SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient

Authors: Lantao Yu, Weinan Zhang, Jun Wang, Yong Yu

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on synthetic data and real-world tasks demonstrate significant improvements over strong baselines. and Extensive experiments based on synthetic and real data are conducted to investigate the efficacy and properties of the proposed Seq GAN.
Researcher Affiliation Academia Lantao Yu, Weinan Zhang, Jun Wang, Yong Yu Shanghai Jiao Tong University, University College London {yulantao,wnzhang,yyu}@apex.sjtu.edu.cn, j.wang@cs.ucl.ac.uk
Pseudocode Yes Algorithm 1 Sequence Generative Adversarial Nets
Open Source Code Yes Experiment code: https://github.com/Lantao Yu/Seq GAN
Open Datasets Yes For music composition, we use Nottingham7 dataset as our training data, which is a collection of 695 music of folk tunes in midi file format. ... 7http://www.iro.umontreal.ca/ lisa/deep/data
Dataset Splits No The paper mentions using a 'training set S' for synthetic data and 'training data' for real-world scenarios, and a 'test stage' for evaluation. However, it does not explicitly provide details about specific training, validation, and test dataset splits (e.g., percentages, sample counts, or explicit mention of a validation set).
Hardware Specification No The paper does not provide specific hardware details (e.g., exact GPU/CPU models, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions architectural components like LSTM and CNN, and optimization algorithms such as Adam and RMSprop, but it does not specify any software dependencies with version numbers (e.g., Python, TensorFlow, PyTorch, CUDA versions).
Experiment Setup Yes To set up the synthetic data experiments, we first initialize the parameters of an LSTM network following the normal distribution N(0, 1)... A curriculum rate ω is used to control the probability of replacing the true tokens with the generated ones. To get a good and stable performance, we decrease ω by 0.002 for every training epoch. ... the kernel size is from 1 to T and the number of each kernel size is between 100 to 2004. Dropout (Srivastava et al. 2014) and L2 regularization are used to avoid over-fitting. ... the g-steps, d-steps and k parameters in Algorithm 1 have a large effect on the convergence and performance of Seq GAN. Figure 3 shows the effect of these parameters.