YAML Metadata Warning:The pipeline tag "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other
Uzbek Text Simplifier (mT5-small fine-tuned)
Fine-tuned google/mt5-small для упрощения сложных узбекских текстов
(юридические документы, официальные новости, гос. инструкции) в текст,
понятный обычному читателю.
Использование
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("username/uz-text-simplifier")
model = AutoModelForSeq2SeqLM.from_pretrained("username/uz-text-simplifier")
text = "Oʻzbekiston Respublikasi Vazirlar Mahkamasining qarori bilan..."
inputs = tokenizer("simplify: " + text, return_tensors="pt", truncation=True, max_length=256)
out = model.generate(**inputs, max_new_tokens=256, num_beams=4)
print(tokenizer.decode(out[0], skip_special_tokens=True))
Важно: на вход всегда добавляй префикс "simplify: " — модель обучена под этот task-prefix.
Данные
~4150 пар (сложный текст → простой текст), собранных с lex.uz (юридические акты), kun.uz, gazeta.uz, uza.uz и uz.wikipedia.org, размеченных teacher-моделью.
Обучение
- База:
google/mt5-small - 8 эпох, learning rate 1e-3, оптимизатор Adafactor
- Метрики на валидации: ROUGE-L ≈ 0.28, chrF ≈ 44.7
Ограничения
- Датасет сильно смещён в сторону юридических текстов (домен
law, ~59%) - На доменах health/technology/education (<100 примеров каждый) качество может быть ниже
- Не предназначена для текстов вне узбекского языка
Model tree for ismailoviskandar02/uzbek-text-simplifier
Base model
google/mt5-small