- **Description:** Updated to remove deprecated parameter penalty_alpha,
and use string variation of prompt rather than json object for better
flexibility. - **Issue:** the issue # it fixes (if applicable),
- **Dependencies:** N/A
- **Tag maintainer:** @eyurtsev
- **Twitter handle:** @symbldotai
---------
Co-authored-by: toshishjawale <toshish@symbl.ai>
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
Instead of using JSON-like syntax to describe node and relationship
properties we changed to a shorter and more concise schema description
Old:
```
Node properties are the following:
[{'properties': [{'property': 'name', 'type': 'STRING'}], 'labels': 'Movie'}, {'properties': [{'property': 'name', 'type': 'STRING'}], 'labels': 'Actor'}]
Relationship properties are the following:
[]
The relationships are the following:
['(:Actor)-[:ACTED_IN]->(:Movie)']
```
New:
```
Node properties are the following:
Movie {name: STRING},Actor {name: STRING}
Relationship properties are the following:
The relationships are the following:
(:Actor)-[:ACTED_IN]->(:Movie)
```
Implements
[#12115](https://github.com/langchain-ai/langchain/issues/12115)
Who can review?
@baskaryan , @eyurtsev , @hwchase17
Integrated Stack Exchange API into Langchain, enabling access to diverse
communities within the platform. This addition enhances Langchain's
capabilities by allowing users to query Stack Exchange for specialized
information and engage in discussions. The integration provides seamless
interaction with Stack Exchange content, offering content from varied
knowledge repositories.
A notebook example and test cases were included to demonstrate the
functionality and reliability of this integration.
- Add StackExchange as a tool.
- Add unit test for the StackExchange wrapper and tool.
- Add documentation for the StackExchange wrapper and tool.
If you have time, could you please review the code and provide any
feedback as necessary! My team is welcome to any suggestions.
---------
Co-authored-by: Yuval Kamani <yuvalkamani@gmail.com>
Co-authored-by: Aryan Thakur <aryanthakur@Aryans-MacBook-Pro.local>
Co-authored-by: Manas1818 <79381912+manas1818@users.noreply.github.com>
Co-authored-by: aryan-thakur <61063777+aryan-thakur@users.noreply.github.com>
Co-authored-by: Bagatur <baskaryan@gmail.com>
- **Description:** The class allows to only select between a few
predefined prompts from the paper. That is not ideal, since other use
cases might need a custom prompt. The changes made allow for this. To be
able to monitor those, I also added functionality to supply a custom
run_manager.
- **Issue:** no issue, but a new feature,
- **Dependencies:** none,
- **Tag maintainer:** @hwchase17,
- **Twitter handle:** @yvesloy
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
- **Description:** Support providing whatever extra parameters you want
to the Mathpix PDF loader API request.
- **Issue:** #12773
- **Dependencies:** None
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
- **Description:** Adds a tqdm progress bar to GooglePalmEmbeddings when
embedding a list.
- **Issue:** #13637
- **Dependencies:** TQDM as a main dependency (instead of extra)
Signed-off-by: ugm2 <unaigaraymaestre@gmail.com>
---------
Signed-off-by: ugm2 <unaigaraymaestre@gmail.com>
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
This PR is fixing an attributeError: object endpoint has no attribute
"_public_match_client" when using gcp matching engine with private VPC
network.
@baskaryan, @eyurtsev, @hwchase17.
---------
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
- **Description:** As of OpenAI's Python package 1.0, the existing
DallEAPIWrapper does not work correctly, so the example in the LangChain
Documentation link below does not work either.
https://python.langchain.com/docs/integrations/tools/dalle_image_generator
Also, since OpenAI only supports DALL-E version 2 or version 3, I
modified the DallEAPIWrapper to support it.
- **Issue:** #13825
- **Twitter handle:** ggeutzzang
- **Description:** According to the document
https://cloud.baidu.com/doc/WENXINWORKSHOP/s/6lp69is2a, add ERNIE-Bot-8K
model support for ErnieBotChat.
- **Dependencies:** Before using the ERNIE-Bot-8K, you should have the
model's access authority.
Replace this entire comment with:
- **Description:** updates `create_llm_result` function within
`openai.py` to consider latest `params`,
- **Issue:** #8928
- **Dependencies:** -,
- **Tag maintainer:** -
- **Twitter handle:** [burkomr](https://twitter.com/burkomr)
<!-- If no one reviews your PR within a few days, please @-mention one
of @baskaryan, @eyurtsev, @hwchase17. -->
---------
Co-authored-by: Burak Ömür <burakomur@retorio.com>
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
Replace this entire comment with:
- **Description:** VertexAI models are now GA, moved away from using
preview ones from the SDK
- **Issue:** #13606
---------
Co-authored-by: Nuno Campos <nuno@boringbits.io>
**Description:**
Repair Wikipedia document loader `load_max_docs` and improve test
coverage.
**Issue:**
The Wikipedia document loader was not respecting the `load_max_docs`
paramater (not reported) and would always return a maximum of 10
documents. This is because the API wrapper (in `utilities/wikipedia.py`)
wasn't passing `top_k_results` to the underlying [Wikipedia
library](https://wikipedia.readthedocs.io/en/latest/code.html#module-wikipedia).
By default this library returns 10 results.
The default number of results for the document loader has been reduced
from 100 to 25. This is because loading 100 results takes a very long
time and is an inconvenient default. It should possibly be 10.
In addition, the documentation for the loader reported that there was a
hard limit (300) on the number of documents returned. In actuality 300
is the maximum Wikipedia query character length set by the API wrapper.
Tests have been added for the document loader (previously missing) and
to test the correct numbers of documents are being returned by each
class, both by default, and when overridden. Also repaired is the
`assert_docs` test which has been updated to correctly test for the
default metadata (which includes `source` in recent releases).
**Dependencies:**
nil
**Tag maintainer:**
@leo-gan
**Twitter handle:**
@queenvictoria
### **Description:**
Previously `python_repl` was a built-in tool, but now it has been moved
to `langchain_experimental`.
When I use `load_tools` I get an error:
```python
In [1]: from langchain.agents import load_tools
In [2]: load_tools(["python_repl"])
---------------------------------------------------------------------------
ImportError Traceback (most recent call last)
Cell In[2], line 1
----> 1 load_tools(["python_repl"])
File ~/workspace/langchain/libs/langchain/langchain/agents/load_tools.py:530, in load_tools(tool_names, llm, callbacks, **kwargs)
528 tool_names.extend(requests_method_tools)
529 elif name in _BASE_TOOLS:
--> 530 tools.append(_BASE_TOOLS[name]())
531 elif name in _LLM_TOOLS:
532 if llm is None:
File ~/workspace/langchain/libs/langchain/langchain/agents/load_tools.py:84, in _get_python_repl()
83 def _get_python_repl() -> BaseTool:
---> 84 raise ImportError(
85 "This tool has been moved to langchain experiment. "
86 "This tool has access to a python REPL. "
87 "For best practices make sure to sandbox this tool. "
88 "Read https://github.com/langchain-ai/langchain/blob/master/SECURITY.md "
89 "To keep using this code as is, install langchain experimental and "
90 "update relevant imports replacing 'langchain' with 'langchain_experimental'"
91 )
ImportError: This tool has been moved to langchain experiment. This tool has access to a python REPL. For best practices make sure to sandbox this tool. Read https://github.com/langchain-ai/langchain/blob/master/SECURITY.md To keep using this code as is, install langchain experimental and update relevant imports replacing 'langchain' with 'langchain_experimental'
```
In this case, it will be very confusing. I think it is no longer a
built-in tool now, so it can be removed from `_BASE_TOOLS`
### **Issue:**
https://github.com/langchain-ai/langchain/issues/13858,
https://github.com/langchain-ai/langchain/issues/13859,
https://github.com/langchain-ai/langchain/issues/13856
### **Twitter handle:**
[lin_bob57617](https://twitter.com/lin_bob57617)
The `integrations/vectorstores/matchingengine.ipynb` example has the
"Google Vertex AI Vector Search" title. This place this Title in the
wrong order in the ToC (it is sorted by the file name).
- Renamed `integrations/vectorstores/matchingengine.ipynb` into
`integrations/vectorstores/google_vertex_ai_vector_search.ipynb`.
- Updated a correspondent comment in docstring
- Rerouted old URL to a new URL
---------
Co-authored-by: Erick Friis <erick@langchain.dev>
It was :
`from langchain.schema.prompts import BasePromptTemplate`
but because of the breaking change in the ns, it is now
`from langchain.schema.prompt_template import BasePromptTemplate`
This bug prevents building the API Reference for the langchain_experimental
There are the following main changes in this PR:
1. Rewrite of the DocugamiLoader to not do any XML parsing of the DGML
format internally, and instead use the `dgml-utils` library we are
separately working on. This is a very lightweight dependency.
2. Added MMR search type as an option to multi-vector retriever, similar
to other retrievers. MMR is especially useful when using Docugami for
RAG since we deal with large sets of documents within which a few might
be duplicates and straight similarity based search doesn't give great
results in many cases.
We are @docugami on twitter, and I am @tjaffri
---------
Co-authored-by: Taqi Jaffri <tjaffri@docugami.com>
- **Description:** Adds a retriever implementation for [Knowledge Bases
for Amazon Bedrock](https://aws.amazon.com/bedrock/knowledge-bases/), a
new service announced at AWS re:Invent, shortly before this PR was
opened. This depends on the `bedrock-agent-runtime` service, which will
be included in a future version of `boto3` and of `botocore`. We will
open a follow-up PR documenting the minimum required versions of `boto3`
and `botocore` after that information is available.
- **Issue:** N/A
- **Dependencies:** `boto3>=1.33.2, botocore>=1.33.2`
- **Tag maintainer:** @baskaryan
- **Twitter handles:** `@pjain7` `@dead_letter_q`
This PR includes a documentation notebook under
`docs/docs/integrations/retrievers`, which I (@dlqqq) have verified
independently.
EDIT: `bedrock-agent-runtime` service is now included in
`boto3>=1.33.2`:
5cf793f493
---------
Co-authored-by: Piyush Jain <piyushjain@duck.com>
Co-authored-by: Erick Friis <erick@langchain.dev>
Co-authored-by: Bagatur <baskaryan@gmail.com>
Addressing incorrect order being sent to callbacks / tracers, due to the
nature of threading
---------
Co-authored-by: Nuno Campos <nuno@boringbits.io>
Add arg to omit streamed_output list, in cases where final_output is
enough this saves bandwidth
<!-- Thank you for contributing to LangChain!
Replace this entire comment with:
- **Description:** a description of the change,
- **Issue:** the issue # it fixes (if applicable),
- **Dependencies:** any dependencies required for this change,
- **Tag maintainer:** for a quicker response, tag the relevant
maintainer (see below),
- **Twitter handle:** we announce bigger features on Twitter. If your PR
gets announced, and you'd like a mention, we'll gladly shout you out!
Please make sure your PR is passing linting and testing before
submitting. Run `make format`, `make lint` and `make test` to check this
locally.
See contribution guidelines for more information on how to write/run
tests, lint, etc:
https://github.com/langchain-ai/langchain/blob/master/.github/CONTRIBUTING.md
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/extras`
directory.
If no one reviews your PR within a few days, please @-mention one of
@baskaryan, @eyurtsev, @hwchase17.
-->
This PR rearranges the docstring for the `AstraDB` vector store class so
as to have all useful information in the _class_ docstring for ease of
reading.
(incidentally, due to an oversight, the docstring that was in the
constructor ended up buried below some lines of code, thereby
disappearing altogether from accessibility. Apologies.)
- **Description:** Updates to `AnthropicFunctions` to be compatible with
the OpenAI `function_call` functionality.
- **Issue:** The functionality to indicate `auto`, `none` and a forced
function_call was not completely implemented in the existing code.
- **Dependencies:** None
- **Tag maintainer:** @baskaryan , and any of the other maintainers if
needed.
- **Twitter handle:** None
I have specifically tested this functionality via AWS Bedrock with the
Claude-2 and Claude-Instant models.
<!-- Thank you for contributing to LangChain!
Replace this entire comment with:
- **Description:** a description of the change,
- **Issue:** the issue # it fixes (if applicable),
- **Dependencies:** any dependencies required for this change,
- **Tag maintainer:** for a quicker response, tag the relevant
maintainer (see below),
- **Twitter handle:** we announce bigger features on Twitter. If your PR
gets announced, and you'd like a mention, we'll gladly shout you out!
Please make sure your PR is passing linting and testing before
submitting. Run `make format`, `make lint` and `make test` to check this
locally.
See contribution guidelines for more information on how to write/run
tests, lint, etc:
https://github.com/langchain-ai/langchain/blob/master/.github/CONTRIBUTING.md
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/extras`
directory.
If no one reviews your PR within a few days, please @-mention one of
@baskaryan, @eyurtsev, @hwchase17.
-->
- **Description:** We are adding functionality to extract message
content from the `attributedBody` field of the database, in case the
content is not in the `text` field.
- **Issue:** Closes#13326 and #10680
- **Dependencies:** None.
- **Tag maintainer:** @eyurtsev, @hwchase17
---------
Co-authored-by: onotate <johnp.pham@mail.utoronto.ca>
- **Description:** Previously `MarkdownHeaderTextSplitter` did not
consider tilde-fenced code blocks
(https://spec.commonmark.org/0.30/#fenced-code-blocks). This PR fixes
that.
````md
# Bug caused by previous implementation:
~~~py
foo()
# This is a comment that would be considered header
bar()
~~~
````
- **Tag maintainer:** @baskaryan
Several bug fixes:
- emails: instead of `bcc` the `cc` is used.
- errors in the truncation descriptions
- no truncation of the `message_search`
Several updates:
- generalized UTC format
- truncation limit can be changed now in _call()
Fixes#13407.
This workaround consists in letting the RunnableLambda create its
self.afunc from its self.func when self.afunc is not provided; the
change has no dependency.
<!-- Thank you for contributing to LangChain!
Replace this entire comment with:
- **Description:** a description of the change,
- **Issue:** the issue # it fixes (if applicable),
- **Dependencies:** any dependencies required for this change,
- **Tag maintainer:** for a quicker response, tag the relevant
maintainer (see below),
- **Twitter handle:** we announce bigger features on Twitter. If your PR
gets announced, and you'd like a mention, we'll gladly shout you out!
Please make sure your PR is passing linting and testing before
submitting. Run `make format`, `make lint` and `make test` to check this
locally.
See contribution guidelines for more information on how to write/run
tests, lint, etc:
https://github.com/langchain-ai/langchain/blob/master/.github/CONTRIBUTING.md
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/extras`
directory.
If no one reviews your PR within a few days, please @-mention one of
@baskaryan, @eyurtsev, @hwchase17.
-->
---------
Co-authored-by: Erick Friis <erick@langchain.dev>
Co-authored-by: Nuno Campos <nuno@langchain.dev>
```
---- chunk 1
{'actions': [AgentActionMessageLog(tool='Search', tool_input="Leo DiCaprio's current girlfriend", log="\nInvoking: `Search` with `Leo DiCaprio's current girlfriend`\n\n\n", message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}})])],
'messages': [AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}})]}
---- chunk 2
{'messages': [FunctionMessage(content="According to Us, the 48-year-old actor is now “exclusively” dating Italian model Vittoria Ceretti. A source told Us that DiCaprio is “completely smitten” with Ceretti, and their relationship is “going so well that Leo's actually being exclusive.”", name='Search')],
'steps': [AgentStep(action=AgentActionMessageLog(tool='Search', tool_input="Leo DiCaprio's current girlfriend", log="\nInvoking: `Search` with `Leo DiCaprio's current girlfriend`\n\n\n", message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}})]), observation="According to Us, the 48-year-old actor is now “exclusively” dating Italian model Vittoria Ceretti. A source told Us that DiCaprio is “completely smitten” with Ceretti, and their relationship is “going so well that Leo's actually being exclusive.”")]}
---- chunk 3
{'actions': [AgentActionMessageLog(tool='Search', tool_input='Vittoria Ceretti age', log='\nInvoking: `Search` with `Vittoria Ceretti age`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}})])],
'messages': [AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}})]}
---- chunk 4
{'messages': [FunctionMessage(content='25 years', name='Search')],
'steps': [AgentStep(action=AgentActionMessageLog(tool='Search', tool_input='Vittoria Ceretti age', log='\nInvoking: `Search` with `Vittoria Ceretti age`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}})]), observation='25 years')]}
---- chunk 5
{'actions': [AgentActionMessageLog(tool='Calculator', tool_input='25^0.43', log='\nInvoking: `Calculator` with `25^0.43`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}})])],
'messages': [AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}})]}
---- chunk 6
{'messages': [FunctionMessage(content='Answer: 3.991298452658078', name='Calculator')],
'steps': [AgentStep(action=AgentActionMessageLog(tool='Calculator', tool_input='25^0.43', log='\nInvoking: `Calculator` with `25^0.43`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}})]), observation='Answer: 3.991298452658078')]}
---- chunk 7
{'messages': [AIMessage(content="Leonardo DiCaprio's current girlfriend is the Italian model Vittoria Ceretti, who is 25 years old. Her age raised to the 0.43 power is approximately 3.99.")],
'output': "Leonardo DiCaprio's current girlfriend is the Italian model "
'Vittoria Ceretti, who is 25 years old. Her age raised to the 0.43 '
'power is approximately 3.99.'}
---- final
{'actions': [AgentActionMessageLog(tool='Search', tool_input="Leo DiCaprio's current girlfriend", log="\nInvoking: `Search` with `Leo DiCaprio's current girlfriend`\n\n\n", message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}})]),
AgentActionMessageLog(tool='Search', tool_input='Vittoria Ceretti age', log='\nInvoking: `Search` with `Vittoria Ceretti age`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}})]),
AgentActionMessageLog(tool='Calculator', tool_input='25^0.43', log='\nInvoking: `Calculator` with `25^0.43`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}})])],
'messages': [AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}}),
FunctionMessage(content="According to Us, the 48-year-old actor is now “exclusively” dating Italian model Vittoria Ceretti. A source told Us that DiCaprio is “completely smitten” with Ceretti, and their relationship is “going so well that Leo's actually being exclusive.”", name='Search'),
AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}}),
FunctionMessage(content='25 years', name='Search'),
AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}}),
FunctionMessage(content='Answer: 3.991298452658078', name='Calculator'),
AIMessage(content="Leonardo DiCaprio's current girlfriend is the Italian model Vittoria Ceretti, who is 25 years old. Her age raised to the 0.43 power is approximately 3.99.")],
'output': "Leonardo DiCaprio's current girlfriend is the Italian model "
'Vittoria Ceretti, who is 25 years old. Her age raised to the 0.43 '
'power is approximately 3.99.',
'steps': [AgentStep(action=AgentActionMessageLog(tool='Search', tool_input="Leo DiCaprio's current girlfriend", log="\nInvoking: `Search` with `Leo DiCaprio's current girlfriend`\n\n\n", message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Leo DiCaprio\'s current girlfriend"\n}'}})]), observation="According to Us, the 48-year-old actor is now “exclusively” dating Italian model Vittoria Ceretti. A source told Us that DiCaprio is “completely smitten” with Ceretti, and their relationship is “going so well that Leo's actually being exclusive.”"),
AgentStep(action=AgentActionMessageLog(tool='Search', tool_input='Vittoria Ceretti age', log='\nInvoking: `Search` with `Vittoria Ceretti age`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Search', 'arguments': '{\n "__arg1": "Vittoria Ceretti age"\n}'}})]), observation='25 years'),
AgentStep(action=AgentActionMessageLog(tool='Calculator', tool_input='25^0.43', log='\nInvoking: `Calculator` with `25^0.43`\n\n\n', message_log=[AIMessageChunk(content='', additional_kwargs={'function_call': {'name': 'Calculator', 'arguments': '{\n "__arg1": "25^0.43"\n}'}})]), observation='Answer: 3.991298452658078')]}
```
- **Description:** Existing model used for Prompt Injection is quite
outdated but we fine-tuned and open-source a new model based on the same
model deberta-v3-base from Microsoft -
[laiyer/deberta-v3-base-prompt-injection](https://huggingface.co/laiyer/deberta-v3-base-prompt-injection).
It supports more up-to-date injections and less prone to
false-positives.
- **Dependencies:** No
- **Tag maintainer:** -
- **Twitter handle:** @alex_yaremchuk
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
