Downloads · 30 days
20
13% of all-time downloads
granite-embedding-97m-multilingual-r2-yat
mlnomad/granite-embedding-97m-multilingual-r2-yat
granite-embedding-97m-multilingual-r2-yat is a sentence similarity model from mlnomad. Use it when you need a score for how close two texts are. It is set up for sentence-transformers. The card lists the license as apache-2.0.
ibm-granite/granite-embedding-97m-multilingual-r2 (ModernBERT, SiLU-gated GLU FFN) with every feed-forward block replaced by a sigmoid-gated Yat-kernel MLP, via phased distillation (random-token + real per-block warm-…
All-time downloads
154
Public
Parameters
97.5M
415 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors390 MB · 94%
At a glance
- Task
- Sentence Similarity
- Library
- sentence-transformers
- License
- apache-2.0
- Model type
- yatmodernbert
- Access
- Public
- Created
- May 31, 2026
- Updated
- May 31, 2026
- SHA
- 65492ea3
Base models
Model card
From the Hugging Face model README
granite-embedding-97m-multilingual-r2-yat
ibm-granite/granite-embedding-97m-multilingual-r2 (ModernBERT, SiLU-gated GLU FFN) with
every feed-forward block replaced by a sigmoid-gated Yat-kernel MLP, via phased
distillation (random-token + real per-block warm-start, then end-to-end last-layer
distillation on English all-nli).
English MTEB STS avg: 0.7268 (teacher 0.764). Distilled on English only; multilingual STS17/STS22 evaluated zero-shot (see repo files / paper).
from sentence_transformers import SentenceTransformer
m = SentenceTransformer("mlnomad/granite-embedding-97m-multilingual-r2-yat", trust_remote_code=True)
m.encode(["hello world"])
Yat FFN: (softplus(ar) * (x.W+b)^2/(||x-W||^2+exp(le)) * sigmoid(gate(x))) @ A + c.
- Task
- Sentence Similarity
- Library
- sentence-transformers
- Type
- yatmodernbert
- License
- apache-2.0
- Created
- May 31, 2026
- Updated
- May 31, 2026