PnHLayman

PnHLayman

This T5-based model is fine-tuned to generate layman-friendly summaries of legal judgments from Punjab and Haryana courts.

Model Description

• Base architecture: T5-base
• Fine-tuned on: 2,100 Indian legal judgments with expert-written simplified summaries (the InLegalLaySum dataset)
• Objective: Given a court judgment, generate a concise, jargon-free summary suitable for lay readers.

Training Data

Each example contains:

• Judgement (full text of the court order)
• summary (expert-written abstract)

Total of ~7000 examples covering diverse case types (civil, criminal, administrative).

Usage

from transformers import T5Tokenizer, T5ForConditionalGeneration

MODEL_ID = "satviksh09/PnHLayman"

tokenizer = T5Tokenizer.from_pretrained(MODEL_ID)
model     = T5ForConditionalGeneration.from_pretrained(MODEL_ID)

text = "Full text of a Punjab court judgment goes here..."

inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=1024)
outputs = model.generate(**inputs, max_length=256, num_beams=5, early_stopping=True)
summary = tokenizer.decode(outputs[0], skip_special_tokens=True)

print("Layman Summary:", summary)