Modelpublic

TartarusXXX/synthetic-n2en-marian-transformers

sourceHugging Faceupdated 3mo agoView on Hugging Face

0likes7downloads

Model Card

synthetic-n2en-marian-transformers

This repository contains a Marian NMT checkpoint packaged in OPUS-MT style.

Model Details

—Source language: multi (Multilingual)
—Target language: en (English)
—Task: translation
—Framework: Marian NMT
—Training data: TartarusXXX/synthetic-parallel-16lang-1-4m-gemini-3-1-flash-lite
—Validation set: bouquet-dev-n2en

Evaluation target languages: tr, kmr, de, es, fr, el, bg, ru, ka, hy, fa, ckb, ur, ar

Validation

—Best bleu: 28.8111
—Epoch: 9
—Update: 318000
—Timestamp: 2026-06-17 05:14:00
—Validation set: bouquet-dev-n2en

Raw best line:

text

[2026-06-17 05:14:00] [valid] Ep. 9 : Up. 318000 : bleu : 28.8111 : new best

Recent validation results:

Timestamp	Epoch	Update	Metric	Score	Status
2026-06-17 03:29:32	8	291500	bleu	28.5341	stalled 2 times (last best: 28.6381)
2026-06-17 05:14:00	9	318000	bleu	28.8111	new best
2026-06-17 06:58:36	10	344500	bleu	28.621	stalled 1 times (last best: 28.8111)
2026-06-17 08:41:49	10	371000	bleu	28.5135	stalled 2 times (last best: 28.8111)
2026-06-17 10:26:14	11	397500	bleu	26.9018	stalled 3 times (last best: 28.8111)
2026-06-17 12:10:50	12	424000	bleu	27.974	stalled 4 times (last best: 28.8111)
2026-06-17 13:54:10	12	450500	bleu	27.0804	stalled 5 times (last best: 28.8111)
2026-06-17 15:38:45	13	477000	bleu	27.3126	stalled 6 times (last best: 28.8111)
2026-06-17 17:23:19	14	503500	bleu	27.4481	stalled 7 times (last best: 28.8111)
2026-06-17 19:06:45	14	530000	bleu	26.43	stalled 8 times (last best: 28.8111)
2026-06-17 20:51:24	15	556500	bleu	26.6759	stalled 9 times (last best: 28.8111)
2026-06-17 22:36:11	16	583000	bleu	26.0265	stalled 10 times (last best: 28.8111)

Files

—synthetic_parallel_16lang_1_4m_gemini_3_1_flash_lite.spm32k-spm32k.transformer.model1.npz.best-bleu.npz
—synthetic_parallel_16lang_1_4m_gemini_3_1_flash_lite.spm32k-spm32k.vocab.yml
—opus.trg.spm32k-model
—opus.src.spm32k-model
—hf-csv-synthetic.n2en.bouquet.transformer.mk
—synthetic_parallel_16lang_1_4m_gemini_3_1_flash_lite.spm32k-spm32k.transformer.model1.npz.yml
—synthetic_parallel_16lang_1_4m_gemini_3_1_flash_lite.spm32k-spm32k.transformer.train1.log
—synthetic_parallel_16lang_1_4m_gemini_3_1_flash_lite.spm32k-spm32k.transformer.valid1.log
—bouquet-dev-n2en.best-bleu.validation-style-spm.tsv
—bouquet-dev-n2en.best-bleu.validation-style-spm.md
—best-bleu.decoder.yml
—translate_with_marian.py

Usage

Install Marian and run:

bash

python translate_with_marian.py --decoder best-bleu.decoder.yml < input.txt > output.txt

The decoder YAML uses paths relative to this repository.

Conversion

Converted from the Marian checkpoint in TartarusXXX/synthetic-n2en-marian using the Transformers Marian converter.

—Source language: multi
—Target languages: en

Transformers Usage

python

from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

model_name = "TartarusXXX/synthetic-n2en-marian-transformers"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

# This model was trained as `multi` -> `en`.
# If the Marian training data uses target-language tags, include the desired
# target tag in the source sentence, following the training convention.
inputs = tokenizer("Hello world!", return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.batch_decode(outputs, skip_special_tokens=True)[0])