This model is a Pali → Sinhala translation model fine-tuned on top of TinyLLaMA, designed to translate Pali Buddhist texts into modern Sinhala.
It is intended for educational, research, and cultural preservation use cases.

Model Details

Model Description

Developed by: Nimantha Perera
Model type: Causal Language Model (LLM)
Base model: TinyLLaMA
Language(s): Pali, Sinhala
Task: Machine Translation (Pali → Sinhala)
License: Apache-2.0
Finetuned from model: TinyLLaMA

Model Sources

Repository: https://huggingface.co/NimanthaPerera/tinyllama-pali-to-sinhala

Uses

Direct Use

Translating Pali verses into Sinhala
Understanding Buddhist scriptures
Educational and academic research
Digital humanities and language preservation

Downstream Use

Fine-tuning for domain-specific Buddhist texts
Integration into Sinhala/Pali learning applications
Chatbots or assistants for religious studies

Out-of-Scope Use

Legal, medical, or financial advice
Authoritative religious rulings
Fully accurate scholarly translation without human review

Bias, Risks, and Limitations

Training data may contain religious or historical bias
Translations may be incomplete or stylistically inconsistent
Not all Pali grammatical structures are perfectly captured
Outputs should be verified by domain experts

Recommendations

Users are advised to:

Use the model as a support tool, not a final authority
Validate translations with scholars or reference texts
Avoid sensitive or high-stakes usage

How to Get Started with the Model

python

from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "NimanthaPerera/tinyllama-pali-to-sinhala"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

prompt = "Pali: Sabbe sankhara anicca"
inputs = tokenizer(prompt, return_tensors="pt")

outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))