kxngluca/DL-Art-School
1
0
Fork 0
forked from mrq/DL-Art-School
Code Pull Requests Activity

Fix bugs with srflow after refactor

Browse Source
This commit is contained in:
James Betker 2020-12-19 10:28:23 -07:00
parent 4328c2f713
commit ae666dc520
2 changed files with 4 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
codes/models/srflow/FlowStep.py
View File

@ -2,6 +2,8 @@ import torch
from torch import nn as nn from torch import nn as nn
import models.srflow.Permutations import models.srflow.Permutations
import models.srflow.FlowAffineCouplingsAblation
import models.srflow.FlowActNorms
def getConditional(rrdbResults, position): def getConditional(rrdbResults, position):

4
recipes/srflow/train_div2k_srflow.yml
View File

@ -24,12 +24,12 @@ datasets:
networks: networks:
generator: generator:
type: generator type: generator
which_model_G: srflow_orig which_model_G: srflow
nf: 64 nf: 64
nb: 23 nb: 23
K: 16 K: 16
scale: 4 scale: 4
initial_stride: 2 initial_stride: 1
flow_scale: 4 flow_scale: 4
train_RRDB: false # <-- Start false. After some time, ~20k-50k steps, set to true. TODO: automate this. train_RRDB: false # <-- Start false. After some time, ~20k-50k steps, set to true. TODO: automate this.
pretrain_rrdb: ../experiments/pretrained_rrdb.pth # <-- Insert path to your pretrained RRDB here. pretrain_rrdb: ../experiments/pretrained_rrdb.pth # <-- Insert path to your pretrained RRDB here.