'lr': 0.001,
'betas': (0.9, 0.999),
'eps': 1e-08,
'weight_decay': 0,
'amsgrad': False
}
loss_params = {'n_critic': 5, 'grad_lambda': 10, 'l1_lambda': (1e4) / 0.05}
paper_opts = adam_opts
paper_opts['betas'] = (0.5, 0.999)
paper_opts['lr'] = 0.0002
range_compress_transform, range_compress_inv_transform = \
create_range_compress_transforms(k_values={"dm": 4, "pressure": 4},
modes={'dm': 'x/(1+x)',
'pressure': 'log'})
transform = chain_transformations([range_compress_transform, atleast_3d])
inv_transform = chain_transformations([squeeze, range_compress_inv_transform])
def Schedule(name,
transform=transform,
inv_transform=inv_transform,
loss_params=loss_params,
paper_opts=paper_opts,
epoch_end=100,
n_test=64):
schedule = {
import torch
import os
from baryon_painter.utils.data_transforms import \
create_range_compress_transforms, chain_transformations, \
atleast_3d, squeeze
from src.tools.data import Data
#TODO
d = Data(os.environ['DATA_DIR'])
redshifts = d.redshifts_list()[0:8]
range_compress_transform, range_compress_inv_transform = \
create_range_compress_transforms(k_values={"dm": [4.0, 1.0],
"pressure": [4.0, 1.0]},
modes={'dm': 'shift-log-cam',
'pressure': 'shift-log-cam'},
eps=1e-4)
transform = chain_transformations([range_compress_transform,
atleast_3d])
inv_transform = chain_transformations([squeeze,
range_compress_inv_transform])
init_params = {
'g': 'kaiming',
'd': 'kaiming',
}
n_scale = 1
n_aux_label = 1
label_fields = ["pressure"]
redshifts = [0.0, 0.125, 0.25, 0.375, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0]
# redshifts = [0.0, 0.125, 0.375, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0]
range_compress_transform, range_compress_inv_transform = data_transforms.create_range_compress_transforms(
k_values={
"dm": 4.0,
"pressure": 4
},
modes={
"dm": "shift-log",
"pressure": "shift-log"
},
# k_values={"dm" : (0.01, 4.0),
# "pressure" : (1.0, 0.3)},
# modes={"dm":"shift-log-2p",
# "pressure" : "shift-log-2p"},
# k_values={"dm" : (2,1),
# "pressure" : (1,0)},
# modes={"dm":"x/(1+x)",
# "pressure" : "x/(1+x)"},
eps=1e-4)
# range_compress_transform, range_compress_inv_transform = data_transforms.create_range_compress_transforms(
# k_values={"dm" : 1.5,
# "pressure" : 1},
# modes={"dm":"x/(1+x)",
# "pressure" : "1/x"})
}
loss_params = {'n_critic': 5, 'grad_lambda': 10, 'l1_lambda': (1e4) / 0.05}
paper_opts = adam_opts
paper_opts['betas'] = (0.5, 0.999)
paper_opts['lr'] = 0.0002
split_scale_transform, inv_split_scale_transform = data_transforms.create_split_scale_transform(
n_scale=2, step_size=8, include_original=False, truncate=2.0)
range_compress_transform, range_compress_inv_transform = data_transforms.create_range_compress_transforms(
k_values={
"dm": 1.5,
"pressure": 4
},
modes={
"dm": "x/(1+x)",
"pressure": "log"
})
transform = data_transforms.chain_transformations([
range_compress_transform,
split_scale_transform,
data_transforms.atleast_3d,
])
inv_transform = data_transforms.chain_transformations([
data_transforms.squeeze,
inv_split_scale_transform,
range_compress_inv_transform,
