LDData / mpmc /dim_2.nsamples_10.nbatch_1.LfL2ctr.b_1.txt
Sou-Cheng's picture
Upload folder using huggingface_hub
2b3e7f3 verified
# MPMC pretrained pointset
# Contains following parameters (refer to QMCPy's MPMC documentation for more details):
# 'loss_fn': L2ctr
# 'lr': 0.001
# 'nlayers': 3
# 'weight_decay': 1e-06
# 'nhid': 32
# 'nbatch': 1
# 'epochs': 200000
# 'start_reduce': 100000
# 'radius': 0.35
# 'weights': None
# 'n_projections': 15
2 #d, the number of dimensions
10 #n, the number of points
1.531098634004592896e-01 1.531193107366561890e-01
4.982938766479492188e-01 4.999959468841552734e-01
1.531102061271667480e-01 8.468897342681884766e-01
8.468816876411437988e-01 1.531084328889846802e-01
6.479213833808898926e-01 6.479219794273376465e-01
5.021970272064208984e-01 4.999971389770507812e-01
8.468926548957824707e-01 8.468828797340393066e-01
3.520831763744354248e-01 6.479098200798034668e-01
3.520871996879577637e-01 3.520870506763458252e-01
6.479202508926391602e-01 3.520825803279876709e-01