You need to agree to share your Hugging Face information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

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

Iban Morphological Analyzer and Generator (ByT5)

Model Description

ryuuu7/Iban_Morphological_Analyzer is a token-free sequence-to-sequence neural Morphological Analyzer and Generator (MAG) for Bahasa Iban (ISO 639-3: iba), an indigenous Austronesian language spoken primarily in Sarawak, Malaysia, and Kalimantan, Indonesia.

Fine-tuned from google/byt5-small this model operates directly on raw UTF-8 byte sequences rather than relying on statistical subword tokenizers (such as BPE or WordPiece). By avoiding subword fragmentation, the model accurately captures Bahasa Iban's agglutinative affixation (prefixes, suffixes, infixes, and circumfixes) as well as complex morphophonemic alternations such as nasal substitution and assimilation—where initial consonants mutate into surface nasals (e.g., bukut -> mukut, peda -> meda, tusui -> nusui)

  • Developed by: Iqbal Qayyim (Universiti Malaysia Sarawak / UNIMAS)
  • Model Type: Byte-level Encoder-Decoder Transformer (Seq2Seq)
  • Language: Bahasa Iban (iba)
  • Base Model: google/byt5-small
  • Demo Repository: GitHub - iban-morphology-demo

Intended Uses

This model is trained in a bidirectional multi-task setup using task-conditioning prefixes:

  1. Morphological Analysis (analyze morphology:): Decomposes an inflected or derived surface Iban word into its underlying canonical root lemma, Part-of-Speech (POS) tag, and morphological affixes.
  2. Morphological Generation (generate surface:): Synthesizes the valid inflected surface word form given an underlying root lemma and target affix tags.

Out-of-Scope Use

  • General open-ended text generation, summarization, or conversational chat in Bahasa Iban.
  • Sentence-level machine translation without word-level segmentation.

How to Use

You can load and run the model directly using the Hugging Face transformers library:

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

model_id = "ryuuu7/Iban_Morphological_Analyzer"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSeq2SeqLM.from_pretrained(model_id)

def predict(prompt: str, max_new_tokens: int = 64) -> str:
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(
        **inputs,
        max_new_tokens=max_new_tokens,
        num_beams=5,
        early_stopping=True
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# 1. Morphological Analysis Example
analysis_input = "analyze morphology: mukut"
print("Analysis:", predict(analysis_input))

# 2. Morphological Generation Example
generation_input = "generate surface: root:bukut aff:N-"
print("Generation:", predict(generation_input))

  author = {Iqbal Qayyim},
  title = {Iban Morphological Analyzer and Generator using ByT5},
  year = {2026},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/ryuuu7/Iban_Morphological_Analyzer}}

Downloads last month
2
Safetensors
Model size
0.3B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ryuuu7/Iban_Morphological_Analyzer

Finetuned
(346)
this model