Total runs:
23.3K
Run Growth:
12.5K
Growth Rate:
53.67%
Sponsored by
Vmake Labs Video Enhancer
ytu-ce-cosmos / modernbert-tr-embed
Model's Last Updated: August 01 2026
sentence-similarity
Introduction of modernbert-tr-embed
Model Details of modernbert-tr-embed
ModernBERT-TR Embed
A 150M-parameter Turkish text-embedding model.
-
Base model:
ytu-ce-cosmos/modernbert-tr-base. -
Distilled from
Qwen/Qwen3-Embedding-8B.
Results
| Model | Params | Retr | Classif | PairCls | Cluster | STS | Bitext | Mean |
|---|---|---|---|---|---|---|---|---|
| ModernBERT-TR-Embed (ours) | 150M | 59.4 | 76.5 | 69.2 | 63.3 | 77.6 | 94.1 | 68.14 |
| ytu-ce-cosmos/turkish-e5-large | 560M | 61.5 | 72.6 | 62.8 | 60.9 | 80.0 | 99.3 | 67.17 |
| microsoft/harrier-oss-v1-0.6b | 600M | 60.1 | 71.1 | 58.6 | 63.3 | 74.5 | 98.6 | 65.57 |
| intfloat/multilingual-e5-large | 560M | 61.7 | 69.2 | 65.6 | 60.8 | 81.0 | 99.0 | 66.56 |
| Qwen/Qwen3-Embedding-4B | 4B | 63.1 | 70.2 | 60.1 | 61.3 | 77.0 | 97.8 | 66.69 |
| Qwen/Qwen3-Embedding-8B | 8B | 64.4 | 72.8 | 62.8 | 62.9 | 80.0 | 98.0 | 68.59 |
How was this model trained?
-
We embedded ~7.9M Turkish passages with the teacher, then trained our model to reproduce those embeddings. A projector maps the teacher's 4096-d vectors down to our model's 768 dimensions. Following Jasper/Stella distillation recipe , a three-term loss aligns the embeddings from both: a per-passage cosine loss, a similarity-matrix loss matching the student's and teacher's Gram matrices within batch, and a CoSENT-style hinge that reproduces the teacher's pairwise-similarity ordering.
-
From the distilled model we ran three fine-tunings, all supervised by teacher embeddings:
- Retrieval: the student ranks the correct passage above hard negatives for a given query. Trained with an InfoNCE contrastive loss over in-batch and hard negatives, plus a KL term matching the teacher's softmax ranking over each query's candidates.
- Multi-task: the retrieval objective plus Turkish language-understanding tasks: NLI, STS as in CoSENT on teacher cosine, supervised contrastive classification as in SupCon , and a replay of the cosine distillation on classification text.
- Cross-lingual: the multi-task recipe with rebalanced task weights and added English retrieval passages, to improve English-Turkish alignment.
-
We weight-average the checkpoints described above into a single model.
Usage
sentence-transformers
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("mrbesher/modernbert-tr-embed")
emb = model.encode(["bir cümle", "başka bir cümle"], normalize_embeddings=True)
q = model.encode(["soru"], prompt_name="query", normalize_embeddings=True)
d = model.encode(["döküman"], normalize_embeddings=True)
The retrieval (see
config_sentence_transformers.json
):
Instruct: Given a web search query, retrieve relevant passages that answer the query\nQuery:{text}
.
ONNX Runtime
The
onnx/
folder has the grapgh for the token embeddings, it includes 3 graphs: token embeddings, mean-pool and L2-norm.
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("mrbesher/modernbert-tr-embed", backend="onnx",
model_kwargs={"file_name": "onnx/model_fp16.onnx"})
Text Embeddings Inference (TEI)
TEI is L2-normalizes by default. Pass per-request prompts for retrieval queries.
text-embeddings-router --model-id mrbesher/modernbert-tr-embed --dtype float16
Encoderfile
See the companion encoderfile repo.
Training data
Turkish retrieval (msmarco-tr, Squad-TR train/dev), Turkish NLI (
boun-tabi/nli_tr
train), Turkish STS-B (train), and Turkish classification-domain text (product reviews, news, social), all teacher-supervised. We check for leaks with text-hash against every MTEB(Turkish) test split.
License & attribution
-
License:
apache-2.0.
Runs of ytu-ce-cosmos modernbert-tr-embed on huggingface.co
987
Total runs
0
24-hour runs
-71
3-day runs
-77
7-day runs
-76
30-day runs
More Information About modernbert-tr-embed huggingface.co Model
Url of modernbert-tr-embed
modernbert-tr-embed huggingface.co Url
Provider of modernbert-tr-embed huggingface.co
Other API from ytu-ce-cosmos
Updated:December 01 2025
Total runs:
3.1K
Run Growth:
-460
Growth Rate:
-14.72%
Total runs:
1.9K
Run Growth:
-9.3K
Growth Rate:
-492.03%
Total runs:
1.7K
Run Growth:
-1.3K
Growth Rate:
-76.82%
Total runs:
1.1K
Run Growth:
-294
Growth Rate:
-26.78%
Total runs:
1.1K
Run Growth:
675
Growth Rate:
61.64%
Total runs:
1.0K
Run Growth:
941
Growth Rate:
93.45%
Total runs:
906
Run Growth:
-17.6K
Growth Rate:
-1947.57%
Total runs:
800
Run Growth:
181
Growth Rate:
22.63%
Total runs:
725
Run Growth:
-51
Growth Rate:
-7.03%
Total runs:
548
Run Growth:
-556
Growth Rate:
-101.46%
Total runs:
477
Run Growth:
-71
Growth Rate:
-14.88%
Total runs:
471
Run Growth:
-242
Growth Rate:
-51.38%
Total runs:
431
Run Growth:
-66
Growth Rate:
-15.31%
Total runs:
281
Run Growth:
249
Growth Rate:
88.61%
Total runs:
262
Run Growth:
-21
Growth Rate:
-8.02%
Total runs:
243
Run Growth:
-193
Growth Rate:
-79.42%
Total runs:
204
Run Growth:
27
Growth Rate:
13.24%
Total runs:
199
Run Growth:
163
Growth Rate:
81.91%
Total runs:
193
Run Growth:
-19
Growth Rate:
-9.84%
Total runs:
187
Run Growth:
29
Growth Rate:
15.51%
Total runs:
179
Run Growth:
-36
Growth Rate:
-20.11%
Total runs:
137
Run Growth:
15
Growth Rate:
10.95%
Total runs:
123
Run Growth:
67
Growth Rate:
54.47%
Total runs:
100
Run Growth:
100
Growth Rate:
100.00%
Total runs:
95
Run Growth:
21
Growth Rate:
22.11%
Total runs:
90
Run Growth:
33
Growth Rate:
36.67%
Total runs:
86
Run Growth:
22
Growth Rate:
25.58%
Total runs:
80
Run Growth:
41
Growth Rate:
51.25%
Total runs:
55
Run Growth:
34
Growth Rate:
61.82%
Total runs:
47
Run Growth:
2
Growth Rate:
4.26%
Total runs:
38
Run Growth:
-465
Growth Rate:
-1223.68%
Total runs:
34
Run Growth:
22
Growth Rate:
64.71%
Total runs:
33
Run Growth:
33
Growth Rate:
100.00%
Total runs:
32
Run Growth:
29
Growth Rate:
90.63%
Total runs:
28
Run Growth:
6
Growth Rate:
21.43%
Total runs:
26
Run Growth:
-124
Growth Rate:
-476.92%
Total runs:
26
Run Growth:
26
Growth Rate:
100.00%
Total runs:
24
Run Growth:
-52
Growth Rate:
-216.67%
Total runs:
24
Run Growth:
18
Growth Rate:
75.00%
Total runs:
24
Run Growth:
24
Growth Rate:
100.00%
Total runs:
21
Run Growth:
21
Growth Rate:
100.00%
Total runs:
19
Run Growth:
9
Growth Rate:
47.37%
Total runs:
19
Run Growth:
-31
Growth Rate:
-163.16%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%