mask_rcnn_R50FPN_CLIP_sigmoid.yaml

MODEL:
  META_ARCHITECTURE: "CustomRCNN"
  WEIGHTS: "detectron2://ImageNetPretrained/MSRA/R-50.pkl"
  MASK_ON: True
  BACKBONE:
    NAME: "build_resnet_fpn_backbone"
  RESNETS:
    OUT_FEATURES: [ "res2", "res3", "res4", "res5" ]
    DEPTH: 50
  FPN:
    IN_FEATURES: [ "res2", "res3", "res4", "res5" ]
  ANCHOR_GENERATOR:
    SIZES: [ [ 32 ], [ 64 ], [ 128 ], [ 256 ], [ 512 ] ]
    ASPECT_RATIOS: [ [ 0.5, 1.0, 2.0 ] ]
  RPN:
    IN_FEATURES: [ "p2", "p3", "p4", "p5", "p6" ]
    PRE_NMS_TOPK_TRAIN: 2000
    PRE_NMS_TOPK_TEST: 1000
    POST_NMS_TOPK_TRAIN: 1000
    POST_NMS_TOPK_TEST: 1000
  ROI_HEADS:
    NAME: "CustomStandardROIHeads"
    IN_FEATURES: [ "p2", "p3", "p4", "p5" ]
    NUM_CLASSES: 1203
    SCORE_THRESH_TEST: 0.0001
  ROI_BOX_HEAD:
    NAME: "FastRCNNConvFCHead"
    NUM_FC: 2
    POOLER_RESOLUTION: 7
    CLS_AGNOSTIC_BBOX_REG: True
    USE_ZEROSHOT_CLS: True
    ZEROSHOT_WEIGHT_PATH: 'datasets/zeroshot_weights/lvis_v1_clip_a+photo+cname.npy'
    CAT_FREQ_PATH: 'datasets/lvis/lvis_v1_train_norare_cat_info.json'
    USE_SIGMOID_CE: True
    USE_FED_LOSS: True
    IGNORE_ZERO_CATS: True
    USE_BIAS: -4.6
  ROI_MASK_HEAD:
    NAME: "MaskRCNNConvUpsampleHead"
    CLS_AGNOSTIC_MASK: True
    NUM_CONV: 4
    POOLER_RESOLUTION: 14
INPUT:
  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
  MAX_SIZE_TRAIN: 1333
DATASETS:
  TRAIN: ("lvis_v1_train_norare",)
  TEST: ("lvis_v1_val",)
TEST:
  DETECTIONS_PER_IMAGE: 300
SOLVER:
  IMS_PER_BATCH: 16 # Overall batch-size
  BASE_LR: 0.02
  STEPS: (60000, 80000)
  MAX_ITER: 90000
  WARMUP_FACTOR: 0.001
  WARMUP_ITERS: 1000
DATALOADER:
  SAMPLER_TRAIN: "RepeatFactorTrainingSampler"
  REPEAT_THRESHOLD: 0.001
VERSION: 2
OUTPUT_DIR: output/lvis_ovd_base