UDA_OH_DANN.yaml

TRAINER: 'DANN'

TASK:
  NAME: 'UDA'

MODEL:
  PRETRAIN: True
  BASENET: 'resnet50'
  DNET: 'Discriminator'
  D_INDIM: 1024
  D_OUTDIM: 1

DATASET:
  NAME: 'office_home'
  ROOT: 'office_home'

METHOD:
  W_ALG: 0.1
  ENT: False

TRAIN:
  BATCH_SIZE_SOURCE: 36
  BATCH_SIZE_TARGET: 36
  LR: 0.001
  TTL_ITE: 8000
  TEST_FREQ: 500
  PRINT_FREQ: 50
  SAVE_FREQ: 5000
  OUTPUT_ROOT: 'exp'