Olive Farm is a cutting-edge web application crafted by the innovative minds at OdiaGenAI. It's designed to effortlessly generate LLM (Language Model) instruction sets in Indic languages. Presently, it offers support for Hindi and Odia, with seamless scalability to incorporate additional languages on the horizon.
This versatile tool accommodates inputs from a variety of sources, including:
- URLs,
- PDF documents,
- Plain text.
Additionally, OliveFarm features a collection of pre-existing templates, powered by ChatGPT, to streamline the process of generating instruction sets. Experience the future of Indic language instruction with OliveFarm!
Contributors:
- AR Kamaldeen (KIIT University, India)
- SK Shahid (Silicon Institute of Technology, India)
- Sambit Sekhar (Odia Generative AI, India)
- Parul Agarwal (Institute of Mathematics, India)
- Dr. Shantipriya Parida (Silo AI, Finland)
Steps to generate an instruction set.
- Select the language.
- Select the input content type.
- Select the number of questions to generate.
- Select the format of instruction.
- Provide your OpenAI key and submit.
- Based on the input content type, input the Text/URL/PDF and submit.
- Click on “Generate Instructions”, which will generate the number of questions selected in Step 3.
- Select the questions by clicking on the checkboxes and click on “Generate Answers” which will generate the answers.
- Save the instruction set (questions and answers) by clicking on the “Save as jsonl” button which will save the instruction set. and make sure to rename the file to .jsonl while saving
- The instruction set will be saved in the system.
- After Each generation set click on clear to format the data
Citation:
If you find this repository useful, please consider giving ⭐ and citing:
@misc{Olivefarm,
author = {AR Kamaldeen and SK Shahid and Sambit Sekhar and Parul Agarwal Shantipriya Parida},
title = {Olive Farm: WebAPP for Indic Instruction Set Generation},
year = {2023},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/shantipriyap/OdiaGenAI}},
}
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.