GMVSE源代码学习记录之“参数配置”

一、参数配置

Parameters:

Name	Type	Description	Note
mode	enum(str)	pretrain, train or evaluate.	模式：预训练、训练、评估
data_filename	str	data file (e.g., ./data/processed_porto.csv).	数据及文件名
map_size	(int, int)	size of the grid map.	网格图大小
model	str	choose a model	GMVSAE
token_dim	int	dimensionality of grid token.	网格标签的维度
rnn_dim	int	dimensionality of rnn hidden state.	RNN隐藏层神经元个数
cluster_num	int	number of Gaussian components.	高斯分量数目
model_dir	str	directory to save/load a model during training or eval.	在训练或者评估期间保存/加载模型的文件夹
pretrain_dir	str	directory to save/load a model during pretraining.	在预训练期间保存/加载模型的文件夹
num_negs	int	number of negative samples during training.	在训练期间负样本的数目
optimizer	enum(str)	training optimizer (e.g., adam or sgd).	训练优化器
learning_rate	float	learning rate for training.	训练时的学习率
batch_size	int	minibatch size	批量数
num_epochs	int	number of passes over the training data.	训练轮数
partial_ratio	float	partial trajectory evaluation	选取轨迹序列的比例
log_steps	int	number of batches to print the log info.	输出日志信息的批量数
gpu_id	str	0	GPU的id号

 

* 论文+代码参考partial_ratio解释——属于个人分析的结果，不知道理解的对不对，欢迎一起讨论一下。