search

CVMI-Lab / SparseKD

(NeurlPS 2022) Towards Efficient 3D Object Detection with Knowledge Distillation
Apache License 2.0
110 stars 10 forks source link

TypeError: _map_state_dict_to_module() missing 1 required positional argument: 'prefix' #3

Closed HuangCongQing closed 1 year ago

HuangCongQing commented 1 year ago

image

I modified this ENABLED parameter to True

    REMAP_PRETRAIN: #
        ENABLED: True # False  True!!!!
        WAY: BN_SCALE # 重映射方式
        BN_SCALE:
            ABS: True # 参数
        OFA:
            l1_norm: max

When I run the code, I get an error. How shall I approach this problom?Thx.

python train.py --cfg_file=cfgs/waymo_models/cp-pillar/cp-pillar-v0.4_sparsekd.yaml  --batch_size=4 --epochs=20
jihanyang commented 1 year ago

Hello, please ignore the REMAP_PRETRAIN in this config as it is not needed for cp-pillar with only voxel size difference (teacher and student have the same architecture). Also, the remapping way BN_SCALE is not yet supported.

If you want to try this part, please refer to: https://github.com/CVMI-Lab/SparseKD/blob/8572093edc77d4217f7596d9e1de68c6dc77e420/tools/cfgs/waymo_models/cp-voxel/cp-voxel-s_sparsekd.yaml#L212-L218

HuangCongQing commented 1 year ago

image RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking arugment for argument index in method wrapper_index_select)

Thx. Then I encountered a new problemRuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking arugment for argument index in method wrapper_index_select). but it has been solved, and the solution code is as follows. You can be updated to the repo.

            # fix: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!
            device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
            _stu_input_dim_idx = _stu_input_dim_idx.to(device)

image

jihanyang commented 1 year ago

Can you provide the full log? I can still run the command successfully.

HuangCongQing commented 1 year ago

This is the full error log.

python train.py --cfg_file=cfgs/waymo_models/cp-pillar/cp-pillar-v0.4_sparsekd.yaml

/home/chongqinghuang/anaconda3/envs/pcdet/bin/python train.py --cfg_file=cfgs/waymo_models/cp-pillar/cp-pillar-v0.4_sparsekd.yaml --batch_size=4 --epochs=20
2023-01-09 10:38:07,662   INFO  **********************Start logging**********************
2023-01-09 10:38:07,662   INFO  CUDA_VISIBLE_DEVICES=ALL
2023-01-09 10:38:07,662   INFO  cfg_file         cfgs/waymo_models/cp-pillar/cp-pillar-v0.4_sparsekd.yaml
2023-01-09 10:38:07,662   INFO  batch_size       4
2023-01-09 10:38:07,662   INFO  epochs           20
2023-01-09 10:38:07,662   INFO  workers          4
2023-01-09 10:38:07,662   INFO  extra_tag        default
2023-01-09 10:38:07,662   INFO  ckpt             None
2023-01-09 10:38:07,662   INFO  pretrained_model ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth
2023-01-09 10:38:07,662   INFO  launcher         none
2023-01-09 10:38:07,662   INFO  tcp_port         18888
2023-01-09 10:38:07,662   INFO  sync_bn          False
2023-01-09 10:38:07,662   INFO  fix_random_seed  False
2023-01-09 10:38:07,662   INFO  ckpt_save_interval 1
2023-01-09 10:38:07,662   INFO  local_rank       0
2023-01-09 10:38:07,662   INFO  max_ckpt_save_num 30
2023-01-09 10:38:07,662   INFO  merge_all_iters_to_one_epoch False
2023-01-09 10:38:07,662   INFO  set_cfgs         None
2023-01-09 10:38:07,662   INFO  max_waiting_mins 0
2023-01-09 10:38:07,662   INFO  start_epoch      0
2023-01-09 10:38:07,662   INFO  save_to_file     False
2023-01-09 10:38:07,663   INFO  teacher_ckpt     ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth
2023-01-09 10:38:07,663   INFO  cfg.ROOT_DIR: /home/chongqinghuang/code/light_weight/SparseKD
2023-01-09 10:38:07,663   INFO  cfg.LOCAL_RANK: 0
2023-01-09 10:38:07,663   INFO  cfg.CLASS_NAMES: ['Vehicle', 'Pedestrian', 'Cyclist']
2023-01-09 10:38:07,663   INFO  cfg.TEACHER_CKPT: ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth
2023-01-09 10:38:07,663   INFO  cfg.PRETRAINED_MODEL: ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth
2023-01-09 10:38:07,663   INFO  
cfg.DATA_CONFIG = edict()
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATASET: WaymoDataset
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATA_PATH: ../data/waymo
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.PROCESSED_DATA_TAG: waymo_processed_data_v0_5_0
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.POINT_CLOUD_RANGE: [-73.6, -73.6, -2, 73.6, 73.6, 4.0]
2023-01-09 10:38:07,663   INFO  
cfg.DATA_CONFIG.DATA_SPLIT = edict()
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATA_SPLIT.train: train
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATA_SPLIT.test: val
2023-01-09 10:38:07,663   INFO  
cfg.DATA_CONFIG.SAMPLED_INTERVAL = edict()
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.train: 5
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.test: 5
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.FILTER_EMPTY_BOXES_FOR_TRAIN: True
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DISABLE_NLZ_FLAG_ON_POINTS: True
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.USE_SHARED_MEMORY: False
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.SHARED_MEMORY_FILE_LIMIT: 35000
2023-01-09 10:38:07,663   INFO  
cfg.DATA_CONFIG.DATA_AUGMENTOR = edict()
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.DISABLE_AUG_LIST: ['placeholder']
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.AUG_CONFIG_LIST: [{'NAME': 'gt_sampling', 'USE_ROAD_PLANE': False, 'DB_INFO_PATH': ['waymo_processed_data_v0_5_0_waymo_dbinfos_train_sampled_1.pkl'], 'USE_SHARED_MEMORY': True, 'DB_DATA_PATH': ['waymo_processed_data_v0_5_0_gt_database_train_sampled_1_global.npy'], 'PREPARE': {'filter_by_min_points': ['Vehicle:5', 'Pedestrian:5', 'Cyclist:5'], 'filter_by_difficulty': [-1]}, 'SAMPLE_GROUPS': ['Vehicle:15', 'Pedestrian:10', 'Cyclist:10'], 'NUM_POINT_FEATURES': 5, 'REMOVE_EXTRA_WIDTH': [0.0, 0.0, 0.0], 'LIMIT_WHOLE_SCENE': True}, {'NAME': 'random_world_flip', 'ALONG_AXIS_LIST': ['x', 'y']}, {'NAME': 'random_world_rotation', 'WORLD_ROT_ANGLE': [-0.78539816, 0.78539816]}, {'NAME': 'random_world_scaling', 'WORLD_SCALE_RANGE': [0.95, 1.05]}]
2023-01-09 10:38:07,663   INFO  
cfg.DATA_CONFIG.POINT_FEATURE_ENCODING = edict()
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.encoding_type: absolute_coordinates_encoding
2023-01-09 10:38:07,663   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.used_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 10:38:07,664   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.src_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 10:38:07,664   INFO  cfg.DATA_CONFIG.DATA_PROCESSOR: [{'NAME': 'mask_points_and_boxes_outside_range', 'REMOVE_OUTSIDE_BOXES': True}, {'NAME': 'shuffle_points', 'SHUFFLE_ENABLED': {'train': True, 'test': True}}, {'NAME': 'transform_points_to_voxels', 'VOXEL_SIZE': [0.4, 0.4, 6.0], 'MAX_POINTS_PER_VOXEL': 28, 'MAX_NUMBER_OF_VOXELS': {'train': 150000, 'test': 150000}}, {'NAME': 'transform_points_to_voxels_tea', 'VOXEL_SIZE': [0.32, 0.32, 6.0], 'MAX_POINTS_PER_VOXEL': 20, 'MAX_NUMBER_OF_VOXELS': {'train': 150000, 'test': 150000}}]
2023-01-09 10:38:07,664   INFO  cfg.DATA_CONFIG._BASE_CONFIG_: cfgs/dataset_configs/waymo_dataset.yaml
2023-01-09 10:38:07,664   INFO  
cfg.MODEL = edict()
2023-01-09 10:38:07,664   INFO  cfg.MODEL.NAME: CenterPoint
2023-01-09 10:38:07,664   INFO  
cfg.MODEL.VFE = edict()
2023-01-09 10:38:07,664   INFO  cfg.MODEL.VFE.NAME: PillarVFE
2023-01-09 10:38:07,664   INFO  cfg.MODEL.VFE.WITH_DISTANCE: False
2023-01-09 10:38:07,664   INFO  cfg.MODEL.VFE.USE_ABSLOTE_XYZ: True
2023-01-09 10:38:07,664   INFO  cfg.MODEL.VFE.USE_NORM: True
2023-01-09 10:38:07,664   INFO  cfg.MODEL.VFE.NUM_FILTERS: [64, 64]
2023-01-09 10:38:07,664   INFO  
cfg.MODEL.MAP_TO_BEV = edict()
2023-01-09 10:38:07,664   INFO  cfg.MODEL.MAP_TO_BEV.NAME: PointPillarScatter
2023-01-09 10:38:07,664   INFO  cfg.MODEL.MAP_TO_BEV.NUM_BEV_FEATURES: 64
2023-01-09 10:38:07,664   INFO  
cfg.MODEL.BACKBONE_2D = edict()
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.NAME: BaseBEVBackbone
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.WIDTH: 1.0
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.LAYER_NUMS: [3, 5, 5]
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.LAYER_STRIDES: [1, 2, 2]
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.NUM_FILTERS: [64, 128, 256]
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.UPSAMPLE_STRIDES: [1, 2, 4]
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.NUM_UPSAMPLE_FILTERS: [128, 128, 128]
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.FOCUS: False
2023-01-09 10:38:07,664   INFO  cfg.MODEL.BACKBONE_2D.ACT_FN: ReLU
2023-01-09 10:38:07,664   INFO  
cfg.MODEL.DENSE_HEAD = edict()
2023-01-09 10:38:07,664   INFO  cfg.MODEL.DENSE_HEAD.NAME: CenterHead
2023-01-09 10:38:07,664   INFO  cfg.MODEL.DENSE_HEAD.CLASS_AGNOSTIC: False
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.CLASS_NAMES_EACH_HEAD: [['Vehicle', 'Pedestrian', 'Cyclist']]
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SHARED_CONV_CHANNEL: 64
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.USE_BIAS_BEFORE_NORM: True
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.NUM_HM_CONV: 2
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_ORDER: ['center', 'center_z', 'dim', 'rot']
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT = edict()
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.out_channels: 2
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.num_conv: 2
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.out_channels: 1
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.num_conv: 2
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.out_channels: 3
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.num_conv: 2
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.out_channels: 2
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.num_conv: 2
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.FEATURE_MAP_STRIDE: 1
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.NUM_MAX_OBJS: 500
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.GAUSSIAN_OVERLAP: 0.1
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.MIN_RADIUS: 2
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.SHARPER: False
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG = edict()
2023-01-09 10:38:07,665   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS = edict()
2023-01-09 10:38:07,665   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.cls_weight: 1.0
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.loc_weight: 2.0
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.SCORE_THRESH: 0.1
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.POST_CENTER_LIMIT_RANGE: [-80, -80, -10.0, 80, 80, 10.0]
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.MAX_OBJ_PER_SAMPLE: 500
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.DENSE_HEAD.LOGIT_KD = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ENABLED: True
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.MODE: raw_pred
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.target: teacher
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 10:38:07,666   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 10:38:07,666   INFO  
cfg.MODEL.POST_PROCESSING = edict()
2023-01-09 10:38:07,666   INFO  cfg.MODEL.POST_PROCESSING.RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
2023-01-09 10:38:07,667   INFO  cfg.MODEL.POST_PROCESSING.EVAL_METRIC: waymo
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.POST_PROCESSING.EVAL_CLASSES = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/AP: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/AP']
2023-01-09 10:38:07,667   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/APH: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/APH']
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD: True
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.KD_LOSS = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.ENABLED: True
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.KD_LOSS.HM_LOSS = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.type: MSELoss
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.weight: 7.0
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.thresh: 0.0
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.fg_mask: True
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.soft_mask: True
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.rank: -1
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.KD_LOSS.REG_LOSS = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.type: RegLossCenterNet
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.weight: 0.0
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.KD_LOSS.FEATURE_LOSS = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.mode: rois
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.type: MSELoss
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.weight: 0.1
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.fg_mask: False
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.score_mask: False
2023-01-09 10:38:07,667   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.score_thresh: 0.3
2023-01-09 10:38:07,667   INFO  
cfg.MODEL.LOGIT_KD = edict()
2023-01-09 10:38:07,667   INFO  cfg.MODEL.LOGIT_KD.ENABLED: True
2023-01-09 10:38:07,667   INFO  cfg.MODEL.LOGIT_KD.MODE: raw_pred
2023-01-09 10:38:07,668   INFO  
cfg.MODEL.LOGIT_KD.ALIGN = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LOGIT_KD.ALIGN.target: teacher
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 10:38:07,668   INFO  
cfg.MODEL.LABEL_ASSIGN_KD = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 10:38:07,668   INFO  cfg.MODEL.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 10:38:07,668   INFO  
cfg.MODEL_TEACHER = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.NAME: CenterPoint
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.IS_TEACHER: True
2023-01-09 10:38:07,668   INFO  
cfg.MODEL_TEACHER.VFE = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.VFE.NAME: PillarVFE
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.VFE.WITH_DISTANCE: False
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.VFE.USE_ABSLOTE_XYZ: True
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.VFE.USE_NORM: True
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.VFE.NUM_FILTERS: [64, 64]
2023-01-09 10:38:07,668   INFO  
cfg.MODEL_TEACHER.MAP_TO_BEV = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.MAP_TO_BEV.NAME: PointPillarScatter
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.MAP_TO_BEV.NUM_BEV_FEATURES: 64
2023-01-09 10:38:07,668   INFO  
cfg.MODEL_TEACHER.BACKBONE_2D = edict()
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NAME: BaseBEVBackbone
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.LAYER_NUMS: [3, 5, 5]
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.LAYER_STRIDES: [1, 2, 2]
2023-01-09 10:38:07,668   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NUM_FILTERS: [64, 128, 256]
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.UPSAMPLE_STRIDES: [1, 2, 4]
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NUM_UPSAMPLE_FILTERS: [128, 128, 128]
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.FOCUS: False
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.ACT_FN: ReLU
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.NAME: CenterHead
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.CLASS_AGNOSTIC: False
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.CLASS_NAMES_EACH_HEAD: [['Vehicle', 'Pedestrian', 'Cyclist']]
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SHARED_CONV_CHANNEL: 64
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.USE_BIAS_BEFORE_NORM: True
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.NUM_HM_CONV: 2
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_ORDER: ['center', 'center_z', 'dim', 'rot']
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT = edict()
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.out_channels: 2
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.num_conv: 2
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.out_channels: 1
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.num_conv: 2
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.out_channels: 3
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.num_conv: 2
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.out_channels: 2
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.num_conv: 2
2023-01-09 10:38:07,669   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG = edict()
2023-01-09 10:38:07,669   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.FEATURE_MAP_STRIDE: 1
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.NUM_MAX_OBJS: 500
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.GAUSSIAN_OVERLAP: 0.1
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.MIN_RADIUS: 2
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.SHARPER: False
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG = edict()
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.cls_weight: 1.0
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.loc_weight: 2.0
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.SCORE_THRESH: 0.1
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.POST_CENTER_LIMIT_RANGE: [-80, -80, -10.0, 80, 80, 10.0]
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.MAX_OBJ_PER_SAMPLE: 500
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.POST_PROCESSING = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_METRIC: waymo
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/AP: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/AP']
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/APH: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/APH']
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.KD: True
2023-01-09 10:38:07,670   INFO  
cfg.MODEL_TEACHER.LOGIT_KD = edict()
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ENABLED: True
2023-01-09 10:38:07,670   INFO  cfg.MODEL_TEACHER.LOGIT_KD.MODE: raw_pred
2023-01-09 10:38:07,671   INFO  
cfg.MODEL_TEACHER.LOGIT_KD.ALIGN = edict()
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.target: teacher
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 10:38:07,671   INFO  
cfg.MODEL_TEACHER.LABEL_ASSIGN_KD = edict()
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 10:38:07,671   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 10:38:07,671   INFO  
cfg.OPTIMIZATION = edict()
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.BATCH_SIZE_PER_GPU: 2
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.NUM_EPOCHS: 30
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.OPTIMIZER: adam_onecycle
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.LR: 0.003
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.WEIGHT_DECAY: 0.01
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.MOMENTUM: 0.9
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.MOMS: [0.95, 0.85]
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.PCT_START: 0.4
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.DIV_FACTOR: 10
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.DECAY_STEP_LIST: [35, 45]
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.LR_DECAY: 0.1
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.LR_CLIP: 1e-07
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.LR_WARMUP: False
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.WARMUP_EPOCH: 1
2023-01-09 10:38:07,671   INFO  cfg.OPTIMIZATION.GRAD_NORM_CLIP: 10
2023-01-09 10:38:07,672   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN = edict()
2023-01-09 10:38:07,672   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.ENABLED: True
2023-01-09 10:38:07,672   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.WAY: FNAV2
2023-01-09 10:38:07,672   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE = edict()
2023-01-09 10:38:07,672   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE.ABS: True
2023-01-09 10:38:07,672   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA = edict()
2023-01-09 10:38:07,672   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA.l1_norm: max
2023-01-09 10:38:07,672   INFO  
cfg.KD = edict()
2023-01-09 10:38:07,672   INFO  cfg.KD.ENABLED: True
2023-01-09 10:38:07,672   INFO  cfg.KD.TEACHER_MODE: train
2023-01-09 10:38:07,672   INFO  cfg.KD.DIFF_VOXEL: True
2023-01-09 10:38:07,672   INFO  
cfg.KD.MASK = edict()
2023-01-09 10:38:07,672   INFO  cfg.KD.MASK.SCORE_MASK: False
2023-01-09 10:38:07,672   INFO  cfg.KD.MASK.FG_MASK: False
2023-01-09 10:38:07,672   INFO  cfg.KD.MASK.BOX_MASK: False
2023-01-09 10:38:07,672   INFO  
cfg.KD.LOGIT_KD = edict()
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ENABLED: True
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.MODE: raw_pred
2023-01-09 10:38:07,672   INFO  
cfg.KD.LOGIT_KD.ALIGN = edict()
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ALIGN.target: teacher
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 10:38:07,672   INFO  cfg.KD.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 10:38:07,672   INFO  
cfg.KD.FEATURE_KD = edict()
2023-01-09 10:38:07,672   INFO  cfg.KD.FEATURE_KD.ENABLED: False
2023-01-09 10:38:07,672   INFO  cfg.KD.FEATURE_KD.FEATURE_NAME: spatial_features_2d
2023-01-09 10:38:07,672   INFO  cfg.KD.FEATURE_KD.FEATURE_NAME_TEA: spatial_features_2d
2023-01-09 10:38:07,673   INFO  
cfg.KD.FEATURE_KD.ALIGN = edict()
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.ENABLED: False
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.MODE: interpolate
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.target: teacher
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.mode: bilinear
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.align_corners: True
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.align_channel: False
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.num_filters: [192, 384]
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.use_norm: True
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.use_act: False
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.kernel_size: 3
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ALIGN.groups: 1
2023-01-09 10:38:07,673   INFO  
cfg.KD.FEATURE_KD.ROI_POOL = edict()
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ROI_POOL.ENABLED: True
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ROI_POOL.GRID_SIZE: 7
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ROI_POOL.DOWNSAMPLE_RATIO: 1
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ROI_POOL.ROI: gt
2023-01-09 10:38:07,673   INFO  cfg.KD.FEATURE_KD.ROI_POOL.THRESH: 0.0
2023-01-09 10:38:07,673   INFO  
cfg.KD.LABEL_ASSIGN_KD = edict()
2023-01-09 10:38:07,673   INFO  cfg.KD.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 10:38:07,673   INFO  cfg.KD.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 10:38:07,673   INFO  cfg.KD.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 10:38:07,673   INFO  cfg.KD.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 10:38:07,673   INFO  cfg.KD.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 10:38:07,673   INFO  
cfg.KD.NMS_CONFIG = edict()
2023-01-09 10:38:07,673   INFO  cfg.KD.NMS_CONFIG.ENABLED: False
2023-01-09 10:38:07,673   INFO  cfg.KD.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 10:38:07,673   INFO  cfg.KD.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 10:38:07,674   INFO  cfg.KD.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 10:38:07,674   INFO  cfg.KD.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 10:38:07,674   INFO  
cfg.KD_LOSS = edict()
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.ENABLED: True
2023-01-09 10:38:07,674   INFO  
cfg.KD_LOSS.HM_LOSS = edict()
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.type: MSELoss
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.weight: 7.0
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.thresh: 0.0
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.fg_mask: True
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.soft_mask: True
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.HM_LOSS.rank: -1
2023-01-09 10:38:07,674   INFO  
cfg.KD_LOSS.REG_LOSS = edict()
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.REG_LOSS.type: RegLossCenterNet
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.REG_LOSS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.REG_LOSS.weight: 0.0
2023-01-09 10:38:07,674   INFO  
cfg.KD_LOSS.FEATURE_LOSS = edict()
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.mode: rois
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.type: MSELoss
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.weight: 0.1
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.fg_mask: False
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.score_mask: False
2023-01-09 10:38:07,674   INFO  cfg.KD_LOSS.FEATURE_LOSS.score_thresh: 0.3
2023-01-09 10:38:07,674   INFO  cfg.TAG: cp-pillar-v0.4_sparsekd
2023-01-09 10:38:07,674   INFO  cfg.EXP_GROUP_PATH: waymo_models/cp-pillar
2023-01-09 10:38:07,766   INFO  Database filter by min points Vehicle: 4430 => 3909
2023-01-09 10:38:07,766   INFO  Database filter by min points Pedestrian: 3967 => 3319
2023-01-09 10:38:07,767   INFO  Database filter by min points Cyclist: 153 => 139
2023-01-09 10:38:07,767   INFO  Database filter by difficulty Vehicle: 3909 => 3909
2023-01-09 10:38:07,768   INFO  Database filter by difficulty Pedestrian: 3319 => 3319
2023-01-09 10:38:07,768   INFO  Database filter by difficulty Cyclist: 139 => 139
2023-01-09 10:38:07,768   INFO  Loading GT database to shared memory
2023-01-09 10:38:07,822   INFO  GT database has been saved to shared memory
2023-01-09 10:38:07,824   INFO  Loading Waymo dataset
2023-01-09 10:38:07,851   INFO  Total skipped info 0
2023-01-09 10:38:07,851   INFO  Total samples for Waymo dataset: 992
2023-01-09 10:38:07,851   INFO  Total sampled samples for Waymo dataset: 199
2023-01-09 10:38:10,171   INFO  Loading teacher parameters >>>>>>
2023-01-09 10:38:10,172   INFO  ==> Loading parameters from checkpoint ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth to GPU
2023-01-09 10:38:10,218   INFO  ==> Checkpoint trained from version: pcdet+0.5.2+e348415
2023-01-09 10:38:10,224   INFO  ==> Done (loaded 179/179)
2023-01-09 10:38:10,232   INFO  Loading pretrained parameters >>>>>>
2023-01-09 10:38:10,232   INFO  ==> Loading parameters from checkpoint ../output/waymo_models/cp-pillar/cp-pillar-v0.4/default/ckpt/checkpoint_epoch_20.pth to GPU
2023-01-09 10:38:10,263   INFO  ==> Checkpoint trained from version: pcdet+0.5.2+e348415
2023-01-09 10:38:10,263   INFO  ==> Remap pretrained model parameters with: fnav2
Traceback (most recent call last):
  File "train.py", line 245, in <module>
    main()
  File "train.py", line 149, in main
    model.load_params_from_file(
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/detectors/detector3d_template.py", line 427, in load_params_from_file
    model_state_disk = self._remap_to_current_model(model_state_disk, remap_cfg) # 返回映射好的权重
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/detectors/detector3d_template.py", line 497, in _remap_to_current_model
    return getattr(kd_tgi_utils, '_remap_to_current_model_by_{}'.format(cfg.WAY.lower()))(self, model_state, cfg)
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/utils/kd_utils/kd_tgi_utils.py", line 224, in _remap_to_current_model_by_fnav2
    curr_v = curr_v.index_select(1, _stu_input_dim_idx)
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking arugment for argument index in method wrapper_index_select)
2023-01-09 10:38:10,358   INFO  Deleting GT database from shared memory
Exception ignored in: <function DataBaseSampler.__del__ at 0x7f76b353b3a0>
Traceback (most recent call last):
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/datasets/augmentor/database_sampler.py", line 63, in __del__
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1446, in info
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1589, in _log
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1599, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1661, in callHandlers
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 954, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1186, in emit
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1176, in _open
NameError: name 'open' is not defined
HuangCongQing commented 1 year ago

I run another yaml file（voxel） and report another error

    teacher_model = build_teacher_network(cfg, args, train_set, dist_train, logger)
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/__init__.py", line 27, in build_teacher_network
    teacher_model.load_params_from_file(filename=args.teacher_ckpt, to_cpu=dist, logger=logger)
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/detectors/detector3d_template.py", line 414, in load_params_from_file
    raise FileNotFoundError
FileNotFoundError

python train --cfg_file=cfgs/waymo_models/cp-voxel/cp-voxel-s_sparsekd.yaml

/home/chongqinghuang/anaconda3/envs/pcdet/bin/python train.py --cfg_file=cfgs/waymo_models/cp-voxel/cp-voxel-s_sparsekd.yaml --batch_size=4 --epochs=20
2023-01-09 11:29:30,417   INFO  **********************Start logging**********************
2023-01-09 11:29:30,417   INFO  CUDA_VISIBLE_DEVICES=ALL
2023-01-09 11:29:30,417   INFO  cfg_file         cfgs/waymo_models/cp-voxel/cp-voxel-s_sparsekd.yaml
2023-01-09 11:29:30,417   INFO  batch_size       4
2023-01-09 11:29:30,417   INFO  epochs           20
2023-01-09 11:29:30,417   INFO  workers          4
2023-01-09 11:29:30,417   INFO  extra_tag        default
2023-01-09 11:29:30,417   INFO  ckpt             None
2023-01-09 11:29:30,417   INFO  pretrained_model ../output/model_zoo/cp-voxel/cp-voxel_6429.pth
2023-01-09 11:29:30,417   INFO  launcher         none
2023-01-09 11:29:30,417   INFO  tcp_port         18888
2023-01-09 11:29:30,417   INFO  sync_bn          False
2023-01-09 11:29:30,417   INFO  fix_random_seed  False
2023-01-09 11:29:30,417   INFO  ckpt_save_interval 1
2023-01-09 11:29:30,417   INFO  local_rank       0
2023-01-09 11:29:30,417   INFO  max_ckpt_save_num 30
2023-01-09 11:29:30,417   INFO  merge_all_iters_to_one_epoch False
2023-01-09 11:29:30,417   INFO  set_cfgs         None
2023-01-09 11:29:30,417   INFO  max_waiting_mins 0
2023-01-09 11:29:30,417   INFO  start_epoch      0
2023-01-09 11:29:30,417   INFO  save_to_file     False
2023-01-09 11:29:30,417   INFO  teacher_ckpt     ../output/model_zoo/cp-voxel/cp-voxel_6429.pth
2023-01-09 11:29:30,417   INFO  cfg.ROOT_DIR: /home/chongqinghuang/code/light_weight/SparseKD
2023-01-09 11:29:30,417   INFO  cfg.LOCAL_RANK: 0
2023-01-09 11:29:30,417   INFO  cfg.CLASS_NAMES: ['Vehicle', 'Pedestrian', 'Cyclist']
2023-01-09 11:29:30,417   INFO  cfg.TEACHER_CKPT: ../output/model_zoo/cp-voxel/cp-voxel_6429.pth
2023-01-09 11:29:30,417   INFO  cfg.PRETRAINED_MODEL: ../output/model_zoo/cp-voxel/cp-voxel_6429.pth
2023-01-09 11:29:30,417   INFO  
cfg.DATA_CONFIG = edict()
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.DATASET: WaymoDataset
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.DATA_PATH: ../data/waymo
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.PROCESSED_DATA_TAG: waymo_processed_data_v0_5_0
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.POINT_CLOUD_RANGE: [-75.2, -75.2, -2, 75.2, 75.2, 4]
2023-01-09 11:29:30,417   INFO  
cfg.DATA_CONFIG.DATA_SPLIT = edict()
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.DATA_SPLIT.train: train
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.DATA_SPLIT.test: val
2023-01-09 11:29:30,417   INFO  
cfg.DATA_CONFIG.SAMPLED_INTERVAL = edict()
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.train: 5
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.test: 5
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.FILTER_EMPTY_BOXES_FOR_TRAIN: True
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.DISABLE_NLZ_FLAG_ON_POINTS: True
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.USE_SHARED_MEMORY: False
2023-01-09 11:29:30,417   INFO  cfg.DATA_CONFIG.SHARED_MEMORY_FILE_LIMIT: 35000
2023-01-09 11:29:30,418   INFO  
cfg.DATA_CONFIG.DATA_AUGMENTOR = edict()
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.DISABLE_AUG_LIST: ['placeholder']
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.AUG_CONFIG_LIST: [{'NAME': 'gt_sampling', 'USE_ROAD_PLANE': False, 'DB_INFO_PATH': ['waymo_processed_data_v0_5_0_waymo_dbinfos_train_sampled_1.pkl'], 'USE_SHARED_MEMORY': True, 'DB_DATA_PATH': ['waymo_processed_data_v0_5_0_gt_database_train_sampled_1_global.npy'], 'PREPARE': {'filter_by_min_points': ['Vehicle:5', 'Pedestrian:5', 'Cyclist:5'], 'filter_by_difficulty': [-1]}, 'SAMPLE_GROUPS': ['Vehicle:15', 'Pedestrian:10', 'Cyclist:10'], 'NUM_POINT_FEATURES': 5, 'REMOVE_EXTRA_WIDTH': [0.0, 0.0, 0.0], 'LIMIT_WHOLE_SCENE': True}, {'NAME': 'random_world_flip', 'ALONG_AXIS_LIST': ['x', 'y']}, {'NAME': 'random_world_rotation', 'WORLD_ROT_ANGLE': [-0.78539816, 0.78539816]}, {'NAME': 'random_world_scaling', 'WORLD_SCALE_RANGE': [0.95, 1.05]}]
2023-01-09 11:29:30,418   INFO  
cfg.DATA_CONFIG.POINT_FEATURE_ENCODING = edict()
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.encoding_type: absolute_coordinates_encoding
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.used_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.src_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG.DATA_PROCESSOR: [{'NAME': 'mask_points_and_boxes_outside_range', 'REMOVE_OUTSIDE_BOXES': True}, {'NAME': 'shuffle_points', 'SHUFFLE_ENABLED': {'train': True, 'test': True}}, {'NAME': 'transform_points_to_voxels', 'VOXEL_SIZE': [0.1, 0.1, 0.15], 'MAX_POINTS_PER_VOXEL': 5, 'MAX_NUMBER_OF_VOXELS': {'train': 150000, 'test': 150000}}]
2023-01-09 11:29:30,418   INFO  cfg.DATA_CONFIG._BASE_CONFIG_: cfgs/dataset_configs/waymo_dataset.yaml
2023-01-09 11:29:30,418   INFO  
cfg.MODEL = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.NAME: CenterPoint
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.VFE = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.VFE.NAME: MeanVFE
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.BACKBONE_3D = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_3D.NAME: VoxelResBackBone8x
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_3D.ACT_FN: ReLU
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_3D.NUM_FILTERS: [16, 16, 32, 64, 128, 128]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_3D.LAYER_NUMS: [1, 2, 3, 3, 3, 1]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_3D.WIDTH: 1.0
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.MAP_TO_BEV = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.MAP_TO_BEV.NAME: HeightCompression
2023-01-09 11:29:30,418   INFO  cfg.MODEL.MAP_TO_BEV.NUM_BEV_FEATURES: 256
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.BACKBONE_2D = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.NAME: BaseBEVBackbone
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.ACT_FN: ReLU
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.NORM_TYPE: BatchNorm2d
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.WIDTH: 0.5
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.LAYER_NUMS: [5, 5]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.LAYER_STRIDES: [1, 2]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.NUM_FILTERS: [128, 256]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.UPSAMPLE_STRIDES: [1, 2]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.BACKBONE_2D.NUM_UPSAMPLE_FILTERS: [256, 256]
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.NAME: CenterHead
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.CLASS_AGNOSTIC: False
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.ACT_FN: ReLU
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.NORM_TYPE: BatchNorm2d
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.CLASS_NAMES_EACH_HEAD: [['Vehicle', 'Pedestrian', 'Cyclist']]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SHARED_CONV_CHANNEL: 32
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.USE_BIAS_BEFORE_NORM: True
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.NUM_HM_CONV: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_ORDER: ['center', 'center_z', 'dim', 'rot']
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT = edict()
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.out_channels: 2
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.num_conv: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.out_channels: 1
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.num_conv: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.out_channels: 3
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.num_conv: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.out_channels: 2
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.num_conv: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.FEATURE_MAP_STRIDE: 8
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.NUM_MAX_OBJS: 500
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.GAUSSIAN_OVERLAP: 0.1
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.MIN_RADIUS: 2
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG = edict()
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.cls_weight: 1.0
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.loc_weight: 2.0
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.SCORE_THRESH: 0.1
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.POST_CENTER_LIMIT_RANGE: [-75.2, -75.2, -2, 75.2, 75.2, 4]
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.MAX_OBJ_PER_SAMPLE: 500
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.LOGIT_KD = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ENABLED: True
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.MODE: raw_pred
2023-01-09 11:29:30,418   INFO  
cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN = edict()
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 11:29:30,418   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.target: teacher
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 11:29:30,419   INFO  cfg.MODEL.DENSE_HEAD.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.POST_PROCESSING = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.POST_PROCESSING.RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
2023-01-09 11:29:30,419   INFO  cfg.MODEL.POST_PROCESSING.EVAL_METRIC: waymo
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.POST_PROCESSING.EVAL_CLASSES = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/AP: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/AP']
2023-01-09 11:29:30,419   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/APH: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/APH']
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD: True
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.KD_LOSS = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.ENABLED: True
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.KD_LOSS.HM_LOSS = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.type: MSELoss
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.weight: 10.0
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.thresh: 0.0
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.fg_mask: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.soft_mask: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.HM_LOSS.rank: -1
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.KD_LOSS.REG_LOSS = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.type: RegLossCenterNet
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.REG_LOSS.weight: 0.2
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.KD_LOSS.FEATURE_LOSS = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.mode: rois
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.type: MSELoss
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.weight: 0.1
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.fg_mask: False
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.score_mask: False
2023-01-09 11:29:30,419   INFO  cfg.MODEL.KD_LOSS.FEATURE_LOSS.score_thresh: 0.3
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.LOGIT_KD = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ENABLED: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.MODE: raw_pred
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.LOGIT_KD.ALIGN = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ALIGN.target: teacher
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 11:29:30,419   INFO  
cfg.MODEL.LABEL_ASSIGN_KD = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 11:29:30,419   INFO  cfg.MODEL.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.NAME: CenterPoint
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.IS_TEACHER: True
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER.VFE = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.VFE.NAME: MeanVFE
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER.BACKBONE_3D = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_3D.NAME: VoxelResBackBone8x
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_3D.ACT_FN: ReLU
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_3D.NUM_FILTERS: [16, 16, 32, 64, 128, 128]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_3D.LAYER_NUMS: [1, 2, 3, 3, 3, 1]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_3D.WIDTH: 1.0
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER.MAP_TO_BEV = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.MAP_TO_BEV.NAME: HeightCompression
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.MAP_TO_BEV.NUM_BEV_FEATURES: 256
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER.BACKBONE_2D = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NAME: BaseBEVBackbone
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.ACT_FN: ReLU
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NORM_TYPE: BatchNorm2d
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.WIDTH: 1.0
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.LAYER_NUMS: [5, 5]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.LAYER_STRIDES: [1, 2]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NUM_FILTERS: [128, 256]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.UPSAMPLE_STRIDES: [1, 2]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.BACKBONE_2D.NUM_UPSAMPLE_FILTERS: [256, 256]
2023-01-09 11:29:30,419   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD = edict()
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.NAME: CenterHead
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.CLASS_AGNOSTIC: False
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.ACT_FN: ReLU
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.NORM_TYPE: BatchNorm2d
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.CLASS_NAMES_EACH_HEAD: [['Vehicle', 'Pedestrian', 'Cyclist']]
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SHARED_CONV_CHANNEL: 64
2023-01-09 11:29:30,419   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.USE_BIAS_BEFORE_NORM: True
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.NUM_HM_CONV: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_ORDER: ['center', 'center_z', 'dim', 'rot']
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT = edict()
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.out_channels: 2
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.num_conv: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.out_channels: 1
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.num_conv: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.out_channels: 3
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.num_conv: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.out_channels: 2
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.num_conv: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.FEATURE_MAP_STRIDE: 8
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.NUM_MAX_OBJS: 500
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.GAUSSIAN_OVERLAP: 0.1
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.MIN_RADIUS: 2
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG = edict()
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.cls_weight: 1.0
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.loc_weight: 2.0
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.SCORE_THRESH: 0.1
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.POST_CENTER_LIMIT_RANGE: [-75.2, -75.2, -2, 75.2, 75.2, 4]
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.MAX_OBJ_PER_SAMPLE: 500
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.POST_PROCESSING = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_METRIC: waymo
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/AP: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/AP']
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/APH: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/APH']
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.KD: True
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.LOGIT_KD = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ENABLED: True
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.MODE: raw_pred
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.LOGIT_KD.ALIGN = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.target: teacher
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 11:29:30,420   INFO  
cfg.MODEL_TEACHER.LABEL_ASSIGN_KD = edict()
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 11:29:30,420   INFO  cfg.MODEL_TEACHER.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 11:29:30,420   INFO  
cfg.OPTIMIZATION = edict()
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.BATCH_SIZE_PER_GPU: 4
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.NUM_EPOCHS: 30
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.OPTIMIZER: adam_onecycle
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.LR: 0.003
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.WEIGHT_DECAY: 0.01
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.MOMENTUM: 0.9
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.MOMS: [0.95, 0.85]
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.PCT_START: 0.4
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.DIV_FACTOR: 10
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.DECAY_STEP_LIST: [35, 45]
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.LR_DECAY: 0.1
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.LR_CLIP: 1e-07
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.LR_WARMUP: False
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.WARMUP_EPOCH: 1
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.GRAD_NORM_CLIP: 10
2023-01-09 11:29:30,420   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN = edict()
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.ENABLED: True
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.WAY: FNAV2
2023-01-09 11:29:30,420   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE = edict()
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE.ABS: True
2023-01-09 11:29:30,420   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA = edict()
2023-01-09 11:29:30,420   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA.l1_norm: max
2023-01-09 11:29:30,420   INFO  
cfg.KD = edict()
2023-01-09 11:29:30,420   INFO  cfg.KD.ENABLED: True
2023-01-09 11:29:30,420   INFO  cfg.KD.TEACHER_MODE: train
2023-01-09 11:29:30,421   INFO  cfg.KD.DIFF_VOXEL: False
2023-01-09 11:29:30,421   INFO  
cfg.KD.MASK = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.MASK.SCORE_MASK: False
2023-01-09 11:29:30,421   INFO  cfg.KD.MASK.FG_MASK: False
2023-01-09 11:29:30,421   INFO  cfg.KD.MASK.BOX_MASK: False
2023-01-09 11:29:30,421   INFO  
cfg.KD.LOGIT_KD = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ENABLED: True
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.MODE: raw_pred
2023-01-09 11:29:30,421   INFO  
cfg.KD.LOGIT_KD.ALIGN = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ALIGN.MODE: interpolate
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ALIGN.target: teacher
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ALIGN.mode: bilinear
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ALIGN.align_corners: True
2023-01-09 11:29:30,421   INFO  cfg.KD.LOGIT_KD.ALIGN.align_channel: False
2023-01-09 11:29:30,421   INFO  
cfg.KD.FEATURE_KD = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ENABLED: False
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.FEATURE_NAME: spatial_features_2d
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.FEATURE_NAME_TEA: spatial_features_2d
2023-01-09 11:29:30,421   INFO  
cfg.KD.FEATURE_KD.ALIGN = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.ENABLED: False
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.MODE: interpolate
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.target: teacher
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.mode: bilinear
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.align_corners: True
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.align_channel: False
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.num_filters: [192, 384]
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.use_norm: True
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.use_act: False
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.kernel_size: 3
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ALIGN.groups: 1
2023-01-09 11:29:30,421   INFO  
cfg.KD.FEATURE_KD.ROI_POOL = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ROI_POOL.ENABLED: True
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ROI_POOL.GRID_SIZE: 7
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ROI_POOL.DOWNSAMPLE_RATIO: 1
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ROI_POOL.ROI: gt
2023-01-09 11:29:30,421   INFO  cfg.KD.FEATURE_KD.ROI_POOL.THRESH: 0.0
2023-01-09 11:29:30,421   INFO  
cfg.KD.LABEL_ASSIGN_KD = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.LABEL_ASSIGN_KD.ENABLED: True
2023-01-09 11:29:30,421   INFO  cfg.KD.LABEL_ASSIGN_KD.SCORE_TYPE: cls
2023-01-09 11:29:30,421   INFO  cfg.KD.LABEL_ASSIGN_KD.USE_GT: True
2023-01-09 11:29:30,421   INFO  cfg.KD.LABEL_ASSIGN_KD.GT_FIRST: False
2023-01-09 11:29:30,421   INFO  cfg.KD.LABEL_ASSIGN_KD.SCORE_THRESH: [0.6, 0.6, 0.6]
2023-01-09 11:29:30,421   INFO  
cfg.KD.NMS_CONFIG = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD.NMS_CONFIG.ENABLED: False
2023-01-09 11:29:30,421   INFO  cfg.KD.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 11:29:30,421   INFO  cfg.KD.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 11:29:30,421   INFO  cfg.KD.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 11:29:30,421   INFO  cfg.KD.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 11:29:30,421   INFO  
cfg.KD_LOSS = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.ENABLED: True
2023-01-09 11:29:30,421   INFO  
cfg.KD_LOSS.HM_LOSS = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.type: MSELoss
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.weight: 10.0
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.thresh: 0.0
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.fg_mask: True
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.soft_mask: True
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.HM_LOSS.rank: -1
2023-01-09 11:29:30,421   INFO  
cfg.KD_LOSS.REG_LOSS = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.REG_LOSS.type: RegLossCenterNet
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.REG_LOSS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.REG_LOSS.weight: 0.2
2023-01-09 11:29:30,421   INFO  
cfg.KD_LOSS.FEATURE_LOSS = edict()
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.mode: rois
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.type: MSELoss
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.weight: 0.1
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.fg_mask: False
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.score_mask: False
2023-01-09 11:29:30,421   INFO  cfg.KD_LOSS.FEATURE_LOSS.score_thresh: 0.3
2023-01-09 11:29:30,421   INFO  cfg.TAG: cp-voxel-s_sparsekd
2023-01-09 11:29:30,421   INFO  cfg.EXP_GROUP_PATH: waymo_models/cp-voxel
2023-01-09 11:29:30,470   INFO  Database filter by min points Vehicle: 4430 => 3909
2023-01-09 11:29:30,471   INFO  Database filter by min points Pedestrian: 3967 => 3319
2023-01-09 11:29:30,471   INFO  Database filter by min points Cyclist: 153 => 139
2023-01-09 11:29:30,471   INFO  Database filter by difficulty Vehicle: 3909 => 3909
2023-01-09 11:29:30,472   INFO  Database filter by difficulty Pedestrian: 3319 => 3319
2023-01-09 11:29:30,472   INFO  Database filter by difficulty Cyclist: 139 => 139
2023-01-09 11:29:30,472   INFO  Loading GT database to shared memory
2023-01-09 11:29:30,480   INFO  GT database has been saved to shared memory
2023-01-09 11:29:30,481   INFO  Loading Waymo dataset
2023-01-09 11:29:30,499   INFO  Total skipped info 0
2023-01-09 11:29:30,499   INFO  Total samples for Waymo dataset: 992
2023-01-09 11:29:30,499   INFO  Total sampled samples for Waymo dataset: 199
2023-01-09 11:29:33,985   INFO  Loading teacher parameters >>>>>>
Traceback (most recent call last):
  File "train.py", line 245, in <module>
    main()
  File "train.py", line 132, in main
    teacher_model = build_teacher_network(cfg, args, train_set, dist_train, logger)
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/__init__.py", line 27, in build_teacher_network
    teacher_model.load_params_from_file(filename=args.teacher_ckpt, to_cpu=dist, logger=logger)
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/models/detectors/detector3d_template.py", line 414, in load_params_from_file
    raise FileNotFoundError
FileNotFoundError
2023-01-09 11:29:34,067   INFO  Deleting GT database from shared memory
Exception ignored in: <function DataBaseSampler.__del__ at 0x7fcb9df4f4c0>
Traceback (most recent call last):
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/datasets/augmentor/database_sampler.py", line 63, in __del__
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1446, in info
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1589, in _log
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1599, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1661, in callHandlers
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 954, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1186, in emit
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1176, in _open
NameError: name 'open' is not defined

Process finished with exit code 1
jihanyang commented 1 year ago

You need to obtain a teacher model first for KD. As you haven't downloaded teacher pretrained weights or train a teacher model by yourself, the KD cannot run without teacher weight. Please refer to https://github.com/CVMI-Lab/SparseKD/blob/master/docs/GETTING_STARTED.md.

HuangCongQing commented 1 year ago

Thx. I didn't download the weights, I trained the weights myself. In addition, want to ask, cp-voxel_6429.pththis weight where to download ah? I didn't find the download link.

HuangCongQing commented 1 year ago

What's more, when I was training the a teacher voxel method, I also encountered a bug,

python train.py --cfg_file=cfgs/waymo_models/cp-voxel/cp-voxel-s.yaml

/home/chongqinghuang/anaconda3/envs/pcdet/bin/python train.py --cfg_file=cfgs/waymo_models/cp-voxel/cp-voxel-s.yaml --batch_size=4 --epochs=20
2023-01-09 17:17:59,597   INFO  **********************Start logging**********************
2023-01-09 17:17:59,597   INFO  CUDA_VISIBLE_DEVICES=ALL
2023-01-09 17:17:59,597   INFO  cfg_file         cfgs/waymo_models/cp-voxel/cp-voxel-s.yaml
2023-01-09 17:17:59,597   INFO  batch_size       4
2023-01-09 17:17:59,597   INFO  epochs           20
2023-01-09 17:17:59,597   INFO  workers          4
2023-01-09 17:17:59,597   INFO  extra_tag        default
2023-01-09 17:17:59,597   INFO  ckpt             None
2023-01-09 17:17:59,597   INFO  pretrained_model None
2023-01-09 17:17:59,597   INFO  launcher         none
2023-01-09 17:17:59,597   INFO  tcp_port         18888
2023-01-09 17:17:59,597   INFO  sync_bn          False
2023-01-09 17:17:59,597   INFO  fix_random_seed  False
2023-01-09 17:17:59,597   INFO  ckpt_save_interval 1
2023-01-09 17:17:59,597   INFO  local_rank       0
2023-01-09 17:17:59,597   INFO  max_ckpt_save_num 30
2023-01-09 17:17:59,597   INFO  merge_all_iters_to_one_epoch False
2023-01-09 17:17:59,597   INFO  set_cfgs         None
2023-01-09 17:17:59,597   INFO  max_waiting_mins 0
2023-01-09 17:17:59,597   INFO  start_epoch      0
2023-01-09 17:17:59,597   INFO  save_to_file     False
2023-01-09 17:17:59,597   INFO  teacher_ckpt     None
2023-01-09 17:17:59,597   INFO  cfg.ROOT_DIR: /home/chongqinghuang/code/light_weight/SparseKD
2023-01-09 17:17:59,597   INFO  cfg.LOCAL_RANK: 0
2023-01-09 17:17:59,597   INFO  cfg.CLASS_NAMES: ['Vehicle', 'Pedestrian', 'Cyclist']
2023-01-09 17:17:59,597   INFO  
cfg.DATA_CONFIG = edict()
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.DATASET: WaymoDataset
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.DATA_PATH: ../data/waymo
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.PROCESSED_DATA_TAG: waymo_processed_data_v0_5_0
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.POINT_CLOUD_RANGE: [-75.2, -75.2, -2, 75.2, 75.2, 4]
2023-01-09 17:17:59,597   INFO  
cfg.DATA_CONFIG.DATA_SPLIT = edict()
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.DATA_SPLIT.train: train
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.DATA_SPLIT.test: val
2023-01-09 17:17:59,597   INFO  
cfg.DATA_CONFIG.SAMPLED_INTERVAL = edict()
2023-01-09 17:17:59,597   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.train: 5
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.SAMPLED_INTERVAL.test: 5
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.FILTER_EMPTY_BOXES_FOR_TRAIN: True
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.DISABLE_NLZ_FLAG_ON_POINTS: True
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.USE_SHARED_MEMORY: False
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.SHARED_MEMORY_FILE_LIMIT: 35000
2023-01-09 17:17:59,598   INFO  
cfg.DATA_CONFIG.DATA_AUGMENTOR = edict()
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.DISABLE_AUG_LIST: ['placeholder']
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.DATA_AUGMENTOR.AUG_CONFIG_LIST: [{'NAME': 'gt_sampling', 'USE_ROAD_PLANE': False, 'DB_INFO_PATH': ['waymo_processed_data_v0_5_0_waymo_dbinfos_train_sampled_1.pkl'], 'USE_SHARED_MEMORY': True, 'DB_DATA_PATH': ['waymo_processed_data_v0_5_0_gt_database_train_sampled_1_global.npy'], 'PREPARE': {'filter_by_min_points': ['Vehicle:5', 'Pedestrian:5', 'Cyclist:5'], 'filter_by_difficulty': [-1]}, 'SAMPLE_GROUPS': ['Vehicle:15', 'Pedestrian:10', 'Cyclist:10'], 'NUM_POINT_FEATURES': 5, 'REMOVE_EXTRA_WIDTH': [0.0, 0.0, 0.0], 'LIMIT_WHOLE_SCENE': True}, {'NAME': 'random_world_flip', 'ALONG_AXIS_LIST': ['x', 'y']}, {'NAME': 'random_world_rotation', 'WORLD_ROT_ANGLE': [-0.78539816, 0.78539816]}, {'NAME': 'random_world_scaling', 'WORLD_SCALE_RANGE': [0.95, 1.05]}]
2023-01-09 17:17:59,598   INFO  
cfg.DATA_CONFIG.POINT_FEATURE_ENCODING = edict()
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.encoding_type: absolute_coordinates_encoding
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.used_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.src_feature_list: ['x', 'y', 'z', 'intensity', 'elongation']
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG.DATA_PROCESSOR: [{'NAME': 'mask_points_and_boxes_outside_range', 'REMOVE_OUTSIDE_BOXES': True}, {'NAME': 'shuffle_points', 'SHUFFLE_ENABLED': {'train': True, 'test': True}}, {'NAME': 'transform_points_to_voxels', 'VOXEL_SIZE': [0.1, 0.1, 0.15], 'MAX_POINTS_PER_VOXEL': 5, 'MAX_NUMBER_OF_VOXELS': {'train': 150000, 'test': 150000}}]
2023-01-09 17:17:59,598   INFO  cfg.DATA_CONFIG._BASE_CONFIG_: cfgs/dataset_configs/waymo_dataset.yaml
2023-01-09 17:17:59,598   INFO  
cfg.MODEL = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.NAME: CenterPoint
2023-01-09 17:17:59,598   INFO  cfg.MODEL.IGNORE_PRETRAIN_MODULES: ['placeholder']
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.VFE = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.VFE.NAME: MeanVFE
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.BACKBONE_3D = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_3D.NAME: VoxelResBackBone8x
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_3D.ACT_FN: ReLU
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_3D.NUM_FILTERS: [16, 16, 32, 64, 128, 128]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_3D.LAYER_NUMS: [1, 2, 3, 3, 3, 1]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_3D.WIDTH: 1.0
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.MAP_TO_BEV = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.MAP_TO_BEV.NAME: HeightCompression
2023-01-09 17:17:59,598   INFO  cfg.MODEL.MAP_TO_BEV.NUM_BEV_FEATURES: 256
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.BACKBONE_2D = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.NAME: BaseBEVBackbone
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.ACT_FN: ReLU
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.NORM_TYPE: BatchNorm2d
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.WIDTH: 0.5
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.LAYER_NUMS: [5, 5]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.LAYER_STRIDES: [1, 2]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.NUM_FILTERS: [128, 256]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.UPSAMPLE_STRIDES: [1, 2]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.BACKBONE_2D.NUM_UPSAMPLE_FILTERS: [256, 256]
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.NAME: CenterHead
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.CLASS_AGNOSTIC: False
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.ACT_FN: ReLU
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.NORM_TYPE: BatchNorm2d
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.CLASS_NAMES_EACH_HEAD: [['Vehicle', 'Pedestrian', 'Cyclist']]
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SHARED_CONV_CHANNEL: 32
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.USE_BIAS_BEFORE_NORM: True
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.NUM_HM_CONV: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_ORDER: ['center', 'center_z', 'dim', 'rot']
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT = edict()
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.out_channels: 2
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center.num_conv: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.out_channels: 1
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.center_z.num_conv: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.out_channels: 3
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.dim.num_conv: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.out_channels: 2
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.SEPARATE_HEAD_CFG.HEAD_DICT.rot.num_conv: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.FEATURE_MAP_STRIDE: 8
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.NUM_MAX_OBJS: 500
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.GAUSSIAN_OVERLAP: 0.1
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.TARGET_ASSIGNER_CONFIG.MIN_RADIUS: 2
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG = edict()
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.cls_weight: 1.0
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.loc_weight: 2.0
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.LOSS_CONFIG.LOSS_WEIGHTS.code_weights: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
2023-01-09 17:17:59,598   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING = edict()
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.SCORE_THRESH: 0.1
2023-01-09 17:17:59,598   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.POST_CENTER_LIMIT_RANGE: [-75.2, -75.2, -2, 75.2, 75.2, 4]
2023-01-09 17:17:59,599   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.MAX_OBJ_PER_SAMPLE: 500
2023-01-09 17:17:59,599   INFO  
cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG = edict()
2023-01-09 17:17:59,599   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_TYPE: nms_gpu
2023-01-09 17:17:59,599   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_THRESH: 0.7
2023-01-09 17:17:59,599   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_PRE_MAXSIZE: 4096
2023-01-09 17:17:59,599   INFO  cfg.MODEL.DENSE_HEAD.POST_PROCESSING.NMS_CONFIG.NMS_POST_MAXSIZE: 500
2023-01-09 17:17:59,599   INFO  
cfg.MODEL.POST_PROCESSING = edict()
2023-01-09 17:17:59,599   INFO  cfg.MODEL.POST_PROCESSING.RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
2023-01-09 17:17:59,599   INFO  cfg.MODEL.POST_PROCESSING.EVAL_METRIC: waymo
2023-01-09 17:17:59,599   INFO  
cfg.MODEL.POST_PROCESSING.EVAL_CLASSES = edict()
2023-01-09 17:17:59,599   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/AP: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/AP', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/AP']
2023-01-09 17:17:59,599   INFO  cfg.MODEL.POST_PROCESSING.EVAL_CLASSES.LEVEL_2/APH: ['OBJECT_TYPE_TYPE_VEHICLE_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_PEDESTRIAN_LEVEL_2/APH', 'OBJECT_TYPE_TYPE_CYCLIST_LEVEL_2/APH']
2023-01-09 17:17:59,599   INFO  
cfg.OPTIMIZATION = edict()
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.BATCH_SIZE_PER_GPU: 4
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.NUM_EPOCHS: 30
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.OPTIMIZER: adam_onecycle
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.LR: 0.003
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.WEIGHT_DECAY: 0.01
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.MOMENTUM: 0.9
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.MOMS: [0.95, 0.85]
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.PCT_START: 0.4
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.DIV_FACTOR: 10
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.DECAY_STEP_LIST: [35, 45]
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.LR_DECAY: 0.1
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.LR_CLIP: 1e-07
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.LR_WARMUP: False
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.WARMUP_EPOCH: 1
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.GRAD_NORM_CLIP: 10
2023-01-09 17:17:59,599   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN = edict()
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.ENABLED: False
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.WAY: BN_SCALE
2023-01-09 17:17:59,599   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE = edict()
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.BN_SCALE.ABS: True
2023-01-09 17:17:59,599   INFO  
cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA = edict()
2023-01-09 17:17:59,599   INFO  cfg.OPTIMIZATION.REMAP_PRETRAIN.OFA.l1_norm: max
2023-01-09 17:17:59,599   INFO  cfg.TAG: cp-voxel-s
2023-01-09 17:17:59,599   INFO  cfg.EXP_GROUP_PATH: waymo_models/cp-voxel
2023-01-09 17:17:59,649   INFO  Database filter by min points Vehicle: 4430 => 3909
2023-01-09 17:17:59,649   INFO  Database filter by min points Pedestrian: 3967 => 3319
2023-01-09 17:17:59,649   INFO  Database filter by min points Cyclist: 153 => 139
2023-01-09 17:17:59,650   INFO  Database filter by difficulty Vehicle: 3909 => 3909
2023-01-09 17:17:59,650   INFO  Database filter by difficulty Pedestrian: 3319 => 3319
2023-01-09 17:17:59,650   INFO  Database filter by difficulty Cyclist: 139 => 139
2023-01-09 17:17:59,650   INFO  Loading GT database to shared memory
2023-01-09 17:17:59,774   INFO  GT database has been saved to shared memory
2023-01-09 17:17:59,776   INFO  Loading Waymo dataset
2023-01-09 17:17:59,794   INFO  Total skipped info 0
2023-01-09 17:17:59,794   INFO  Total samples for Waymo dataset: 992
2023-01-09 17:17:59,794   INFO  Total sampled samples for Waymo dataset: 199
2023-01-09 17:18:03,252   INFO  CenterPoint(
  (vfe): MeanVFE()
  (backbone_3d): VoxelResBackBone8x(
    (conv_input): SparseSequential(
      (0): SubMConv3d(5, 16, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], bias=False, algo=ConvAlgo.MaskImplicitGemm)
      (1): BatchNorm1d(16, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      (2): ReLU()
    )
    (conv1): SparseSequential(
      (0): SparseBasicBlock(
        (conv1): SubMConv3d(16, 16, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(16, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(16, 16, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(16, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
      (1): SparseBasicBlock(
        (conv1): SubMConv3d(16, 16, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(16, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(16, 16, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(16, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
    )
    (conv2): SparseSequential(
      (0): SparseSequential(
        (0): SparseConv3d(16, 32, kernel_size=[3, 3, 3], stride=[2, 2, 2], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], bias=False, algo=ConvAlgo.MaskImplicitGemm)
        (1): BatchNorm1d(32, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (2): ReLU()
      )
      (1): SparseBasicBlock(
        (conv1): SubMConv3d(32, 32, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(32, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(32, 32, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(32, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
      (2): SparseBasicBlock(
        (conv1): SubMConv3d(32, 32, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(32, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(32, 32, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(32, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
    )
    (conv3): SparseSequential(
      (0): SparseSequential(
        (0): SparseConv3d(32, 64, kernel_size=[3, 3, 3], stride=[2, 2, 2], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], bias=False, algo=ConvAlgo.MaskImplicitGemm)
        (1): BatchNorm1d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (2): ReLU()
      )
      (1): SparseBasicBlock(
        (conv1): SubMConv3d(64, 64, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(64, 64, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
      (2): SparseBasicBlock(
        (conv1): SubMConv3d(64, 64, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(64, 64, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
    )
    (conv4): SparseSequential(
      (0): SparseSequential(
        (0): SparseConv3d(64, 128, kernel_size=[3, 3, 3], stride=[2, 2, 2], padding=[0, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], bias=False, algo=ConvAlgo.MaskImplicitGemm)
        (1): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (2): ReLU()
      )
      (1): SparseBasicBlock(
        (conv1): SubMConv3d(128, 128, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(128, 128, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
      (2): SparseBasicBlock(
        (conv1): SubMConv3d(128, 128, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn1): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (relu): ReLU()
        (conv2): SubMConv3d(128, 128, kernel_size=[3, 3, 3], stride=[1, 1, 1], padding=[1, 1, 1], dilation=[1, 1, 1], output_padding=[0, 0, 0], algo=ConvAlgo.MaskImplicitGemm)
        (bn2): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
      )
    )
    (conv_out): SparseSequential(
      (0): SparseConv3d(128, 128, kernel_size=[3, 1, 1], stride=[2, 1, 1], padding=[0, 0, 0], dilation=[1, 1, 1], output_padding=[0, 0, 0], bias=False, algo=ConvAlgo.MaskImplicitGemm)
      (1): BatchNorm1d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
    )
    (final_act): ReLU()
  )
  (map_to_bev_module): HeightCompression()
  (pillar_adaptor): None
  (pfe): None
  (backbone_2d): BaseBEVBackbone(
    (blocks): ModuleList(
      (0): Sequential(
        (0): Identity()
        (1): Conv2d(256, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (2): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (3): ReLU()
        (4): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (5): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (6): ReLU()
        (7): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (8): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (9): ReLU()
        (10): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (11): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (12): ReLU()
        (13): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (14): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (15): ReLU()
        (16): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (17): BatchNorm2d(64, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (18): ReLU()
      )
      (1): Sequential(
        (0): Identity()
        (1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (2): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (3): ReLU()
        (4): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (5): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (6): ReLU()
        (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (8): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (9): ReLU()
        (10): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (11): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (12): ReLU()
        (13): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (14): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (15): ReLU()
        (16): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (17): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (18): ReLU()
      )
    )
    (deblocks): ModuleList(
      (0): Sequential(
        (0): ConvTranspose2d(64, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (1): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (2): ReLU()
      )
      (1): Sequential(
        (0): ConvTranspose2d(128, 128, kernel_size=(2, 2), stride=(2, 2), bias=False)
        (1): BatchNorm2d(128, eps=0.001, momentum=0.01, affine=True, track_running_stats=True)
        (2): ReLU()
      )
    )
  )
  (dense_head): CenterHead(
    (shared_conv): Sequential(
      (0): Conv2d(256, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
      (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
      (2): ReLU()
    )
    (heads_list): ModuleList(
      (0): SeparateHead(
        (center): Sequential(
          (0): Sequential(
            (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
            (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (2): ReLU()
          )
          (1): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
        )
        (center_z): Sequential(
          (0): Sequential(
            (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
            (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (2): ReLU()
          )
          (1): Conv2d(32, 1, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
        )
        (dim): Sequential(
          (0): Sequential(
            (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
            (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (2): ReLU()
          )
          (1): Conv2d(32, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
        )
        (rot): Sequential(
          (0): Sequential(
            (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
            (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (2): ReLU()
          )
          (1): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
        )
        (hm): Sequential(
          (0): Sequential(
            (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
            (1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
            (2): ReLU()
          )
          (1): Conv2d(32, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
        )
      )
    )
    (hm_loss_func): FocalLossCenterNet()
    (reg_loss_func): RegLossCenterNet()
  )
  (dense_head_aux): None
  (kd_adapt_block): None
  (point_head): None
  (roi_head): None
)
2023-01-09 17:18:03,254   INFO  **********************Start training waymo_models/cp-voxel/cp-voxel-s(default)**********************
epochs:   0%|          | 0/20 [00:00<?, ?it/s]
epochs:   0%|          | 0/20 [00:02<?, ?it/s]
Traceback (most recent call last):
  File "train.py", line 245, in <module>
    main()
  File "train.py", line 183, in main
    train_func(
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/train_utils/train_utils.py", line 134, in train_model
    accumulated_iter = train_one_epoch(
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/train_utils/train_utils.py", line 27, in train_one_epoch
    batch = next(dataloader_iter)
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 521, in __next__
    data = self._next_data()
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1186, in _next_data
    idx, data = self._get_data()
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1142, in _get_data
    success, data = self._try_get_data()
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 990, in _try_get_data
    data = self._data_queue.get(timeout=timeout)
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/queue.py", line 179, in get
    self.not_empty.wait(remaining)
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/threading.py", line 306, in wait
    gotit = waiter.acquire(True, timeout)
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/site-packages/torch/utils/data/_utils/signal_handling.py", line 66, in handler
    _error_if_any_worker_fails()
RuntimeError: DataLoader worker (pid 4109366) is killed by signal: Killed. 

                                             2023-01-09 17:18:06,202   INFO  Deleting GT database from shared memory
Exception ignored in: <function DataBaseSampler.__del__ at 0x7fc78fe894c0>
Traceback (most recent call last):
  File "/home/chongqinghuang/code/light_weight/SparseKD/tools/../pcdet/datasets/augmentor/database_sampler.py", line 63, in __del__
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1446, in info
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1589, in _log
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1599, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1661, in callHandlers
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 954, in handle
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1186, in emit
  File "/home/chongqinghuang/anaconda3/envs/pcdet/lib/python3.8/logging/__init__.py", line 1176, in _open
NameError: name 'open' is not defined
jihanyang commented 1 year ago

For the pretrained weights, you can refer to README.md

We could not publicly provide the above pretrained models due to Waymo Dataset License Agreement. To access these pretrained models, please email us your name, institute, a screenshot of the Waymo dataset registration confirmation mail, and your intended usage. Please send a second email if we don't get back to you in two days. Please note that Waymo open dataset is under strict non-commercial license, so we are not allowed to share the model with you if it will use for any profit-oriented activities.