sinanalyuruk/trmteb-trnews-v1-mlx

trmteb-trnews — MLX build

Apple-Silicon MLX float16 build of the Turkish-news retrieval fine-tune of trmteb/turkish-embedding-model. Same weights (mean pooling, 768-d, L2-normalized), converted with mlx-embeddings; verified at parity with the PyTorch source (per-text cosine 1.00000; identical R@1/MRR).

• Format: mlx-embeddings, float16 (~211 MB) · max_seq_length 512
• Evaluation: the full before/after table lives on the main model card → trmteb-trnews-v1 (the PyTorch build this was converted from).

Usage

import numpy as np, mlx_embeddings
model, tok = mlx_embeddings.load("<this-repo-or-path>")
o = mlx_embeddings.generate(model, tok, ["Örnek başlık.", "İkinci haber."])
E = np.array(o.text_embeds)   # already L2-normalized; cosine == dot product

Requires mlx>=0.30, mlx-embeddings>=0.1.0.

License & attribution

Released under the MIT license — matching the verified upstream root (dbmdz/bert-base-turkish-uncased, MIT).

NOTICE — this is a derivative work. Attribution chain:

• Fine-tuned from trmteb/turkish-embedding-model — no license is declared upstream (as of 2026-07-03 the source model card states none); attribution retained.
• Whose base is dbmdz/bert-base-turkish-uncased (BERTurk) — MIT. Retain BERTurk's copyright and permission notice (see its model card).
• Changes (documented for provenance): fine-tuned on 100,000 TR-News (title, abstract) pairs with MultipleNegativesRankingLoss; published at max_seq_length 512.

trmteb/turkish-embedding-model itself declares no license, so the MIT terms above inherit from the BERTurk root and cover this fine-tune's contribution; confirm the intermediate model's terms with its author before commercial redistribution.

Citation — training data (TR-News)

@article{10.1007/s10579-021-09568-y,
    year = {2022},
    title = {{Abstractive text summarization and new large-scale datasets for agglutinative languages Turkish and Hungarian}},
    author = {Baykara, Batuhan and Güngör, Tunga},
    journal = {Language Resources and Evaluation},
    issn = {1574-020X},
    doi = {10.1007/s10579-021-09568-y},
    pages = {1--35}
}