## Description
Memory return could be set as `str` or `message` by `return_messages`
flag as mentioned in
https://python.langchain.com/docs/modules/memory/#whether-memory-is-a-string-or-a-list-of-messages,
where
`langchain.chains.conversation.memory.ConversationSummaryBufferMemory`
did not implement that.
This commit added `buffer_as_str` and `buffer_as_messages` function, and
`buffer` now affected by `return_messages` flag.
## Example Test Code and Output
```python
# Fix: ConversationSummaryBufferMemory with return_messages flag function
# Test code
from langchain.chains.conversation.memory import ConversationSummaryBufferMemory
from langchain_community.llms.ollama import Ollama
llm = Ollama()
# Create an instance of ConversationSummaryBufferMemory with return_messages set to True
memory = ConversationSummaryBufferMemory(return_messages=True, llm=llm)
# Add user and AI messages to the chat memory
memory.chat_memory.add_user_message("hi!")
memory.chat_memory.add_ai_message("what's up?")
# Print the buffer
print("Buffer:")
print(*map(type, memory.buffer), sep="\n")
print(memory.buffer, "\n")
# Print the buffer as a string
print("Buffer as String:")
print(type(memory.buffer_as_str))
print(memory.buffer_as_str, "\n")
# Print the buffer as messages
print("Buffer as Messages:")
print(*map(type, memory.buffer_as_messages), sep="\n")
print(memory.buffer_as_messages, "\n")
# Print the buffer after setting return_messages to False
memory.return_messages = False
print("Buffer after setting return_messages to False:")
print(type(memory.buffer))
print(memory.buffer, "\n")
```
```plaintext
Buffer:
<class 'langchain_core.messages.human.HumanMessage'>
<class 'langchain_core.messages.ai.AIMessage'>
[HumanMessage(content='hi!'), AIMessage(content="what's up?")]
Buffer as String:
<class 'str'>
Human: hi!
AI: what's up?
Buffer as Messages:
<class 'langchain_core.messages.human.HumanMessage'>
<class 'langchain_core.messages.ai.AIMessage'>
[HumanMessage(content='hi!'), AIMessage(content="what's up?")]
Buffer after setting return_messages to False:
<class 'str'>
Human: hi!
AI: what's up?
```
---------
Co-authored-by: Bagatur <22008038+baskaryan@users.noreply.github.com>
Issue: we have several helper functions to import third-party libraries
like tools.gmail.utils.import_google in
[community.tools](https://api.python.langchain.com/en/latest/community_api_reference.html#id37).
And we have core.utils.utils.guard_import that works exactly for this
purpose.
The import_<package> functions work inconsistently and rather be private
functions.
Change: replaced these functions with the guard_import function.
Related to #21133
Issues (nit):
1. `utils.guard_import` prints wrong error message when there is an
import `error.` It prints the whole `module_name` but should be only the
first part as the pip package name. E.i. `langchain_core.utils` -> print
not `langchain-core` but `langchain_core.utils`. Also replace '_' with
'-' in the pip package name.
2. it does not handle the `ModuleNotFoundError` which raised if
`guard_import("wrong_module")`
Fixed issues; added ut-s. Controversial: I've reraised
`ModuleNotFoundError` as `ImportError`, since in case of the error, the
proposed action is the same - we need to install a missed package.
Thank you for contributing to LangChain!
- [ ] **PR title**: "package: description"
- Where "package" is whichever of langchain, community, core,
experimental, etc. is being modified. Use "docs: ..." for purely docs
changes, "templates: ..." for template changes, "infra: ..." for CI
changes.
- Example: "community: add foobar LLM"
- [ ] **PR message**: ***Delete this entire checklist*** and replace
with
- **Description:** a description of the change
- **Issue:** the issue # it fixes, if applicable
- **Dependencies:** any dependencies required for this change
- **Twitter handle:** if your PR gets announced, and you'd like a
mention, we'll gladly shout you out!
- [ ] **Add tests and docs**: If you're adding a new integration, please
include
1. a test for the integration, preferably unit tests that do not rely on
network access,
2. an example notebook showing its use. It lives in
`docs/docs/integrations` directory.
- [ ] **Lint and test**: Run `make format`, `make lint` and `make test`
from the root of the package(s) you've modified. See contribution
guidelines for more: https://python.langchain.com/docs/contributing/
Additional guidelines:
- Make sure optional dependencies are imported within a function.
- Please do not add dependencies to pyproject.toml files (even optional
ones) unless they are required for unit tests.
- Most PRs should not touch more than one package.
- Changes should be backwards compatible.
- If you are adding something to community, do not re-import it in
langchain.
If no one reviews your PR within a few days, please @-mention one of
baskaryan, efriis, eyurtsev, hwchase17.
Issue: `load_summarize_chain` is placed in the __init__.py file. As a
result, it doesn't listed in the API Reference docs.
Change: moved code from __init__.py into a new file.
# Newline Characters breaking formatting
**Description**:
As you can see in the image below, the formatting in the documentation
is broken. As far as I can see the two added `\n` characters are
breaking the documentation. Therefore I would propose to remove those
**Dependencies**:
None
**Twitter Handle**
- epu9byj
---------
Co-authored-by: gere <gere@kapo.zh.ch>
Co-authored-by: Chester Curme <chester.curme@gmail.com>
**PR message**:
- **Description:** Corrected a syntax error in the code comments within
the `create_tool_calling_agent` function in the langchain package.
- **Issue:** N/A
- **Dependencies:** No additional dependencies required.
- **Twitter handle:** N/A
This PR fixes#21196.
The error was occurring when calling chat completion API with a chat
history. Indeed, the Mistral API does not accept both `content` and
`tool_calls` in the same body.
This PR removes one of theses variables depending on the necessity.
---------
Co-authored-by: Maxime Perrin <mperrin@doing.fr>
Co-authored-by: Chester Curme <chester.curme@gmail.com>
* Introduce individual `fetch_` methods for easier typing.
* Rework some docstrings to google style
* Move some logic to the tool
* Merge the 2 cassandra utility files
Description:Added documentation on Anthropic models on Vertex
@lkuligin for review
---------
Co-authored-by: adityarane@google.com <adityarane@google.com>
- support two-tuples of any sequence type (eg. json.loads never produces
tuples)
- support type alias for role key
- if id is passed in in dict form use it
- if tool_calls passed in in dict form use them
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
Refactors the docs build in order to:
- run the same `make build` command in both vercel and local build
- incrementally build artifacts in 2 distinct steps, instead of building
all docs in-place (in vercel) or in a _dist dir (locally)
Highlights:
- introduces `make build` in order to build the docs
- collects and generates all files for the build in
`docs/build/intermediate`
- renders those jupyter notebook + markdown files into
`docs/build/outputs`
And now the outputs to host are in `docs/build/outputs`, which will need
a vercel settings change.
Todo:
- [ ] figure out how to point the right directory (right now deleting
and moving docs dir in vercel_build.sh isn't great)
