README.md

metadata

dataset_info:
  features:
    - name: Colloquial_Telugu
      dtype: string
    - name: Standard_Telugu
      dtype: string
    - name: English
      dtype: string
    - name: Source
      dtype: string
    - name: Notes
      dtype: string
    - name: Type
      dtype: string
    - name: Answer
      dtype: string
    - name: Meaning
      dtype: string
    - name: Author
      dtype: string
  splits:
    - name: train
      num_bytes: 79681
      num_examples: 304
  download_size: 39845
  dataset_size: 79681
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Colloquial Telugu Dataset

This dataset contains colloquial Telugu phrases along with their standard Telugu translations and English equivalents. It is designed to help train Telugu language models with a better understanding of informal and conversational Telugu.

---

📌 Dataset Overview

• Name: Colloquial Telugu Dataset
• Source: Collected from native Telugu speakers across different regions.
• Fields:
  • Colloquial_Telugu: Informal Telugu phrase.
  • Standard_Telugu: Formal/standard Telugu translation.
  • English: English translation with context.
  • Source: Where the phrase was collected from.
  • Notes: Additional context (e.g., politeness level, usage scenarios).
  • Type: (Optional) Category of the phrase.
  • Answer: (Optional) Possible response.
  • Meaning: (Optional) Detailed meaning or breakdown.
  • Author: (Optional) Contributor's name.

---

📥 How to Load the Dataset

You can access the dataset directly from Hugging Face using:

from datasets import load_dataset

dataset = load_dataset("ankitha29/telugu-colloquial-corpus")

##🔹 Usage
This dataset can be used for:

Training Telugu Chatbots to understand informal speech.
Fine-tuning LLMs for better Telugu conversational understanding.
Text Normalization Tasks converting colloquial to formal Telugu.
Machine Translation improvements for Telugu.

##🏆 Contribution
If you want to add more phrases or improve translations, feel free to contribute by submitting a pull request!

##contacts
ravellaankitha2006@gmail.com
6305604176-ph.no