Sentence Transformers v5.0 - Sparse Encoder Models (#2924)

* Add an initial draft for Sentence Transformers v5.0

* Finish the Trainer section, add evaluation section

* Add a section on what Sparse Embedding models are

* Add table of contents

* Relocate Contrastive Sparse Representation (CSR) section and Inference-free Splade for clarity

* Update table of contents

* Add Vector Database Integration section with Qdrant example + cite CSR paper

* Add training examples and resources for sparse embedding models

* Add some commas

* Unindent inference script

* Use human-first titles in additional resources

* Add section on pretrained sparse encoders and SPLADE Models collection

* Update inference-free Splade implementation to use SparseStaticEmbedding module and chnage lambda associated code

* corpus_regularizer -> document_regularizer

* Add _blog.yml with release date

* Remove incorrect documentation list item

* Update model link

* Update SPLADE Models collection link

* CSRSparsity -> SparseAutoEncoder

* "generally" use expansion

* Update variable name

* Fix link

* Update link to Loss Overview

* Update HPO link

---------

Co-authored-by: arthurbr11 <[email protected]>

Author: tomaarsen

_blog.yml

@@ -6228,3 +6228,14 @@
     - community
     - research
     - multimodal
+
+- local: train-sparse-encoder
+  title: "Training and Finetuning Sparse Embedding Models with Sentence Transformers v5"
+  author: tomaarsen
+  thumbnail: /blog/assets/train-sentence-transformers/st-hf-thumbnail.png
+  date: Jul 1, 2025
+  tags:
+    - nlp
+    - guide
+    - community
+    - open-source