You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Plugin for vanilla pytorch with features to load, save and version models
Support for HF accelerate
Support for Deepspeed for transformer inference
Support for manual sharding based on a device map file
Guides:
Fine-tuning embedding models
Improve retrieval systems
Undocumented Features :
Pydantic optional vector field
Huggingface integratoin
Doc improvement tips based on onboarding feedback :
Concepts
What is the preferred way to ingest data
- We support various ways to ingest data - pylist, pydict, arrow, recordbatch, pydantic model list.~
- We don't have an official preferred format.~
What is the preferred way to define LanceDB table schema**~
- We allow defining schema explicitly via pyarrow or Pydantic. But we don't say which is preferred ~
- Pydantic is required for using EMbedding API, which is what we want the users to use ~ Better document drop_table
…cement (#1326)
- Tried to address some onboarding feedbacks listed in
#1224
- Improve visibility of pydantic integration and embedding API. (Based
on onboarding feedback - Many ways of ingesting data, defining schema
but not sure what to use in a specific use-case)
- Add a guide that takes users through testing and improving retriever
performance using built-in utilities like hybrid-search and reranking
- Add some benchmarks for the above
- Add missing cohere docs
---------
Co-authored-by: Weston Pace <weston.pace@gmail.com>
Description
Experiments
Guides:
Undocumented Features :
Doc improvement tips based on onboarding feedback :
Concepts
What is the preferred way to ingest data
- We support various ways to ingest data - pylist, pydict, arrow, recordbatch, pydantic model list.~
- We don't have an official preferred format.~
What is the preferred way to define LanceDB table schema**~
- We allow defining schema explicitly via pyarrow or Pydantic. But we don't say which is preferred ~
- Pydantic is required for using EMbedding API, which is what we want the users to use ~
Better document drop_tableEnrich integrations pages with examples:
Docs typos/bug
-Links
The text was updated successfully, but these errors were encountered: