Pydantic type conversion to SecreteStr for ChatAnthropic anthropic_api_key keyword value raises attribute error in Anthropic client

[RubenTurciosTensoriot]

System Info

Amazon Linux, langchain version 0.0.304, Docker Image

Who can help?

@hwchase17
@agola11

Information

• The official example notebooks/scripts
• My own modified scripts

Related Components

• LLMs/Chat Models
• Embedding Models
• Prompts / Prompt Templates / Prompt Selectors
• Output Parsers
• Document Loaders
• Vector Stores / Retrievers
• Memory
• Agents / Agent Executors
• Tools / Toolkits
• Chains
• Callbacks/Tracing
• Async

Reproduction

Steps to reproduce:

1. Initialize ChatAnthropic model, with keyword "anthropic_api_key", as documented: https://api.python.langchain.com/en/latest/chat_models/langchain.chat_models.anthropic.ChatAnthropic.html#langchain.chat_models.anthropic.ChatAnthropic

2. Call chat model via call method, e.g.:
   """
   chat = ChatAnthropic(..., anthropic_api_key=os.environ["ANTHROPIC_API_KEY"])
   messages = [....]
   response = chat(messages)
   """

Expected behavior

A normal response for a chat model is expected, but instead we get:

Traceback (most recent call last):
  File "/var/task/app/src/utils/error_route.py", line 27, in custom_route_handler
    return await original_route_handler(request)
  File "/var/lang/lib/python3.10/site-packages/fastapi/routing.py", line 273, in app
    raw_response = await run_endpoint_function(
  File "/var/lang/lib/python3.10/site-packages/fastapi/routing.py", line 190, in run_endpoint_function
    return await dependant.call(**values)
  File "/var/task/app/app.py", line 184, in chat
    reply = chat(data["question"])
  File "/var/task/app/src/app_modules/chat.py", line 192, in __call__
    res = self.chatbot(messages)
  File "/var/lang/lib/python3.10/site-packages/langchain/chat_models/base.py", line 595, in __call__
    generation = self.generate(
  File "/var/lang/lib/python3.10/site-packages/langchain/chat_models/base.py", line 348, in generate
    raise e
  File "/var/lang/lib/python3.10/site-packages/langchain/chat_models/base.py", line 338, in generate
    self._generate_with_cache(
  File "/var/lang/lib/python3.10/site-packages/langchain/chat_models/base.py", line 490, in _generate_with_cache
    return self._generate(
  File "/var/lang/lib/python3.10/site-packages/langchain/chat_models/anthropic.py", line 184, in _generate
    response = self.client.completions.create(**params)
  File "/var/lang/lib/python3.10/site-packages/anthropic/_utils/_utils.py", line 250, in wrapper
    return func(*args, **kwargs)
  File "/var/lang/lib/python3.10/site-packages/anthropic/resources/completions.py", line 223, in create
    return self._post(
  File "/var/lang/lib/python3.10/site-packages/anthropic/_base_client.py", line 949, in post
    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
  File "/var/lang/lib/python3.10/site-packages/anthropic/_base_client.py", line 748, in request
    return self._request(
  File "/var/lang/lib/python3.10/site-packages/anthropic/_base_client.py", line 766, in _request
    request = self._build_request(options)
  File "/var/lang/lib/python3.10/site-packages/anthropic/_base_client.py", line 397, in _build_request
    headers = self._build_headers(options)
  File "/var/lang/lib/python3.10/site-packages/anthropic/_base_client.py", line 373, in _build_headers
    headers = httpx.Headers(headers_dict)
  File "/var/lang/lib/python3.10/site-packages/httpx/_models.py", line 71, in __init__
    self._list = [
  File "/var/lang/lib/python3.10/site-packages/httpx/_models.py", line 75, in <listcomp>
    normalize_header_value(v, encoding),
  File "/var/lang/lib/python3.10/site-packages/httpx/_utils.py", line 53, in normalize_header_value
    return value.encode(encoding or "ascii")
AttributeError: 'SecretStr' object has no attribute 'encode'

Corresponding httpx code:

def normalize_header_value(
    value: typing.Union[str, bytes], encoding: typing.Optional[str] = None
) -> bytes:
    """
    Coerce str/bytes into a strictly byte-wise HTTP header value.
    """
    if isinstance(value, bytes):
        return value
    return value.encode(encoding or "ascii")

[dosubot]

🤖

Hello,

Thank you for providing detailed information about the issue you're facing. The error AttributeError: 'SecretStr' object has no attribute 'encode' is likely due to the fact that the SecretStr object is being used where a regular string is expected.

In the LangChain codebase, the anthropic_api_key is indeed passed as a SecretStr object when initializing the ChatAnthropic model. This is done in the validate_environment method in the _AnthropicCommon class where the anthropic_api_key is being set as a SecretStr object. Here is the relevant code:

@root_validator()
def validate_environment(cls, values: Dict) -> Dict:
    """Validate that api key and python package exists in environment."""
    values["anthropic_api_key"] = SecretStr(
        get_from_dict_or_env(values, "anthropic_api_key", "ANTHROPIC_API_KEY")
    )
    ...
    return values

When initializing the Anthropic and AsyncAnthropic clients, the get_secret_value() method is used to retrieve the actual string value of the API key:

values["client"] = anthropic.Anthropic(
    base_url=values["anthropic_api_url"],
    api_key=values["anthropic_api_key"].get_secret_value(),
    timeout=values["default_request_timeout"],
)
values["async_client"] = anthropic.AsyncAnthropic(
    base_url=values["anthropic_api_url"],
    api_key=values["anthropic_api_key"].get_secret_value(),
    timeout=values["default_request_timeout"],
)

If the get_secret_value() method is not used when the SecretStr object is passed to a function or method that expects a regular string, it could result in the AttributeError you're seeing.

Please ensure that you're using the get_secret_value() method to retrieve the string value of the anthropic_api_key wherever it's used in your code. If the error persists, please provide more specific details about where and how the anthropic_api_key is being used when the error occurs.

I hope this helps! If you have any other questions or need further clarification, feel free to ask.

Best,
Dosu

Sources

• libs/langchain/langchain/llms/anthropic.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[RubenTurciosTensoriot]

@dosu-bot This comment does not apply, as it is an Anthropic third party dependency using the SecretStr value is an input.

[dosubot]

🤖

Hello,

Thank you for reaching out and providing a detailed description of the issue you're facing. It seems like the problem is related to the SecretStr object from the pydantic library, which is used for secure handling of sensitive data like API keys. The SecretStr object is not directly compatible with the httpx library, which expects a regular string.

The SecretStr object has a get_secret_value() method that returns the original string. You can use this method to extract the string value of the anthropic_api_key before passing it to the ChatAnthropic model. Here's how you can modify your code:

from pydantic import SecretStr

anthropic_api_key = SecretStr(os.environ["ANTHROPIC_API_KEY"])
chat = ChatAnthropic(..., anthropic_api_key=anthropic_api_key.get_secret_value())
messages = [....]
response = chat(messages)

This should resolve the AttributeError: 'SecretStr' object has no attribute 'encode' issue you're facing. Please try this solution and let me know if it works for you or if you encounter any other issues.

Best regards.

Sources

• libs/langchain/langchain/llms/anthropic.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[eyurtsev]

Resolved on master and will be released soon!