Some custom agents might continue to iterate until they find the correct
answer, getting stuck on loops that generate request after request and
are really expensive for the end user. Putting an upper bound for the
number of iterations
by default controls this and can be explicitly tweaked by the user if
necessary.
Co-authored-by: Francisco Ingham <>
Referring to #687, I implemented the functionality to reduce K if it
exceeds the token limit.
Edit: I should have ran make lint locally. Also, this only applies to
`StuffDocumentChain`
* add implementations of `BaseCallbackHandler` to support tracing:
`SharedTracer` which is thread-safe and `Tracer` which is not and is
meant to be used locally.
* Tracers persist runs to locally running `langchain-server`
Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
If `distance_func` and `collection_name` are in `kwargs` they are sent
to the `QdrantClient` which results in an error being raised.
Co-authored-by: Francisco Ingham <>
`SentenceTransformer` returns a NumPy array, not a `List[List[float]]`
or `List[float]` as specified in the interface of `Embeddings`. That PR
makes it consistent with the interface.
I'm providing a hotfix for Qdrant integration. Calculating a single
embedding to obtain the vector size was great idea. However, that change
introduced a bug trying to put only that single embedding into the
database. It's fixed. Right now all the embeddings will be pushed to
Qdrant.
Now that OpenAI has deprecated all embeddings models except
text-embedding-ada-002, we should stop specifying a legacy embedding
model in the example. This will also avoid confusion from people (like
me) trying to specify model="text-embedding-ada-002" and having that
erroneously expanded to text-search-text-embedding-ada-002-query-001
Since the tokenizer and model are constructed manually, model_kwargs
needs to
be passed to their constructors. Additionally, the pipeline has a
specific
named parameter to pass these with, which can provide forward
compatibility if
they are used for something other than tokenizer or model construction.
On the [Getting Started
page](https://langchain.readthedocs.io/en/latest/modules/prompts/getting_started.html)
for prompt templates, I believe the very last example
```python
print(dynamic_prompt.format(adjective=long_string))
```
should actually be
```python
print(dynamic_prompt.format(input=long_string))
```
The existing example produces `KeyError: 'input'` as expected
***
On the [Create a custom prompt
template](https://langchain.readthedocs.io/en/latest/modules/prompts/examples/custom_prompt_template.html#id1)
page, I believe the line
```python
Function Name: {kwargs["function_name"]}
```
should actually be
```python
Function Name: {kwargs["function_name"].__name__}
```
The existing example produces the prompt:
```
Given the function name and source code, generate an English language explanation of the function.
Function Name: <function get_source_code at 0x7f907bc0e0e0>
Source Code:
def get_source_code(function_name):
# Get the source code of the function
return inspect.getsource(function_name)
Explanation:
```
***
On the [Example
Selectors](https://langchain.readthedocs.io/en/latest/modules/prompts/examples/example_selectors.html)
page, the first example does not define `example_prompt`, which is also
subtly different from previous example prompts used. For user
convenience, I suggest including
```python
example_prompt = PromptTemplate(
input_variables=["input", "output"],
template="Input: {input}\nOutput: {output}",
)
```
in the code to be copy-pasted
- This uses the faiss built-in `write_index` and `load_index` to save
and load faiss indexes locally
- Also fixes#674
- The save/load functions also use the faiss library, so I refactored
the dependency into a function
