Work in Progress.
WIP
Not ready...
Adds Document Loader support for
[Geopandas.GeoDataFrames](https://geopandas.org/)
Example:
- [x] stub out `GeoDataFrameLoader` class
- [x] stub out integration tests
- [ ] Experiment with different geometry text representations
- [ ] Verify CRS is successfully added in metadata
- [ ] Test effectiveness of searches on geometries
- [ ] Test with different geometry types (point, line, polygon with
multi-variants).
- [ ] Add documentation
---------
Co-authored-by: Lance Martin <lance@langchain.dev>
Co-authored-by: Bagatur <baskaryan@gmail.com>
Co-authored-by: Lance Martin <122662504+rlancemartin@users.noreply.github.com>
Removing **kwargs argument from add_texts method in DeepLake vectorstore
as it confuses users and doesn't fail when user is typing incorrect
parameters.
Also added small test to ensure the change is applies correctly.
Guys could pls take a look: @rlancemartin, @eyurtsev, this is a small
PR.
Thx so much!
Description: This PR adds the option to retrieve scores and explanations
in the WeaviateHybridSearchRetriever. This feature improves the
usability of the retriever by allowing users to understand the scoring
logic behind the search results and further refine their search queries.
Issue: This PR is a solution to the issue #7855
Dependencies: This PR does not introduce any new dependencies.
Tag maintainer: @rlancemartin, @eyurtsev
I have included a unit test for the added feature, ensuring that it
retrieves scores and explanations correctly. I have also included an
example notebook demonstrating its use.
Motivation, it seems that when dealing with a long context and "big"
number of relevant documents we must avoid using out of the box score
ordering from vector stores.
See: https://arxiv.org/pdf/2306.01150.pdf
So, I added an additional parameter that allows you to reorder the
retrieved documents so we can work around this performance degradation.
The relevance respect the original search score but accommodates the
lest relevant document in the middle of the context.
Extract from the paper (one image speaks 1000 tokens):
This seems to be common to all diff arquitectures. SO I think we need a
good generic way to implement this reordering and run some test in our
already running retrievers.
It could be that my approach is not the best one from the architecture
point of view, happy to have a discussion about that.
For me this was the best place to introduce the change and start
retesting diff implementations.
@rlancemartin, @eyurtsev
---------
Co-authored-by: Lance Martin <lance@langchain.dev>
Some docstring / small nits to #6003
---------
Co-authored-by: BoazWasserman <49598618+boazwasserman@users.noreply.github.com>
Co-authored-by: HippoTerrific <49598618+HippoTerrific@users.noreply.github.com>
Co-authored-by: Or Raz <orraz1994@gmail.com>
- Description: Add a BM25 Retriever that do not need Elastic search
- Dependencies: rank_bm25(if it is not installed it will be install by
using pip, just like TFIDFRetriever do)
- Tag maintainer: @rlancemartin, @eyurtsev
- Twitter handle: DayuanJian21687
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
Description:
Add LLM for ChatGLM-6B & ChatGLM2-6B API
Related Issue:
Will the langchain support ChatGLM? #4766
Add support for selfhost models like ChatGLM or transformer models #1780
Dependencies:
No extra library install required.
It wraps api call to a ChatGLM(2)-6B server(start with api.py), so api
endpoint is required to run.
Tag maintainer: @mlot
Any comments on this PR would be appreciated.
---------
Co-authored-by: mlot <limpo2000@gmail.com>
Co-authored-by: Bagatur <baskaryan@gmail.com>
- [Xorbits](https://doc.xorbits.io/en/latest/) is an open-source
computing framework that makes it easy to scale data science and machine
learning workloads in parallel. Xorbits can leverage multi cores or GPUs
to accelerate computation on a single machine, or scale out up to
thousands of machines to support processing terabytes of data.
- This PR added support for the Xorbits agent, which allows langchain to
interact with Xorbits Pandas dataframe and Xorbits Numpy array.
- Dependencies: This change requires the Xorbits library to be installed
in order to be used.
`pip install xorbits`
- Request for review: @hinthornw
- Twitter handle: https://twitter.com/Xorbitsio
Starting over from #5654 because I utterly borked the poetry.lock file.
Adds new paramerters for to the MWDumpLoader class:
* skip_redirecst (bool) Tells the loader to skip articles that redirect
to other articles. False by default.
* stop_on_error (bool) Tells the parser to skip any page that causes a
parse error. True by default.
* namespaces (List[int]) Tells the parser which namespaces to parse.
Contains namespaces from -2 to 15 by default.
Default values are chosen to preserve backwards compatibility.
Sample dump XML and full unit test coverage (with extended tests that
pass!) also included!
---------
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
Co-authored-by: Bagatur <baskaryan@gmail.com>
Inspired by #5550, I implemented full async API support in Qdrant. The
docs were extended to mention the existence of asynchronous operations
in Langchain. I also used that chance to restructure the tests of Qdrant
and provided a suite of tests for the async version. Async API requires
the GRPC protocol to be enabled. Thus, it doesn't work on local mode
yet, but we're considering including the support to be consistent.
<!-- Thank you for contributing to LangChain!
Replace this 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!
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.
Maintainer responsibilities:
- General / Misc / if you don't know who to tag: @baskaryan
- DataLoaders / VectorStores / Retrievers: @rlancemartin, @eyurtsev
- Models / Prompts: @hwchase17, @baskaryan
- Memory: @hwchase17
- Agents / Tools / Toolkits: @hinthornw
- Tracing / Callbacks: @agola11
- Async: @agola11
If no one reviews your PR within a few days, feel free to @-mention the
same people again.
See contribution guidelines for more information on how to write/run
tests, lint, etc:
https://github.com/hwchase17/langchain/blob/master/.github/CONTRIBUTING.md
-->
Integrate [Rockset](https://rockset.com/docs/) as a document loader.
Issue: None
Dependencies: Nothing new (rockset's dependency was already added
[here](https://github.com/hwchase17/langchain/pull/6216))
Tag maintainer: @rlancemartin
I have added a test for the integration and an example notebook showing
its use. I ran `make lint` and everything looks good.
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
- Add langchain.llms.Tonyi for text completion, in examples into the
Tonyi Text API,
- Add system tests.
Note async completion for the Text API is not yet supported and will be
included in a future PR.
Dependencies: dashscope. It will be installed manually cause it is not
need by everyone.
Happy for feedback on any aspect of this PR @hwchase17 @baskaryan.
Multiple people have asked in #5081 for a way to limit the documents
returned from an AzureCognitiveSearchRetriever. This PR adds the `top_n`
parameter to allow that.
Twitter handle:
[@UmerHAdil](twitter.com/umerHAdil)
<!-- Thank you for contributing to LangChain!
Replace this comment with:
- Description:
- 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!
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.
Maintainer responsibilities:
- General / Misc / if you don't know who to tag: @baskaryan
- DataLoaders / VectorStores / Retrievers: @rlancemartin, @eyurtsev
- Models / Prompts: @hwchase17, @baskaryan
- Memory: @hwchase17
- Agents / Tools / Toolkits: @hinthornw
- Tracing / Callbacks: @agola11
- Async: @agola11
If no one reviews your PR within a few days, feel free to @-mention the
same people again.
See contribution guidelines for more information on how to write/run
tests, lint, etc:
https://github.com/hwchase17/langchain/blob/master/.github/CONTRIBUTING.md
-->
## Description
This PR addresses a bug in the RecursiveUrlLoader class where absolute
URLs were being treated as relative URLs, causing malformed URLs to be
produced. The fix involves using the urljoin function from the
urllib.parse module to correctly handle both absolute and relative URLs.
@rlancemartin @eyurtsev
---------
Co-authored-by: Lance Martin <lance@langchain.dev>
Fixes # (issue)
The existing PlaywrightURLLoader load() function uses a synchronous
browser which is not compatible with jupyter.
This PR adds a sister function aload() which can be run insisde a
notebook.
---------
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
- Migrate from deprecated langchainplus_sdk to `langsmith` package
- Update the `run_on_dataset()` API to use an eval config
- Update a number of evaluators, as well as the loading logic
- Update docstrings / reference docs
- Update tracer to share single HTTP session
Sometimes the score responded by chatgpt would be like 'Respone
example\nScore: 90 (fully answers the question, but could provide more
detail on the specific error message)'
For the score contains not only numbers, it raise a ValueError like
Update the RegexParser from `.*` to `\d*` would help us to ignore the
text after number.
Co-authored-by: Bagatur <baskaryan@gmail.com>
Fixed#6768.
This is a workaround only. I think a better longer-term solution is for
chains to declare how many input variables they *actually* need (as
opposed to ones that are in the prompt, where some may be satisfied by
the memory). Then, a wrapping chain can check the input match against
the actual input variables.
@hwchase17
- Description: Add two new document transformers that translates
documents into different languages and converts documents into q&a
format to improve vector search results. Uses OpenAI function calling
via the [doctran](https://github.com/psychic-api/doctran/tree/main)
library.
- Issue: N/A
- Dependencies: `doctran = "^0.0.5"`
- Tag maintainer: @rlancemartin @eyurtsev @hwchase17
- Twitter handle: @psychicapi or @jfan001
Notes
- Adheres to the `DocumentTransformer` abstraction set by @dev2049 in
#3182
- refactored `EmbeddingsRedundantFilter` to put it in a file under a new
`document_transformers` module
- Added basic docs for `DocumentInterrogator`, `DocumentTransformer` as
well as the existing `EmbeddingsRedundantFilter`
---------
Co-authored-by: Lance Martin <lance@langchain.dev>
Co-authored-by: Bagatur <baskaryan@gmail.com>
Probably the most boring PR to review ;)
Individual commits might be easier to digest
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
Co-authored-by: Bagatur <22008038+baskaryan@users.noreply.github.com>
- Description: Adds a new chain that acts as a wrapper around Sympy to
give LLMs the ability to do some symbolic math.
- Dependencies: SymPy
---------
Co-authored-by: sreiswig <sreiswig@github.com>
Co-authored-by: Bagatur <baskaryan@gmail.com>
When using callbacks, there are times when callbacks can be added
redundantly: for instance sometimes you might need to create an llm with
specific callbacks, but then also create and agent that uses a chain
that has those callbacks already set. This means that "callbacks" might
get passed down again to the llm at predict() time, resulting in
duplicate calls to the `on_llm_start` callback.
For the sake of simplicity, I made it so that langchain never adds an
exact handler/callbacks object in `add_handler`, thus avoiding the
duplicate handler issue.
Tagging @hwchase17 for callback review
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
Currently `ChatOutputParser` extracts actions by splitting the text on
"```", and then load the second part as a json string.
But sometimes the LLM will wrap the action in markdown code block like:
````markdown
```json
{
"action": "foo",
"action_input": "bar"
}
```
````
Splitting text on "```" will cause `OutputParserException` in such case.
This PR changes the behaviour to extract the `$JSON_BLOB` by regex, so
that it can handle both ` ``` ``` ` and ` ```json ``` `
@hinthornw
---------
Co-authored-by: Junlin Zhou <jlzhou@zjuici.com>
This PR changes the behavior of `Qdrant.from_texts` so the collection is
reused if not requested to recreate it. Previously, calling
`Qdrant.from_texts` or `Qdrant.from_documents` resulted in removing the
old data which was confusing for many.
# Causal program-aided language (CPAL) chain
## Motivation
This builds on the recent [PAL](https://arxiv.org/abs/2211.10435) to
stop LLM hallucination. The problem with the
[PAL](https://arxiv.org/abs/2211.10435) approach is that it hallucinates
on a math problem with a nested chain of dependence. The innovation here
is that this new CPAL approach includes causal structure to fix
hallucination.
For example, using the below word problem, PAL answers with 5, and CPAL
answers with 13.
"Tim buys the same number of pets as Cindy and Boris."
"Cindy buys the same number of pets as Bill plus Bob."
"Boris buys the same number of pets as Ben plus Beth."
"Bill buys the same number of pets as Obama."
"Bob buys the same number of pets as Obama."
"Ben buys the same number of pets as Obama."
"Beth buys the same number of pets as Obama."
"If Obama buys one pet, how many pets total does everyone buy?"
The CPAL chain represents the causal structure of the above narrative as
a causal graph or DAG, which it can also plot, as shown below.
.
The two major sections below are:
1. Technical overview
2. Future application
Also see [this jupyter
notebook](https://github.com/borisdev/langchain/blob/master/docs/extras/modules/chains/additional/cpal.ipynb)
doc.
## 1. Technical overview
### CPAL versus PAL
Like [PAL](https://arxiv.org/abs/2211.10435), CPAL intends to reduce
large language model (LLM) hallucination.
The CPAL chain is different from the PAL chain for a couple of reasons.
* CPAL adds a causal structure (or DAG) to link entity actions (or math
expressions).
* The CPAL math expressions are modeling a chain of cause and effect
relations, which can be intervened upon, whereas for the PAL chain math
expressions are projected math identities.
PAL's generated python code is wrong. It hallucinates when complexity
increases.
```python
def solution():
"""Tim buys the same number of pets as Cindy and Boris.Cindy buys the same number of pets as Bill plus Bob.Boris buys the same number of pets as Ben plus Beth.Bill buys the same number of pets as Obama.Bob buys the same number of pets as Obama.Ben buys the same number of pets as Obama.Beth buys the same number of pets as Obama.If Obama buys one pet, how many pets total does everyone buy?"""
obama_pets = 1
tim_pets = obama_pets
cindy_pets = obama_pets + obama_pets
boris_pets = obama_pets + obama_pets
total_pets = tim_pets + cindy_pets + boris_pets
result = total_pets
return result # math result is 5
```
CPAL's generated python code is correct.
```python
story outcome data
name code value depends_on
0 obama pass 1.0 []
1 bill bill.value = obama.value 1.0 [obama]
2 bob bob.value = obama.value 1.0 [obama]
3 ben ben.value = obama.value 1.0 [obama]
4 beth beth.value = obama.value 1.0 [obama]
5 cindy cindy.value = bill.value + bob.value 2.0 [bill, bob]
6 boris boris.value = ben.value + beth.value 2.0 [ben, beth]
7 tim tim.value = cindy.value + boris.value 4.0 [cindy, boris]
query data
{
"question": "how many pets total does everyone buy?",
"expression": "SELECT SUM(value) FROM df",
"llm_error_msg": ""
}
# query result is 13
```
Based on the comments below, CPAL's intended location in the library is
`experimental/chains/cpal` and PAL's location is`chains/pal`.
### CPAL vs Graph QA
Both the CPAL chain and the Graph QA chain extract entity-action-entity
relations into a DAG.
The CPAL chain is different from the Graph QA chain for a few reasons.
* Graph QA does not connect entities to math expressions
* Graph QA does not associate actions in a sequence of dependence.
* Graph QA does not decompose the narrative into these three parts:
1. Story plot or causal model
4. Hypothetical question
5. Hypothetical condition
### Evaluation
Preliminary evaluation on simple math word problems shows that this CPAL
chain generates less hallucination than the PAL chain on answering
questions about a causal narrative. Two examples are in [this jupyter
notebook](https://github.com/borisdev/langchain/blob/master/docs/extras/modules/chains/additional/cpal.ipynb)
doc.
## 2. Future application
### "Describe as Narrative, Test as Code"
The thesis here is that the Describe as Narrative, Test as Code approach
allows you to represent a causal mental model both as code and as a
narrative, giving you the best of both worlds.
#### Why describe a causal mental mode as a narrative?
The narrative form is quick. At a consensus building meeting, people use
narratives to persuade others of their causal mental model, aka. plan.
You can share, version control and index a narrative.
#### Why test a causal mental model as a code?
Code is testable, complex narratives are not. Though fast, narratives
are problematic as their complexity increases. The problem is LLMs and
humans are prone to hallucination when predicting the outcomes of a
narrative. The cost of building a consensus around the validity of a
narrative outcome grows as its narrative complexity increases. Code does
not require tribal knowledge or social power to validate.
Code is composable, complex narratives are not. The answer of one CPAL
chain can be the hypothetical conditions of another CPAL Chain. For
stochastic simulations, a composable plan can be integrated with the
[DoWhy library](https://github.com/py-why/dowhy). Lastly, for the
futuristic folk, a composable plan as code allows ordinary community
folk to design a plan that can be integrated with a blockchain for
funding.
An explanation of a dependency planning application is
[here.](https://github.com/borisdev/cpal-llm-chain-demo)
---
Twitter handle: @boris_dev
---------
Co-authored-by: Boris Dev <borisdev@Boriss-MacBook-Air.local>
Description: Current `_call` function in the
`langchain.llms.HuggingFaceEndpoint` class truncates response when
`task=text-generation`. Same error discussed a few days ago on Hugging
Face: https://huggingface.co/tiiuae/falcon-40b-instruct/discussions/51
Issue: Fixes#7353
Tag maintainer: @hwchase17 @baskaryan @hinthornw
---------
Co-authored-by: Bagatur <baskaryan@gmail.com>
Description: This pull request aims to support generating the correct
generic relevancy scores for different vector stores by refactoring the
relevance score functions and their selection in the base class and
subclasses of VectorStore. This is especially relevant with VectorStores
that require a distance metric upon initialization. Note many of the
current implenetations of `_similarity_search_with_relevance_scores` are
not technically correct, as they just return
`self.similarity_search_with_score(query, k, **kwargs)` without applying
the relevant score function
Also includes changes associated with:
https://github.com/hwchase17/langchain/pull/6564 and
https://github.com/hwchase17/langchain/pull/6494
See more indepth discussion in thread in #6494
Issue:
https://github.com/hwchase17/langchain/issues/6526https://github.com/hwchase17/langchain/issues/6481https://github.com/hwchase17/langchain/issues/6346
Dependencies: None
The changes include:
- Properly handling score thresholding in FAISS
`similarity_search_with_score_by_vector` for the corresponding distance
metric.
- Refactoring the `_similarity_search_with_relevance_scores` method in
the base class and removing it from the subclasses for incorrectly
implemented subclasses.
- Adding a `_select_relevance_score_fn` method in the base class and
implementing it in the subclasses to select the appropriate relevance
score function based on the distance strategy.
- Updating the `__init__` methods of the subclasses to set the
`relevance_score_fn` attribute.
- Removing the `_default_relevance_score_fn` function from the FAISS
class and using the base class's `_euclidean_relevance_score_fn`
instead.
- Adding the `DistanceStrategy` enum to the `utils.py` file and updating
the imports in the vector store classes.
- Updating the tests to import the `DistanceStrategy` enum from the
`utils.py` file.
---------
Co-authored-by: Hanit <37485638+hanit-com@users.noreply.github.com>
<!--
Thank you for contributing to LangChain! Your PR will appear in our
release under the title you set. Please make sure it highlights your
valuable contribution.
Replace this with a description of the change, the issue it fixes (if
applicable), and relevant context. List any dependencies required for
this change.
After you're done, someone will review your PR. They may suggest
improvements. If no one reviews your PR within a few days, feel free to
@-mention the same people again, as notifications can get lost.
Finally, we'd love to show appreciation for your contribution - if you'd
like us to shout you out on Twitter, please also include your handle!
-->
<!-- Remove if not applicable -->
Fixes # (issue)
#### Before submitting
<!-- If you're adding a new integration, please include:
1. a test for the integration - favor unit tests that does not rely on
network access.
2. an example notebook showing its use
See contribution guidelines for more information on how to write tests,
lint
etc:
https://github.com/hwchase17/langchain/blob/master/.github/CONTRIBUTING.md
-->
#### Who can review?
Tag maintainers/contributors who might be interested:
<!-- For a quicker response, figure out the right person to tag with @
@hwchase17 - project lead
Tracing / Callbacks
- @agola11
Async
- @agola11
DataLoaders
- @eyurtsev
Models
- @hwchase17
- @agola11
Agents / Tools / Toolkits
- @hwchase17
VectorStores / Retrievers / Memory
- @dev2049
-->
1. Added use cases of the new features
2. Done some code refactoring
---------
Co-authored-by: Ivo Stranic <istranic@gmail.com>
- [Xorbits](https://doc.xorbits.io/en/latest/) is an open-source
computing framework that makes it easy to scale data science and machine
learning workloads in parallel. Xorbits can leverage multi cores or GPUs
to accelerate computation on a single machine, or scale out up to
thousands of machines to support processing terabytes of data.
- This PR added support for the Xorbits document loader, which allows
langchain to leverage Xorbits to parallelize and distribute the loading
of data.
- Dependencies: This change requires the Xorbits library to be installed
in order to be used.
`pip install xorbits`
- Request for review: @rlancemartin, @eyurtsev
- Twitter handle: https://twitter.com/Xorbitsio
Co-authored-by: Bagatur <baskaryan@gmail.com>
- Description: Adding async method for CTransformers
- Issue: I've found impossible without this code to run Websockets
inside a FastAPI micro service and a CTransformers model.
- Tag maintainer: Not necessary yet, I don't like to mention directly
- Twitter handle: @_semoal
Adding a maximal_marginal_relevance method to the
MongoDBAtlasVectorSearch vectorstore enhances the user experience by
providing more diverse search results
Issue: #7304
### Summary
Adds an `UnstructuredTSVLoader` for TSV files. Also updates the doc
strings for `UnstructuredCSV` and `UnstructuredExcel` loaders.
### Testing
```python
from langchain.document_loaders.tsv import UnstructuredTSVLoader
loader = UnstructuredTSVLoader(
file_path="example_data/mlb_teams_2012.csv", mode="elements"
)
docs = loader.load()
```
`SpacyTextSplitter` currently uses spacy's statistics-based
`en_core_web_sm` model for sentence splitting. This is a good splitter,
but it's also pretty slow, and in this case it's doing a lot of work
that's not needed given that the spacy parse is then just thrown away.
However, there is also a simple rules-based spacy sentencizer. Using
this is at least an order of magnitude faster than using
`en_core_web_sm` according to my local tests.
Also, spacy sentence tokenization based on `en_core_web_sm` can be sped
up in this case by not doing the NER stage. This shaves some cycles too,
both when loading the model and when parsing the text.
Consequently, this PR adds the option to use the basic spacy
sentencizer, and it disables the NER stage for the current approach,
*which is kept as the default*.
Lastly, when extracting the tokenized sentences, the `text` attribute is
called directly instead of doing the string conversion, which is IMO a
bit more idiomatic.
