hongyehu / rg-flow

This is project page for the paper "RG-Flow: a hierarchical and explainable flow model based on renormalization group and sparse prior". Paper link: https://arxiv.org/abs/2010.00029

My name is Hong-Ye Hu. I am currently an HQI Postdoctoral Fellow at Harvard University. I am currently working on the interface between quantum computation, machine learning and many-body physics.

🖥️: Working Experience:

• September 2022 - Present, Harvard Quantum Initiative Fellow @ Harvard Physics, and Harvard Quantum Science & Engineering
• May-August 2022, Quantum Algorithm Intern @ QuEra Computing Inc.
• June-September 2021 & March-May 2022, Feynman Research Intern @ NASA quantum AI Lab, Ames Research Center, supported by NAMS Student R&D program.
• Sept 2016-March 2018, Research Intern @ Salk Institute for biological studies. Worked on information theory and vision systems.

📖 Education:

• 2018 March - 2022 February University of California, San Diego, Department of Physics. Advisor: Prof. Yi-Zhuang You.
• 2012 September - 2016 June Peking University, Department of Physics. Advisor: Prof. Biao Wu

📋 You can find more info at my Harvard webpage

rg-flow's Issues

run "python main.py --data minst32" and return error

Dear developers,
I used this code to test on Windows, however I got following message, after I run command " python main.py --data mnist32"

L = 32
batch_size = 64
clip_grad = 1
cuda =
data = mnist32
data_path = ./data
depth = 8
device = cpu
device_count = 1
dtype = float32
epoch = 500
keep_epoch = 10
kernel_size = 4
lr = 0.001
nchannels = 3
net_name = laplace_nl8,6,4,2_nr4_nm2_nh512
nhidden = 512
nhidden_list = [512, 512, 512, 512, 512, 512, 512, 512]
nlayers = 8,6,4,2
nlayers_list = [8, 8, 6, 6, 4, 4, 2, 2]
nmlp = 2
nmlp_list = [2, 2, 2, 2, 2, 2, 2, 2]
no_stdout = False
nresblocks = 4
nresblocks_list = [4, 4, 4, 4, 4, 4, 4, 4]
out_dir = ./saved_model
out_filename = ./saved_model\mnist32\laplace_nl8,6,4,2_nr4_nm2_nh512\out
out_infix =
plot_epoch = 1
plot_filename = ./saved_model\mnist32\laplace_nl8,6,4,2_nr4_nm2_nh512\epoch_sample
print_step = 1
prior = laplace
save_epoch = 1
subnet = rnvp
weight_decay = 5e-05

nparams in each RG layer: [20129536, 20129536, 15097152, 15097152, 10064768, 10064768, 5032384, 5032384]
Total nparams: 100647680
Traceback (most recent call last):
File "main.py", line 246, in
main()
File "main.py", line 182, in main
train_split, val_split, data_info = utils.load_dataset()
File "D:\Research Data\Coding\Python\RG-Flow-master\code\utils\data_utils.py", line 97, in load_dataset
assert data_info.channel == args.nchannels
AssertionError
(pytorch) PS D:\Research Data\Coding\Python\RG-Flow-master\code>