paper-with-me

Papers

Phased Instruction Fine-Tuning for Large Language Models

2024-06-01 · Wei Pang, Chuan Zhou, Xiao-Hua Zhou, Xiaojie Wang

Instruction Fine-Tuning enhances pre-trained language models from basic next-word prediction to complex instruction-following. However, existing One-off Instruction Fine-Tuning (One-off IFT) method, applied on a diverse instruction, may not effectively boost models' adherence to instructions due to the simultaneous handling of varying instruction complexities. To improve this, Phased Instruction Fine-Tuning (Phased IFT) is proposed, based on the idea that learning to follow instructions is a gradual process. It assesses instruction difficulty using GPT-4, divides the instruction data into subsets of increasing difficulty, and uptrains the model sequentially on these subsets. Experiments with Llama-2 7B/13B/70B, Llama3 8/70B and Mistral-7B models using Alpaca data show that Phased IFT significantly outperforms One-off IFT, supporting the progressive alignment hypothesis and providing a simple and efficient way to enhance large language models. Codes and datasets from our experiments are freely available at https://github.com/xubuvd/PhasedSFT.

📄 PDF Abstract BibTeX arXiv:2406.04371

Code (1)

xubuvd/phasedsft 공식 구현 pytorch

Tasks

Instruction Following

Methods 이 논문이 사용한 방법론

Residual Connection 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Adam 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.

Similar Papers 제목 키워드 기반

Unveiling the Secret Recipe: A Guide For Supervised Fine-Tuning Small LLMs

2024-12-17 · Aldo Pareja, Nikhil Shivakumar Nayak, Hao Wang, KrishnaTeja Killamsetty 외

The rise of large language models (LLMs) has created a significant disparity: industrial research labs with their computational resources, expert teams, and advanced infrastructures, can effectively fine-tune LLMs, while…

MMLU

Instruction-Free Tuning of Large Vision Language Models for Medical Instruction Following

2026-03-19 · Myeongkyun Kang, Soopil Kim, Xiaoxiao Li, Sang Hyun Park arxiv

Large vision language models (LVLMs) have demonstrated impressive performance across a wide range of tasks. These capabilities largely stem from visual instruction tuning, which fine-tunes models on datasets consisting o…

Visual Question AnsweringInstruction Following

Building Instruction-Tuning Datasets from Human-Written Instructions with Open-Weight Large Language Models

2025-03-31 · Youmi Ma, Sakae Mizuki, Kazuki Fujii, Taishi Nakamura 외

Instruction tuning is crucial for enabling Large Language Models (LLMs) to solve real-world tasks. Prior work has shown the effectiveness of instruction-tuning data synthesized solely from LLMs, raising a fundamental que…

Demystifying Instruction Mixing for Fine-tuning Large Language Models

2023-12-17 · Renxi Wang, Haonan Li, Minghao Wu, Yuxia Wang 외

Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly …

Language ModellingLarge Language Model

Scaling Instruction-Finetuned Language Models

2022-10-20 · Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph 외

Finetuning language models on a collection of datasets phrased as instructions has been shown to improve model performance and generalization to unseen tasks. In this paper we explore instruction finetuning with a partic…

Coreference ResolutionCross-Lingual Question AnsweringMMLUMulti-task Language Understanding+2