different spec_min and spec_max values

Hi,

I use the codes in diffsinger_task.py to calculate the spec_min and spec_max for popcs train set. But I got the different values from the usr/config/popcs_ds_beta6.yaml. Especially for spec_min, there is a big gap between the calculate value and popcs_ds_beta6.yaml.
Could you please help if I miss something in the calculation step?
Thank you very much.

The codes I used is:
def build_tts_model(self):
import torch
from tqdm import tqdm
v_min = torch.ones([80]) * 100
v_max = torch.ones([80]) * -100
for i, ds in enumerate(tqdm(self.dataset_cls('train'))):
v_max = torch.max(torch.max(ds['mel'].reshape(-1, 80), 0)[0], v_max)
v_min = torch.min(torch.min(ds['mel'].reshape(-1, 80), 0)[0], v_min)
if i % 100 == 0:
print(i, v_min, v_max)
print('final', v_min, v_max)

The outputs of spec_mix and spec_max by diffsinger_task.py are the followings:
final tensor([-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10., -10.,
-10., -10., -10., -10., -10., -10., -10., -10.]) tensor([ 0.2640, 0.0904, -0.1892, 0.0053, 0.1487, 0.1805, 0.1374, 0.1493,
0.2692, 0.2099, 0.2410, 0.3416, 0.3569, 0.3552, 0.2916, 0.2683,
0.2851, 0.2865, 0.2468, 0.2173, 0.1341, 0.1747, 0.1961, 0.2825,
0.1238, 0.1087, 0.0675, 0.1542, -0.0440, 0.0215, 0.0748, -0.0377,
0.0039, -0.0455, -0.1376, -0.0589, -0.2367, -0.1709, -0.1267, -0.1893,
-0.3908, -0.5237, -0.3885, -0.4496, -0.6173, -0.5266, -0.3997, -0.3948,
-0.4007, -0.2719, -0.3427, -0.3948, -0.4103, -0.2829, -0.4692, -0.4576,
-0.4587, -0.4601, -0.5434, -0.5312, -0.6687, -0.7729, -0.8233, -0.8344,
-0.8734, -0.9413, -0.8186, -0.7304, -0.7915, -0.8383, -0.6616, -0.8759,
-0.8764, -0.8546, -0.9282, -0.8728, -0.9349, -1.0676, -1.1571, -1.5266])

while the spec_mix and spec_max in popcs_ds_beta6.yaml are the followings:

spec_min: [-6.8276, -7.0270, -6.8142, -7.1429, -7.6669, -7.6000, -7.1148, -6.9640,
-6.8414, -6.6596, -6.6880, -6.7439, -6.7986, -7.4940, -7.7845, -7.6586,
-6.9288, -6.7639, -6.9118, -6.8246, -6.7183, -7.1769, -6.9794, -7.4513,
-7.3422, -7.5623, -6.9610, -6.8158, -6.9595, -6.8403, -6.5688, -6.6356,
-7.0209, -6.5002, -6.7819, -6.5232, -6.6927, -6.5701, -6.5531, -6.7069,
-6.6462, -6.4523, -6.5954, -6.4264, -6.4487, -6.7070, -6.4025, -6.3042,
-6.4008, -6.3857, -6.3903, -6.3094, -6.2491, -6.3518, -6.3566, -6.4168,
-6.2481, -6.3624, -6.2858, -6.2575, -6.3638, -6.4520, -6.1835, -6.2754,
-6.1253, -6.1645, -6.0638, -6.1262, -6.0710, -6.1039, -6.4428, -6.1363,
-6.1054, -6.1252, -6.1797, -6.0235, -6.0758, -5.9453, -6.0213, -6.0446]
spec_max: [ 0.2645, 0.0583, -0.2344, -0.0184, 0.1227, 0.1533, 0.1103, 0.1212,
0.2421, 0.1809, 0.2134, 0.3161, 0.3301, 0.3289, 0.2667, 0.2421,
0.2581, 0.2600, 0.1394, 0.1907, 0.1082, 0.1474, 0.1680, 0.2550,
0.1057, 0.0826, 0.0423, 0.1203, -0.0701, -0.0056, 0.0477, -0.0639,
-0.0272, -0.0728, -0.1648, -0.0855, -0.2652, -0.1998, -0.1547, -0.2167,
-0.4181, -0.5463, -0.4161, -0.4733, -0.6518, -0.5387, -0.4290, -0.4191,
-0.4151, -0.3042, -0.3810, -0.4160, -0.4496, -0.2847, -0.4676, -0.4658,
-0.4931, -0.4885, -0.5547, -0.5481, -0.6948, -0.7968, -0.8455, -0.8392,
-0.8770, -0.9520, -0.8749, -0.7297, -0.8374, -0.8667, -0.7157, -0.9035,
-0.9219, -0.8801, -0.9298, -0.9009, -0.9604, -1.0537, -1.0781, -1.3766]

Start with the calculation in diffsinger_task.py and compare its dataset_cls('train') mel values with usr/config/popcs_ds_beta6.yaml. Trace how the training data and configuration values are produced, then document or correct the source of the spec_min discrepancy while confirming spec_max behavior.