39Z | 2024-03-14T18:14:28Z | https://github.com/langchain-ai/langchain/issues/19067 | 2,185,712,911 | 19,067 |
[
"langchain-ai",
"langchain"
] | ### 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
The patches are listed below:
```
diff -x *.pyc -r ~/venv/lib/python3.9/site-packages/langchain_openai/llms/base.py langchain_openai/llms/base.py
207c207,210
< values["client"] = openai.OpenAI(**client_params).completions
---
> client = openai.OpenAI(**client_params)
> completion = client.chat.completions
> values["client"] = completion
> #values["client"] = openai.OpenAI(**client_params).completions
340c343,352
< response = self.client.create(prompt=_prompts, **params)
---
> from openai.types.chat.chat_completion_user_message_param import ChatCompletionUserMessageParam
>
> umessage = ChatCompletionUserMessageParam(content=_prompts[0], role = "user")
> messages=[umessage,]
> response = self.client.create(messages=messages, **params)
> #response = self.client.create(prompt=_prompts, **params)
454c466,467
< text=choice["text"],
---
> #text=choice["text"],
> text=choice["message"]["content"],
```
### Error Message and Stack Trace (if applicable)
```Entering new SQLDatabaseChain chain...
Given an input question, first create a syntactically correct postgresql query to run, then look at the results of the query and return the answer.
Use the following format:
Question: Question here
SQLQuery: SQL Query to run
SQLResult: Result of the SQLQuery
Answer: Final answer here
how many tasks?
SQLQuery:SELECT COUNT(*) FROM tasks
SQLResult:
count
-----
3
Answer: There are 3 tasks in the tasks table.(psycopg2.errors.SyntaxError) syntax error at or near ":"
LINE 2: SQLResult:
^
[SQL: SELECT COUNT(*) FROM tasks
SQLResult:
count
-----
3
Answer: There are 3 tasks in the tasks table.]
(Background on this error at: https://sqlalche.me/e/20/f405)
Traceback (most recent call last):
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1969, in _exec_single_context
self.dialect.do_execute(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/default.py", line 922, in do_execute
cursor.execute(statement, parameters)
psycopg2.errors.SyntaxError: syntax error at or near ":"
LINE 2: SQLResult:
^
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/Users/marvel/Nutstore Files/programs/ai/xiaoyi-robot/codes/postgres_agent.py", line 61, in get_prompt
print(db_chain.run(question))
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_core/_api/deprecation.py", line 145, in warning_emitting_wrapper
return wrapped(*args, **kwargs)
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain/chains/base.py", line 538, in run
return self(args[0], callbacks=callbacks, tags=tags, metadata=metadata)[
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_core/_api/deprecation.py", line 145, in warning_emitting_wrapper
return wrapped(*args, **kwargs)
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain/chains/base.py", line 363, in __call__
return self.invoke(
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain/chains/base.py", line 162, in invoke
raise e
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain/chains/base.py", line 156, in invoke
self._call(inputs, run_manager=run_manager)
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_experimental/sql/base.py", line 201, in _call
raise exc
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_experimental/sql/base.py", line 146, in _call
result = self.database.run(sql_cmd)
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_community/utilities/sql_database.py", line 436, in run
result = self._execute(command, fetch)
File "/Users/marvel/venv/lib/python3.9/site-packages/langchain_community/utilities/sql_database.py", line 413, in _execute
cursor = connection.execute(text(command))
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1416, in execute
return meth(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/sql/elements.py", line 517, in _execute_on_connection
return connection._execute_clauseelement(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1639, in _execute_clauseelement
ret = self._execute_context(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1848, in _execute_context
return self._exec_single_context(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1988, in _exec_single_context
self._handle_dbapi_exception(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 2344, in _handle_dbapi_exception
raise sqlalchemy_exception.with_traceback(exc_info[2]) from e
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/base.py", line 1969, in _exec_single_context
self.dialect.do_execute(
File "/Users/marvel/venv/lib/python3.9/site-packages/sqlalchemy/engine/default.py", line 922, in do_execute
cursor.execute(statement, parameters)
sqlalchemy.exc.ProgrammingError: (psycopg2.errors.SyntaxError) syntax error at or near ":"
LINE 2: SQLResult:
^
[SQL: SELECT COUNT(*) FROM tasks
SQLResult:
count
-----
3
Answer: There are 3 tasks in the tasks table.]
```
<img width="1563" alt="error" src="https://github.com/langchain-ai/langchain/assets/905594/3c9ff8a4-d1b0-482f-ac46-70bd15a240b4">
### Description
I followed the article (https://coinsbench.com/chat-with-your-databases-using-langchain-bb7d31ed2e76) to make langchain interact with postgres DB, but I have got error response from OpenAI recently, error message are listed below.
<img width="1488" alt="chat-comp" src="https://github.com/langchain-ai/langchain/assets/905594/7616ecf9-1ea0-4726-910e-8ace52825fae">
<img width="1559" alt="comp" src="https://github.com/langchain-ai/langchain/assets/905594/ab5386d8-8c5c-431a-abde-11427630f63a">
It seems that OpenAI has stopped supporting Completions API and stop words in the request fail to work. OpenAI's suggestions are(https://openai.com/blog/gpt-4-api-general-availability):
> Starting today, all paying API customers have access to GPT-4. In March, we [introduced the ChatGPT API](https://openai.com/blog/introducing-chatgpt-and-whisper-apis), and earlier this month we [released our first updates](https://openai.com/blog/function-calling-and-other-api-updates) to the chat-based models. We envision a future where chat-based models can support any use case. Today we’re announcing a deprecation plan for older models of the Completions API, and recommend that users adopt the Chat Completions API.
So I did some research of current codes and decided to modify langchain_openai/llms/base.py, the modifications are list above. finally it works
<img width="849" alt="finally" src="https://github.com/langchain-ai/langchain/assets/905594/bd0e33be-6921-4ac0-b264-322f928ed3c7">
### System Info
% pip list|grep langchain
langchain 0.1.5
langchain-community 0.0.17
langchain-core 0.1.18
langchain-experimental 0.0.50
langchain-openai 0.0.5
% pip list|grep openai
langchain-openai 0.0.5
openai 1.11.1 | OpenAI's legacy completion API causes that stop words do not work, patches are attached. | https://api.github.com/repos/langchain-ai/langchain/issues/19062/comments | 1 | 2024-03-14T06:45:32Z | 2024-03-14T18:37:47Z | https://github.com/langchain-ai/langchain/issues/19062 | 2,185,552,198 | 19,062 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
import os
from langchain.vectorstores.qdrant import Qdrant
from langchain.document_loaders.pdf import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings.huggingface import HuggingFaceBgeEmbeddings
def load_pdf(
file: str,
collection_name: str,
chunk_size: int = 512,
chunk_overlap: int = 32,
):
loader = PyPDFLoader(file)
documents = loader.load()
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=chunk_size, chunk_overlap=chunk_overlap
)
texts = text_splitter.split_documents(documents)
embeddings = HuggingFaceBgeEmbeddings(
model_name=os.environ.get("MODEL_NAME", "microsoft/phi-2"),
model_kwargs=dict(device="cpu"),
encode_kwargs=dict(normalize_embeddings=False),
)
url = os.environ.get("VDB_URL", "http://localhost:6333")
qdrant = Qdrant.from_documents(
texts,
embeddings,
url=url,
collection_name=collection_name,
prefer_grpc=False,
)
return qdrant
load_pdf("/Users/alvynabranches/Downloads/bh1.pdf", "buget2425")
```
### Error Message and Stack Trace (if applicable)
```bash
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Traceback (most recent call last):
File "/Users/alvynabranches/rag/rag/ingest.py", line 39, in <module>
load_pdf("/Users/alvynabranches/Downloads/bh1.pdf", "buget2425")
File "/Users/alvynabranches/rag/rag/ingest.py", line 29, in load_pdf
qdrant = Qdrant.from_documents(
^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/langchain_core/vectorstores.py", line 528, in from_documents
return cls.from_texts(texts, embedding, metadatas=metadatas, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/langchain_community/vectorstores/qdrant.py", line 1334, in from_texts
qdrant = cls.construct_instance(
^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/langchain_community/vectorstores/qdrant.py", line 1591, in construct_instance
partial_embeddings = embedding.embed_documents(texts[:1])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/langchain_community/embeddings/huggingface.py", line 257, in embed_documents
embeddings = self.client.encode(texts, **self.encode_kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/sentence_transformers/SentenceTransformer.py", line 345, in encode
features = self.tokenize(sentences_batch)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/sentence_transformers/SentenceTransformer.py", line 553, in tokenize
return self._first_module().tokenize(texts)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/sentence_transformers/models/Transformer.py", line 146, in tokenize
self.tokenizer(
File "/opt/homebrew/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 2829, in __call__
encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 2915, in _call_one
return self.batch_encode_plus(
^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 3097, in batch_encode_plus
padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 2734, in _get_padding_truncation_strategies
raise ValueError(
ValueError: Asking to pad but the tokenizer does not have a padding token. Please select a token to use as `pad_token` `(tokenizer.pad_token = tokenizer.eos_token e.g.)` or add a new pad token via `tokenizer.add_special_tokens({'pad_token': '[PAD]'})`.
```
### Description
I am using the langchain library to store pdf on qdrant. The model which I am using is phi-2. I tried to solve it by changing the kwargs but it is still not working.
### System Info
```bash
pip3 freeze | grep langchain
```
```
langchain==0.1.12
langchain-community==0.0.28
langchain-core==0.1.31
langchain-text-splitters==0.0.1
```
#### Platform
Apple M3 Max
macOS Sonoma
Version 14.1
```bash
python3 -V
```
```
Python 3.12.2
``` | Asking to pad but the tokenizer does not have a padding token. Please select a token to use as `pad_token` `(tokenizer.pad_token = tokenizer.eos_token e.g.)` or add a new pad token via `tokenizer.add_special_tokens({'pad_token': '[PAD]'})` | https://api.github.com/repos/langchain-ai/langchain/issues/19061/comments | 0 | 2024-03-14T06:39:58Z | 2024-06-20T16:08:46Z | https://github.com/langchain-ai/langchain/issues/19061 | 2,185,545,359 | 19,061 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_experimental.openai_assistant import OpenAIAssistantRunnable
interpreter_assistant = OpenAIAssistantRunnable.create_assistant(
name="langchain assistant",
instructions="You are a personal math tutor. Write and run code to answer math questions.",
tools=[{"type": "code_interpreter"}],
model="gpt-4-1106-preview"
)
output = interpreter_assistant.invoke({"content": "What's 10 - 4 raised to the 2.7"})
```
### Error Message and Stack Trace (if applicable)
File "xxx/t.py", line 43, in <module>
interpreter_assistant = OpenAIAssistantRunnable.create_assistant(
File "xxx/site-packages/langchain/agents/openai_assistant/base.py", line 213, in create_assistant
tools=[convert_to_openai_tool(tool) for tool in tools], # type: ignore
File "xxx/langchain/agents/openai_assistant/base.py", line 213, in <listcomp>
tools=[convert_to_openai_tool(tool) for tool in tools], # type: ignore
File "xxx/langchain_core/utils/function_calling.py", line 329, in convert_to_openai_tool
function = convert_to_openai_function(tool)
File "xxx/langchain_core/utils/function_calling.py", line 304, in convert_to_openai_function
raise ValueError(
ValueError: Unsupported function
{'type': 'code_interpreter'}
Functions must be passed in as Dict, pydantic.BaseModel, or Callable. If they're a dict they must either be in OpenAI function format or valid JSON schema with top-level 'title' and 'description' keys.
### Description
I used the example code, but it threw an error when executed.
example code from https://python.langchain.com/docs/modules/agents/agent_types/openai_assistants#using-only-openai-tools
### System Info
$ pip list | grep langchain
langchain 0.1.12
langchain-community 0.0.28
langchain-core 0.1.31
langchain-openai 0.0.8
langchain-text-splitters 0.0.1 | The validation of tools within OpenAIAssistantRunnable.create_assistant does not account for `{"type": "code_interpreter"}`. | https://api.github.com/repos/langchain-ai/langchain/issues/19057/comments | 2 | 2024-03-14T04:01:10Z | 2024-06-28T16:07:18Z | https://github.com/langchain-ai/langchain/issues/19057 | 2,185,358,514 | 19,057 |
[
"langchain-ai",
"langchain"
] | ### 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
pass
### Error Message and Stack Trace (if applicable)
_No response_
### Description
When calling the `_stream_log_implementation` from the `astream_log` method in the `Runnable` class, it is not handing over the `kwargs` argument. Therefore, even if i want to customize APIHandler and implement additional features with additional arguments, it is not possible. Conversely, the `astream_events` method normally handing over the `kwargs` argument.
### System Info
pass | Runnable astream_log does not pass kwargs to _astream_log_implementation | https://api.github.com/repos/langchain-ai/langchain/issues/19054/comments | 0 | 2024-03-14T03:03:13Z | 2024-03-14T19:53:48Z | https://github.com/langchain-ai/langchain/issues/19054 | 2,185,283,399 | 19,054 |
[
"langchain-ai",
"langchain"
] | ### 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
```
conda install langchain=0.1.12 zeep=4.2.1
```
### Error Message and Stack Trace (if applicable)
```
Could not solve for environment specs
The following packages are incompatible
├─ langchain 0.1.12** is installable and it requires
│ ├─ langchain-text-splitters >=0.0.1,<0.1 , which requires
│ │ └─ lxml >=5.1.0,<6.0.0 , which can be installed;
│ └─ lxml >=4.9.2,<5.0.0 , which can be installed;
└─ zeep 4.2.1** is not installable because it requires
└─ lxml >=4.6.0 , which conflicts with any installable versions previously reported.
```
### Description
I'm trying to install latest langchain along with some other packages depending on lxml. It appears that `langchain` requires `lxml >=4.9.2,<5.0.0` but `langchain-text-splitters` requires `lxml >=5.1.0,<6.0.0` which are incompatible. Since the former depends on the latter, it makes dependency resolution fail.
### System Info
System Information
------------------
> OS: Windows
> OS Version: 10.0.19045
> Python Version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:27:34) [MSC v.1937 64 bit (AMD64)]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.24
> langchain_text_splitters: 0.0.1
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | Inconsistent lxml dependency between `langchain` and `langchain-text-splitters` | https://api.github.com/repos/langchain-ai/langchain/issues/19040/comments | 5 | 2024-03-13T18:23:55Z | 2024-04-07T19:23:10Z | https://github.com/langchain-ai/langchain/issues/19040 | 2,184,636,140 | 19,040 |
[
"langchain-ai",
"langchain"
] | ### 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
'''
from langchain import hub
# Get the prompt to use - you can modify this!
prompt = hub.pull("hwchase17/openai-functions-agent")
# Choose the LLM that will drive the agent
# Only certain models support this
model = ChatOpenAI(model="gpt-4-0125-preview", temperature=0)
# Construct the OpenAI Tools agent
agent = create_openai_tools_agent(model, tools, prompt)
# Create an agent executor by passing in the agent and tools
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True)
'''
### Error Message and Stack Trace (if applicable)
AttributeError: 'Client' object has no attribute 'pull_repo'
### Description
I am unable to import resource from LangChain hub, I am getting error while doing so. I uninstalled LangChain and installed it back but the error persist.
### System Info
"langchain==0.1.12" | Unable to import resource from LangChain hub | https://api.github.com/repos/langchain-ai/langchain/issues/19038/comments | 1 | 2024-03-13T17:25:15Z | 2024-03-13T17:33:05Z | https://github.com/langchain-ai/langchain/issues/19038 | 2,184,536,534 | 19,038 |
[
"langchain-ai",
"langchain"
] | ### 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
``` python
template = """<s>[INST] <<SYS>>
- Bạn là một trợ lí Tiếng Việt nhiệt tình và trung thực.
- Hãy luôn trả lời một cách hữu ích nhất có thể, đồng thời giữ an toàn.
- Câu trả lời của bạn không nên chứa bất kỳ nội dung gây hại, phân biệt chủng tộc, phân biệt giới tính, độc hại, nguy hiểm hoặc bất hợp pháp nào.
- Hãy đảm bảo rằng các câu trả lời của bạn không có thiên kiến xã hội và mang tính tích cực.
- Nếu một câu hỏi không có ý nghĩa hoặc không hợp lý về mặt thông tin, hãy giải thích tại sao thay vì trả lời một điều gì đó không chính xác.
- Nếu bạn không biết câu trả lời cho một câu hỏi, hãy trẳ lời là bạn không biết và vui lòng không chia sẻ thông tin sai lệch.
- Hãy trả lời một cách ngắn gọn, súc tích và chỉ trả lời chi tiết nếu được yêu cầu.
<</SYS>>
{question} [/INST]"""
@tool
def time(text: str) -> str:
"""Returns todays date, use this for any \
questions related to knowing todays date. \
The input should always be an empty string, \
and this function will always return todays \
date - any date mathmatics should occur \
outside this function."""
return str(date.today())
prompt = PromptTemplate(template=template, input_variables=["question"])
callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
llm = ChatOllama(model="vistral-7b-q8", temperature=0.0, callback_manager=callback_manager)
llm_chain = LLMChain(prompt=prompt, llm=llm, output_parser=StrOutputParser())
tools = load_tools(["ddg-search", "wikipedia"], llm=llm_chain)
agent= initialize_agent(
tools + [time],
llm_chain,
agent=AgentType.CHAT_ZERO_SHOT_REACT_DESCRIPTION,
handle_parsing_errors=True,
verbose = True)
agent("Hôm nay ngày mấy?")
```
### Error Message and Stack Trace (if applicable)
File [~/miniconda3/envs/stt/lib/python3.9/site-packages/langchain_core/_api/deprecation.py:145](https://vscode-remote+ssh-002dremote-002bvietmapserver.vscode-resource.vscode-cdn.net/home/vuongnt/workspace/mimi/~/miniconda3/envs/stt/lib/python3.9/site-packages/langchain_core/_api/deprecation.py:145), in deprecated.<locals>.deprecate.<locals>.warning_emitting_wrapper(*args, **kwargs)
143 warned = True
144 emit_warning()
--> 145 return wrapped(*args, **kwargs)
File [~/miniconda3/envs/stt/lib/python3.9/site-packages/langchain/chains/base.py:378](https://vscode-remote+ssh-002dremote-002bvietmapserver.vscode-resource.vscode-cdn.net/home/vuongnt/workspace/mimi/~/miniconda3/envs/stt/lib/python3.9/site-packages/langchain/chains/base.py:378), in Chain.__call__(self, inputs, return_only_outputs, callbacks, tags, metadata, run_name, include_run_info)
346 """Execute the chain.
347
348 Args:
(...)
369 `Chain.output_keys`.
370 """
371 config = {
372 "callbacks": callbacks,
373 "tags": tags,
374 "metadata": metadata,
375 "run_name": run_name,
376 }
--> 378 return self.invoke(
379 inputs,
...
343 object_setattr(__pydantic_self__, '__dict__', values)
ValidationError: 1 validation error for Generation
text
str type expected (type=type_error.str)
### Description
I got problem while using Langchain with LLM model to get output.
### System Info
System Information
------------------
> OS: Linux
> OS Version: #50-Ubuntu SMP PREEMPT_DYNAMIC Mon Jul 10 18:24:29 UTC 2023
> Python Version: 3.9.18 (main, Sep 11 2023, 13:41:44)
[GCC 11.2.0]
Package Information
-------------------
> langchain_core: 0.1.31
> langchain: 0.1.12
> langchain_community: 0.0.28
> langsmith: 0.1.23
> langchain_experimental: 0.0.54
> langchain_openai: 0.0.8
> langchain_text_splitters: 0.0.1
> langchainhub: 0.1.15
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve | Langchain ValidationError: str type expected | https://api.github.com/repos/langchain-ai/langchain/issues/19037/comments | 1 | 2024-03-13T16:37:20Z | 2024-06-20T16:08:47Z | https://github.com/langchain-ai/langchain/issues/19037 | 2,184,445,658 | 19,037 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
In [the discussion of partial prompts](https://python.langchain.com/docs/modules/model_io/prompts/) the examples are all given in terms of `format`, and there is no discussion of partial prompts in the context of pipelines.
### Idea or request for content:
I can easily imagine a pipeline in which a prompt is first partialed, then filled, and then passed to an LLM. But it's not at all clear how to do that, and this page only discusses `format`, not `invoke`.
It would be very helpful to add an example (or, if an example exists elsewhere, a link) that shows how to put partial filling of a prompt into a pipeline. If this is not possible, then it would save programmers lots of time if the documentation would say so. | DOC: RFE - Extend the discussion of Partial Prmpts with Pipeline example | https://api.github.com/repos/langchain-ai/langchain/issues/19033/comments | 3 | 2024-03-13T15:05:55Z | 2024-06-21T16:37:26Z | https://github.com/langchain-ai/langchain/issues/19033 | 2,184,245,090 | 19,033 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
I was reading the docs on [Prompts](https://python.langchain.com/docs/modules/model_io/prompts/) and stumbled accross the section [Example Selector Types](https://python.langchain.com/docs/modules/model_io/prompts/example_selector_types/) without really understanding what was about and how it connected to the other topics.
Only after moving to the next section, on [Example selectors](https://python.langchain.com/docs/modules/model_io/prompts/example_selectors), concepts seemed to align and make sense. I'm led to believe I'm not the only one getting lost when reading the documentation sequentially this way.
### Idea or request for content:
My suggestion is to simply move section [Example selectors](https://python.langchain.com/docs/modules/model_io/prompts/example_selectors) before [Example Selector Types](https://python.langchain.com/docs/modules/model_io/prompts/example_selector_types/), so the concepts are presented in a sequential way. | DOC: Swap "example selector types" and "example selectors" in the docs to make reading smoother | https://api.github.com/repos/langchain-ai/langchain/issues/19031/comments | 0 | 2024-03-13T13:58:10Z | 2024-06-19T16:08:33Z | https://github.com/langchain-ai/langchain/issues/19031 | 2,184,088,940 | 19,031 |
[
"langchain-ai",
"langchain"
] | ### 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
from langchain_community.document_loaders import UnstructuredFileLoader
import json
import os
msFiles = ["BV24-006-0_BV_Laufbahnmodell.docx"]
for filename in msFiles:
loader = UnstructuredFileLoader(filename, mode='elements') # elements | single
docs = loader.load() # Document
# Convert each document to a dictionary
data = [doc.dict() for doc in docs]
print(f"document will be serialized with {len(data)} elements!")
with open(f"{filename}.txt", 'w', encoding="utf-8") as f:
json.dump(data, f, ensure_ascii=False)
### Error Message and Stack Trace (if applicable)
_No response_
### Description
The page number of returned list of documents is wrong:
View of DOC File - page 55 of 70
View of parsed Document: page 16
### System Info
windows
python 11 | UnstructuredFileLoader loads wrong "page number" metadata of word documents | https://api.github.com/repos/langchain-ai/langchain/issues/19029/comments | 3 | 2024-03-13T10:37:07Z | 2024-06-23T16:09:14Z | https://github.com/langchain-ai/langchain/issues/19029 | 2,183,659,724 | 19,029 |
[
"langchain-ai",
"langchain"
] | ### 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
```python ` # Create the semantic settings with the configuration
semantic_search = None
if semantic_configurations is None and semantic_configuration_name is not None:
semantic_configuration = SemanticConfiguration(
name=semantic_configuration_name,
prioritized_fields=SemanticPrioritizedFields(
content_fields=[SemanticField(field_name=FIELDS_CONTENT)],
),
)
semantic_search = SemanticSearch(configurations=[semantic_configuration]) ```
### Error Message and Stack Trace (if applicable)
no error message
### Description
I am trying to create a semantic_search object with custom semantic_configuration as you can see in https://github.com/Azure/azure-search-vector-samples/blob/main/demo-python/code/basic-vector-workflow/azure-search-vector-python-sample.ipynb, ```python semantic_config = SemanticConfiguration(
name="my-semantic-config",
prioritized_fields=SemanticPrioritizedFields(
title_field=SemanticField(field_name="title"),
keywords_fields=[SemanticField(field_name="category")],
content_fields=[SemanticField(field_name="content")]
)
) ``` However with the langchain code it is not possible to define this since the this line semantic_search = None prevent to create any semantic_search object from outside although i can create a semantic_configurations, i cannot create a semantic_search object
### System Info
langchain==0.1.9
langchain-community==0.0.24
langchain-core==0.1.27
langchain-openai==0.0.7
langchainhub==0.1.14
windows 11
Python 3.11.8
conda environment | Bug Report: Custom Semantic Search Functionality Issue in Azure Search | https://api.github.com/repos/langchain-ai/langchain/issues/18998/comments | 1 | 2024-03-13T00:05:52Z | 2024-06-19T16:08:23Z | https://github.com/langchain-ai/langchain/issues/18998 | 2,182,874,556 | 18,998 |
[
"langchain-ai",
"langchain"
] | ### 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
Code isn't causing the problem here. Its the conflicting deps
### Error Message and Stack Trace (if applicable)
ERROR: Cannot install -r requirements.txt (line 6) and langchain because these package versions have conflicting dependencies.
The conflict is caused by:
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.28 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.27 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.26 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.25 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.24 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.23 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts
langchain-community 0.0.22 depends on langsmith<0.2.0 and >=0.1.0
langchain 0.1.6 depends on langsmith<0.1 and >=0.0.83
langchain-community 0.0.21 depends on langsmith<0.2.0 and >=0.1.0
To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict
### Description
I'm trying to get a python scrip I wrote to run with a github action. When the script tries to install langchain as specified in my requirements.txt It fails.
### System Info
> langchain_core: 0.1.23
> langchain: 0.1.6
> langchain_community: 0.0.19
> langsmith: 0.0.87
> langchain_experimental: 0.0.50
> langchain_mistralai: 0.0.5
> langchain_openai: 0.0.5 | Dep Conflict with Langchain | https://api.github.com/repos/langchain-ai/langchain/issues/18996/comments | 6 | 2024-03-12T22:57:03Z | 2024-04-30T18:33:20Z | https://github.com/langchain-ai/langchain/issues/18996 | 2,182,820,017 | 18,996 |
[
"langchain-ai",
"langchain"
] | ### 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
```
from langchain_elasticsearch import ElasticsearchStore
from langchain_community.embeddings import HuggingFaceEmbeddings
import torch
from langchain_community.document_loaders import TextLoader
from langchain_text_splitters import CharacterTextSplitter
import pandas as pd
from langchain_community.document_loaders import DataFrameLoader
import logging
logging.getLogger().setLevel(logging.ERROR)
device = "cuda:0" if torch.cuda.is_available() else "cpu"
print(device)
model_name = "sentence-transformers/all-mpnet-base-v2"
model_kwargs = {'device': device}
encode_kwargs = {'normalize_embeddings': False}
hf_embedding_model = HuggingFaceEmbeddings(
model_name=model_name,
model_kwargs=model_kwargs,
encode_kwargs=encode_kwargs
)
df = pd.DataFrame({'CONTENT':["abc","def"]})
loader = DataFrameLoader(df, page_content_column="CONTENT")
documents = loader.load()
text_splitter = CharacterTextSplitter(chunk_size=500, chunk_overlap=0)
docs = text_splitter.split_documents(documents)
db = ElasticsearchStore.from_documents(
docs,
hf_embedding_model,
es_url="http://localhost:9200",
index_name="test-index",
strategy=ElasticsearchStore.ApproxRetrievalStrategy(
hybrid=True,
)
)
db.add_documents(docs)
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
When using add_documents, it dumps all doc ids to STDOUT when complete.
For example:
['fbfa0f76-5b3c-4dd7-93e0-6e592a5c11b4',
'bfc8bh96-c6a2-46d3-a0e0-3d12abf24948',
'bf8a6ae6-6c14-4691-b9ba-e7987ad68e65']
### System Info
langchain==0.1.10
langchain-community==0.0.25
langchain-core==0.1.28
langchain-elasticsearch==0.1.0
langchain-openai==0.0.8
langchain-text-splitters==0.0.1 | ElasticsearchStore.add_documents() prints out document IDs to STDOUT after job completion | https://api.github.com/repos/langchain-ai/langchain/issues/18986/comments | 1 | 2024-03-12T19:10:54Z | 2024-03-12T19:13:14Z | https://github.com/langchain-ai/langchain/issues/18986 | 2,182,489,252 | 18,986 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [x] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
There is a spelling mistake in line 289.
### Idea or request for content:
I can correct this error. Please assign this issue to me. | DOC: Typo error in "https://github.com/langchain-ai/langchain/blob/master/docs/docs/get_started/quickstart.mdx", line 289. | https://api.github.com/repos/langchain-ai/langchain/issues/18981/comments | 1 | 2024-03-12T17:49:54Z | 2024-06-14T08:56:09Z | https://github.com/langchain-ai/langchain/issues/18981 | 2,182,320,367 | 18,981 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_community.agent_toolkits import create_sql_agent
from langchain_community.llms import GPT4All
from langchain_community.utilities.sql_database import SQLDatabase
# path to the gpt4all .gguf
from local_config import MODEL_PATH
llm = GPT4All(
model=MODEL_PATH,
max_tokens=2048,
)
db = SQLDatabase.from_uri("sqlite:///Chinook.db")
agent_executor = create_sql_agent(llm, db=db, agent_type="openai-tools", verbose=True)
agent_executor.invoke(
"List the total sales per country. Which country's customers spent the most?"
)
```
### Error Message and Stack Trace (if applicable)
```
TypeError Traceback (most recent call last)
Cell In[11], line 1
----> 1 agent_executor.invoke(
2 "List the total sales per country. Which country's customers spent the most?"
3 )
File c:\Users\Bob\.virtualenvs\langchain_local-QYN7rPyV\lib\site-packages\langchain\chains\base.py:163, in Chain.invoke(self, input, config, **kwargs)
161 except BaseException as e:
162 run_manager.on_chain_error(e)
--> 163 raise e
164 run_manager.on_chain_end(outputs)
166 if include_run_info:
File c:\Users\Bob\.virtualenvs\langchain_local-QYN7rPyV\lib\site-packages\langchain\chains\base.py:153, in Chain.invoke(self, input, config, **kwargs)
150 try:
151 self._validate_inputs(inputs)
152 outputs = (
--> 153 self._call(inputs, run_manager=run_manager)
154 if new_arg_supported
155 else self._call(inputs)
156 )
158 final_outputs: Dict[str, Any] = self.prep_outputs(
159 inputs, outputs, return_only_outputs
160 )
...
--> 207 for token in self.client.generate(prompt, **params):
208 if text_callback:
209 text_callback(token)
TypeError: generate() got an unexpected keyword argument 'tools'
```
### Description
I'm following the tutorial under Langchain/Components/Toolkits/[SQL Database](https://python.langchain.com/docs/integrations/toolkits/sql_database) but subbing with a local GPT4All model.
I'v successfully installed the Chinook.db and am able to execute a test sqlite query.
GPT4all loads as well. I'm expecting that when I run `agent_executor.invoke`, it doesn't error out.
### System Info
Python 3.9.6
Langchain 0.1.11
Langchain-community 0.0.27
Langchain-core 0.1.30
Langchain-experimental 0.0.53 | TypeError running `agent_executor.invoke` with open-source LM | https://api.github.com/repos/langchain-ai/langchain/issues/18979/comments | 6 | 2024-03-12T16:51:28Z | 2024-06-20T16:07:42Z | https://github.com/langchain-ai/langchain/issues/18979 | 2,182,190,543 | 18,979 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
from langchain_openai import AzureChatOpenAI
aoai_chat = AzureChatOpenAI(deployment_name='gpt4',
openai_api_key="...",
azure_endpoint="...",
openai_api_version='2023-12-01-preview')
aoai_chat.stream("Hello")
```
### Error Message and Stack Trace (if applicable)
``` bash
AttributeError("'NoneType' object has no attribute 'get'")Traceback (most recent call last): File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1592, in _atransform_stream_with_config chunk: Output = await asyncio.create_task( # type: ignore[call-arg] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 287, in await
yield self # This tells Task to wait for completion.
^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 203, in result
raise self._exception.with_traceback(self._exception_tb)
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 267, in __step
result = coro.send(None)
^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2405, in _atransform
async for output in final_pipeline:
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 4176, in atransform
async for item in self.bound.atransform(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 4176, in atransform
async for item in self.bound.atransform(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2435, in atransform
async for chunk in self._atransform_stream_with_config(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1592, in _atransform_stream_with_config
chunk: Output = await asyncio.create_task( # type: ignore[call-arg]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 287, in await
yield self # This tells Task to wait for completion.
^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 339, in __wakeup
future.result()
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 203, in result
raise self._exception.with_traceback(self._exception_tb)
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 267, in __step
result = coro.send(None)
^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2405, in _atransform
async for output in final_pipeline:
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 4176, in atransform
async for item in self.bound.atransform(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2435, in atransform
async for chunk in self._atransform_stream_with_config(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1592, in _atransform_stream_with_config
chunk: Output = await asyncio.create_task( # type: ignore[call-arg]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 287, in await
yield self # This tells Task to wait for completion.
^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 339, in __wakeup
future.result()
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 203, in result
raise self._exception.with_traceback(self._exception_tb)
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 267, in __step
result = coro.send(None)
^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 2405, in _atransform
async for output in final_pipeline:
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/output_parsers/transform.py", line 60, in atransform
async for chunk in self._atransform_stream_with_config(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1592, in _atransform_stream_with_config
chunk: Output = await asyncio.create_task( # type: ignore[call-arg]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 287, in await
yield self # This tells Task to wait for completion.
^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 339, in __wakeup
future.result()
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/futures.py", line 203, in result
raise self._exception.with_traceback(self._exception_tb)
File "/Users/xxx/anaconda3/envs//lib/python3.11/asyncio/tasks.py", line 267, in __step
result = coro.send(None)
^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/output_parsers/transform.py", line 38, in _atransform
async for chunk in input:
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/utils/aiter.py", line 97, in tee_peer
item = await iterator.anext()
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/runnables/base.py", line 1068, in atransform
async for output in self.astream(final, config, **kwargs):
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 308, in astream
raise e
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_core/language_models/chat_models.py", line 292, in astream
async for chunk in self._astream(
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_openai/chat_models/base.py", line 519, in _astream
chunk = _convert_delta_to_message_chunk(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/xxx/anaconda3/envs//lib/python3.11/site-packages/langchain_openai/chat_models/base.py", line 174, in _convert_delta_to_message_chunk
role = cast(str, _dict.get("role"))
^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'get'
```
### Description
### I have found the cause of this Issue.
Because when I was using Azure OpenAI, in order to stream rapidly, I enabled the [Content filtering streaming mode](
https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/content-filter?tabs=warning%2Cpython#content-streaming), Azure would return a chunk of data at the end of the streaming transmission.
```json
{
"id": "",
"object": "",
"created": 0,
"model": "",
"choices": [
{
"finish_reason": null,
"index": 0,
"content_filter_results": {
"hate": {
"filtered": false,
"severity": "safe"
},
"self_harm": {
"filtered": false,
"severity": "safe"
},
"sexual": {
"filtered": false,
"severity": "safe"
},
"violence": {
"filtered": false,
"severity": "safe"
}
},
"content_filter_offsets": {
"check_offset": 4522,
"start_offset": 4522,
"end_offset": 4686
}
}
]
}
```
**This chunk of data, compared to a normal chunk of data, lacks the 'delta' field, causing this piece of code to throw an error.**
<img width="1090" alt="image" src="https://github.com/langchain-ai/langchain/assets/38649663/03b148bd-4eb5-4ce0-8664-2cfef34d42e8">
### System Info
langchain==0.1.11
langchain-community==0.0.25
langchain-core==0.1.29
langchain-experimental==0.0.52
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
**platform = mac**
**python = 3.11.5** | An error occurs when enabling Content filtering streaming while using Azure OpenAI | https://api.github.com/repos/langchain-ai/langchain/issues/18977/comments | 0 | 2024-03-12T16:36:39Z | 2024-03-28T21:46:28Z | https://github.com/langchain-ai/langchain/issues/18977 | 2,182,155,882 | 18,977 |
[
"langchain-ai",
"langchain"
] | ### 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
```python
def len_in_words(text: str) -> int:
return len(re.findall(r'\b\w+\b', text))
splitter = RecursiveCharacterTextSplitter(
chunk_size=2,
chunk_overlap=1,
length_function=len_in_words, # custom len fun
add_start_index=True,
)
doc = Document(page_content='test test test test', metadata={})
chunks = splitter.split_documents([doc])
assert chunks[0].metadata['start_index'] == 0
assert chunks[1].metadata['start_index'] == 5 # fails: actual is 10
assert chunks[2].metadata['start_index'] == 10 # fails: actual is -1
```
### Error Message and Stack Trace (if applicable)
_No response_
### Description
There is a bug in the way `TextSplitter. create_documents` calculates the offset when a custom length function is used: https://github.com/langchain-ai/langchain/blob/471f2ed40abbf9ea02ccf5b384db2e8580ed1cbb/libs/text-splitters/langchain_text_splitters/base.py#L81
Indeed, `previous_chunk_len` is in chars, while `self._chunk_overlap` can be in any custom unit (I'm using words in my example). This leads to a wrong offset, which in turns means a wrong `start_index`.
Solution: convert the chunk overlap in chars too, so that the formula uses the same unit.
### System Info
System Information
------------------
> OS: Darwin
> OS Version: Darwin Kernel Version 22.6.0: Wed Jul 5 22:22:05 PDT 2023; root:xnu-8796.141.3~6/RELEASE_ARM64_T6000
> Python Version: 3.10.5 (main, Jul 22 2022, 10:15:54) [Clang 13.1.6 (clang-1316.0.21.2.3)]
Package Information
-------------------
> langchain_core: 0.1.18
> langchain: 0.1.5
> langchain_community: 0.0.17
Packages not installed (Not Necessarily a Problem)
--------------------------------------------------
The following packages were not found:
> langgraph
> langserve
| TextSplitter sets wrong start_index in case of custom len function | https://api.github.com/repos/langchain-ai/langchain/issues/18972/comments | 1 | 2024-03-12T14:25:53Z | 2024-03-12T15:09:18Z | https://github.com/langchain-ai/langchain/issues/18972 | 2,181,759,866 | 18,972 |
[
"langchain-ai",
"langchain"
] | ### 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
from langchain_community.document_loaders import WebBaseLoader
loader = WebBaseLoader("https://docs.smith.langchain.com/user_guide")
docs = loader.load()
### Error Message and Stack Trace (if applicable)
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Cell In[1], line 1
----> 1 from langchain_community.document_loaders import WebBaseLoader
2 loader = WebBaseLoader("https://docs.smith.langchain.com/user_guide")
4 docs = loader.load()
File ~\miniconda3\envs\python38\lib\site-packages\langchain_community\document_loaders\__init__.py:191
189 from langchain_community.document_loaders.snowflake_loader import SnowflakeLoader
190 from langchain_community.document_loaders.spreedly import SpreedlyLoader
--> 191 from langchain_community.document_loaders.sql_database import SQLDatabaseLoader
192 from langchain_community.document_loaders.srt import SRTLoader
193 from langchain_community.document_loaders.stripe import StripeLoader
File ~\miniconda3\envs\python38\lib\site-packages\langchain_community\document_loaders\sql_database.py:10
6 from langchain_community.document_loaders.base import BaseLoader
7 from langchain_community.utilities.sql_database import SQLDatabase
---> 10 class SQLDatabaseLoader(BaseLoader):
11 """
12 Load documents by querying database tables supported by SQLAlchemy.
13
(...)
17 Each document represents one row of the result.
18 """
20 def __init__(
21 self,
22 query: Union[str, sa.Select],
(...)
30 include_query_into_metadata: bool = False,
31 ):
File ~\miniconda3\envs\python38\lib\site-packages\langchain_community\document_loaders\sql_database.py:22, in SQLDatabaseLoader()
10 class SQLDatabaseLoader(BaseLoader):
11 """
12 Load documents by querying database tables supported by SQLAlchemy.
13
(...)
17 Each document represents one row of the result.
18 """
20 def __init__(
21 self,
---> 22 query: Union[str, sa.Select],
23 db: SQLDatabase,
24 *,
25 parameters: Optional[Dict[str, Any]] = None,
26 page_content_mapper: Optional[Callable[..., str]] = None,
27 metadata_mapper: Optional[Callable[..., Dict[str, Any]]] = None,
28 source_columns: Optional[Sequence[str]] = None,
29 include_rownum_into_metadata: bool = False,
30 include_query_into_metadata: bool = False,
31 ):
32 """
33 Args:
34 query: The query to execute.
(...)
49 expression into the metadata dictionary. Default: False.
50 """
51 self.query = query
AttributeError: module 'sqlalchemy' has no attribute 'Select'
### Description
Hi All,
New to this project so apologies if I have misunderstood/missed anything. I am simply trying to follow the official documentation of langchain to familiarise myself with the functionality and encountering an error on the "Retrieval" example steps.
The community functions are supposed to be built on SQLAlchemy >=1.4 <1.4.3, but I'm getting the above error with SQLAlchemy 1.4 and 1.4.1. I believe the code needs to be updated? Or I am doing something wrong?
Any help much appreciated.
Thanks
### System Info
aenum==3.1.15
aiohttp==3.9.3
aiosignal==1.3.1
altair==5.2.0
annotated-types==0.6.0
anyio==3.7.1
argon2-cffi==23.1.0
argon2-cffi-bindings==21.2.0
asn1crypto==1.5.1
asteval==0.9.23
astor==0.8.1
asttokens==2.4.1
async-timeout==4.0.3
attrs==23.2.0
Babel==2.14.0
backcall==0.2.0
backoff==2.2.1
backports.zoneinfo==0.2.1
beautifulsoup4==4.11.2
bleach==6.1.0
blinker==1.7.0
cachetools==5.3.3
certifi==2024.2.2
cffi==1.16.0
charset-normalizer==2.1.1
ChromeController==0.3.26
click==8.1.7
cloudpickle==2.0.0
colorama==0.4.6
comm==0.2.1
cryptography==36.0.2
cssselect==1.2.0
cssutils==2.9.0
cycler==0.12.1
dataclasses-json==0.6.4
dataframe-image==0.1.14
debugpy==1.8.1
decorator==5.1.1
defusedxml==0.7.1
deprecation==2.1.0
distro==1.9.0
docopt==0.6.2
entrypoints==0.4
et-xmlfile==1.1.0
exceptiongroup==1.2.0
exchange-calendars==4.2.8
executing==2.0.1
fastjsonschema==2.19.1
fds.analyticsapi.engines==5.6.0
fds.protobuf.stach==1.0.0
fds.protobuf.stach.extensions==1.3.1
fds.protobuf.stach.v2==1.0.2
filelock==3.13.1
fonttools==4.49.0
frozenlist==1.4.1
func-timeout==4.3.5
funcsigs==1.0.2
future==1.0.0
gitdb==4.0.11
GitPython==3.1.42
greenlet==3.0.3
gs-quant==0.9.108
h11==0.14.0
h2o==3.44.0.3
html2image==2.0.4.3
httpcore==1.0.4
httpx==0.27.0
idna==3.6
importlib-metadata==6.11.0
importlib_resources==6.1.3
inflection==0.5.1
ipykernel==6.29.3
ipython==8.12.3
ipython-genutils==0.2.0
ipywidgets==8.0.7
jedi==0.19.1
Jinja2==3.1.3
joblib==1.3.2
json5==0.9.22
jsonpatch==1.33
jsonpointer==2.4
jsonschema==4.21.1
jsonschema-specifications==2023.12.1
jupyter==1.0.0
jupyter-console==6.6.3
jupyter-server==1.24.0
jupyter_client==7.4.9
jupyter_core==5.7.1
jupyterlab==3.4.8
jupyterlab_pygments==0.3.0
jupyterlab_server==2.24.0
jupyterlab_widgets==3.0.10
kiwisolver==1.4.5
korean-lunar-calendar==0.3.1
langchain==0.1.11
langchain-community==0.0.27
langchain-core==0.1.30
langchain-openai==0.0.8
langchain-text-splitters==0.0.1
langsmith==0.1.23
lmfit==1.0.2
lxml==5.1.0
markdown-it-py==3.0.0
MarkupSafe==2.1.5
marshmallow==3.21.1
matplotlib==3.5.3
matplotlib-inline==0.1.6
mdurl==0.1.2
mistune==3.0.2
more-itertools==10.2.0
MouseInfo==0.1.3
msgpack==1.0.8
multidict==6.0.5
mypy-extensions==1.0.0
nbclassic==1.0.0
nbclient==0.9.0
nbconvert==7.16.2
nbformat==5.9.2
nest-asyncio==1.6.0
notebook==6.5.6
notebook_shim==0.2.4
numpy==1.23.5
openai==1.13.3
openpyxl==3.0.10
opentracing==2.4.0
orjson==3.9.15
oscrypto==1.3.0
packaging==23.2
pandas==1.4.4
pandas_market_calendars==4.3.3
pandocfilters==1.5.1
parso==0.8.3
patsy==0.5.6
pdfkit==1.0.0
pendulum==2.1.2
pickleshare==0.7.5
Pillow==9.5.0
pkgutil_resolve_name==1.3.10
platformdirs==4.2.0
premailer==3.10.0
prometheus_client==0.20.0
prompt-toolkit==3.0.43
protobuf==3.20.3
psutil==5.9.8
pure-eval==0.2.2
pyarrow==8.0.0
PyAutoGUI==0.9.54
pycparser==2.21
pycryptodomex==3.20.0
pydantic==2.6.3
pydantic_core==2.16.3
pydash==7.0.7
pydeck==0.8.1b0
PyGetWindow==0.0.9
Pygments==2.17.2
PyJWT==2.8.0
pyluach==2.2.0
Pympler==1.0.1
PyMsgBox==1.0.9
pyOpenSSL==22.0.0
pyparsing==3.1.2
pyperclip==1.8.2
pypiwin32==223
PyRect==0.2.0
PyScreeze==0.1.30
python-dateutil==2.8.2
python-decouple==3.8
python-dotenv==1.0.1
pytweening==1.2.0
pytz==2024.1
pytz-deprecation-shim==0.1.0.post0
pytzdata==2020.1
pywin32==306
pywinpty==2.0.13
PyYAML==6.0.1
pyzmq==24.0.1
qtconsole==5.5.1
QtPy==2.4.1
referencing==0.33.0
regex==2023.12.25
requests==2.28.2
rich==13.7.1
rpds-py==0.18.0
scikit-learn==1.3.2
scipy==1.9.3
seaborn==0.12.2
Send2Trash==1.8.2
six==1.16.0
slackclient==2.9.4
smmap==5.0.1
sniffio==1.3.1
snowflake-connector-python==2.7.12
snowflake-snowpark-python==0.9.0
snowflake-sqlalchemy==1.4.7
soupsieve==2.5
SQLAlchemy==1.4.0
stack-data==0.6.3
statsmodels==0.13.5
streamlit==1.23.1
streamlit-aggrid==0.3.4.post3
tabulate==0.9.0
tenacity==8.2.3
terminado==0.18.0
threadpoolctl==3.3.0
tiktoken==0.6.0
tinycss2==1.2.1
toml==0.10.2
tomli==2.0.1
toolz==0.12.1
tornado==6.4
tqdm==4.64.1
traitlets==5.14.1
typing-inspect==0.9.0
typing_extensions==4.10.0
tzdata==2024.1
tzlocal==4.3.1
uncertainties==3.1.7
urllib3==1.26.18
validators==0.22.0
watchdog==4.0.0
wcwidth==0.2.13
webencodings==0.5.1
websocket-client==1.7.0
websockets==12.0
widgetsnbextension==4.0.10
xlrd==2.0.1
yarl==1.9.4
zipp==3.17.0 | Issue with the use of sqlalchemy in community.document_loaders? | https://api.github.com/repos/langchain-ai/langchain/issues/18968/comments | 2 | 2024-03-12T11:23:06Z | 2024-03-12T11:33:43Z | https://github.com/langchain-ai/langchain/issues/18968 | 2,181,373,166 | 18,968 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [x] I added a very descriptive title to this issue.
- [ ] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
I am utilizing the bedrock Llama2 LLM and I aim to integrate a stop feature. How can this be achieved?
### Idea or request for content:
_No response_ | DOC:I am utilizing the bedrock Llama2 LLM and I aim to integrate a stop feature. How can this be achieved? | https://api.github.com/repos/langchain-ai/langchain/issues/18966/comments | 3 | 2024-03-12T11:03:45Z | 2024-03-19T04:08:22Z | https://github.com/langchain-ai/langchain/issues/18966 | 2,181,335,351 | 18,966 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
Custom tool parameter defined within the notebook is for **Python_REPL** tool.
and it does not show the usage of the tool in an agent.
[SerpAPI Tool Docs](https://python.langchain.com/docs/integrations/tools/serpapi#custom-parameters)
### Idea or request for content:
- Fix Custom tool Definition of SerpAPI tool.
- Add more information about being used with an agent.
| DOC: Inaccurate Tool Custom parameters of SerpAPI tool Documentation | https://api.github.com/repos/langchain-ai/langchain/issues/18959/comments | 1 | 2024-03-12T07:16:24Z | 2024-06-18T16:09:45Z | https://github.com/langchain-ai/langchain/issues/18959 | 2,180,899,756 | 18,959 |
[
"langchain-ai",
"langchain"
] | ### Privileged issue
- [X] I am a LangChain maintainer, or was asked directly by a LangChain maintainer to create an issue here.
### Issue Content
# Goal
When using `astream()`, LLMs should fallback to sync streaming if an async streaming implementation is not available.
# Context
Implementation of LLMs often include a sync implementation of streaming, but are missing an async implementation.
LLMs currently do not fallback on the sync streaming implementation.
For reference here's the [BaseLLM](https://github.com/langchain-ai/langchain/blob/43db4cd20e0e718f368267528706f92bf604bac9/libs/core/langchain_core/language_models/llms.py#L464-L464) implementation.
The current fallback sequence is:
1) If _astream is defined use it
2) if _astream is not defined fallback on ainvoke
The fallback sequence should be:
1) if _astream is defined use it
2) if _stream is defined fallback to it
3) Finally if neither _astream or _stream are defined, fallback to ainvoke
This PR shows how the same problem was fixed for chat models: https://github.com/langchain-ai/langchain/pull/18748
## Acceptance criteria
* Fallback sequence is correctly implemented
* Unit-tests confirm that the fallback sequence works correctly (see the PR for the unit-tests)
This PR will not be accepted without unit-tests since this is critical functionality! | Allow LLMs async streaming to fallback on sync streaming | https://api.github.com/repos/langchain-ai/langchain/issues/18920/comments | 5 | 2024-03-11T15:09:40Z | 2024-03-20T15:43:07Z | https://github.com/langchain-ai/langchain/issues/18920 | 2,179,384,439 | 18,920 |
[
"langchain-ai",
"langchain"
] | ### 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
```
url = f"https://{ES_USER}:{ES_PASSWORD}@localhost:9200"
client = Elasticsearch(url, ca_certs = "./http_ca.crt", verify_certs = True)
print(client.info())
import elastic_transport
elastic_transport.debug_logging()
es = ElasticsearchStore.from_documents(
docs,
strategy=ElasticsearchStore.SparseVectorRetrievalStrategy(),
es_url = url,
es_connection = client,
index_name = elastic_index_name,
es_user = ES_USER,
es_password = ES_PASSWORD)
```
### Error Message and Stack Trace (if applicable)
Error adding texts: 116 document(s) failed to index.
First error reason: Could not find trained model [.elser_model_1]
### Description
The reason for this is because I deployed .elser_mode_2 on my Elasticsearch cluster. Later on, I used the following version of the code:
url = f"https://{ES_USER}:{ES_PASSWORD}@localhost:9200"
client = Elasticsearch(url, ca_certs = "./http_ca.crt", verify_certs = True)
print(client.info())
```
import elastic_transport
elastic_transport.debug_logging()
es = ElasticsearchStore.from_documents(
docs,
strategy=ElasticsearchStore.SparseVectorRetrievalStrategy(model_id=".elser_model_2"),
es_url = url,
es_connection = client,
index_name = elastic_index_name,
es_user = ES_USER,
es_password = ES_PASSWORD)
```
This time, I set the model_id correctly to .elser_model_2. After I run the above code, I still got the same error message. It is still looking for .elser_model_1.
After consulting it with our colleague, and we found the root cause of the problem: In the first step, "elastic_index_name" index has already been created, and changing the model_id without deleting the created index in the first step won't work for the updated model_id.
The solution to this problem is:
1) delete the created index in the first step
2) run the code again with the updated model_id parameter.
Improvement:
This error correct cannot be easily resolved without looking into the code. Can we add some comments into the function? or can we recreate the pipeline when model_id is change?
Thanks
### System Info
$ pip freeze | grep langchain
langchain==0.1.11
langchain-community==0.0.27
langchain-core==0.1.30
langchain-text-splitters==0.0.1 | Could not find trained model [.elser_model_1] for ElasticsearchStore | https://api.github.com/repos/langchain-ai/langchain/issues/18917/comments | 3 | 2024-03-11T14:44:02Z | 2024-06-18T16:09:46Z | https://github.com/langchain-ai/langchain/issues/18917 | 2,179,318,114 | 18,917 |
[
"langchain-ai",
"langchain"
] | ### Checklist
- [X] I added a very descriptive title to this issue.
- [X] I included a link to the documentation page I am referring to (if applicable).
### Issue with current documentation:
LangChainDeprecationWarning: The function `__call__` was deprecated in LangChain 0.1.0 and will be removed in 0.2.0. Use invoke instead.
warn_deprecated(
#!/usr/bin/env python
# coding: utf-8
# In[4]:
from langchain.document_loaders import PyPDFLoader, DirectoryLoader
from langchain.prompts import PromptTemplate
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
from langchain.llms import CTransformers
from langchain.chains import RetrievalQA
#import chainlit as cl
import streamlit as st
DB_FAISS_PATH = 'vectorstore/db_faiss'
custom_prompt_template = """Use the following pieces of information to answer the user's question.
If you don't know the answer, just say that you don't know, don't try to make up an answer.
Context: {context}
Question: {question}
Only return the helpful answer below and nothing else.
Helpful answer:
"""
def set_custom_prompt():
"""
Prompt template for QA retrieval for each vectorstore
"""
prompt = PromptTemplate(template=custom_prompt_template,
input_variables=['context', 'question'])
return prompt
# Retrieval QA Chain
def retrieval_qa_chain(llm, prompt, db):
qa_chain = RetrievalQA.from_chain_type(llm=llm,
chain_type='stuff',
retriever=db.as_retriever(search_kwargs={'k': 2}),
return_source_documents=True,
chain_type_kwargs={'prompt': prompt})
return qa_chain
# Loading the model
import torch
def load_llm():
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
llm = CTransformers(
model="D:\internships\Rajyug ITSolutions\LLM\model\llama-2-7b-chat.ggmlv3.q8_0.bin",
model_type="llama",
max_new_tokens=512,
temperature=0.5,
device=device # Specify device here
)
return llm
# QA Model Function
def qa_bot():
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
embeddings = HuggingFaceEmbeddings(
model_name="sentence-transformers/all-MiniLM-L6-v2",
model_kwargs={'device': device} # Specify device here
)
db = FAISS.load_local(DB_FAISS_PATH, embeddings, allow_dangerous_deserialization=True)
# db = FAISS.load_local(DB_FAISS_PATH, embeddings)
llm = load_llm()
qa_prompt = set_custom_prompt()
qa = retrieval_qa_chain(llm, qa_prompt, db)
return qa
# Output function
def final_result(query):
qa_result = qa_bot()
response = qa_result({'query': query})
return response['result'] # Extract only the 'result' field from the response
def main():
st.title("Medical Chatbot")
query = st.text_input("Enter your medical query:")
if st.button("Get Answer"):
if query:
answer = final_result(query)
st.write("Bot's Response:")
st.write(answer) # Print only the 'result'
else:
st.write("Please enter a query.")
# Call qa_bot and store the returned chain
qa_chain = qa_bot()
# Assuming you have a chain named 'my_chain' (commented out)
# Assuming you have a chain named 'my_chain'
# Old (deprecated):
# result = my_chain()
# New (recommended):
result = qa_chain.invoke(input=query)
# Verbosity (if needed)
from langchain.globals import set_verbose, get_verbose
# Set verbosity to True
langchain.globals.set_verbose(True)
# Check current verbosity
langchain.globals.get_verbose(True)
# ... (code using qa_chain, if applicable)
# Use the 'invoke' method to execute the chain (fix for deprecation warning)
# result = qa_chain.invoke() # Uncomment if you need the result
# Verbosity section (commented out for clarity)
# from langchain.globals import set_verbose, get_verbose
#
# # Set verbosity to True (optional)
# # langchain.globals.set_verbose(True)
#
# # Check current verbosity
# # current_verbosity = get_verbose()
if __name__ == "__main__":
main()
### Idea or request for content:
_No response_ | DOC: <PleLangChainDeprecationWarning: The function `__call__` was deprecated in LangChain 0.1.0 and will be removed in 0.2.0. Use invoke instead. warn_deprecated(ase write a comprehensive title after the 'DOC: ' prefix> | https://api.github.com/repos/langchain-ai/langchain/issues/18911/comments | 1 | 2024-03-11T12:12:06Z | 2024-06-21T16:37:55Z | https://github.com/langchain-ai/langchain/issues/18911 | 2,178,980,605 | 18,911 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.