Add paper link and improve metadata

#14
by nielsr HF staff - opened
Files changed (1) hide show
  1. README.md +14 -8
README.md CHANGED
@@ -1,13 +1,6 @@
1
  ---
2
  base_model:
3
  - nomic-ai/nomic-embed-text-v2-moe-unsupervised
4
- library_name: sentence-transformers
5
- pipeline_tag: sentence-similarity
6
- tags:
7
- - sentence-transformers
8
- - sentence-similarity
9
- - feature-extraction
10
- license: apache-2.0
11
  language:
12
  - en
13
  - es
@@ -110,12 +103,25 @@ language:
110
  - gn
111
  - gv
112
  - hy
 
 
 
 
 
 
113
  ---
114
 
115
  # nomic-embed-text-v2-moe: Multilingual Mixture of Experts Text Embeddings
116
 
 
 
 
 
 
 
 
117
  ## Model Overview
118
- `nomic-embed-text-v2-moe` is SoTA multilingual MoE text embedding model that excels at multilingual retrieval:
119
 
120
  - **High Performance**: SoTA Multilingual performance compared to ~300M parameter models, competitive with models 2x in size
121
  - **Multilinguality**: Supports ~100 languages and trained on over 1.6B pairs
 
1
  ---
2
  base_model:
3
  - nomic-ai/nomic-embed-text-v2-moe-unsupervised
 
 
 
 
 
 
 
4
  language:
5
  - en
6
  - es
 
103
  - gn
104
  - gv
105
  - hy
106
+ library_name: sentence-transformers
107
+ license: apache-2.0
108
+ pipeline_tag: feature-extraction
109
+ tags:
110
+ - sentence-transformers
111
+ - feature-extraction
112
  ---
113
 
114
  # nomic-embed-text-v2-moe: Multilingual Mixture of Experts Text Embeddings
115
 
116
+ This model was presented in the paper [Training Sparse Mixture Of Experts Text Embedding Models](https://huggingface.co/papers/2502.07972).
117
+
118
+ ## Paper Abstract
119
+
120
+ Transformer-based text embedding models have improved their performance on benchmarks like MIRACL and BEIR by increasing their parameter counts. However, this scaling approach introduces significant deployment challenges, including increased inference latency and memory usage. These challenges are particularly severe in retrieval-augmented generation (RAG) applications, where large models' increased memory requirements constrain dataset ingestion capacity, and their higher latency directly impacts query-time performance. While causal language models have addressed similar efficiency challenges using Mixture of Experts (MoE) architectures, this approach hasn't been successfully adapted to the general text embedding setting. In this paper, we introduce Nomic Embed v2, the first general purpose MoE text embedding model. Our model outperforms models in the same parameter class on both monolingual and multilingual benchmarks while also maintaining competitive performance with models twice its size. We open-source all code, models, and evaluation data to ensure full reproducibility of our training pipeline at https://github.com/nomic-ai/contrastors.
121
+
122
+
123
  ## Model Overview
124
+ `nomic-embed-text-v2-moe` is a SoTA multilingual MoE text embedding model that excels at multilingual retrieval:
125
 
126
  - **High Performance**: SoTA Multilingual performance compared to ~300M parameter models, competitive with models 2x in size
127
  - **Multilinguality**: Supports ~100 languages and trained on over 1.6B pairs