arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.09643cs.LGeess.SP
收藏

尝试盲约束下降实验

Attempted Blind Constrained Descent Experiments

Carnegie Mellon University(卡内基梅隆大学)

Prasad N R

摘要:

Blind Descent使用受约束但有指导的方法来学习权重。概率密度函数在维度的无限空间中是非零的(例子:高斯和正态概率分布函数)。在Blind Descent论文中,一些隐含的想法,包括逐层训练和逐过滤器训练(具有不同的批量大小),被提出作为可能的贪婪解决方案。对类似实验的结果进行了讨论。本文实验的Octave(和提出的PyTorch变体)源代码可以在https://github.com/PrasadNR/Attempted-Blind-Constrained-Descent-Experiments-ABCDE-上找到。这与https://github.com/akshat57/Blind-Descent的原始PyTorch源代码的ABCDE衍生物进行了比较。

英文摘要:

Blind Descent uses constrained but, guided approach to learn the weights. The probability density function is non-zero in the infinite space of the dimension (case in point: Gaussians and normal probability distribution functions). In Blind Descent paper, some of the implicit ideas involving layer by layer training and filter by filter training (with different batch sizes) were proposed as probable greedy solutions. The results of similar experiments are discussed. Octave (and proposed PyTorch variants) source code of the experiments of this paper can be found at https://github.com/PrasadNR/Attempted-Blind-Constrained-Descent-Experiments-ABCDE- . This is compared against the ABCDE derivatives of the original PyTorch source code of https://github.com/akshat57/Blind-Descent .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑