translate-windy-hplt-en-mk — en to mk

CTranslate2 INT8 bilingual translation model, derived from HPLT/translate-en-mk-v2.0-hplt_opus (the HPLT project, CC-BY-4.0).

CC-BY-4.0 permits commercial use and modification and requires attribution; this notice provides it. Windstorm Labs did not train the base model.

Recovery note

HPLT publishes native Marian (.npz + .spm + .vocab) and marks its repos inference: false. An earlier conversion of this model to transformers format corrupted the weights, producing untranslatable output, and the model was written off. It was recovered on 2026-07-26 by converting the original .npz with the official Marian converter. The model was always fine; the conversion was not.

What was changed

A deliberately minimal LoRA fine-tune ("windy hallmark") merged into the base weights.

Rank / alpha	8 / 16
Learning rate / steps	2.5e-06 / 50
Target modules	q_proj, v_proj
Seed	42 (reproducible)
Training data	opus100_en-mk (clinic cache), 2000 sentence pairs
Tensors modified	36
Max absolute weight delta	4.155e-05

Provenance

base   model.bin sha256  b04eb312b9d62e45f5e68d7faec8a3bcd52d4baee637e7f427753aef63a3e71c
this   model.bin sha256  69d92935ad32c26bfc67f35404b78952c832d2f467594a8a01525ac8d6fc08c9

Verified on the quantized artifact you download, not merely internal weights.

Evaluation

FLORES-200 devtest, 200 sentences, beam 4, CTranslate2 INT8 on CPU. Metrics: spBLEU (flores200 tokenizer) + chrF (word_order=0).

	spBLEU	chrF
this model	40.77	61.93
upstream, pre-fog	39.58	—

Limitations

• One language pair only (en to mk).
• FLORES-200 is news and encyclopedic prose; it says little about conversational register.
• No human evaluation was performed.