Sentence Similarity
sentence-transformers
Safetensors
Turkish
modernbert
feature-extraction
turkish
mteb
text-embeddings-inference
Instructions to use moganai/MoganBERT-Embed with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- sentence-transformers
How to use moganai/MoganBERT-Embed with sentence-transformers:
from sentence_transformers import SentenceTransformer model = SentenceTransformer("moganai/MoganBERT-Embed") sentences = [ "The weather is lovely today.", "It's so sunny outside!", "He drove to the stadium." ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [3, 3] - Notebooks
- Google Colab
- Kaggle
Fix image extensions (png -> jpg)
Browse files
README.md
CHANGED
|
@@ -58,7 +58,7 @@ The released checkpoint is a weighted average of Phase 1 and two Phase 2 runs.
|
|
| 58 |
|
| 59 |
Evaluated on MTEB(Turkish) — 26 tasks, fp16, cosine similarity, max 2,048 tokens.
|
| 60 |
|
| 61 |
-
<img src="assets/mteb_en.
|
| 62 |
|
| 63 |
Retrieval is the relative weak spot — for retrieval-first workloads see [MoganColBERT-TR](https://huggingface.co/moganai/MoganColBERT-TR). Full per-task numbers are in the [paper](https://huggingface.co/papers/2608.25768).
|
| 64 |
|
|
@@ -122,7 +122,7 @@ Yayınlanan checkpoint, Faz 1 ve iki Faz 2 koşusunun ağırlıklı ortalamasıd
|
|
| 122 |
|
| 123 |
MTEB(Turkish) üzerinde ölçüldü — 26 görev, fp16, kosinüs benzerliği, azami 2.048 token.
|
| 124 |
|
| 125 |
-
<img src="assets/mteb.
|
| 126 |
|
| 127 |
Retrieval göreli zayıf noktadır — retrieval öncelikli işler için [MoganColBERT-TR](https://huggingface.co/moganai/MoganColBERT-TR) modeline bakın. Görev bazlı tüm sayılar [makalede](https://huggingface.co/papers/2608.25768).
|
| 128 |
|
|
|
|
| 58 |
|
| 59 |
Evaluated on MTEB(Turkish) — 26 tasks, fp16, cosine similarity, max 2,048 tokens.
|
| 60 |
|
| 61 |
+
<img src="assets/mteb_en.jpg" alt="MTEB(Turkish)" width="850"/>
|
| 62 |
|
| 63 |
Retrieval is the relative weak spot — for retrieval-first workloads see [MoganColBERT-TR](https://huggingface.co/moganai/MoganColBERT-TR). Full per-task numbers are in the [paper](https://huggingface.co/papers/2608.25768).
|
| 64 |
|
|
|
|
| 122 |
|
| 123 |
MTEB(Turkish) üzerinde ölçüldü — 26 görev, fp16, kosinüs benzerliği, azami 2.048 token.
|
| 124 |
|
| 125 |
+
<img src="assets/mteb.jpg" alt="MTEB(Turkish)" width="850"/>
|
| 126 |
|
| 127 |
Retrieval göreli zayıf noktadır — retrieval öncelikli işler için [MoganColBERT-TR](https://huggingface.co/moganai/MoganColBERT-TR) modeline bakın. Görev bazlı tüm sayılar [makalede](https://huggingface.co/papers/2608.25768).
|
| 128 |
|