paper-with-me

Papers

Dive into Big Model Training

2022-07-25 · Qinghua Liu, Yuxiang Jiang

The increasing scale of model size and continuous improvement of performance herald the arrival of the Big Model era. In this report, we explore what and how the big model training works by diving into training objectives and training methodologies. Specifically,training objectives describe how to leverage web-scale data to develop extremely capable and incredibly large models based on self-supervised learning, and training methodologies which are based on distributed training describe how to make big model training a reality. We summarize the existing training methodologies into three main categories: training parallelism, memory-saving technologies, and model sparsity design. Training parallelism can be categorized into data, pipeline, and tensor parallelism according to the dimension of parallelism that takes place. Memory-saving technologies are orthogonal and complementary to training parallelism. And model sparsity design further scales up the model size with a constant computational cost. A continuously updated paper list of big model training is provided at https://github.com/qhliu26/BM-Training.

📄 PDF Abstract BibTeX arXiv:2207.11912

Code (1)

qhliu26/bm-training 공식 구현 pytorch

Tasks

modelSelf-Supervised Learning

Similar Papers 제목 키워드 기반

Diversity-Aware Policy Optimization for Large Language Model Reasoning

2025-05-29 · Jian Yao, Ran Cheng, Xingyu Wu, Jibin Wu 외

The reasoning capabilities of large language models (LLMs) have advanced rapidly, particularly following the release of DeepSeek R1, which has inspired a surge of research into data quality and reinforcement learning (RL…

DiversityLanguage ModelingLanguage ModellingLarge Language Model+2

An Empirical Study on Noisy Data and LLM Pretraining Loss Divergence

2026-02-02 · Qizhen Zhang, Ankush Garg, Jakob Foerster, Niladri Chatterji 외 arxiv

Large-scale pretraining datasets drive the success of large language models (LLMs). However, these web-scale corpora inevitably contain large amounts of noisy data due to unregulated web content or randomness inherent in…

Learning GANs and Ensembles Using Discrepancy

2019-10-20 · NeurIPS 2019 12 · Ben Adlam, Corinna Cortes, Mehryar Mohri, Ningshan Zhang

Generative adversarial networks (GANs) generate data based on minimizing a divergence between two distributions. The choice of that divergence is therefore critical. We argue that the divergence must take into account th…

Domain Adaptation

State Diversity Matters in Offline Behavior Distillation

2025-12-07 · Shiye Lei, Zhihao Cheng, Dacheng Tao arxiv

Offline Behavior Distillation (OBD), which condenses massive offline RL data into a compact synthetic behavioral dataset, offers a promising approach for efficient policy training and can be applied across various downst…

Offline RL

Controllable Dual Skew Divergence Loss for Neural Machine Translation

2019-08-22 · Zuchao Li, Hai Zhao, Yingting Wu, Fengshun Xiao 외

In sequence prediction tasks like neural machine translation, training with cross-entropy loss often leads to models that overgeneralize and plunge into local optima. In this paper, we propose an extended loss function c…

Machine TranslationNMTTranslation