Sparse and Dense Embedding
==========================

Overview
--------

The ``zvec_db.embedders`` sub-package provides sparse and dense embedding models for text vectorization.

**Sparse Embedders:**

All sparse embedders return dictionaries ``{index: score, ...}`` compatible with zvec's ``SPARSE_FP32`` format.

+------------------+------------------------------------------------+
| Embedder         | When to use                                    |
+==================+================================================+
| CountEmbedder    | Baseline, documents of similar length          |
+------------------+------------------------------------------------+
| BM25Embedder     | General use, good IR performance               |
+------------------+------------------------------------------------+
| BM25LEmbedder    | Documents with very variable lengths           |
+------------------+------------------------------------------------+
| BM25PlusEmbedder | Many rare terms, need recall                   |
+------------------+------------------------------------------------+
| DisMaxEmbedder   | Multi-field, match any field                   |
+------------------+------------------------------------------------+
| TfidfEmbedder    | Relative term importance in corpus             |
+------------------+------------------------------------------------+

**Dense Embedders:**

+----------------------------+------------------------------------------------+
| Embedder                   | When to use                                    |
+============================+================================================+
| SentenceTransformersEmbedder | Local models (e.g., all-MiniLM-L6-v2)        |
+----------------------------+------------------------------------------------+
| OpenAIEmbedder             | OpenAI API or compatible endpoints (vLLM)      |
+----------------------------+------------------------------------------------+

CountEmbedder
-------------

.. autoclass:: zvec_db.embedders.CountEmbedder
   :members:
   :undoc-members:
   :inherited-members:

BM25Embedder
------------

.. autoclass:: zvec_db.embedders.BM25Embedder
   :members:
   :undoc-members:
   :inherited-members:

BM25LEmbedder
-------------

.. autoclass:: zvec_db.embedders.BM25LEmbedder
   :members:
   :undoc-members:
   :inherited-members:

BM25PlusEmbedder
----------------

.. autoclass:: zvec_db.embedders.BM25PlusEmbedder
   :members:
   :undoc-members:
   :inherited-members:

DisMaxEmbedder
--------------

.. autoclass:: zvec_db.embedders.DisMaxEmbedder
   :members:
   :undoc-members:
   :inherited-members:

TfidfEmbedder
-------------

.. autoclass:: zvec_db.embedders.TfidfEmbedder
   :members:
   :undoc-members:
   :inherited-members:

Dense Embedding
---------------

.. autoclass:: zvec_db.embedders.SentenceTransformersEmbedder
   :members:
   :undoc-members:
   :inherited-members:

.. autoclass:: zvec_db.embedders.OpenAIEmbedder
   :members:
   :undoc-members:
   :inherited-members: