Hungarian morphological generator model with mT5

For further models, scripts and details, see our demo site.

• Pretrained model used: mT5
• Prefix: "morph: "
• emMorph-based generation

Limitations

• max_source_length = 64
• max_target_length = 32

Results

Usage with pipeline

from transformers import pipeline

text2text_generator = pipeline(task="text2text-generation", model="NYTK/morphological-generator-emmorph-mt5-hungarian")

print(text2text_generator("morph: munka [/N][Acc]")[0]["generated_text"])

Citation

If you use this model, please cite the following paper:

@inproceedings {morph-generator,
    title = {Neural Morphological Generators for Hungarian},
	booktitle = {XIX. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2023)},
	year = {2023},
	publisher = {Szegedi Tudományegyetem, Informatikai Intézet},
	address = {Szeged, Hungary},
	author = {Laki, László János and Ligeti-Nagy, Noémi and Vadász, Noémi and Yang, Zijian Győző},
	pages = {331--340}
}