Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs

• URL: http://arxiv.org/abs/2406.18629v1
• Date: Wed, 26 Jun 2024 17:43:06 GMT
• Title: Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs
• Authors: Xin Lai, Zhuotao Tian, Yukang Chen, Senqiao Yang, Xiangru Peng, Jiaya Jia,
• Abstract summary: We propose a simple, effective, and data-efficient method called Step-DPO. Step-DPO treats individual reasoning steps as units for preference optimization rather than evaluating answers holistically. Our findings demonstrate that as few as 10K preference data pairs and fewer than 500 Step-DPO training steps can yield a nearly 3% gain in accuracy on MATH for models with over 70B parameters.
• Score: 54.05511925104712
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Mathematical reasoning presents a significant challenge for Large Language Models (LLMs) due to the extensive and precise chain of reasoning required for accuracy. Ensuring the correctness of each reasoning step is critical. To address this, we aim to enhance the robustness and factuality of LLMs by learning from human feedback. However, Direct Preference Optimization (DPO) has shown limited benefits for long-chain mathematical reasoning, as models employing DPO struggle to identify detailed errors in incorrect answers. This limitation stems from a lack of fine-grained process supervision. We propose a simple, effective, and data-efficient method called Step-DPO, which treats individual reasoning steps as units for preference optimization rather than evaluating answers holistically. Additionally, we have developed a data construction pipeline for Step-DPO, enabling the creation of a high-quality dataset containing 10K step-wise preference pairs. We also observe that in DPO, self-generated data is more effective than data generated by humans or GPT-4, due to the latter's out-of-distribution nature. Our findings demonstrate that as few as 10K preference data pairs and fewer than 500 Step-DPO training steps can yield a nearly 3% gain in accuracy on MATH for models with over 70B parameters. Notably, Step-DPO, when applied to Qwen2-72B-Instruct, achieves scores of 70.8% and 94.0% on the test sets of MATH and GSM8K, respectively, surpassing a series of closed-source models, including GPT-4-1106, Claude-3-Opus, and Gemini-1.5-Pro. Our code, data, and models are available at https://github.com/dvlab-research/Step-DPO.

Related papers

• Minor DPO reject penalty to increase training robustness [8.971332948872185]
  Learning from human preference is a paradigm used in large-scale language model (LLM) fine-tuning step to better align pretrained LLM to human preference for downstream task. Recently, Direct Preference Optimization (DPO) has been proposed to solve the alignment problem with a simplified RL-free method. In this article, we analyze the working mechanism of $beta$ in DPO, disclose its syntax difference between RL algorithm and DPO, and understand the potential shortage brought by the DPO simplification.
  arXiv  Detail & Related papers  (2024-08-19T09:29:31Z)
• Bridging and Modeling Correlations in Pairwise Data for Direct Preference Optimization [75.1240295759264]
  We propose an effective framework for Bridging and Modeling Correlations in pairwise data, named BMC. We increase the consistency and informativeness of the pairwise preference signals through targeted modifications. We identify that DPO alone is insufficient to model these correlations and capture nuanced variations.
  arXiv  Detail & Related papers  (2024-08-14T11:29:47Z)
• Crafting Efficient Fine-Tuning Strategies for Large Language Models [2.633490094119608]
  Fine-tuning large language models (LLMs) with as few as 200 samples can improve model accuracy from 70% to 88% in a product attribute extraction task. A bayesian hyperparameter optimization method, which evaluates models at 20% of total training time, correlates strongly with final model performance. This approach led to a 2% improvement in accuracy over baseline models when evaluated on an independent test set.
  arXiv  Detail & Related papers  (2024-07-18T21:36:00Z)
• Step-Controlled DPO: Leveraging Stepwise Error for Enhanced Mathematical Reasoning [38.127313175508746]
  Step-Controlled DPO creates negative samples of mathematical reasoning rationales that start making errors at a specified step. By applying these samples in DPO training, SCDPO can better align the model to understand reasoning errors and output accurate reasoning steps.
  arXiv  Detail & Related papers  (2024-06-30T17:59:07Z)
• Iterative Length-Regularized Direct Preference Optimization: A Case Study on Improving 7B Language Models to GPT-4 Level [50.897438358317686]
  We show that iLR-DPO can enhance a 7B model to perform on par with GPT-4 without increasing verbosity. Specifically, our 7B model achieves a $50.5%$ length-controlled win rate against $texttGPT-4 Preview$ on AlpacaEval 2.0.
  arXiv  Detail & Related papers  (2024-06-17T17:55:38Z)
• Self-Play Preference Optimization for Language Model Alignment [75.83359213697854]
  Recent advancements suggest that directly working with preference probabilities can yield a more accurate reflection of human preferences. We propose a self-play-based method for language model alignment, which treats the problem as a constant-sum two-player game. Our approach, dubbed Self-Play Preference Optimization (SPPO), utilizes iterative policy updates to provably approximate the Nash equilibrium.
  arXiv  Detail & Related papers  (2024-05-01T17:59:20Z)
• Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning [55.96599486604344]
  We introduce an approach aimed at enhancing the reasoning capabilities of Large Language Models (LLMs) through an iterative preference learning process. We use Monte Carlo Tree Search (MCTS) to iteratively collect preference data, utilizing its look-ahead ability to break down instance-level rewards into more granular step-level signals. The proposed algorithm employs Direct Preference Optimization (DPO) to update the LLM policy using this newly generated step-level preference data.
  arXiv  Detail & Related papers  (2024-05-01T11:10:24Z)
• Smaug: Fixing Failure Modes of Preference Optimisation with DPO-Positive [15.066029556877721]
  We show theoretically that the standard DPO loss can lead to a reduction of the model's likelihood of the preferred examples. We design DPO-Positive (DPOP), a new loss function and training procedure which avoids this failure mode. Surprisingly, we find that DPOP outperforms DPO and other fine-tuning procedures across a wide variety of datasets and downstream tasks.
  arXiv  Detail & Related papers  (2024-02-20T18:42:34Z)
• Pruning Large Language Models via Accuracy Predictor [0.0]
  Large language models (LLMs) containing tens of billions of parameters (or even more) have demonstrated impressive capabilities in various NLP tasks. We propose a novel pruning approach: firstly, a training set of a certain number of architecture-accuracy pairs is established, and then a non-neural model is trained as an accuracy predictor.
  arXiv  Detail & Related papers  (2023-09-18T06:38:24Z)
• A Meta-Learning Approach to Predicting Performance and Data Requirements [163.4412093478316]
  We propose an approach to estimate the number of samples required for a model to reach a target performance. We find that the power law, the de facto principle to estimate model performance, leads to large error when using a small dataset. We introduce a novel piecewise power law (PPL) that handles the two data differently.
  arXiv  Detail & Related papers  (2023-03-02T21:48:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.