Fearnworks/EDI_241210

该数据集包含《质量效应》游戏中EDI角色的语音数据。WAV文件总时长为35.51分钟，其中最长的片段为Mass_Effect_EDI_72.wav，时长为26.68秒，帧数为533.58帧。训练列表train_list.txt包含88行，验证列表val_list.txt包含10行。标记和音素的长度统计显示，最小标记长度为409，最大为509，平均为438.46。总标记数为38513，总音素数为5763，每个文件的平均标记数为437.65，平均字符数为65.49。

This dataset contains voice data of the character EDI from the game Mass Effect. The total duration of WAV files is 35.51 minutes, with the longest snippet being Mass_Effect_EDI_72.wav, lasting 26.68 seconds and consisting of 533.58 frames. The training list train_list.txt includes 88 lines, and the validation list val_list.txt includes 10 lines. Token and phoneme length statistics show a minimum token length of 409, a maximum of 509, and an average of 438.46. The total number of tokens is 38513, and the total number of phonemes is 5763. The average number of tokens per file is 437.65, and the average number of characters per file is 65.49.