Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
[NAACL'24] Self-data filtering of LLM instruction-tuning data using a novel perplexity-based difficulty score, without using any other models
| Date | Stars |
|---|---|
| 2026-07-31 | 417 |
| 2026-08-06 | 417 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning (NAACL'24) [From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning](https://arxiv.org/abs/2308.12032)<br>(NAACL'24)<br> Chinese Version: [[知乎]](https://zhuanlan.zhihu.com/p/718117800) <p align="center" width="40%"> <a ><img src="images/cherry.jpeg" alt="overview" style="width: 40%; min-width: 300px; display: block; margin: auto;"></a> </p> This is the repo for the Cherry Data Selection project, which introduces a self-guided methodology for LLMs to autonomously discern and select cherry samples from vast open-source datasets, effectively minimizing manual curation and potential cost for instruction tuning an LLM. The repo contains: - The cherry data used for fine-tuning the model, cherry_data_v1 represents the cherry data obtained based on the llama-1 model. - The model checkpoints that were trained using our cherry data. - The code for selecting cherry data from the existing instruction-tuning dataset. (Feel free to email Ming ([Homepage](https://mingliiii.github.io/), [Email]([email protected])) for any questions or feedback.) ## News - [2024/03] Our paper has been accepted to the **NAACL 2024** main conference! - [2024/02] We released the [Superfiltering](https://github.com/tianyi-lab/Superfiltering), which reveals the strong consistency between small and large LLMs in perceiving and evaluating the difficulty of instruction tuning data and utilizes a small LM, e.g., GPT-2 (124M), to effectively and efficiently select data for instruction tuning. - [2023/12] An updated code for calculating the statistics for IFD scores, please check [Reflection-Tuning Code for Selection](https://github.com/tianyi-lab/Reflection_Tuning?tab=readme-ov-file#code-for-selection). - [2023/12] The statistics necessary for calculating IFD scores on Alpaca and WizardLM on llama2-7b and llama2-13b were released, please check: [Alpaca llama2 7b](https://huggingface.co/datasets/MingLiiii/Alpaca_Analysis_llama2_7b), [Alpaca llama2 13b](https://huggingface.co/datasets/MingLiiii/Alpaca_Analysis_llama2_13b), [WizardLM70k llama2 7b](https://huggingface.co/datasets/MingLiiii/Wiz70_Analysis_llama2_7b), [WizardLM70k llama2 13b](https://huggingface.co/datasets/MingLiiii/Wiz70_Analysis_llama2_13b). - [2023/11] We added some results on llama2-7b and llama2-13b, further showing the generalizability of our method. - [2023/09] We partially reconstructed the repo structure and added some results on llama2. - [2023/09] We released codes for evaluating the performance between two LLMs by using GPT4 or chatGPT. - [2023/09] We released codes for this project. ## Contents - [Overview](#overview) - [Highlights](#highlights) - [Install](#install) - [Run Code](#run-code) - [Data and Model Weights V1](#data-and-model-weights-v1) - [Data and Model Weights V2](#data-and-model-weights-v2) - [Evaluation](#evaluation) - [Performance Comparison ](#performance-comparison) - [Prompt](#prompt) - [Hyperparameters](#hyperparameters) - [ToDo](#todo) - [Citation](#citation) - [Our Related Works](#our-related-works) ## Overview Our study puts forth a method for autonomously sifting through expansive open-source datasets to discover the most impactful training samples. We coin these samples as "cherry data", designating those data fragments that hold the potential to exponentially enhance LLM instruction tuning. At the heart of our research is the hypothesis that during their preliminary training stages with carefully chosen instruction data, LLMs can develop an intrinsic capability to discern instructions. This foundational understanding equips them with the discernment to assess the quality of broader datasets thus making it possible to estimate the instruction-following difficulty in a self-guided manner. <p align="center" width="70%"> <a ><img src="images/method_overview.png" alt="overview" style="width: 50%; mi
Excerpt of 22,510 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:f473cdc37a1d92f4, desc:instruction tuning