Skip to content

Commit

Permalink
Docs: table for Pebblo safe loader and safe retriever args/parameters…
Browse files Browse the repository at this point in the history
… description (#568)

* Added table for Pebblo Safe loader args/parameters desciption

* PebbloSafeLoader paramters

* PebbloRetrievalQA parameters

* Minor updates in the descriptions

* Incorporated review comments:
- Updated the classifier_url description with a link to the Pebblo Server Configuration.
- Fixed values for classifier_location.

---------

Co-authored-by: Rajendra Kadam <[email protected]>
  • Loading branch information
Raj725 and Rajendra Kadam authored Oct 8, 2024
1 parent 7a60288 commit a1d598b
Show file tree
Hide file tree
Showing 2 changed files with 35 additions and 0 deletions.
16 changes: 16 additions & 0 deletions docs/gh_pages/docs/rag.md
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,22 @@ Note: By default Pebblo Server runs at localhost:8000. If your Pebblo Server is
export PEBBLO_CLASSIFIER_URL="<pebblo-server-host:pebblo-server-port>"
```

## Parameters
PebbloSafeLoader takes the following parameters:

| Parameter | Type | Description |
|:-------------------|:--------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| name | str | Name of the application; should be unique across the loader and retriever applications. |
| owner | str | (**Optional**, Default: None) Owner of the application. |
| description | str | (**Optional**, Default: None) Description of the application. |
| loader | DocumentLoader| Langchain DocumentLoader. |
| api_key | str | (**Optional**, Default: None) API Key for Pebblo Cloud; if not provided, PebbloSafeLoader will look for `PEBBLO_API_KEY` in the environment. If found, documents will be sent to Pebblo Cloud. |
| load_semantic | bool | (**Optional**, Default: False) Indicates whether to include semantic metadata in the documents being loaded into VectorDB. |
| classifier_url | str | (**Optional**, Default: http://localhost:8000) URL of the Pebblo Classifier Server. For more details on configuring the server URL, see [Pebblo Server Configuration](https://daxa-ai.github.io/pebblo/config#server). |
| classifier_location| str | (**Optional**, Default: local) Location of the classifier, 'local' or 'pebblo-cloud'. |
| anonymize_snippets | bool | (**Optional**, Default: False) Indicates whether to anonymize snippets in the document. |


## Supported Document Loaders

The following Langchain DocumentLoaders are currently supported.
Expand Down
19 changes: 19 additions & 0 deletions docs/gh_pages/docs/retrieval_chain.md
Original file line number Diff line number Diff line change
Expand Up @@ -147,6 +147,25 @@ def ask(question: str, auth_context: dict):
return qa_chain.invoke(chain_input_obj.dict())
```

### Parameters
PebbloRetrievalQA takes the following parameters:

| Parameter | Type | Description |
|:-------------------|:---------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| llm | BaseLanguageModel | Langchain LLM instance. |
| app_name | str | Name of the application; should be unique across the loader and retriever applications. |
| owner | str | (**Optional**, Default: None) Owner of the application. |
| description | str | (**Optional**, Default: None) Description of the application. |
| chain_type | str | Type of document combining chain to use. Should be one of "stuff", "map_reduce", "map_rerank", and "refine". |
| retriever | VectorStoreRetriever | Vector database retriever. |
| verbose | bool | (**Optional**, Default: False) Whether chains should be run in verbose mode or not. |
| api_key | str | (**Optional**, Default: None) API Key for Pebblo Cloud; if not provided, PebbloRetrievalQA will look for `PEBBLO_API_KEY` in the environment. If found, retrieval data will be sent to Pebblo Cloud. |
| classifier_url | str | (**Optional**, Default: http://localhost:8000) URL of the Pebblo Classifier Server. For more details on configuring the server URL, see [Pebblo Server Configuration](https://daxa-ai.github.io/pebblo/config#server). |
| classifier_location| str | (**Optional**, Default: local) Location of the classifier, 'local' or 'pebblo-cloud'. |

\* _In addition to the above-mentioned parameters, `PebbloRetrievalQA` also supports the keyword arguments that are supported by the [Langchain Chain class](https://python.langchain.com/api_reference/langchain/chains/langchain.chains.base.Chain.html#langchain.chains.base.Chain)._


### Questions by Authorized User

Data has been ingested for the authorized identities ["hr-support", "hr-leadership"].
Expand Down

0 comments on commit a1d598b

Please sign in to comment.