diff --git a/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD.log b/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD.log new file mode 100644 index 0000000000000000000000000000000000000000..7876a66933ff322b7979530b82c401ce223c13e8 --- /dev/null +++ b/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD.log @@ -0,0 +1,2541 @@ +[04/24 14:33:58] ModelNet40Ply2048 INFO: dist_url: tcp://localhost:8888 +dist_backend: nccl +multiprocessing_distributed: False +ngpus_per_node: 1 +world_size: 1 +launcher: mp +local_rank: 0 +use_gpu: True +seed: 1234 +epoch: 0 +epochs: 600 +ignore_index: None +val_fn: validate +deterministic: False +sync_bn: False +criterion_args: + NAME: SmoothCrossEntropy + label_smoothing: 0.2 +use_mask: False +grad_norm_clip: 1 +layer_decay: 0 +step_per_update: 1 +start_epoch: 1 +sched_on_epoch: True +wandb: + use_wandb: False + project: PointNeXt-ModelNet40Ply2048 + tags: ['modelnet40ply2048', 'train', 'ppv2-s', 'ngpus1', 'seed1234'] + name: modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD +use_amp: False +use_voting: False +val_freq: 1 +resume: False +test: False +finetune: False +mode: train +logname: None +load_path: None +print_freq: 10 +save_freq: -1 +root_dir: log/modelnet40ply2048 +pretrained_path: None +datatransforms: + train: ['PointsToTensor', 'PointCloudScaleAndTranslate'] + val: ['PointsToTensor'] + vote: ['PointCloudScaleAndTranslate'] + kwargs: + shift: [0.2, 0.2, 0.2] +feature_keys: pos +num_points: 1024 +dataset: + common: + NAME: ModelNet40Ply2048 + data_dir: ./data/ModelNet40Ply2048 + train: + split: train + num_points: 1024 + val: + split: test + num_points: 1024 +batch_size: 32 +dataloader: + num_workers: 6 +num_classes: 40 +sched: cosine +warmup_epochs: 0 +min_lr: None +lr: 0.001 +optimizer: + NAME: adamw + weight_decay: 0.05 +log_dir: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD +val_batch_size: 64 +model: + NAME: DpnCls + encoder_args: + NAME: PPV2Encoder + blocks: [1, 1, 1, 1, 1, 1] + strides: [1, 2, 2, 2, 2, 1] + width: 32 + in_channels: 3 + radius: 0.15 + flag: 0 + radius_scaling: 1.5 + sa_layers: 2 + sa_use_res: True + nsample: 32 + expansion: 4 + aggr_args: + feature_type: dp_fj + reduction: max + group_args: + NAME: ballquery + normalize_dp: True + conv_args: + order: conv-norm-act + act_args: + act: relu + norm_args: + norm: bn + cls_args: + NAME: DpnClsHead + num_classes: 40 + mlps: [512, 256] + norm_args: + norm: bn1d +rank: 0 +distributed: False +mp: False +task_name: modelnet40ply2048 +exp_name: ppv2-s +opts: +run_name: modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD +run_dir: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD +exp_dir: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD +ckpt_dir: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint +log_path: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD.log +cfg_path: log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/cfg.yaml +[04/24 14:33:58] ModelNet40Ply2048 INFO: radius: [[0.15], [0.15], [0.22499999999999998], [0.33749999999999997], [0.50625], [0.7593749999999999]], + nsample: [[32], [32], [32], [32], [32], [32]] +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.15 +nsample: 32 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.15 +nsample: 32 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.525 +nsample: 64 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.22499999999999998 +nsample: 32 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.7874999999999999 +nsample: 64 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.33749999999999997 +nsample: 32 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: 0.50625 +nsample: 32 +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: NAME: ballquery +normalize_dp: True +radius: None +nsample: None +return_idx: True +[04/24 14:33:58] ModelNet40Ply2048 INFO: DpnCls( + (encoder): PPV2Encoder( + (grouper0): QueryAndGroup() + (encoder): Sequential( + (0): Sequential( + (0): SetAbstractionCls( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(11, 32, kernel_size=(1,), stride=(1,)) + ) + ) + ) + ) + (1): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (preconv): Sequential( + (0): Conv1d(32, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (scorenet_global): Sequential( + (0): Conv1d(8, 1, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(1, 1, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 64, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 64, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_v): Sequential( + (0): Conv1d(8, 64, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(64, 64, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + ) + ) + (2): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(8, 3, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(3, 3, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(64, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=24, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 128, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=24, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_v): Sequential( + (0): Conv1d(24, 128, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(128, 128, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + ) + ) + (3): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(24, 9, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(9, 9, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(128, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=72, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(27, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(27, 256, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=72, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_v): Sequential( + (0): Conv1d(72, 256, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(256, 256, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + ) + ) + (4): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(72, 27, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(27, 27, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(256, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=216, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(75, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(75, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(75, 512, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=216, out_features=216, bias=True) + (linear_k): Linear(in_features=216, out_features=216, bias=True) + (linear_v): Sequential( + (0): Conv1d(216, 512, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(512, 512, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + ) + ) + (5): Sequential( + (0): SetAbstractionCls( + (grouper): GroupAll() + (preconv): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + ) + (prediction): DpnClsHead( + (head): Sequential( + (0): Sequential( + (0): Linear(in_features=512, out_features=512, bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Dropout(p=0.5, inplace=False) + (2): Sequential( + (0): Linear(in_features=512, out_features=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (3): Dropout(p=0.5, inplace=False) + (4): Sequential( + (0): Linear(in_features=256, out_features=40, bias=True) + ) + ) + ) + (criterion): SmoothCrossEntropy() +) +[04/24 14:33:58] ModelNet40Ply2048 INFO: Number of params: 2.5667 M +[04/24 14:33:58] ModelNet40Ply2048 INFO: Param groups = { + "decay": { + "weight_decay": 0.05, + "params": [ + "encoder.encoder.0.0.convs.0.0.weight", + "encoder.encoder.1.0.skipconv.0.weight", + "encoder.encoder.1.0.preconv.0.weight", + "encoder.encoder.1.0.scorenet_global.0.weight", + "encoder.encoder.1.0.scorenet_global.3.weight", + "encoder.encoder.1.0.pt.linear_q.weight", + "encoder.encoder.1.0.pt.linear_k.weight", + "encoder.encoder.1.0.pt.linear_p.0.weight", + "encoder.encoder.1.0.pt.linear_p.3.weight", + "encoder.encoder.1.0.pt.w.2.weight", + "encoder.encoder.1.0.pt.v.0.weight", + "encoder.encoder.1.0.pt.v.3.weight", + "encoder.encoder.1.0.pt.conv_p.0.weight", + "encoder.encoder.1.0.pt.conv_p.3.weight", + "encoder.encoder.1.0.conv_finanal.0.weight", + "encoder.encoder.1.0.selfattention.linear_q.weight", + "encoder.encoder.1.0.selfattention.linear_k.weight", + "encoder.encoder.1.0.selfattention.linear_v.0.weight", + "encoder.encoder.1.0.selfattention.linear_v.3.weight", + "encoder.encoder.2.0.skipconv.0.weight", + "encoder.encoder.2.0.scorenet_global.0.weight", + "encoder.encoder.2.0.scorenet_global.3.weight", + "encoder.encoder.2.0.preconv.0.weight", + "encoder.encoder.2.0.pt.linear_q.weight", + "encoder.encoder.2.0.pt.linear_k.weight", + "encoder.encoder.2.0.pt.linear_p.0.weight", + "encoder.encoder.2.0.pt.linear_p.3.weight", + "encoder.encoder.2.0.pt.w.2.weight", + "encoder.encoder.2.0.pt.v.0.weight", + "encoder.encoder.2.0.pt.v.3.weight", + "encoder.encoder.2.0.pt.conv_p.0.weight", + "encoder.encoder.2.0.pt.conv_p.3.weight", + "encoder.encoder.2.0.conv_finanal.0.weight", + "encoder.encoder.2.0.selfattention.linear_q.weight", + "encoder.encoder.2.0.selfattention.linear_k.weight", + "encoder.encoder.2.0.selfattention.linear_v.0.weight", + "encoder.encoder.2.0.selfattention.linear_v.3.weight", + "encoder.encoder.3.0.skipconv.0.weight", + "encoder.encoder.3.0.scorenet_global.0.weight", + "encoder.encoder.3.0.scorenet_global.3.weight", + "encoder.encoder.3.0.preconv.0.weight", + "encoder.encoder.3.0.pt.linear_q.weight", + "encoder.encoder.3.0.pt.linear_k.weight", + "encoder.encoder.3.0.pt.linear_p.0.weight", + "encoder.encoder.3.0.pt.linear_p.3.weight", + "encoder.encoder.3.0.pt.w.2.weight", + "encoder.encoder.3.0.pt.v.0.weight", + "encoder.encoder.3.0.pt.v.3.weight", + "encoder.encoder.3.0.pt.conv_p.0.weight", + "encoder.encoder.3.0.pt.conv_p.3.weight", + "encoder.encoder.3.0.conv_finanal.0.weight", + "encoder.encoder.3.0.selfattention.linear_q.weight", + "encoder.encoder.3.0.selfattention.linear_k.weight", + "encoder.encoder.3.0.selfattention.linear_v.0.weight", + "encoder.encoder.3.0.selfattention.linear_v.3.weight", + "encoder.encoder.4.0.skipconv.0.weight", + "encoder.encoder.4.0.scorenet_global.0.weight", + "encoder.encoder.4.0.scorenet_global.3.weight", + "encoder.encoder.4.0.preconv.0.weight", + "encoder.encoder.4.0.pt.linear_q.weight", + "encoder.encoder.4.0.pt.linear_k.weight", + "encoder.encoder.4.0.pt.linear_p.0.weight", + "encoder.encoder.4.0.pt.linear_p.3.weight", + "encoder.encoder.4.0.pt.w.2.weight", + "encoder.encoder.4.0.pt.v.0.weight", + "encoder.encoder.4.0.pt.v.3.weight", + "encoder.encoder.4.0.pt.conv_p.0.weight", + "encoder.encoder.4.0.pt.conv_p.3.weight", + "encoder.encoder.4.0.conv_finanal.0.weight", + "encoder.encoder.4.0.selfattention.linear_q.weight", + "encoder.encoder.4.0.selfattention.linear_k.weight", + "encoder.encoder.4.0.selfattention.linear_v.0.weight", + "encoder.encoder.4.0.selfattention.linear_v.3.weight", + "encoder.encoder.5.0.preconv.0.weight", + "prediction.head.0.0.weight", + "prediction.head.2.0.weight", + "prediction.head.4.0.weight" + ], + "lr_scale": 1.0 + }, + "no_decay": { + "weight_decay": 0.0, + "params": [ + "encoder.encoder.0.0.convs.0.0.bias", + "encoder.encoder.1.0.beta", + "encoder.encoder.1.0.skipconv.0.bias", + "encoder.encoder.1.0.preconv.1.weight", + "encoder.encoder.1.0.preconv.1.bias", + "encoder.encoder.1.0.scorenet_global.0.bias", + "encoder.encoder.1.0.scorenet_global.1.weight", + "encoder.encoder.1.0.scorenet_global.1.bias", + "encoder.encoder.1.0.scorenet_global.3.bias", + "encoder.encoder.1.0.scorenet_global.4.weight", + "encoder.encoder.1.0.scorenet_global.4.bias", + "encoder.encoder.1.0.pt.linear_q.bias", + "encoder.encoder.1.0.pt.linear_k.bias", + "encoder.encoder.1.0.pt.linear_p.0.bias", + "encoder.encoder.1.0.pt.linear_p.1.weight", + "encoder.encoder.1.0.pt.linear_p.1.bias", + "encoder.encoder.1.0.pt.linear_p.3.bias", + "encoder.encoder.1.0.pt.w.0.weight", + "encoder.encoder.1.0.pt.w.0.bias", + "encoder.encoder.1.0.pt.w.2.bias", + "encoder.encoder.1.0.pt.w.3.weight", + "encoder.encoder.1.0.pt.w.3.bias", + "encoder.encoder.1.0.pt.v.0.bias", + "encoder.encoder.1.0.pt.v.1.weight", + "encoder.encoder.1.0.pt.v.1.bias", + "encoder.encoder.1.0.pt.v.3.bias", + "encoder.encoder.1.0.pt.conv_p.0.bias", + "encoder.encoder.1.0.pt.conv_p.1.weight", + "encoder.encoder.1.0.pt.conv_p.1.bias", + "encoder.encoder.1.0.pt.conv_p.3.bias", + "encoder.encoder.1.0.conv_finanal.1.weight", + "encoder.encoder.1.0.conv_finanal.1.bias", + "encoder.encoder.1.0.selfattention.linear_q.bias", + "encoder.encoder.1.0.selfattention.linear_k.bias", + "encoder.encoder.1.0.selfattention.linear_v.0.bias", + "encoder.encoder.1.0.selfattention.linear_v.1.weight", + "encoder.encoder.1.0.selfattention.linear_v.1.bias", + "encoder.encoder.1.0.selfattention.linear_v.3.bias", + "encoder.encoder.1.0.selfattention.linear_v.4.weight", + "encoder.encoder.1.0.selfattention.linear_v.4.bias", + "encoder.encoder.2.0.beta", + "encoder.encoder.2.0.skipconv.0.bias", + "encoder.encoder.2.0.scorenet_global.0.bias", + "encoder.encoder.2.0.scorenet_global.1.weight", + "encoder.encoder.2.0.scorenet_global.1.bias", + "encoder.encoder.2.0.scorenet_global.3.bias", + "encoder.encoder.2.0.scorenet_global.4.weight", + "encoder.encoder.2.0.scorenet_global.4.bias", + "encoder.encoder.2.0.preconv.1.weight", + "encoder.encoder.2.0.preconv.1.bias", + "encoder.encoder.2.0.pt.linear_q.bias", + "encoder.encoder.2.0.pt.linear_k.bias", + "encoder.encoder.2.0.pt.linear_p.0.bias", + "encoder.encoder.2.0.pt.linear_p.1.weight", + "encoder.encoder.2.0.pt.linear_p.1.bias", + "encoder.encoder.2.0.pt.linear_p.3.bias", + "encoder.encoder.2.0.pt.w.0.weight", + "encoder.encoder.2.0.pt.w.0.bias", + "encoder.encoder.2.0.pt.w.2.bias", + "encoder.encoder.2.0.pt.w.3.weight", + "encoder.encoder.2.0.pt.w.3.bias", + "encoder.encoder.2.0.pt.v.0.bias", + "encoder.encoder.2.0.pt.v.1.weight", + "encoder.encoder.2.0.pt.v.1.bias", + "encoder.encoder.2.0.pt.v.3.bias", + "encoder.encoder.2.0.pt.conv_p.0.bias", + "encoder.encoder.2.0.pt.conv_p.1.weight", + "encoder.encoder.2.0.pt.conv_p.1.bias", + "encoder.encoder.2.0.pt.conv_p.3.bias", + "encoder.encoder.2.0.conv_finanal.1.weight", + "encoder.encoder.2.0.conv_finanal.1.bias", + "encoder.encoder.2.0.selfattention.linear_q.bias", + "encoder.encoder.2.0.selfattention.linear_k.bias", + "encoder.encoder.2.0.selfattention.linear_v.0.bias", + "encoder.encoder.2.0.selfattention.linear_v.1.weight", + "encoder.encoder.2.0.selfattention.linear_v.1.bias", + "encoder.encoder.2.0.selfattention.linear_v.3.bias", + "encoder.encoder.2.0.selfattention.linear_v.4.weight", + "encoder.encoder.2.0.selfattention.linear_v.4.bias", + "encoder.encoder.3.0.beta", + "encoder.encoder.3.0.skipconv.0.bias", + "encoder.encoder.3.0.scorenet_global.0.bias", + "encoder.encoder.3.0.scorenet_global.1.weight", + "encoder.encoder.3.0.scorenet_global.1.bias", + "encoder.encoder.3.0.scorenet_global.3.bias", + "encoder.encoder.3.0.scorenet_global.4.weight", + "encoder.encoder.3.0.scorenet_global.4.bias", + "encoder.encoder.3.0.preconv.1.weight", + "encoder.encoder.3.0.preconv.1.bias", + "encoder.encoder.3.0.pt.linear_q.bias", + "encoder.encoder.3.0.pt.linear_k.bias", + "encoder.encoder.3.0.pt.linear_p.0.bias", + "encoder.encoder.3.0.pt.linear_p.1.weight", + "encoder.encoder.3.0.pt.linear_p.1.bias", + "encoder.encoder.3.0.pt.linear_p.3.bias", + "encoder.encoder.3.0.pt.w.0.weight", + "encoder.encoder.3.0.pt.w.0.bias", + "encoder.encoder.3.0.pt.w.2.bias", + "encoder.encoder.3.0.pt.w.3.weight", + "encoder.encoder.3.0.pt.w.3.bias", + "encoder.encoder.3.0.pt.v.0.bias", + "encoder.encoder.3.0.pt.v.1.weight", + "encoder.encoder.3.0.pt.v.1.bias", + "encoder.encoder.3.0.pt.v.3.bias", + "encoder.encoder.3.0.pt.conv_p.0.bias", + "encoder.encoder.3.0.pt.conv_p.1.weight", + "encoder.encoder.3.0.pt.conv_p.1.bias", + "encoder.encoder.3.0.pt.conv_p.3.bias", + "encoder.encoder.3.0.conv_finanal.1.weight", + "encoder.encoder.3.0.conv_finanal.1.bias", + "encoder.encoder.3.0.selfattention.linear_q.bias", + "encoder.encoder.3.0.selfattention.linear_k.bias", + "encoder.encoder.3.0.selfattention.linear_v.0.bias", + "encoder.encoder.3.0.selfattention.linear_v.1.weight", + "encoder.encoder.3.0.selfattention.linear_v.1.bias", + "encoder.encoder.3.0.selfattention.linear_v.3.bias", + "encoder.encoder.3.0.selfattention.linear_v.4.weight", + "encoder.encoder.3.0.selfattention.linear_v.4.bias", + "encoder.encoder.4.0.beta", + "encoder.encoder.4.0.skipconv.0.bias", + "encoder.encoder.4.0.scorenet_global.0.bias", + "encoder.encoder.4.0.scorenet_global.1.weight", + "encoder.encoder.4.0.scorenet_global.1.bias", + "encoder.encoder.4.0.scorenet_global.3.bias", + "encoder.encoder.4.0.scorenet_global.4.weight", + "encoder.encoder.4.0.scorenet_global.4.bias", + "encoder.encoder.4.0.preconv.1.weight", + "encoder.encoder.4.0.preconv.1.bias", + "encoder.encoder.4.0.pt.linear_q.bias", + "encoder.encoder.4.0.pt.linear_k.bias", + "encoder.encoder.4.0.pt.linear_p.0.bias", + "encoder.encoder.4.0.pt.linear_p.1.weight", + "encoder.encoder.4.0.pt.linear_p.1.bias", + "encoder.encoder.4.0.pt.linear_p.3.bias", + "encoder.encoder.4.0.pt.w.0.weight", + "encoder.encoder.4.0.pt.w.0.bias", + "encoder.encoder.4.0.pt.w.2.bias", + "encoder.encoder.4.0.pt.w.3.weight", + "encoder.encoder.4.0.pt.w.3.bias", + "encoder.encoder.4.0.pt.v.0.bias", + "encoder.encoder.4.0.pt.v.1.weight", + "encoder.encoder.4.0.pt.v.1.bias", + "encoder.encoder.4.0.pt.v.3.bias", + "encoder.encoder.4.0.pt.conv_p.0.bias", + "encoder.encoder.4.0.pt.conv_p.1.weight", + "encoder.encoder.4.0.pt.conv_p.1.bias", + "encoder.encoder.4.0.pt.conv_p.3.bias", + "encoder.encoder.4.0.conv_finanal.1.weight", + "encoder.encoder.4.0.conv_finanal.1.bias", + "encoder.encoder.4.0.selfattention.linear_q.bias", + "encoder.encoder.4.0.selfattention.linear_k.bias", + "encoder.encoder.4.0.selfattention.linear_v.0.bias", + "encoder.encoder.4.0.selfattention.linear_v.1.weight", + "encoder.encoder.4.0.selfattention.linear_v.1.bias", + "encoder.encoder.4.0.selfattention.linear_v.3.bias", + "encoder.encoder.4.0.selfattention.linear_v.4.weight", + "encoder.encoder.4.0.selfattention.linear_v.4.bias", + "encoder.encoder.5.0.preconv.1.weight", + "encoder.encoder.5.0.preconv.1.bias", + "prediction.head.0.1.weight", + "prediction.head.0.1.bias", + "prediction.head.2.1.weight", + "prediction.head.2.1.bias", + "prediction.head.4.0.bias" + ], + "lr_scale": 1.0 + } +} +[04/24 14:33:59] ModelNet40Ply2048 INFO: ==> sucessfully loaded test data +[04/24 14:33:59] ModelNet40Ply2048 INFO: length of validation dataset: 2468 +[04/24 14:34:00] ModelNet40Ply2048 INFO: ==> sucessfully loaded test data +[04/24 14:34:00] ModelNet40Ply2048 INFO: number of classes of the dataset: 40, number of points sampled from dataset: 1024, number of points as model input: 1024 +[04/24 14:34:00] ModelNet40Ply2048 INFO: Training from scratch +[04/24 14:34:02] ModelNet40Ply2048 INFO: ==> sucessfully loaded train data +[04/24 14:34:02] ModelNet40Ply2048 INFO: length of training dataset: 9840 +[04/24 14:34:54] ModelNet40Ply2048 INFO: Find a better ckpt @E1 +[04/24 14:34:54] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 99.00% +bathtub : 0.00% +bed : 87.00% +bench : 0.00% +bookshelf : 96.00% +bottle : 98.00% +bowl : 0.00% +car : 95.00% +chair : 95.00% +cone : 85.00% +cup : 0.00% +curtain : 0.00% +desk : 31.40% +door : 70.00% +dresser : 58.14% +flower_pot: 0.00% +glass_box : 40.00% +guitar : 78.00% +keyboard : 100.00% +lamp : 25.00% +laptop : 90.00% +mantel : 2.00% +monitor : 92.00% +night_stand: 2.33% +person : 30.00% +piano : 17.00% +plant : 55.00% +radio : 0.00% +range_hood: 0.00% +sink : 0.00% +sofa : 94.00% +stairs : 0.00% +stool : 0.00% +table : 95.00% +tent : 0.00% +toilet : 92.00% +tv_stand : 49.00% +vase : 49.00% +wardrobe : 10.00% +xbox : 0.00% +E@1 OA: 56.48 mAcc: 43.37 + +[04/24 14:34:54] ModelNet40Ply2048 INFO: Epoch 1 LR 0.001000 train_oa 32.86, val_oa 56.48, best val oa 56.48 +[04/24 14:34:54] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:35:41] ModelNet40Ply2048 INFO: Find a better ckpt @E2 +[04/24 14:35:41] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 99.00% +bathtub : 22.00% +bed : 90.00% +bench : 5.00% +bookshelf : 58.00% +bottle : 81.00% +bowl : 45.00% +car : 84.00% +chair : 96.00% +cone : 80.00% +cup : 0.00% +curtain : 60.00% +desk : 29.07% +door : 35.00% +dresser : 52.33% +flower_pot: 0.00% +glass_box : 25.00% +guitar : 91.00% +keyboard : 95.00% +lamp : 80.00% +laptop : 95.00% +mantel : 58.00% +monitor : 90.00% +night_stand: 17.44% +person : 65.00% +piano : 10.00% +plant : 90.00% +radio : 0.00% +range_hood: 48.00% +sink : 35.00% +sofa : 75.00% +stairs : 0.00% +stool : 25.00% +table : 93.00% +tent : 25.00% +toilet : 100.00% +tv_stand : 13.00% +vase : 90.00% +wardrobe : 30.00% +xbox : 0.00% +E@2 OA: 61.67 mAcc: 52.17 + +[04/24 14:35:41] ModelNet40Ply2048 INFO: Epoch 2 LR 0.001000 train_oa 62.33, val_oa 61.67, best val oa 61.67 +[04/24 14:35:41] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:36:30] ModelNet40Ply2048 INFO: Find a better ckpt @E3 +[04/24 14:36:30] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 46.00% +bed : 96.00% +bench : 30.00% +bookshelf : 94.00% +bottle : 95.00% +bowl : 65.00% +car : 67.00% +chair : 100.00% +cone : 85.00% +cup : 5.00% +curtain : 55.00% +desk : 67.44% +door : 50.00% +dresser : 70.93% +flower_pot: 0.00% +glass_box : 81.00% +guitar : 100.00% +keyboard : 95.00% +lamp : 70.00% +laptop : 95.00% +mantel : 91.00% +monitor : 99.00% +night_stand: 43.02% +person : 80.00% +piano : 76.00% +plant : 83.00% +radio : 0.00% +range_hood: 85.00% +sink : 65.00% +sofa : 100.00% +stairs : 30.00% +stool : 60.00% +table : 89.00% +tent : 90.00% +toilet : 100.00% +tv_stand : 58.00% +vase : 68.00% +wardrobe : 5.00% +xbox : 25.00% +E@3 OA: 78.69 mAcc: 67.86 + +[04/24 14:36:30] ModelNet40Ply2048 INFO: Epoch 3 LR 0.001000 train_oa 71.46, val_oa 78.69, best val oa 78.69 +[04/24 14:36:30] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:37:17] ModelNet40Ply2048 INFO: Epoch 4 LR 0.001000 train_oa 75.60, val_oa 77.19, best val oa 78.69 +[04/24 14:38:05] ModelNet40Ply2048 INFO: Find a better ckpt @E5 +[04/24 14:38:05] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 74.00% +bed : 98.00% +bench : 45.00% +bookshelf : 97.00% +bottle : 88.00% +bowl : 60.00% +car : 92.00% +chair : 96.00% +cone : 85.00% +cup : 5.00% +curtain : 100.00% +desk : 89.53% +door : 65.00% +dresser : 68.60% +flower_pot: 0.00% +glass_box : 82.00% +guitar : 99.00% +keyboard : 95.00% +lamp : 80.00% +laptop : 100.00% +mantel : 89.00% +monitor : 96.00% +night_stand: 33.72% +person : 40.00% +piano : 80.00% +plant : 92.00% +radio : 25.00% +range_hood: 95.00% +sink : 60.00% +sofa : 100.00% +stairs : 50.00% +stool : 80.00% +table : 76.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 90.00% +vase : 88.00% +wardrobe : 70.00% +xbox : 55.00% +E@5 OA: 84.36 mAcc: 75.85 + +[04/24 14:38:05] ModelNet40Ply2048 INFO: Epoch 5 LR 0.001000 train_oa 79.03, val_oa 84.36, best val oa 84.36 +[04/24 14:38:05] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:38:51] ModelNet40Ply2048 INFO: Find a better ckpt @E6 +[04/24 14:38:51] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 84.00% +bed : 97.00% +bench : 40.00% +bookshelf : 97.00% +bottle : 97.00% +bowl : 95.00% +car : 96.00% +chair : 97.00% +cone : 90.00% +cup : 65.00% +curtain : 60.00% +desk : 81.40% +door : 95.00% +dresser : 90.70% +flower_pot: 0.00% +glass_box : 91.00% +guitar : 97.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 90.00% +monitor : 98.00% +night_stand: 38.37% +person : 85.00% +piano : 81.00% +plant : 80.00% +radio : 10.00% +range_hood: 87.00% +sink : 70.00% +sofa : 99.00% +stairs : 70.00% +stool : 75.00% +table : 91.00% +tent : 90.00% +toilet : 100.00% +tv_stand : 86.00% +vase : 68.00% +wardrobe : 15.00% +xbox : 50.00% +E@6 OA: 85.70 mAcc: 78.66 + +[04/24 14:38:51] ModelNet40Ply2048 INFO: Epoch 6 LR 0.001000 train_oa 80.19, val_oa 85.70, best val oa 85.70 +[04/24 14:38:51] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:39:42] ModelNet40Ply2048 INFO: Epoch 7 LR 0.001000 train_oa 82.03, val_oa 85.25, best val oa 85.70 +[04/24 14:40:33] ModelNet40Ply2048 INFO: Find a better ckpt @E8 +[04/24 14:40:33] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 82.00% +bed : 99.00% +bench : 65.00% +bookshelf : 98.00% +bottle : 97.00% +bowl : 70.00% +car : 99.00% +chair : 100.00% +cone : 90.00% +cup : 15.00% +curtain : 100.00% +desk : 89.53% +door : 85.00% +dresser : 80.23% +flower_pot: 0.00% +glass_box : 93.00% +guitar : 99.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 93.00% +monitor : 97.00% +night_stand: 63.95% +person : 90.00% +piano : 76.00% +plant : 89.00% +radio : 5.00% +range_hood: 89.00% +sink : 55.00% +sofa : 100.00% +stairs : 60.00% +stool : 50.00% +table : 75.00% +tent : 90.00% +toilet : 100.00% +tv_stand : 62.00% +vase : 92.00% +wardrobe : 55.00% +xbox : 70.00% +E@8 OA: 86.55 mAcc: 78.84 + +[04/24 14:40:33] ModelNet40Ply2048 INFO: Epoch 8 LR 0.001000 train_oa 82.94, val_oa 86.55, best val oa 86.55 +[04/24 14:40:33] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:41:26] ModelNet40Ply2048 INFO: Find a better ckpt @E9 +[04/24 14:41:26] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 86.00% +bed : 98.00% +bench : 60.00% +bookshelf : 96.00% +bottle : 97.00% +bowl : 80.00% +car : 98.00% +chair : 100.00% +cone : 90.00% +cup : 50.00% +curtain : 90.00% +desk : 97.67% +door : 95.00% +dresser : 80.23% +flower_pot: 25.00% +glass_box : 92.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 92.00% +monitor : 98.00% +night_stand: 55.81% +person : 95.00% +piano : 83.00% +plant : 57.00% +radio : 30.00% +range_hood: 92.00% +sink : 60.00% +sofa : 100.00% +stairs : 65.00% +stool : 75.00% +table : 57.00% +tent : 100.00% +toilet : 100.00% +tv_stand : 88.00% +vase : 89.00% +wardrobe : 55.00% +xbox : 80.00% +E@9 OA: 86.99 mAcc: 82.17 + +[04/24 14:41:26] ModelNet40Ply2048 INFO: Epoch 9 LR 0.001000 train_oa 83.67, val_oa 86.99, best val oa 86.99 +[04/24 14:41:26] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:42:16] ModelNet40Ply2048 INFO: Find a better ckpt @E10 +[04/24 14:42:16] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 84.00% +bed : 99.00% +bench : 60.00% +bookshelf : 95.00% +bottle : 94.00% +bowl : 90.00% +car : 96.00% +chair : 99.00% +cone : 90.00% +cup : 45.00% +curtain : 100.00% +desk : 83.72% +door : 90.00% +dresser : 89.53% +flower_pot: 5.00% +glass_box : 90.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 94.00% +monitor : 97.00% +night_stand: 59.30% +person : 100.00% +piano : 91.00% +plant : 78.00% +radio : 45.00% +range_hood: 92.00% +sink : 70.00% +sofa : 99.00% +stairs : 90.00% +stool : 80.00% +table : 71.00% +tent : 95.00% +toilet : 97.00% +tv_stand : 91.00% +vase : 83.00% +wardrobe : 60.00% +xbox : 75.00% +E@10 OA: 88.53 mAcc: 84.19 + +[04/24 14:42:16] ModelNet40Ply2048 INFO: Epoch 10 LR 0.000999 train_oa 84.81, val_oa 88.53, best val oa 88.53 +[04/24 14:42:16] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:43:09] ModelNet40Ply2048 INFO: Epoch 11 LR 0.000999 train_oa 85.35, val_oa 88.49, best val oa 88.53 +[04/24 14:44:00] ModelNet40Ply2048 INFO: Epoch 12 LR 0.000999 train_oa 85.53, val_oa 87.48, best val oa 88.53 +[04/24 14:44:53] ModelNet40Ply2048 INFO: Find a better ckpt @E13 +[04/24 14:44:53] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 92.00% +bed : 99.00% +bench : 70.00% +bookshelf : 99.00% +bottle : 98.00% +bowl : 85.00% +car : 99.00% +chair : 99.00% +cone : 90.00% +cup : 65.00% +curtain : 90.00% +desk : 91.86% +door : 95.00% +dresser : 88.37% +flower_pot: 0.00% +glass_box : 93.00% +guitar : 99.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 90.00% +monitor : 99.00% +night_stand: 54.65% +person : 90.00% +piano : 86.00% +plant : 92.00% +radio : 45.00% +range_hood: 92.00% +sink : 70.00% +sofa : 100.00% +stairs : 75.00% +stool : 85.00% +table : 90.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 83.00% +vase : 84.00% +wardrobe : 60.00% +xbox : 55.00% +E@13 OA: 89.99 mAcc: 84.60 + +[04/24 14:44:53] ModelNet40Ply2048 INFO: Epoch 13 LR 0.000999 train_oa 86.22, val_oa 89.99, best val oa 89.99 +[04/24 14:44:54] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:45:40] ModelNet40Ply2048 INFO: Epoch 14 LR 0.000999 train_oa 86.47, val_oa 87.20, best val oa 89.99 +[04/24 14:46:28] ModelNet40Ply2048 INFO: Epoch 15 LR 0.000999 train_oa 87.01, val_oa 88.17, best val oa 89.99 +[04/24 14:47:16] ModelNet40Ply2048 INFO: Epoch 16 LR 0.000998 train_oa 87.47, val_oa 89.95, best val oa 89.99 +[04/24 14:48:09] ModelNet40Ply2048 INFO: Epoch 17 LR 0.000998 train_oa 88.01, val_oa 89.51, best val oa 89.99 +[04/24 14:48:58] ModelNet40Ply2048 INFO: Epoch 18 LR 0.000998 train_oa 88.16, val_oa 89.02, best val oa 89.99 +[04/24 14:49:44] ModelNet40Ply2048 INFO: Epoch 19 LR 0.000998 train_oa 88.30, val_oa 89.51, best val oa 89.99 +[04/24 14:50:33] ModelNet40Ply2048 INFO: Epoch 20 LR 0.000998 train_oa 88.45, val_oa 89.47, best val oa 89.99 +[04/24 14:51:18] ModelNet40Ply2048 INFO: Epoch 21 LR 0.000997 train_oa 89.02, val_oa 89.10, best val oa 89.99 +[04/24 14:52:05] ModelNet40Ply2048 INFO: Find a better ckpt @E22 +[04/24 14:52:05] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 92.00% +bed : 98.00% +bench : 80.00% +bookshelf : 98.00% +bottle : 98.00% +bowl : 85.00% +car : 97.00% +chair : 100.00% +cone : 95.00% +cup : 45.00% +curtain : 100.00% +desk : 91.86% +door : 95.00% +dresser : 82.56% +flower_pot: 10.00% +glass_box : 95.00% +guitar : 99.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 95.00% +monitor : 99.00% +night_stand: 66.28% +person : 90.00% +piano : 83.00% +plant : 76.00% +radio : 70.00% +range_hood: 93.00% +sink : 85.00% +sofa : 100.00% +stairs : 95.00% +stool : 75.00% +table : 91.00% +tent : 95.00% +toilet : 99.00% +tv_stand : 86.00% +vase : 86.00% +wardrobe : 45.00% +xbox : 70.00% +E@22 OA: 90.40 mAcc: 86.27 + +[04/24 14:52:05] ModelNet40Ply2048 INFO: Epoch 22 LR 0.000997 train_oa 89.14, val_oa 90.40, best val oa 90.40 +[04/24 14:52:05] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:52:56] ModelNet40Ply2048 INFO: Epoch 23 LR 0.000997 train_oa 89.51, val_oa 89.71, best val oa 90.40 +[04/24 14:53:45] ModelNet40Ply2048 INFO: Epoch 24 LR 0.000996 train_oa 88.96, val_oa 90.32, best val oa 90.40 +[04/24 14:54:32] ModelNet40Ply2048 INFO: Epoch 25 LR 0.000996 train_oa 90.12, val_oa 90.15, best val oa 90.40 +[04/24 14:55:17] ModelNet40Ply2048 INFO: Find a better ckpt @E26 +[04/24 14:55:17] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 92.00% +bed : 99.00% +bench : 70.00% +bookshelf : 98.00% +bottle : 97.00% +bowl : 85.00% +car : 100.00% +chair : 97.00% +cone : 90.00% +cup : 55.00% +curtain : 100.00% +desk : 87.21% +door : 90.00% +dresser : 84.88% +flower_pot: 20.00% +glass_box : 86.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 96.00% +monitor : 97.00% +night_stand: 73.26% +person : 95.00% +piano : 92.00% +plant : 76.00% +radio : 65.00% +range_hood: 96.00% +sink : 85.00% +sofa : 100.00% +stairs : 80.00% +stool : 90.00% +table : 75.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 81.00% +vase : 90.00% +wardrobe : 80.00% +xbox : 90.00% +E@26 OA: 90.48 mAcc: 87.43 + +[04/24 14:55:17] ModelNet40Ply2048 INFO: Epoch 26 LR 0.000996 train_oa 89.65, val_oa 90.48, best val oa 90.48 +[04/24 14:55:17] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:56:04] ModelNet40Ply2048 INFO: Find a better ckpt @E27 +[04/24 14:56:04] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 94.00% +bed : 98.00% +bench : 80.00% +bookshelf : 99.00% +bottle : 97.00% +bowl : 100.00% +car : 99.00% +chair : 99.00% +cone : 85.00% +cup : 70.00% +curtain : 90.00% +desk : 91.86% +door : 95.00% +dresser : 88.37% +flower_pot: 5.00% +glass_box : 90.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 94.00% +monitor : 99.00% +night_stand: 59.30% +person : 90.00% +piano : 89.00% +plant : 89.00% +radio : 70.00% +range_hood: 93.00% +sink : 80.00% +sofa : 100.00% +stairs : 90.00% +stool : 85.00% +table : 85.00% +tent : 95.00% +toilet : 99.00% +tv_stand : 90.00% +vase : 81.00% +wardrobe : 40.00% +xbox : 75.00% +E@27 OA: 90.80 mAcc: 86.74 + +[04/24 14:56:04] ModelNet40Ply2048 INFO: Epoch 27 LR 0.000995 train_oa 89.99, val_oa 90.80, best val oa 90.80 +[04/24 14:56:04] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:56:58] ModelNet40Ply2048 INFO: Find a better ckpt @E28 +[04/24 14:56:58] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 98.00% +bed : 98.00% +bench : 75.00% +bookshelf : 93.00% +bottle : 97.00% +bowl : 80.00% +car : 98.00% +chair : 99.00% +cone : 90.00% +cup : 30.00% +curtain : 90.00% +desk : 95.35% +door : 100.00% +dresser : 89.53% +flower_pot: 5.00% +glass_box : 92.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 75.00% +laptop : 100.00% +mantel : 92.00% +monitor : 99.00% +night_stand: 69.77% +person : 90.00% +piano : 92.00% +plant : 90.00% +radio : 80.00% +range_hood: 94.00% +sink : 80.00% +sofa : 100.00% +stairs : 90.00% +stool : 80.00% +table : 81.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 82.00% +vase : 91.00% +wardrobe : 65.00% +xbox : 85.00% +E@28 OA: 91.09 mAcc: 86.52 + +[04/24 14:56:58] ModelNet40Ply2048 INFO: Epoch 28 LR 0.000995 train_oa 90.31, val_oa 91.09, best val oa 91.09 +[04/24 14:56:58] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 14:57:53] ModelNet40Ply2048 INFO: Epoch 29 LR 0.000995 train_oa 90.28, val_oa 90.84, best val oa 91.09 +[04/24 14:58:42] ModelNet40Ply2048 INFO: Epoch 30 LR 0.000994 train_oa 90.49, val_oa 89.63, best val oa 91.09 +[04/24 14:59:29] ModelNet40Ply2048 INFO: Find a better ckpt @E31 +[04/24 14:59:29] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 90.00% +bed : 98.00% +bench : 80.00% +bookshelf : 96.00% +bottle : 98.00% +bowl : 100.00% +car : 100.00% +chair : 99.00% +cone : 95.00% +cup : 75.00% +curtain : 95.00% +desk : 95.35% +door : 100.00% +dresser : 79.07% +flower_pot: 10.00% +glass_box : 94.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 94.00% +monitor : 99.00% +night_stand: 76.74% +person : 100.00% +piano : 89.00% +plant : 86.00% +radio : 80.00% +range_hood: 94.00% +sink : 80.00% +sofa : 100.00% +stairs : 95.00% +stool : 85.00% +table : 81.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 84.00% +vase : 81.00% +wardrobe : 25.00% +xbox : 90.00% +E@31 OA: 91.25 mAcc: 88.10 + +[04/24 14:59:29] ModelNet40Ply2048 INFO: Epoch 31 LR 0.000994 train_oa 90.82, val_oa 91.25, best val oa 91.25 +[04/24 14:59:30] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:00:19] ModelNet40Ply2048 INFO: Epoch 32 LR 0.000993 train_oa 91.36, val_oa 90.88, best val oa 91.25 +[04/24 15:01:02] ModelNet40Ply2048 INFO: Epoch 33 LR 0.000993 train_oa 91.03, val_oa 90.52, best val oa 91.25 +[04/24 15:01:46] ModelNet40Ply2048 INFO: Find a better ckpt @E34 +[04/24 15:01:46] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 86.00% +bed : 98.00% +bench : 75.00% +bookshelf : 98.00% +bottle : 99.00% +bowl : 85.00% +car : 98.00% +chair : 99.00% +cone : 90.00% +cup : 90.00% +curtain : 100.00% +desk : 95.35% +door : 100.00% +dresser : 72.09% +flower_pot: 0.00% +glass_box : 93.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 93.00% +monitor : 98.00% +night_stand: 84.88% +person : 100.00% +piano : 91.00% +plant : 87.00% +radio : 90.00% +range_hood: 92.00% +sink : 85.00% +sofa : 100.00% +stairs : 95.00% +stool : 80.00% +table : 78.00% +tent : 90.00% +toilet : 100.00% +tv_stand : 86.00% +vase : 80.00% +wardrobe : 60.00% +xbox : 95.00% +E@34 OA: 91.29 mAcc: 88.58 + +[04/24 15:01:46] ModelNet40Ply2048 INFO: Epoch 34 LR 0.000993 train_oa 91.23, val_oa 91.29, best val oa 91.29 +[04/24 15:01:47] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:02:36] ModelNet40Ply2048 INFO: Epoch 35 LR 0.000992 train_oa 91.09, val_oa 91.17, best val oa 91.29 +[04/24 15:03:22] ModelNet40Ply2048 INFO: Epoch 36 LR 0.000992 train_oa 91.57, val_oa 90.40, best val oa 91.29 +[04/24 15:04:08] ModelNet40Ply2048 INFO: Epoch 37 LR 0.000991 train_oa 91.24, val_oa 90.52, best val oa 91.29 +[04/24 15:04:53] ModelNet40Ply2048 INFO: Epoch 38 LR 0.000991 train_oa 91.78, val_oa 90.72, best val oa 91.29 +[04/24 15:05:38] ModelNet40Ply2048 INFO: Epoch 39 LR 0.000990 train_oa 91.67, val_oa 90.88, best val oa 91.29 +[04/24 15:06:21] ModelNet40Ply2048 INFO: Epoch 40 LR 0.000990 train_oa 92.39, val_oa 90.92, best val oa 91.29 +[04/24 15:07:09] ModelNet40Ply2048 INFO: Find a better ckpt @E41 +[04/24 15:07:09] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 94.00% +bed : 98.00% +bench : 80.00% +bookshelf : 97.00% +bottle : 98.00% +bowl : 85.00% +car : 99.00% +chair : 99.00% +cone : 80.00% +cup : 65.00% +curtain : 100.00% +desk : 93.02% +door : 100.00% +dresser : 90.70% +flower_pot: 0.00% +glass_box : 91.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 96.00% +monitor : 98.00% +night_stand: 67.44% +person : 95.00% +piano : 91.00% +plant : 89.00% +radio : 75.00% +range_hood: 95.00% +sink : 75.00% +sofa : 100.00% +stairs : 90.00% +stool : 85.00% +table : 83.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 89.00% +vase : 86.00% +wardrobe : 55.00% +xbox : 95.00% +E@41 OA: 91.73 mAcc: 87.85 + +[04/24 15:07:09] ModelNet40Ply2048 INFO: Epoch 41 LR 0.000989 train_oa 92.30, val_oa 91.73, best val oa 91.73 +[04/24 15:07:09] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:07:55] ModelNet40Ply2048 INFO: Epoch 42 LR 0.000989 train_oa 92.01, val_oa 91.17, best val oa 91.73 +[04/24 15:08:48] ModelNet40Ply2048 INFO: Epoch 43 LR 0.000988 train_oa 92.66, val_oa 91.53, best val oa 91.73 +[04/24 15:09:41] ModelNet40Ply2048 INFO: Epoch 44 LR 0.000987 train_oa 92.19, val_oa 91.13, best val oa 91.73 +[04/24 15:10:29] ModelNet40Ply2048 INFO: Epoch 45 LR 0.000987 train_oa 92.62, val_oa 90.80, best val oa 91.73 +[04/24 15:11:18] ModelNet40Ply2048 INFO: Epoch 46 LR 0.000986 train_oa 92.65, val_oa 91.37, best val oa 91.73 +[04/24 15:12:06] ModelNet40Ply2048 INFO: Epoch 47 LR 0.000986 train_oa 92.83, val_oa 91.25, best val oa 91.73 +[04/24 15:12:52] ModelNet40Ply2048 INFO: Epoch 48 LR 0.000985 train_oa 92.92, val_oa 90.40, best val oa 91.73 +[04/24 15:13:41] ModelNet40Ply2048 INFO: Find a better ckpt @E49 +[04/24 15:13:41] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 96.00% +bed : 99.00% +bench : 75.00% +bookshelf : 97.00% +bottle : 97.00% +bowl : 100.00% +car : 99.00% +chair : 100.00% +cone : 95.00% +cup : 70.00% +curtain : 95.00% +desk : 93.02% +door : 95.00% +dresser : 94.19% +flower_pot: 10.00% +glass_box : 92.00% +guitar : 100.00% +keyboard : 95.00% +lamp : 75.00% +laptop : 100.00% +mantel : 96.00% +monitor : 99.00% +night_stand: 58.14% +person : 100.00% +piano : 92.00% +plant : 84.00% +radio : 80.00% +range_hood: 97.00% +sink : 85.00% +sofa : 100.00% +stairs : 95.00% +stool : 80.00% +table : 85.00% +tent : 90.00% +toilet : 100.00% +tv_stand : 89.00% +vase : 83.00% +wardrobe : 65.00% +xbox : 85.00% +E@49 OA: 91.82 mAcc: 88.51 + +[04/24 15:13:41] ModelNet40Ply2048 INFO: Epoch 49 LR 0.000984 train_oa 93.26, val_oa 91.82, best val oa 91.82 +[04/24 15:13:41] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:14:26] ModelNet40Ply2048 INFO: Epoch 50 LR 0.000984 train_oa 92.93, val_oa 90.64, best val oa 91.82 +[04/24 15:15:12] ModelNet40Ply2048 INFO: Epoch 51 LR 0.000983 train_oa 93.16, val_oa 90.72, best val oa 91.82 +[04/24 15:15:59] ModelNet40Ply2048 INFO: Epoch 52 LR 0.000982 train_oa 92.95, val_oa 91.29, best val oa 91.82 +[04/24 15:16:46] ModelNet40Ply2048 INFO: Epoch 53 LR 0.000982 train_oa 93.24, val_oa 91.69, best val oa 91.82 +[04/24 15:17:33] ModelNet40Ply2048 INFO: Epoch 54 LR 0.000981 train_oa 93.36, val_oa 91.21, best val oa 91.82 +[04/24 15:18:23] ModelNet40Ply2048 INFO: Epoch 55 LR 0.000980 train_oa 93.27, val_oa 91.77, best val oa 91.82 +[04/24 15:19:10] ModelNet40Ply2048 INFO: Epoch 56 LR 0.000979 train_oa 93.41, val_oa 91.69, best val oa 91.82 +[04/24 15:20:05] ModelNet40Ply2048 INFO: Epoch 57 LR 0.000979 train_oa 93.20, val_oa 91.49, best val oa 91.82 +[04/24 15:20:55] ModelNet40Ply2048 INFO: Find a better ckpt @E58 +[04/24 15:20:55] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 98.00% +bed : 98.00% +bench : 80.00% +bookshelf : 98.00% +bottle : 96.00% +bowl : 90.00% +car : 98.00% +chair : 99.00% +cone : 95.00% +cup : 80.00% +curtain : 100.00% +desk : 91.86% +door : 95.00% +dresser : 83.72% +flower_pot: 0.00% +glass_box : 95.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 95.00% +monitor : 100.00% +night_stand: 63.95% +person : 100.00% +piano : 93.00% +plant : 91.00% +radio : 70.00% +range_hood: 97.00% +sink : 90.00% +sofa : 100.00% +stairs : 95.00% +stool : 85.00% +table : 88.00% +tent : 95.00% +toilet : 98.00% +tv_stand : 81.00% +vase : 83.00% +wardrobe : 85.00% +xbox : 90.00% +E@58 OA: 92.10 mAcc: 89.69 + +[04/24 15:20:55] ModelNet40Ply2048 INFO: Epoch 58 LR 0.000978 train_oa 93.08, val_oa 92.10, best val oa 92.10 +[04/24 15:20:55] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:21:42] ModelNet40Ply2048 INFO: Epoch 59 LR 0.000977 train_oa 93.88, val_oa 91.45, best val oa 92.10 +[04/24 15:22:29] ModelNet40Ply2048 INFO: Epoch 60 LR 0.000976 train_oa 93.80, val_oa 90.48, best val oa 92.10 +[04/24 15:23:17] ModelNet40Ply2048 INFO: Epoch 61 LR 0.000976 train_oa 93.45, val_oa 91.49, best val oa 92.10 +[04/24 15:23:59] ModelNet40Ply2048 INFO: Epoch 62 LR 0.000975 train_oa 94.24, val_oa 91.41, best val oa 92.10 +[04/24 15:24:46] ModelNet40Ply2048 INFO: Epoch 63 LR 0.000974 train_oa 93.45, val_oa 91.82, best val oa 92.10 +[04/24 15:25:33] ModelNet40Ply2048 INFO: Epoch 64 LR 0.000973 train_oa 94.00, val_oa 91.90, best val oa 92.10 +[04/24 15:26:24] ModelNet40Ply2048 INFO: Epoch 65 LR 0.000972 train_oa 93.71, val_oa 91.25, best val oa 92.10 +[04/24 15:27:11] ModelNet40Ply2048 INFO: Epoch 66 LR 0.000971 train_oa 94.13, val_oa 91.21, best val oa 92.10 +[04/24 15:28:00] ModelNet40Ply2048 INFO: Epoch 67 LR 0.000970 train_oa 94.30, val_oa 91.73, best val oa 92.10 +[04/24 15:28:53] ModelNet40Ply2048 INFO: Epoch 68 LR 0.000970 train_oa 93.96, val_oa 91.57, best val oa 92.10 +[04/24 15:29:36] ModelNet40Ply2048 INFO: Epoch 69 LR 0.000969 train_oa 94.20, val_oa 89.99, best val oa 92.10 +[04/24 15:30:21] ModelNet40Ply2048 INFO: Epoch 70 LR 0.000968 train_oa 94.32, val_oa 91.61, best val oa 92.10 +[04/24 15:31:10] ModelNet40Ply2048 INFO: Epoch 71 LR 0.000967 train_oa 94.46, val_oa 91.94, best val oa 92.10 +[04/24 15:31:57] ModelNet40Ply2048 INFO: Epoch 72 LR 0.000966 train_oa 94.46, val_oa 91.21, best val oa 92.10 +[04/24 15:32:47] ModelNet40Ply2048 INFO: Epoch 73 LR 0.000965 train_oa 94.50, val_oa 91.29, best val oa 92.10 +[04/24 15:33:35] ModelNet40Ply2048 INFO: Epoch 74 LR 0.000964 train_oa 94.54, val_oa 91.61, best val oa 92.10 +[04/24 15:34:22] ModelNet40Ply2048 INFO: Find a better ckpt @E75 +[04/24 15:34:22] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 100.00% +bed : 99.00% +bench : 80.00% +bookshelf : 98.00% +bottle : 98.00% +bowl : 95.00% +car : 99.00% +chair : 99.00% +cone : 95.00% +cup : 75.00% +curtain : 95.00% +desk : 93.02% +door : 90.00% +dresser : 79.07% +flower_pot: 5.00% +glass_box : 96.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 95.00% +monitor : 99.00% +night_stand: 77.91% +person : 100.00% +piano : 94.00% +plant : 90.00% +radio : 70.00% +range_hood: 93.00% +sink : 95.00% +sofa : 100.00% +stairs : 90.00% +stool : 85.00% +table : 83.00% +tent : 95.00% +toilet : 99.00% +tv_stand : 86.00% +vase : 83.00% +wardrobe : 60.00% +xbox : 90.00% +E@75 OA: 92.26 mAcc: 89.15 + +[04/24 15:34:22] ModelNet40Ply2048 INFO: Epoch 75 LR 0.000963 train_oa 94.46, val_oa 92.26, best val oa 92.26 +[04/24 15:34:22] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:35:14] ModelNet40Ply2048 INFO: Epoch 76 LR 0.000962 train_oa 94.55, val_oa 91.41, best val oa 92.26 +[04/24 15:36:09] ModelNet40Ply2048 INFO: Epoch 77 LR 0.000961 train_oa 94.53, val_oa 90.88, best val oa 92.26 +[04/24 15:37:04] ModelNet40Ply2048 INFO: Epoch 78 LR 0.000960 train_oa 94.84, val_oa 91.09, best val oa 92.26 +[04/24 15:37:55] ModelNet40Ply2048 INFO: Epoch 79 LR 0.000959 train_oa 94.74, val_oa 91.57, best val oa 92.26 +[04/24 15:38:49] ModelNet40Ply2048 INFO: Epoch 80 LR 0.000958 train_oa 94.94, val_oa 90.56, best val oa 92.26 +[04/24 15:39:33] ModelNet40Ply2048 INFO: Epoch 81 LR 0.000957 train_oa 94.82, val_oa 91.13, best val oa 92.26 +[04/24 15:40:23] ModelNet40Ply2048 INFO: Epoch 82 LR 0.000956 train_oa 94.77, val_oa 90.92, best val oa 92.26 +[04/24 15:41:11] ModelNet40Ply2048 INFO: Epoch 83 LR 0.000955 train_oa 94.87, val_oa 91.77, best val oa 92.26 +[04/24 15:41:58] ModelNet40Ply2048 INFO: Epoch 84 LR 0.000954 train_oa 94.91, val_oa 91.41, best val oa 92.26 +[04/24 15:42:43] ModelNet40Ply2048 INFO: Epoch 85 LR 0.000952 train_oa 95.02, val_oa 90.64, best val oa 92.26 +[04/24 15:43:37] ModelNet40Ply2048 INFO: Epoch 86 LR 0.000951 train_oa 94.90, val_oa 91.61, best val oa 92.26 +[04/24 15:44:28] ModelNet40Ply2048 INFO: Epoch 87 LR 0.000950 train_oa 95.16, val_oa 91.57, best val oa 92.26 +[04/24 15:45:20] ModelNet40Ply2048 INFO: Epoch 88 LR 0.000949 train_oa 95.12, val_oa 91.29, best val oa 92.26 +[04/24 15:46:13] ModelNet40Ply2048 INFO: Epoch 89 LR 0.000948 train_oa 95.21, val_oa 91.69, best val oa 92.26 +[04/24 15:47:03] ModelNet40Ply2048 INFO: Epoch 90 LR 0.000947 train_oa 95.22, val_oa 91.53, best val oa 92.26 +[04/24 15:47:54] ModelNet40Ply2048 INFO: Epoch 91 LR 0.000946 train_oa 95.29, val_oa 91.13, best val oa 92.26 +[04/24 15:48:44] ModelNet40Ply2048 INFO: Epoch 92 LR 0.000944 train_oa 95.43, val_oa 91.41, best val oa 92.26 +[04/24 15:49:38] ModelNet40Ply2048 INFO: Epoch 93 LR 0.000943 train_oa 95.00, val_oa 91.45, best val oa 92.26 +[04/24 15:50:25] ModelNet40Ply2048 INFO: Epoch 94 LR 0.000942 train_oa 95.27, val_oa 91.25, best val oa 92.26 +[04/24 15:51:14] ModelNet40Ply2048 INFO: Epoch 95 LR 0.000941 train_oa 95.28, val_oa 91.45, best val oa 92.26 +[04/24 15:51:59] ModelNet40Ply2048 INFO: Epoch 96 LR 0.000939 train_oa 95.30, val_oa 91.94, best val oa 92.26 +[04/24 15:52:49] ModelNet40Ply2048 INFO: Epoch 97 LR 0.000938 train_oa 95.54, val_oa 91.37, best val oa 92.26 +[04/24 15:53:42] ModelNet40Ply2048 INFO: Epoch 98 LR 0.000937 train_oa 95.19, val_oa 91.45, best val oa 92.26 +[04/24 15:54:35] ModelNet40Ply2048 INFO: Epoch 99 LR 0.000936 train_oa 95.59, val_oa 91.53, best val oa 92.26 +[04/24 15:55:28] ModelNet40Ply2048 INFO: Epoch 100 LR 0.000934 train_oa 95.50, val_oa 91.29, best val oa 92.26 +[04/24 15:56:14] ModelNet40Ply2048 INFO: Epoch 101 LR 0.000933 train_oa 95.69, val_oa 91.25, best val oa 92.26 +[04/24 15:57:00] ModelNet40Ply2048 INFO: Epoch 102 LR 0.000932 train_oa 95.60, val_oa 91.33, best val oa 92.26 +[04/24 15:57:47] ModelNet40Ply2048 INFO: Epoch 103 LR 0.000930 train_oa 95.75, val_oa 91.90, best val oa 92.26 +[04/24 15:58:41] ModelNet40Ply2048 INFO: Find a better ckpt @E104 +[04/24 15:58:41] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 96.00% +bed : 99.00% +bench : 75.00% +bookshelf : 96.00% +bottle : 96.00% +bowl : 95.00% +car : 99.00% +chair : 99.00% +cone : 95.00% +cup : 75.00% +curtain : 100.00% +desk : 93.02% +door : 100.00% +dresser : 84.88% +flower_pot: 15.00% +glass_box : 96.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 95.00% +monitor : 97.00% +night_stand: 75.58% +person : 95.00% +piano : 93.00% +plant : 81.00% +radio : 75.00% +range_hood: 96.00% +sink : 85.00% +sofa : 100.00% +stairs : 95.00% +stool : 80.00% +table : 92.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 88.00% +vase : 83.00% +wardrobe : 75.00% +xbox : 90.00% +E@104 OA: 92.42 mAcc: 89.61 + +[04/24 15:58:41] ModelNet40Ply2048 INFO: Epoch 104 LR 0.000929 train_oa 95.93, val_oa 92.42, best val oa 92.42 +[04/24 15:58:42] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 15:59:32] ModelNet40Ply2048 INFO: Epoch 105 LR 0.000928 train_oa 95.82, val_oa 91.61, best val oa 92.42 +[04/24 16:00:21] ModelNet40Ply2048 INFO: Epoch 106 LR 0.000926 train_oa 95.86, val_oa 92.30, best val oa 92.42 +[04/24 16:01:06] ModelNet40Ply2048 INFO: Epoch 107 LR 0.000925 train_oa 95.85, val_oa 91.77, best val oa 92.42 +[04/24 16:01:55] ModelNet40Ply2048 INFO: Epoch 108 LR 0.000924 train_oa 96.04, val_oa 91.57, best val oa 92.42 +[04/24 16:02:41] ModelNet40Ply2048 INFO: Epoch 109 LR 0.000922 train_oa 95.99, val_oa 91.17, best val oa 92.42 +[04/24 16:03:29] ModelNet40Ply2048 INFO: Epoch 110 LR 0.000921 train_oa 95.65, val_oa 91.17, best val oa 92.42 +[04/24 16:04:23] ModelNet40Ply2048 INFO: Epoch 111 LR 0.000919 train_oa 96.08, val_oa 91.41, best val oa 92.42 +[04/24 16:05:14] ModelNet40Ply2048 INFO: Epoch 112 LR 0.000918 train_oa 95.79, val_oa 92.02, best val oa 92.42 +[04/24 16:06:04] ModelNet40Ply2048 INFO: Epoch 113 LR 0.000917 train_oa 95.83, val_oa 91.82, best val oa 92.42 +[04/24 16:06:53] ModelNet40Ply2048 INFO: Epoch 114 LR 0.000915 train_oa 95.76, val_oa 91.94, best val oa 92.42 +[04/24 16:07:40] ModelNet40Ply2048 INFO: Epoch 115 LR 0.000914 train_oa 96.22, val_oa 91.69, best val oa 92.42 +[04/24 16:08:26] ModelNet40Ply2048 INFO: Epoch 116 LR 0.000912 train_oa 95.72, val_oa 91.86, best val oa 92.42 +[04/24 16:09:15] ModelNet40Ply2048 INFO: Epoch 117 LR 0.000911 train_oa 96.21, val_oa 90.19, best val oa 92.42 +[04/24 16:10:01] ModelNet40Ply2048 INFO: Epoch 118 LR 0.000909 train_oa 95.96, val_oa 91.29, best val oa 92.42 +[04/24 16:10:48] ModelNet40Ply2048 INFO: Find a better ckpt @E119 +[04/24 16:10:48] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 96.00% +bed : 99.00% +bench : 80.00% +bookshelf : 95.00% +bottle : 97.00% +bowl : 100.00% +car : 100.00% +chair : 98.00% +cone : 95.00% +cup : 80.00% +curtain : 95.00% +desk : 90.70% +door : 95.00% +dresser : 80.23% +flower_pot: 30.00% +glass_box : 96.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 97.00% +monitor : 97.00% +night_stand: 81.40% +person : 100.00% +piano : 95.00% +plant : 84.00% +radio : 85.00% +range_hood: 95.00% +sink : 85.00% +sofa : 100.00% +stairs : 100.00% +stool : 85.00% +table : 81.00% +tent : 95.00% +toilet : 99.00% +tv_stand : 90.00% +vase : 82.00% +wardrobe : 70.00% +xbox : 90.00% +E@119 OA: 92.50 mAcc: 90.46 + +[04/24 16:10:48] ModelNet40Ply2048 INFO: Epoch 119 LR 0.000908 train_oa 96.30, val_oa 92.50, best val oa 92.50 +[04/24 16:10:48] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 16:11:36] ModelNet40Ply2048 INFO: Epoch 120 LR 0.000906 train_oa 96.19, val_oa 92.10, best val oa 92.50 +[04/24 16:12:25] ModelNet40Ply2048 INFO: Epoch 121 LR 0.000905 train_oa 96.21, val_oa 92.14, best val oa 92.50 +[04/24 16:13:15] ModelNet40Ply2048 INFO: Epoch 122 LR 0.000903 train_oa 96.29, val_oa 91.69, best val oa 92.50 +[04/24 16:14:03] ModelNet40Ply2048 INFO: Epoch 123 LR 0.000902 train_oa 96.00, val_oa 91.29, best val oa 92.50 +[04/24 16:14:49] ModelNet40Ply2048 INFO: Epoch 124 LR 0.000900 train_oa 95.98, val_oa 91.53, best val oa 92.50 +[04/24 16:15:36] ModelNet40Ply2048 INFO: Epoch 125 LR 0.000898 train_oa 96.18, val_oa 92.34, best val oa 92.50 +[04/24 16:16:27] ModelNet40Ply2048 INFO: Epoch 126 LR 0.000897 train_oa 96.29, val_oa 92.06, best val oa 92.50 +[04/24 16:17:15] ModelNet40Ply2048 INFO: Epoch 127 LR 0.000895 train_oa 96.29, val_oa 91.94, best val oa 92.50 +[04/24 16:18:03] ModelNet40Ply2048 INFO: Epoch 128 LR 0.000894 train_oa 96.47, val_oa 91.90, best val oa 92.50 +[04/24 16:18:50] ModelNet40Ply2048 INFO: Epoch 129 LR 0.000892 train_oa 96.77, val_oa 91.82, best val oa 92.50 +[04/24 16:19:38] ModelNet40Ply2048 INFO: Epoch 130 LR 0.000890 train_oa 96.25, val_oa 91.57, best val oa 92.50 +[04/24 16:20:30] ModelNet40Ply2048 INFO: Epoch 131 LR 0.000889 train_oa 96.36, val_oa 92.30, best val oa 92.50 +[04/24 16:21:25] ModelNet40Ply2048 INFO: Epoch 132 LR 0.000887 train_oa 96.42, val_oa 91.69, best val oa 92.50 +[04/24 16:22:12] ModelNet40Ply2048 INFO: Epoch 133 LR 0.000885 train_oa 96.48, val_oa 91.94, best val oa 92.50 +[04/24 16:22:58] ModelNet40Ply2048 INFO: Epoch 134 LR 0.000884 train_oa 96.38, val_oa 92.22, best val oa 92.50 +[04/24 16:23:44] ModelNet40Ply2048 INFO: Epoch 135 LR 0.000882 train_oa 96.27, val_oa 91.37, best val oa 92.50 +[04/24 16:24:30] ModelNet40Ply2048 INFO: Epoch 136 LR 0.000880 train_oa 96.41, val_oa 91.49, best val oa 92.50 +[04/24 16:25:17] ModelNet40Ply2048 INFO: Epoch 137 LR 0.000879 train_oa 96.32, val_oa 91.82, best val oa 92.50 +[04/24 16:26:09] ModelNet40Ply2048 INFO: Epoch 138 LR 0.000877 train_oa 96.41, val_oa 92.02, best val oa 92.50 +[04/24 16:26:59] ModelNet40Ply2048 INFO: Epoch 139 LR 0.000875 train_oa 96.24, val_oa 92.10, best val oa 92.50 +[04/24 16:27:45] ModelNet40Ply2048 INFO: Epoch 140 LR 0.000873 train_oa 96.51, val_oa 91.37, best val oa 92.50 +[04/24 16:28:30] ModelNet40Ply2048 INFO: Epoch 141 LR 0.000872 train_oa 96.47, val_oa 91.49, best val oa 92.50 +[04/24 16:29:17] ModelNet40Ply2048 INFO: Epoch 142 LR 0.000870 train_oa 96.55, val_oa 91.41, best val oa 92.50 +[04/24 16:30:03] ModelNet40Ply2048 INFO: Epoch 143 LR 0.000868 train_oa 96.30, val_oa 91.82, best val oa 92.50 +[04/24 16:30:48] ModelNet40Ply2048 INFO: Find a better ckpt @E144 +[04/24 16:30:48] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 96.00% +bed : 99.00% +bench : 80.00% +bookshelf : 99.00% +bottle : 97.00% +bowl : 100.00% +car : 100.00% +chair : 99.00% +cone : 95.00% +cup : 75.00% +curtain : 95.00% +desk : 95.35% +door : 100.00% +dresser : 86.05% +flower_pot: 5.00% +glass_box : 96.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 90.00% +laptop : 100.00% +mantel : 96.00% +monitor : 100.00% +night_stand: 72.09% +person : 95.00% +piano : 95.00% +plant : 88.00% +radio : 65.00% +range_hood: 95.00% +sink : 85.00% +sofa : 99.00% +stairs : 90.00% +stool : 85.00% +table : 81.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 87.00% +vase : 83.00% +wardrobe : 80.00% +xbox : 85.00% +E@144 OA: 92.54 mAcc: 89.59 + +[04/24 16:30:48] ModelNet40Ply2048 INFO: Epoch 144 LR 0.000866 train_oa 96.76, val_oa 92.54, best val oa 92.54 +[04/24 16:30:49] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 16:31:42] ModelNet40Ply2048 INFO: Find a better ckpt @E145 +[04/24 16:31:42] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 98.00% +bed : 99.00% +bench : 80.00% +bookshelf : 94.00% +bottle : 96.00% +bowl : 95.00% +car : 99.00% +chair : 98.00% +cone : 100.00% +cup : 65.00% +curtain : 95.00% +desk : 94.19% +door : 85.00% +dresser : 88.37% +flower_pot: 5.00% +glass_box : 95.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 95.00% +monitor : 100.00% +night_stand: 77.91% +person : 100.00% +piano : 95.00% +plant : 86.00% +radio : 90.00% +range_hood: 95.00% +sink : 85.00% +sofa : 99.00% +stairs : 90.00% +stool : 90.00% +table : 81.00% +tent : 95.00% +toilet : 98.00% +tv_stand : 86.00% +vase : 88.00% +wardrobe : 85.00% +xbox : 95.00% +E@145 OA: 92.59 mAcc: 90.06 + +[04/24 16:31:42] ModelNet40Ply2048 INFO: Epoch 145 LR 0.000865 train_oa 96.59, val_oa 92.59, best val oa 92.59 +[04/24 16:31:43] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 16:32:28] ModelNet40Ply2048 INFO: Epoch 146 LR 0.000863 train_oa 96.54, val_oa 91.53, best val oa 92.59 +[04/24 16:33:15] ModelNet40Ply2048 INFO: Epoch 147 LR 0.000861 train_oa 96.61, val_oa 91.65, best val oa 92.59 +[04/24 16:34:01] ModelNet40Ply2048 INFO: Epoch 148 LR 0.000859 train_oa 96.77, val_oa 91.73, best val oa 92.59 +[04/24 16:34:48] ModelNet40Ply2048 INFO: Epoch 149 LR 0.000857 train_oa 96.63, val_oa 92.34, best val oa 92.59 +[04/24 16:35:33] ModelNet40Ply2048 INFO: Epoch 150 LR 0.000856 train_oa 96.85, val_oa 91.98, best val oa 92.59 +[04/24 16:36:21] ModelNet40Ply2048 INFO: Epoch 151 LR 0.000854 train_oa 96.55, val_oa 91.33, best val oa 92.59 +[04/24 16:37:07] ModelNet40Ply2048 INFO: Epoch 152 LR 0.000852 train_oa 96.49, val_oa 91.17, best val oa 92.59 +[04/24 16:37:52] ModelNet40Ply2048 INFO: Epoch 153 LR 0.000850 train_oa 96.83, val_oa 92.22, best val oa 92.59 +[04/24 16:38:46] ModelNet40Ply2048 INFO: Epoch 154 LR 0.000848 train_oa 96.57, val_oa 91.73, best val oa 92.59 +[04/24 16:39:38] ModelNet40Ply2048 INFO: Epoch 155 LR 0.000846 train_oa 97.07, val_oa 91.77, best val oa 92.59 +[04/24 16:40:24] ModelNet40Ply2048 INFO: Epoch 156 LR 0.000844 train_oa 96.56, val_oa 91.86, best val oa 92.59 +[04/24 16:41:10] ModelNet40Ply2048 INFO: Epoch 157 LR 0.000842 train_oa 96.93, val_oa 90.64, best val oa 92.59 +[04/24 16:42:01] ModelNet40Ply2048 INFO: Epoch 158 LR 0.000841 train_oa 96.80, val_oa 91.65, best val oa 92.59 +[04/24 16:42:50] ModelNet40Ply2048 INFO: Epoch 159 LR 0.000839 train_oa 96.92, val_oa 92.02, best val oa 92.59 +[04/24 16:43:35] ModelNet40Ply2048 INFO: Epoch 160 LR 0.000837 train_oa 96.56, val_oa 91.65, best val oa 92.59 +[04/24 16:44:29] ModelNet40Ply2048 INFO: Epoch 161 LR 0.000835 train_oa 96.63, val_oa 91.86, best val oa 92.59 +[04/24 16:45:16] ModelNet40Ply2048 INFO: Find a better ckpt @E162 +[04/24 16:45:16] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 98.00% +bed : 99.00% +bench : 75.00% +bookshelf : 98.00% +bottle : 98.00% +bowl : 95.00% +car : 99.00% +chair : 99.00% +cone : 95.00% +cup : 80.00% +curtain : 95.00% +desk : 90.70% +door : 100.00% +dresser : 81.40% +flower_pot: 15.00% +glass_box : 95.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 96.00% +monitor : 99.00% +night_stand: 84.88% +person : 95.00% +piano : 96.00% +plant : 87.00% +radio : 85.00% +range_hood: 97.00% +sink : 90.00% +sofa : 99.00% +stairs : 95.00% +stool : 85.00% +table : 83.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 87.00% +vase : 80.00% +wardrobe : 75.00% +xbox : 95.00% +E@162 OA: 92.91 mAcc: 90.55 + +[04/24 16:45:16] ModelNet40Ply2048 INFO: Epoch 162 LR 0.000833 train_oa 96.97, val_oa 92.91, best val oa 92.91 +[04/24 16:45:16] ModelNet40Ply2048 INFO: Found the best model and saved in log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 16:46:02] ModelNet40Ply2048 INFO: Epoch 163 LR 0.000831 train_oa 96.79, val_oa 91.25, best val oa 92.91 +[04/24 16:46:50] ModelNet40Ply2048 INFO: Epoch 164 LR 0.000829 train_oa 97.22, val_oa 91.41, best val oa 92.91 +[04/24 16:47:36] ModelNet40Ply2048 INFO: Epoch 165 LR 0.000827 train_oa 96.89, val_oa 91.57, best val oa 92.91 +[04/24 16:48:22] ModelNet40Ply2048 INFO: Epoch 166 LR 0.000825 train_oa 97.19, val_oa 92.02, best val oa 92.91 +[04/24 16:49:09] ModelNet40Ply2048 INFO: Epoch 167 LR 0.000823 train_oa 96.90, val_oa 91.17, best val oa 92.91 +[04/24 16:50:02] ModelNet40Ply2048 INFO: Epoch 168 LR 0.000821 train_oa 96.85, val_oa 92.34, best val oa 92.91 +[04/24 16:50:50] ModelNet40Ply2048 INFO: Epoch 169 LR 0.000819 train_oa 97.07, val_oa 92.18, best val oa 92.91 +[04/24 16:51:36] ModelNet40Ply2048 INFO: Epoch 170 LR 0.000817 train_oa 96.85, val_oa 91.49, best val oa 92.91 +[04/24 16:52:26] ModelNet40Ply2048 INFO: Epoch 171 LR 0.000815 train_oa 96.96, val_oa 91.25, best val oa 92.91 +[04/24 16:53:18] ModelNet40Ply2048 INFO: Epoch 172 LR 0.000813 train_oa 96.79, val_oa 91.49, best val oa 92.91 +[04/24 16:54:13] ModelNet40Ply2048 INFO: Epoch 173 LR 0.000811 train_oa 97.08, val_oa 90.15, best val oa 92.91 +[04/24 16:55:00] ModelNet40Ply2048 INFO: Epoch 174 LR 0.000809 train_oa 97.10, val_oa 92.14, best val oa 92.91 +[04/24 16:55:51] ModelNet40Ply2048 INFO: Epoch 175 LR 0.000807 train_oa 97.18, val_oa 92.26, best val oa 92.91 +[04/24 16:56:45] ModelNet40Ply2048 INFO: Epoch 176 LR 0.000805 train_oa 96.86, val_oa 92.46, best val oa 92.91 +[04/24 16:57:33] ModelNet40Ply2048 INFO: Epoch 177 LR 0.000802 train_oa 97.16, val_oa 91.86, best val oa 92.91 +[04/24 16:58:20] ModelNet40Ply2048 INFO: Epoch 178 LR 0.000800 train_oa 97.01, val_oa 91.69, best val oa 92.91 +[04/24 16:59:12] ModelNet40Ply2048 INFO: Epoch 179 LR 0.000798 train_oa 97.20, val_oa 91.73, best val oa 92.91 +[04/24 17:00:05] ModelNet40Ply2048 INFO: Epoch 180 LR 0.000796 train_oa 97.00, val_oa 91.57, best val oa 92.91 +[04/24 17:00:51] ModelNet40Ply2048 INFO: Epoch 181 LR 0.000794 train_oa 97.06, val_oa 91.05, best val oa 92.91 +[04/24 17:01:39] ModelNet40Ply2048 INFO: Epoch 182 LR 0.000792 train_oa 97.12, val_oa 92.02, best val oa 92.91 +[04/24 17:02:28] ModelNet40Ply2048 INFO: Epoch 183 LR 0.000790 train_oa 97.00, val_oa 92.50, best val oa 92.91 +[04/24 17:03:22] ModelNet40Ply2048 INFO: Epoch 184 LR 0.000788 train_oa 97.20, val_oa 91.94, best val oa 92.91 +[04/24 17:04:15] ModelNet40Ply2048 INFO: Epoch 185 LR 0.000786 train_oa 97.32, val_oa 91.65, best val oa 92.91 +[04/24 17:05:06] ModelNet40Ply2048 INFO: Epoch 186 LR 0.000783 train_oa 97.40, val_oa 91.73, best val oa 92.91 +[04/24 17:05:57] ModelNet40Ply2048 INFO: Epoch 187 LR 0.000781 train_oa 97.28, val_oa 91.49, best val oa 92.91 +[04/24 17:06:43] ModelNet40Ply2048 INFO: Epoch 188 LR 0.000779 train_oa 97.27, val_oa 91.33, best val oa 92.91 +[04/24 17:07:31] ModelNet40Ply2048 INFO: Epoch 189 LR 0.000777 train_oa 97.31, val_oa 92.14, best val oa 92.91 +[04/24 17:08:18] ModelNet40Ply2048 INFO: Epoch 190 LR 0.000775 train_oa 97.28, val_oa 91.65, best val oa 92.91 +[04/24 17:09:07] ModelNet40Ply2048 INFO: Epoch 191 LR 0.000773 train_oa 97.06, val_oa 91.69, best val oa 92.91 +[04/24 17:09:51] ModelNet40Ply2048 INFO: Epoch 192 LR 0.000770 train_oa 97.31, val_oa 91.45, best val oa 92.91 +[04/24 17:10:35] ModelNet40Ply2048 INFO: Epoch 193 LR 0.000768 train_oa 97.52, val_oa 92.87, best val oa 92.91 +[04/24 17:11:19] ModelNet40Ply2048 INFO: Epoch 194 LR 0.000766 train_oa 97.43, val_oa 91.82, best val oa 92.91 +[04/24 17:12:07] ModelNet40Ply2048 INFO: Epoch 195 LR 0.000764 train_oa 97.53, val_oa 92.18, best val oa 92.91 +[04/24 17:12:53] ModelNet40Ply2048 INFO: Epoch 196 LR 0.000761 train_oa 97.29, val_oa 92.02, best val oa 92.91 +[04/24 17:13:45] ModelNet40Ply2048 INFO: Epoch 197 LR 0.000759 train_oa 97.63, val_oa 92.34, best val oa 92.91 +[04/24 17:14:34] ModelNet40Ply2048 INFO: Epoch 198 LR 0.000757 train_oa 97.12, val_oa 90.92, best val oa 92.91 +[04/24 17:15:20] ModelNet40Ply2048 INFO: Epoch 199 LR 0.000755 train_oa 97.22, val_oa 91.25, best val oa 92.91 +[04/24 17:16:04] ModelNet40Ply2048 INFO: Epoch 200 LR 0.000753 train_oa 97.02, val_oa 90.60, best val oa 92.91 +[04/24 17:16:47] ModelNet40Ply2048 INFO: Epoch 201 LR 0.000750 train_oa 97.25, val_oa 91.49, best val oa 92.91 +[04/24 17:17:31] ModelNet40Ply2048 INFO: Epoch 202 LR 0.000748 train_oa 97.42, val_oa 91.57, best val oa 92.91 +[04/24 17:18:16] ModelNet40Ply2048 INFO: Epoch 203 LR 0.000746 train_oa 97.29, val_oa 91.45, best val oa 92.91 +[04/24 17:19:09] ModelNet40Ply2048 INFO: Epoch 204 LR 0.000743 train_oa 97.56, val_oa 91.73, best val oa 92.91 +[04/24 17:19:56] ModelNet40Ply2048 INFO: Epoch 205 LR 0.000741 train_oa 97.31, val_oa 90.56, best val oa 92.91 +[04/24 17:20:43] ModelNet40Ply2048 INFO: Epoch 206 LR 0.000739 train_oa 97.00, val_oa 90.96, best val oa 92.91 +[04/24 17:21:36] ModelNet40Ply2048 INFO: Epoch 207 LR 0.000737 train_oa 97.63, val_oa 91.17, best val oa 92.91 +[04/24 17:22:23] ModelNet40Ply2048 INFO: Epoch 208 LR 0.000734 train_oa 97.14, val_oa 91.45, best val oa 92.91 +[04/24 17:23:16] ModelNet40Ply2048 INFO: Epoch 209 LR 0.000732 train_oa 97.59, val_oa 91.33, best val oa 92.91 +[04/24 17:24:06] ModelNet40Ply2048 INFO: Epoch 210 LR 0.000730 train_oa 97.51, val_oa 91.90, best val oa 92.91 +[04/24 17:24:53] ModelNet40Ply2048 INFO: Epoch 211 LR 0.000727 train_oa 97.51, val_oa 91.33, best val oa 92.91 +[04/24 17:25:43] ModelNet40Ply2048 INFO: Epoch 212 LR 0.000725 train_oa 97.47, val_oa 90.72, best val oa 92.91 +[04/24 17:26:31] ModelNet40Ply2048 INFO: Epoch 213 LR 0.000723 train_oa 97.59, val_oa 92.02, best val oa 92.91 +[04/24 17:27:19] ModelNet40Ply2048 INFO: Epoch 214 LR 0.000720 train_oa 97.51, val_oa 90.88, best val oa 92.91 +[04/24 17:28:06] ModelNet40Ply2048 INFO: Epoch 215 LR 0.000718 train_oa 97.55, val_oa 92.50, best val oa 92.91 +[04/24 17:29:00] ModelNet40Ply2048 INFO: Epoch 216 LR 0.000716 train_oa 97.49, val_oa 91.61, best val oa 92.91 +[04/24 17:29:47] ModelNet40Ply2048 INFO: Epoch 217 LR 0.000713 train_oa 97.39, val_oa 92.14, best val oa 92.91 +[04/24 17:30:32] ModelNet40Ply2048 INFO: Epoch 218 LR 0.000711 train_oa 97.65, val_oa 91.98, best val oa 92.91 +[04/24 17:31:21] ModelNet40Ply2048 INFO: Epoch 219 LR 0.000708 train_oa 97.37, val_oa 91.65, best val oa 92.91 +[04/24 17:32:09] ModelNet40Ply2048 INFO: Epoch 220 LR 0.000706 train_oa 97.46, val_oa 92.10, best val oa 92.91 +[04/24 17:32:52] ModelNet40Ply2048 INFO: Epoch 221 LR 0.000704 train_oa 97.48, val_oa 91.73, best val oa 92.91 +[04/24 17:33:38] ModelNet40Ply2048 INFO: Epoch 222 LR 0.000701 train_oa 97.66, val_oa 91.94, best val oa 92.91 +[04/24 17:34:25] ModelNet40Ply2048 INFO: Epoch 223 LR 0.000699 train_oa 97.58, val_oa 92.26, best val oa 92.91 +[04/24 17:35:11] ModelNet40Ply2048 INFO: Epoch 224 LR 0.000696 train_oa 97.57, val_oa 92.14, best val oa 92.91 +[04/24 17:36:01] ModelNet40Ply2048 INFO: Epoch 225 LR 0.000694 train_oa 97.51, val_oa 91.90, best val oa 92.91 +[04/24 17:36:52] ModelNet40Ply2048 INFO: Epoch 226 LR 0.000692 train_oa 97.82, val_oa 92.46, best val oa 92.91 +[04/24 17:37:46] ModelNet40Ply2048 INFO: Epoch 227 LR 0.000689 train_oa 97.72, val_oa 91.49, best val oa 92.91 +[04/24 17:38:38] ModelNet40Ply2048 INFO: Epoch 228 LR 0.000687 train_oa 97.59, val_oa 92.14, best val oa 92.91 +[04/24 17:39:26] ModelNet40Ply2048 INFO: Epoch 229 LR 0.000684 train_oa 97.66, val_oa 91.77, best val oa 92.91 +[04/24 17:40:16] ModelNet40Ply2048 INFO: Epoch 230 LR 0.000682 train_oa 97.74, val_oa 92.26, best val oa 92.91 +[04/24 17:41:00] ModelNet40Ply2048 INFO: Epoch 231 LR 0.000680 train_oa 97.61, val_oa 91.41, best val oa 92.91 +[04/24 17:41:45] ModelNet40Ply2048 INFO: Epoch 232 LR 0.000677 train_oa 97.81, val_oa 92.22, best val oa 92.91 +[04/24 17:42:35] ModelNet40Ply2048 INFO: Epoch 233 LR 0.000675 train_oa 97.81, val_oa 91.13, best val oa 92.91 +[04/24 17:43:25] ModelNet40Ply2048 INFO: Epoch 234 LR 0.000672 train_oa 97.73, val_oa 91.90, best val oa 92.91 +[04/24 17:44:16] ModelNet40Ply2048 INFO: Epoch 235 LR 0.000670 train_oa 97.69, val_oa 92.10, best val oa 92.91 +[04/24 17:45:03] ModelNet40Ply2048 INFO: Epoch 236 LR 0.000667 train_oa 97.95, val_oa 90.88, best val oa 92.91 +[04/24 17:45:52] ModelNet40Ply2048 INFO: Epoch 237 LR 0.000665 train_oa 97.98, val_oa 91.82, best val oa 92.91 +[04/24 17:46:39] ModelNet40Ply2048 INFO: Epoch 238 LR 0.000662 train_oa 98.00, val_oa 91.21, best val oa 92.91 +[04/24 17:47:25] ModelNet40Ply2048 INFO: Epoch 239 LR 0.000660 train_oa 98.10, val_oa 92.42, best val oa 92.91 +[04/24 17:48:11] ModelNet40Ply2048 INFO: Epoch 240 LR 0.000657 train_oa 97.73, val_oa 92.59, best val oa 92.91 +[04/24 17:48:58] ModelNet40Ply2048 INFO: Epoch 241 LR 0.000655 train_oa 97.81, val_oa 92.02, best val oa 92.91 +[04/24 17:49:47] ModelNet40Ply2048 INFO: Epoch 242 LR 0.000652 train_oa 97.88, val_oa 92.54, best val oa 92.91 +[04/24 17:50:38] ModelNet40Ply2048 INFO: Epoch 243 LR 0.000650 train_oa 97.82, val_oa 91.69, best val oa 92.91 +[04/24 17:51:25] ModelNet40Ply2048 INFO: Epoch 244 LR 0.000647 train_oa 97.80, val_oa 92.22, best val oa 92.91 +[04/24 17:52:16] ModelNet40Ply2048 INFO: Epoch 245 LR 0.000645 train_oa 97.64, val_oa 91.49, best val oa 92.91 +[04/24 17:53:04] ModelNet40Ply2048 INFO: Epoch 246 LR 0.000642 train_oa 97.93, val_oa 90.96, best val oa 92.91 +[04/24 17:53:47] ModelNet40Ply2048 INFO: Epoch 247 LR 0.000640 train_oa 97.98, val_oa 91.77, best val oa 92.91 +[04/24 17:54:34] ModelNet40Ply2048 INFO: Epoch 248 LR 0.000637 train_oa 98.11, val_oa 91.41, best val oa 92.91 +[04/24 17:55:29] ModelNet40Ply2048 INFO: Epoch 249 LR 0.000635 train_oa 98.16, val_oa 92.30, best val oa 92.91 +[04/24 17:56:20] ModelNet40Ply2048 INFO: Epoch 250 LR 0.000632 train_oa 97.89, val_oa 91.17, best val oa 92.91 +[04/24 17:57:06] ModelNet40Ply2048 INFO: Epoch 251 LR 0.000630 train_oa 98.20, val_oa 91.65, best val oa 92.91 +[04/24 17:57:52] ModelNet40Ply2048 INFO: Epoch 252 LR 0.000627 train_oa 97.94, val_oa 91.49, best val oa 92.91 +[04/24 17:58:43] ModelNet40Ply2048 INFO: Epoch 253 LR 0.000625 train_oa 97.96, val_oa 91.77, best val oa 92.91 +[04/24 17:59:33] ModelNet40Ply2048 INFO: Epoch 254 LR 0.000622 train_oa 98.07, val_oa 91.86, best val oa 92.91 +[04/24 18:00:23] ModelNet40Ply2048 INFO: Epoch 255 LR 0.000620 train_oa 98.13, val_oa 91.09, best val oa 92.91 +[04/24 18:01:14] ModelNet40Ply2048 INFO: Epoch 256 LR 0.000617 train_oa 97.99, val_oa 91.61, best val oa 92.91 +[04/24 18:02:04] ModelNet40Ply2048 INFO: Epoch 257 LR 0.000615 train_oa 98.00, val_oa 91.77, best val oa 92.91 +[04/24 18:02:51] ModelNet40Ply2048 INFO: Epoch 258 LR 0.000612 train_oa 97.88, val_oa 91.57, best val oa 92.91 +[04/24 18:03:38] ModelNet40Ply2048 INFO: Epoch 259 LR 0.000609 train_oa 98.00, val_oa 90.52, best val oa 92.91 +[04/24 18:04:24] ModelNet40Ply2048 INFO: Epoch 260 LR 0.000607 train_oa 98.16, val_oa 91.73, best val oa 92.91 +[04/24 18:05:07] ModelNet40Ply2048 INFO: Epoch 261 LR 0.000604 train_oa 97.90, val_oa 92.26, best val oa 92.91 +[04/24 18:05:57] ModelNet40Ply2048 INFO: Epoch 262 LR 0.000602 train_oa 98.02, val_oa 91.69, best val oa 92.91 +[04/24 18:06:44] ModelNet40Ply2048 INFO: Epoch 263 LR 0.000599 train_oa 98.11, val_oa 90.96, best val oa 92.91 +[04/24 18:07:31] ModelNet40Ply2048 INFO: Epoch 264 LR 0.000597 train_oa 98.02, val_oa 92.22, best val oa 92.91 +[04/24 18:08:18] ModelNet40Ply2048 INFO: Epoch 265 LR 0.000594 train_oa 98.05, val_oa 92.22, best val oa 92.91 +[04/24 18:09:07] ModelNet40Ply2048 INFO: Epoch 266 LR 0.000592 train_oa 98.13, val_oa 92.34, best val oa 92.91 +[04/24 18:09:59] ModelNet40Ply2048 INFO: Epoch 267 LR 0.000589 train_oa 98.08, val_oa 92.54, best val oa 92.91 +[04/24 18:10:46] ModelNet40Ply2048 INFO: Epoch 268 LR 0.000586 train_oa 97.91, val_oa 91.90, best val oa 92.91 +[04/24 18:11:31] ModelNet40Ply2048 INFO: Epoch 269 LR 0.000584 train_oa 98.08, val_oa 92.22, best val oa 92.91 +[04/24 18:12:22] ModelNet40Ply2048 INFO: Epoch 270 LR 0.000581 train_oa 98.22, val_oa 92.02, best val oa 92.91 +[04/24 18:13:12] ModelNet40Ply2048 INFO: Epoch 271 LR 0.000579 train_oa 97.88, val_oa 92.14, best val oa 92.91 +[04/24 18:14:02] ModelNet40Ply2048 INFO: Epoch 272 LR 0.000576 train_oa 98.21, val_oa 92.14, best val oa 92.91 +[04/24 18:14:50] ModelNet40Ply2048 INFO: Epoch 273 LR 0.000573 train_oa 98.06, val_oa 92.50, best val oa 92.91 +[04/24 18:15:43] ModelNet40Ply2048 INFO: Epoch 274 LR 0.000571 train_oa 98.20, val_oa 92.26, best val oa 92.91 +[04/24 18:16:31] ModelNet40Ply2048 INFO: Epoch 275 LR 0.000568 train_oa 98.32, val_oa 91.57, best val oa 92.91 +[04/24 18:17:16] ModelNet40Ply2048 INFO: Epoch 276 LR 0.000566 train_oa 98.21, val_oa 90.96, best val oa 92.91 +[04/24 18:18:01] ModelNet40Ply2048 INFO: Epoch 277 LR 0.000563 train_oa 98.17, val_oa 92.18, best val oa 92.91 +[04/24 18:18:47] ModelNet40Ply2048 INFO: Epoch 278 LR 0.000561 train_oa 98.30, val_oa 91.86, best val oa 92.91 +[04/24 18:19:34] ModelNet40Ply2048 INFO: Epoch 279 LR 0.000558 train_oa 98.25, val_oa 91.53, best val oa 92.91 +[04/24 18:20:25] ModelNet40Ply2048 INFO: Epoch 280 LR 0.000555 train_oa 98.12, val_oa 91.73, best val oa 92.91 +[04/24 18:21:18] ModelNet40Ply2048 INFO: Epoch 281 LR 0.000553 train_oa 98.12, val_oa 91.57, best val oa 92.91 +[04/24 18:22:03] ModelNet40Ply2048 INFO: Epoch 282 LR 0.000550 train_oa 98.28, val_oa 91.29, best val oa 92.91 +[04/24 18:22:53] ModelNet40Ply2048 INFO: Epoch 283 LR 0.000548 train_oa 98.24, val_oa 91.69, best val oa 92.91 +[04/24 18:23:43] ModelNet40Ply2048 INFO: Epoch 284 LR 0.000545 train_oa 98.27, val_oa 91.86, best val oa 92.91 +[04/24 18:24:34] ModelNet40Ply2048 INFO: Epoch 285 LR 0.000542 train_oa 98.45, val_oa 92.02, best val oa 92.91 +[04/24 18:25:21] ModelNet40Ply2048 INFO: Epoch 286 LR 0.000540 train_oa 98.32, val_oa 91.25, best val oa 92.91 +[04/24 18:26:08] ModelNet40Ply2048 INFO: Epoch 287 LR 0.000537 train_oa 98.54, val_oa 91.49, best val oa 92.91 +[04/24 18:26:59] ModelNet40Ply2048 INFO: Epoch 288 LR 0.000534 train_oa 98.36, val_oa 92.06, best val oa 92.91 +[04/24 18:27:46] ModelNet40Ply2048 INFO: Epoch 289 LR 0.000532 train_oa 98.29, val_oa 92.63, best val oa 92.91 +[04/24 18:28:33] ModelNet40Ply2048 INFO: Epoch 290 LR 0.000529 train_oa 98.34, val_oa 91.49, best val oa 92.91 +[04/24 18:29:19] ModelNet40Ply2048 INFO: Epoch 291 LR 0.000527 train_oa 98.31, val_oa 91.82, best val oa 92.91 +[04/24 18:30:07] ModelNet40Ply2048 INFO: Epoch 292 LR 0.000524 train_oa 98.32, val_oa 92.02, best val oa 92.91 +[04/24 18:30:54] ModelNet40Ply2048 INFO: Epoch 293 LR 0.000521 train_oa 98.48, val_oa 92.46, best val oa 92.91 +[04/24 18:31:45] ModelNet40Ply2048 INFO: Epoch 294 LR 0.000519 train_oa 98.28, val_oa 91.98, best val oa 92.91 +[04/24 18:32:32] ModelNet40Ply2048 INFO: Epoch 295 LR 0.000516 train_oa 98.51, val_oa 91.05, best val oa 92.91 +[04/24 18:33:16] ModelNet40Ply2048 INFO: Epoch 296 LR 0.000514 train_oa 98.51, val_oa 91.86, best val oa 92.91 +[04/24 18:34:00] ModelNet40Ply2048 INFO: Epoch 297 LR 0.000511 train_oa 98.35, val_oa 91.61, best val oa 92.91 +[04/24 18:34:47] ModelNet40Ply2048 INFO: Epoch 298 LR 0.000508 train_oa 98.28, val_oa 91.94, best val oa 92.91 +[04/24 18:35:34] ModelNet40Ply2048 INFO: Epoch 299 LR 0.000506 train_oa 98.34, val_oa 92.34, best val oa 92.91 +[04/24 18:36:27] ModelNet40Ply2048 INFO: Epoch 300 LR 0.000503 train_oa 98.35, val_oa 92.63, best val oa 92.91 +[04/24 18:37:15] ModelNet40Ply2048 INFO: Epoch 301 LR 0.000501 train_oa 98.34, val_oa 90.88, best val oa 92.91 +[04/24 18:38:01] ModelNet40Ply2048 INFO: Epoch 302 LR 0.000498 train_oa 98.36, val_oa 91.61, best val oa 92.91 +[04/24 18:38:50] ModelNet40Ply2048 INFO: Epoch 303 LR 0.000495 train_oa 98.47, val_oa 91.82, best val oa 92.91 +[04/24 18:39:44] ModelNet40Ply2048 INFO: Epoch 304 LR 0.000493 train_oa 98.64, val_oa 92.14, best val oa 92.91 +[04/24 18:40:34] ModelNet40Ply2048 INFO: Epoch 305 LR 0.000490 train_oa 98.36, val_oa 91.49, best val oa 92.91 +[04/24 18:41:24] ModelNet40Ply2048 INFO: Epoch 306 LR 0.000487 train_oa 98.60, val_oa 91.25, best val oa 92.91 +[04/24 18:42:15] ModelNet40Ply2048 INFO: Epoch 307 LR 0.000485 train_oa 98.59, val_oa 91.82, best val oa 92.91 +[04/24 18:43:07] ModelNet40Ply2048 INFO: Epoch 308 LR 0.000482 train_oa 98.56, val_oa 91.41, best val oa 92.91 +[04/24 18:43:53] ModelNet40Ply2048 INFO: Epoch 309 LR 0.000480 train_oa 98.48, val_oa 92.18, best val oa 92.91 +[04/24 18:44:39] ModelNet40Ply2048 INFO: Epoch 310 LR 0.000477 train_oa 98.47, val_oa 91.41, best val oa 92.91 +[04/24 18:45:29] ModelNet40Ply2048 INFO: Epoch 311 LR 0.000474 train_oa 98.56, val_oa 91.61, best val oa 92.91 +[04/24 18:46:17] ModelNet40Ply2048 INFO: Epoch 312 LR 0.000472 train_oa 98.59, val_oa 91.98, best val oa 92.91 +[04/24 18:47:11] ModelNet40Ply2048 INFO: Epoch 313 LR 0.000469 train_oa 98.45, val_oa 92.30, best val oa 92.91 +[04/24 18:48:04] ModelNet40Ply2048 INFO: Epoch 314 LR 0.000467 train_oa 98.62, val_oa 91.53, best val oa 92.91 +[04/24 18:48:50] ModelNet40Ply2048 INFO: Epoch 315 LR 0.000464 train_oa 98.65, val_oa 91.86, best val oa 92.91 +[04/24 18:49:41] ModelNet40Ply2048 INFO: Epoch 316 LR 0.000461 train_oa 98.54, val_oa 91.82, best val oa 92.91 +[04/24 18:50:33] ModelNet40Ply2048 INFO: Epoch 317 LR 0.000459 train_oa 98.77, val_oa 91.57, best val oa 92.91 +[04/24 18:51:19] ModelNet40Ply2048 INFO: Epoch 318 LR 0.000456 train_oa 98.53, val_oa 92.02, best val oa 92.91 +[04/24 18:52:05] ModelNet40Ply2048 INFO: Epoch 319 LR 0.000453 train_oa 98.43, val_oa 91.82, best val oa 92.91 +[04/24 18:52:52] ModelNet40Ply2048 INFO: Epoch 320 LR 0.000451 train_oa 98.74, val_oa 91.82, best val oa 92.91 +[04/24 18:53:40] ModelNet40Ply2048 INFO: Epoch 321 LR 0.000448 train_oa 98.67, val_oa 91.49, best val oa 92.91 +[04/24 18:54:22] ModelNet40Ply2048 INFO: Epoch 322 LR 0.000446 train_oa 98.52, val_oa 92.18, best val oa 92.91 +[04/24 18:55:10] ModelNet40Ply2048 INFO: Epoch 323 LR 0.000443 train_oa 98.66, val_oa 91.98, best val oa 92.91 +[04/24 18:55:58] ModelNet40Ply2048 INFO: Epoch 324 LR 0.000440 train_oa 98.79, val_oa 91.37, best val oa 92.91 +[04/24 18:56:43] ModelNet40Ply2048 INFO: Epoch 325 LR 0.000438 train_oa 98.73, val_oa 90.92, best val oa 92.91 +[04/24 18:57:26] ModelNet40Ply2048 INFO: Epoch 326 LR 0.000435 train_oa 98.69, val_oa 91.53, best val oa 92.91 +[04/24 18:58:13] ModelNet40Ply2048 INFO: Epoch 327 LR 0.000433 train_oa 99.05, val_oa 91.49, best val oa 92.91 +[04/24 18:58:59] ModelNet40Ply2048 INFO: Epoch 328 LR 0.000430 train_oa 98.47, val_oa 92.38, best val oa 92.91 +[04/24 18:59:45] ModelNet40Ply2048 INFO: Epoch 329 LR 0.000428 train_oa 98.59, val_oa 92.06, best val oa 92.91 +[04/24 19:00:34] ModelNet40Ply2048 INFO: Epoch 330 LR 0.000425 train_oa 98.86, val_oa 91.90, best val oa 92.91 +[04/24 19:01:27] ModelNet40Ply2048 INFO: Epoch 331 LR 0.000422 train_oa 98.82, val_oa 91.94, best val oa 92.91 +[04/24 19:02:14] ModelNet40Ply2048 INFO: Epoch 332 LR 0.000420 train_oa 98.57, val_oa 92.26, best val oa 92.91 +[04/24 19:03:01] ModelNet40Ply2048 INFO: Epoch 333 LR 0.000417 train_oa 98.65, val_oa 91.45, best val oa 92.91 +[04/24 19:03:48] ModelNet40Ply2048 INFO: Epoch 334 LR 0.000415 train_oa 98.70, val_oa 91.90, best val oa 92.91 +[04/24 19:04:37] ModelNet40Ply2048 INFO: Epoch 335 LR 0.000412 train_oa 98.89, val_oa 91.61, best val oa 92.91 +[04/24 19:05:23] ModelNet40Ply2048 INFO: Epoch 336 LR 0.000409 train_oa 98.93, val_oa 91.90, best val oa 92.91 +[04/24 19:06:13] ModelNet40Ply2048 INFO: Epoch 337 LR 0.000407 train_oa 98.75, val_oa 92.10, best val oa 92.91 +[04/24 19:07:07] ModelNet40Ply2048 INFO: Epoch 338 LR 0.000404 train_oa 98.77, val_oa 92.18, best val oa 92.91 +[04/24 19:07:59] ModelNet40Ply2048 INFO: Epoch 339 LR 0.000402 train_oa 98.71, val_oa 91.77, best val oa 92.91 +[04/24 19:08:49] ModelNet40Ply2048 INFO: Epoch 340 LR 0.000399 train_oa 98.76, val_oa 92.34, best val oa 92.91 +[04/24 19:09:41] ModelNet40Ply2048 INFO: Epoch 341 LR 0.000397 train_oa 98.94, val_oa 92.02, best val oa 92.91 +[04/24 19:10:34] ModelNet40Ply2048 INFO: Epoch 342 LR 0.000394 train_oa 98.63, val_oa 92.10, best val oa 92.91 +[04/24 19:11:23] ModelNet40Ply2048 INFO: Epoch 343 LR 0.000392 train_oa 98.81, val_oa 92.10, best val oa 92.91 +[04/24 19:12:09] ModelNet40Ply2048 INFO: Epoch 344 LR 0.000389 train_oa 98.87, val_oa 91.53, best val oa 92.91 +[04/24 19:12:48] ModelNet40Ply2048 INFO: Epoch 345 LR 0.000386 train_oa 98.99, val_oa 91.65, best val oa 92.91 +[04/24 19:13:33] ModelNet40Ply2048 INFO: Epoch 346 LR 0.000384 train_oa 98.84, val_oa 91.45, best val oa 92.91 +[04/24 19:14:21] ModelNet40Ply2048 INFO: Epoch 347 LR 0.000381 train_oa 98.92, val_oa 91.69, best val oa 92.91 +[04/24 19:15:16] ModelNet40Ply2048 INFO: Epoch 348 LR 0.000379 train_oa 98.87, val_oa 91.57, best val oa 92.91 +[04/24 19:16:10] ModelNet40Ply2048 INFO: Epoch 349 LR 0.000376 train_oa 98.89, val_oa 92.14, best val oa 92.91 +[04/24 19:17:00] ModelNet40Ply2048 INFO: Epoch 350 LR 0.000374 train_oa 98.86, val_oa 91.69, best val oa 92.91 +[04/24 19:17:50] ModelNet40Ply2048 INFO: Epoch 351 LR 0.000371 train_oa 98.79, val_oa 91.53, best val oa 92.91 +[04/24 19:18:37] ModelNet40Ply2048 INFO: Epoch 352 LR 0.000369 train_oa 98.94, val_oa 91.98, best val oa 92.91 +[04/24 19:19:26] ModelNet40Ply2048 INFO: Epoch 353 LR 0.000366 train_oa 98.96, val_oa 92.02, best val oa 92.91 +[04/24 19:20:14] ModelNet40Ply2048 INFO: Epoch 354 LR 0.000364 train_oa 98.86, val_oa 92.42, best val oa 92.91 +[04/24 19:20:56] ModelNet40Ply2048 INFO: Epoch 355 LR 0.000361 train_oa 99.04, val_oa 91.98, best val oa 92.91 +[04/24 19:21:41] ModelNet40Ply2048 INFO: Epoch 356 LR 0.000359 train_oa 99.01, val_oa 91.82, best val oa 92.91 +[04/24 19:22:33] ModelNet40Ply2048 INFO: Epoch 357 LR 0.000356 train_oa 98.95, val_oa 91.61, best val oa 92.91 +[04/24 19:23:19] ModelNet40Ply2048 INFO: Epoch 358 LR 0.000354 train_oa 98.89, val_oa 91.98, best val oa 92.91 +[04/24 19:24:06] ModelNet40Ply2048 INFO: Epoch 359 LR 0.000351 train_oa 99.00, val_oa 91.45, best val oa 92.91 +[04/24 19:25:00] ModelNet40Ply2048 INFO: Epoch 360 LR 0.000349 train_oa 98.95, val_oa 91.86, best val oa 92.91 +[04/24 19:25:49] ModelNet40Ply2048 INFO: Epoch 361 LR 0.000346 train_oa 98.98, val_oa 91.86, best val oa 92.91 +[04/24 19:26:36] ModelNet40Ply2048 INFO: Epoch 362 LR 0.000344 train_oa 98.97, val_oa 92.38, best val oa 92.91 +[04/24 19:27:29] ModelNet40Ply2048 INFO: Epoch 363 LR 0.000341 train_oa 98.85, val_oa 91.41, best val oa 92.91 +[04/24 19:28:15] ModelNet40Ply2048 INFO: Epoch 364 LR 0.000339 train_oa 99.05, val_oa 92.22, best val oa 92.91 +[04/24 19:29:06] ModelNet40Ply2048 INFO: Epoch 365 LR 0.000336 train_oa 99.01, val_oa 92.42, best val oa 92.91 +[04/24 19:29:58] ModelNet40Ply2048 INFO: Epoch 366 LR 0.000334 train_oa 98.96, val_oa 92.46, best val oa 92.91 +[04/24 19:30:48] ModelNet40Ply2048 INFO: Epoch 367 LR 0.000331 train_oa 99.01, val_oa 91.82, best val oa 92.91 +[04/24 19:31:33] ModelNet40Ply2048 INFO: Epoch 368 LR 0.000329 train_oa 99.11, val_oa 91.94, best val oa 92.91 +[04/24 19:32:26] ModelNet40Ply2048 INFO: Epoch 369 LR 0.000326 train_oa 99.05, val_oa 91.69, best val oa 92.91 +[04/24 19:33:10] ModelNet40Ply2048 INFO: Epoch 370 LR 0.000324 train_oa 98.94, val_oa 92.06, best val oa 92.91 +[04/24 19:33:55] ModelNet40Ply2048 INFO: Epoch 371 LR 0.000321 train_oa 99.00, val_oa 92.10, best val oa 92.91 +[04/24 19:34:42] ModelNet40Ply2048 INFO: Epoch 372 LR 0.000319 train_oa 99.24, val_oa 91.90, best val oa 92.91 +[04/24 19:35:28] ModelNet40Ply2048 INFO: Epoch 373 LR 0.000317 train_oa 99.14, val_oa 91.98, best val oa 92.91 +[04/24 19:36:14] ModelNet40Ply2048 INFO: Epoch 374 LR 0.000314 train_oa 99.05, val_oa 92.38, best val oa 92.91 +[04/24 19:37:00] ModelNet40Ply2048 INFO: Epoch 375 LR 0.000312 train_oa 99.11, val_oa 91.33, best val oa 92.91 +[04/24 19:37:52] ModelNet40Ply2048 INFO: Epoch 376 LR 0.000309 train_oa 99.07, val_oa 91.57, best val oa 92.91 +[04/24 19:38:39] ModelNet40Ply2048 INFO: Epoch 377 LR 0.000307 train_oa 99.23, val_oa 91.86, best val oa 92.91 +[04/24 19:39:27] ModelNet40Ply2048 INFO: Epoch 378 LR 0.000305 train_oa 99.06, val_oa 92.02, best val oa 92.91 +[04/24 19:40:21] ModelNet40Ply2048 INFO: Epoch 379 LR 0.000302 train_oa 99.13, val_oa 92.10, best val oa 92.91 +[04/24 19:41:06] ModelNet40Ply2048 INFO: Epoch 380 LR 0.000300 train_oa 99.08, val_oa 92.06, best val oa 92.91 +[04/24 19:41:52] ModelNet40Ply2048 INFO: Epoch 381 LR 0.000297 train_oa 99.05, val_oa 91.57, best val oa 92.91 +[04/24 19:42:38] ModelNet40Ply2048 INFO: Epoch 382 LR 0.000295 train_oa 99.21, val_oa 91.73, best val oa 92.91 +[04/24 19:43:28] ModelNet40Ply2048 INFO: Epoch 383 LR 0.000293 train_oa 99.12, val_oa 92.18, best val oa 92.91 +[04/24 19:44:16] ModelNet40Ply2048 INFO: Epoch 384 LR 0.000290 train_oa 99.05, val_oa 91.86, best val oa 92.91 +[04/24 19:45:04] ModelNet40Ply2048 INFO: Epoch 385 LR 0.000288 train_oa 99.08, val_oa 91.82, best val oa 92.91 +[04/24 19:45:55] ModelNet40Ply2048 INFO: Epoch 386 LR 0.000285 train_oa 99.14, val_oa 92.30, best val oa 92.91 +[04/24 19:46:40] ModelNet40Ply2048 INFO: Epoch 387 LR 0.000283 train_oa 99.25, val_oa 92.42, best val oa 92.91 +[04/24 19:47:27] ModelNet40Ply2048 INFO: Epoch 388 LR 0.000281 train_oa 99.33, val_oa 91.82, best val oa 92.91 +[04/24 19:48:16] ModelNet40Ply2048 INFO: Epoch 389 LR 0.000278 train_oa 99.18, val_oa 92.22, best val oa 92.91 +[04/24 19:49:06] ModelNet40Ply2048 INFO: Epoch 390 LR 0.000276 train_oa 99.29, val_oa 91.90, best val oa 92.91 +[04/24 19:50:00] ModelNet40Ply2048 INFO: Epoch 391 LR 0.000274 train_oa 99.27, val_oa 91.53, best val oa 92.91 +[04/24 19:50:53] ModelNet40Ply2048 INFO: Epoch 392 LR 0.000271 train_oa 99.27, val_oa 91.41, best val oa 92.91 +[04/24 19:51:41] ModelNet40Ply2048 INFO: Epoch 393 LR 0.000269 train_oa 99.37, val_oa 91.98, best val oa 92.91 +[04/24 19:52:30] ModelNet40Ply2048 INFO: Epoch 394 LR 0.000267 train_oa 99.21, val_oa 92.34, best val oa 92.91 +[04/24 19:53:20] ModelNet40Ply2048 INFO: Epoch 395 LR 0.000264 train_oa 99.24, val_oa 91.73, best val oa 92.91 +[04/24 19:54:10] ModelNet40Ply2048 INFO: Epoch 396 LR 0.000262 train_oa 99.30, val_oa 92.83, best val oa 92.91 +[04/24 19:54:57] ModelNet40Ply2048 INFO: Epoch 397 LR 0.000260 train_oa 99.16, val_oa 91.53, best val oa 92.91 +[04/24 19:55:40] ModelNet40Ply2048 INFO: Epoch 398 LR 0.000258 train_oa 99.22, val_oa 91.90, best val oa 92.91 +[04/24 19:56:32] ModelNet40Ply2048 INFO: Epoch 399 LR 0.000255 train_oa 99.34, val_oa 92.26, best val oa 92.91 +[04/24 19:57:25] ModelNet40Ply2048 INFO: Epoch 400 LR 0.000253 train_oa 99.31, val_oa 91.90, best val oa 92.91 +[04/24 19:58:12] ModelNet40Ply2048 INFO: Epoch 401 LR 0.000251 train_oa 99.31, val_oa 91.77, best val oa 92.91 +[04/24 19:58:55] ModelNet40Ply2048 INFO: Epoch 402 LR 0.000248 train_oa 99.53, val_oa 91.77, best val oa 92.91 +[04/24 19:59:48] ModelNet40Ply2048 INFO: Epoch 403 LR 0.000246 train_oa 99.25, val_oa 91.57, best val oa 92.91 +[04/24 20:00:38] ModelNet40Ply2048 INFO: Epoch 404 LR 0.000244 train_oa 99.37, val_oa 92.10, best val oa 92.91 +[04/24 20:01:30] ModelNet40Ply2048 INFO: Epoch 405 LR 0.000242 train_oa 99.38, val_oa 91.49, best val oa 92.91 +[04/24 20:02:19] ModelNet40Ply2048 INFO: Epoch 406 LR 0.000240 train_oa 99.26, val_oa 91.98, best val oa 92.91 +[04/24 20:03:06] ModelNet40Ply2048 INFO: Epoch 407 LR 0.000237 train_oa 99.42, val_oa 92.02, best val oa 92.91 +[04/24 20:03:53] ModelNet40Ply2048 INFO: Epoch 408 LR 0.000235 train_oa 99.41, val_oa 91.69, best val oa 92.91 +[04/24 20:04:43] ModelNet40Ply2048 INFO: Epoch 409 LR 0.000233 train_oa 99.32, val_oa 91.77, best val oa 92.91 +[04/24 20:05:31] ModelNet40Ply2048 INFO: Epoch 410 LR 0.000231 train_oa 99.38, val_oa 92.02, best val oa 92.91 +[04/24 20:06:20] ModelNet40Ply2048 INFO: Epoch 411 LR 0.000228 train_oa 99.33, val_oa 92.18, best val oa 92.91 +[04/24 20:07:09] ModelNet40Ply2048 INFO: Epoch 412 LR 0.000226 train_oa 99.48, val_oa 92.06, best val oa 92.91 +[04/24 20:08:04] ModelNet40Ply2048 INFO: Epoch 413 LR 0.000224 train_oa 99.30, val_oa 92.34, best val oa 92.91 +[04/24 20:08:54] ModelNet40Ply2048 INFO: Epoch 414 LR 0.000222 train_oa 99.32, val_oa 91.90, best val oa 92.91 +[04/24 20:09:42] ModelNet40Ply2048 INFO: Epoch 415 LR 0.000220 train_oa 99.43, val_oa 91.69, best val oa 92.91 +[04/24 20:10:29] ModelNet40Ply2048 INFO: Epoch 416 LR 0.000218 train_oa 99.37, val_oa 91.57, best val oa 92.91 +[04/24 20:11:22] ModelNet40Ply2048 INFO: Epoch 417 LR 0.000215 train_oa 99.44, val_oa 91.90, best val oa 92.91 +[04/24 20:12:10] ModelNet40Ply2048 INFO: Epoch 418 LR 0.000213 train_oa 99.49, val_oa 91.94, best val oa 92.91 +[04/24 20:12:58] ModelNet40Ply2048 INFO: Epoch 419 LR 0.000211 train_oa 99.49, val_oa 91.53, best val oa 92.91 +[04/24 20:13:44] ModelNet40Ply2048 INFO: Epoch 420 LR 0.000209 train_oa 99.46, val_oa 91.73, best val oa 92.91 +[04/24 20:14:34] ModelNet40Ply2048 INFO: Epoch 421 LR 0.000207 train_oa 99.32, val_oa 92.34, best val oa 92.91 +[04/24 20:15:20] ModelNet40Ply2048 INFO: Epoch 422 LR 0.000205 train_oa 99.39, val_oa 92.06, best val oa 92.91 +[04/24 20:16:09] ModelNet40Ply2048 INFO: Epoch 423 LR 0.000203 train_oa 99.42, val_oa 91.61, best val oa 92.91 +[04/24 20:16:57] ModelNet40Ply2048 INFO: Epoch 424 LR 0.000201 train_oa 99.52, val_oa 92.14, best val oa 92.91 +[04/24 20:17:45] ModelNet40Ply2048 INFO: Epoch 425 LR 0.000199 train_oa 99.52, val_oa 92.42, best val oa 92.91 +[04/24 20:18:38] ModelNet40Ply2048 INFO: Epoch 426 LR 0.000196 train_oa 99.45, val_oa 91.29, best val oa 92.91 +[04/24 20:19:25] ModelNet40Ply2048 INFO: Epoch 427 LR 0.000194 train_oa 99.45, val_oa 92.06, best val oa 92.91 +[04/24 20:20:12] ModelNet40Ply2048 INFO: Epoch 428 LR 0.000192 train_oa 99.43, val_oa 92.22, best val oa 92.91 +[04/24 20:21:02] ModelNet40Ply2048 INFO: Epoch 429 LR 0.000190 train_oa 99.36, val_oa 91.09, best val oa 92.91 +[04/24 20:21:49] ModelNet40Ply2048 INFO: Epoch 430 LR 0.000188 train_oa 99.56, val_oa 91.77, best val oa 92.91 +[04/24 20:22:37] ModelNet40Ply2048 INFO: Epoch 431 LR 0.000186 train_oa 99.47, val_oa 92.18, best val oa 92.91 +[04/24 20:23:23] ModelNet40Ply2048 INFO: Epoch 432 LR 0.000184 train_oa 99.50, val_oa 92.06, best val oa 92.91 +[04/24 20:24:07] ModelNet40Ply2048 INFO: Epoch 433 LR 0.000182 train_oa 99.49, val_oa 92.30, best val oa 92.91 +[04/24 20:24:55] ModelNet40Ply2048 INFO: Epoch 434 LR 0.000180 train_oa 99.51, val_oa 92.02, best val oa 92.91 +[04/24 20:25:44] ModelNet40Ply2048 INFO: Epoch 435 LR 0.000178 train_oa 99.42, val_oa 91.90, best val oa 92.91 +[04/24 20:26:32] ModelNet40Ply2048 INFO: Epoch 436 LR 0.000176 train_oa 99.52, val_oa 92.38, best val oa 92.91 +[04/24 20:27:22] ModelNet40Ply2048 INFO: Epoch 437 LR 0.000174 train_oa 99.51, val_oa 92.06, best val oa 92.91 +[04/24 20:28:13] ModelNet40Ply2048 INFO: Epoch 438 LR 0.000172 train_oa 99.52, val_oa 92.06, best val oa 92.91 +[04/24 20:29:05] ModelNet40Ply2048 INFO: Epoch 439 LR 0.000170 train_oa 99.51, val_oa 92.10, best val oa 92.91 +[04/24 20:29:54] ModelNet40Ply2048 INFO: Epoch 440 LR 0.000168 train_oa 99.51, val_oa 91.82, best val oa 92.91 +[04/24 20:30:46] ModelNet40Ply2048 INFO: Epoch 441 LR 0.000166 train_oa 99.62, val_oa 91.41, best val oa 92.91 +[04/24 20:31:36] ModelNet40Ply2048 INFO: Epoch 442 LR 0.000164 train_oa 99.54, val_oa 92.26, best val oa 92.91 +[04/24 20:32:27] ModelNet40Ply2048 INFO: Epoch 443 LR 0.000162 train_oa 99.56, val_oa 91.65, best val oa 92.91 +[04/24 20:33:12] ModelNet40Ply2048 INFO: Epoch 444 LR 0.000160 train_oa 99.56, val_oa 91.90, best val oa 92.91 +[04/24 20:34:00] ModelNet40Ply2048 INFO: Epoch 445 LR 0.000159 train_oa 99.54, val_oa 91.65, best val oa 92.91 +[04/24 20:34:54] ModelNet40Ply2048 INFO: Epoch 446 LR 0.000157 train_oa 99.59, val_oa 91.57, best val oa 92.91 +[04/24 20:35:39] ModelNet40Ply2048 INFO: Epoch 447 LR 0.000155 train_oa 99.42, val_oa 91.41, best val oa 92.91 +[04/24 20:36:25] ModelNet40Ply2048 INFO: Epoch 448 LR 0.000153 train_oa 99.59, val_oa 91.73, best val oa 92.91 +[04/24 20:37:13] ModelNet40Ply2048 INFO: Epoch 449 LR 0.000151 train_oa 99.64, val_oa 92.22, best val oa 92.91 +[04/24 20:38:04] ModelNet40Ply2048 INFO: Epoch 450 LR 0.000149 train_oa 99.50, val_oa 92.02, best val oa 92.91 +[04/24 20:38:49] ModelNet40Ply2048 INFO: Epoch 451 LR 0.000147 train_oa 99.64, val_oa 91.45, best val oa 92.91 +[04/24 20:39:34] ModelNet40Ply2048 INFO: Epoch 452 LR 0.000145 train_oa 99.63, val_oa 91.61, best val oa 92.91 +[04/24 20:40:18] ModelNet40Ply2048 INFO: Epoch 453 LR 0.000144 train_oa 99.63, val_oa 91.37, best val oa 92.91 +[04/24 20:41:05] ModelNet40Ply2048 INFO: Epoch 454 LR 0.000142 train_oa 99.48, val_oa 92.14, best val oa 92.91 +[04/24 20:41:57] ModelNet40Ply2048 INFO: Epoch 455 LR 0.000140 train_oa 99.67, val_oa 91.65, best val oa 92.91 +[04/24 20:42:43] ModelNet40Ply2048 INFO: Epoch 456 LR 0.000138 train_oa 99.57, val_oa 91.77, best val oa 92.91 +[04/24 20:43:30] ModelNet40Ply2048 INFO: Epoch 457 LR 0.000136 train_oa 99.64, val_oa 92.42, best val oa 92.91 +[04/24 20:44:23] ModelNet40Ply2048 INFO: Epoch 458 LR 0.000135 train_oa 99.70, val_oa 91.77, best val oa 92.91 +[04/24 20:45:09] ModelNet40Ply2048 INFO: Epoch 459 LR 0.000133 train_oa 99.77, val_oa 91.82, best val oa 92.91 +[04/24 20:46:01] ModelNet40Ply2048 INFO: Epoch 460 LR 0.000131 train_oa 99.69, val_oa 91.86, best val oa 92.91 +[04/24 20:46:49] ModelNet40Ply2048 INFO: Epoch 461 LR 0.000129 train_oa 99.64, val_oa 91.82, best val oa 92.91 +[04/24 20:47:38] ModelNet40Ply2048 INFO: Epoch 462 LR 0.000128 train_oa 99.61, val_oa 91.86, best val oa 92.91 +[04/24 20:48:30] ModelNet40Ply2048 INFO: Epoch 463 LR 0.000126 train_oa 99.54, val_oa 91.69, best val oa 92.91 +[04/24 20:49:19] ModelNet40Ply2048 INFO: Epoch 464 LR 0.000124 train_oa 99.67, val_oa 91.61, best val oa 92.91 +[04/24 20:50:04] ModelNet40Ply2048 INFO: Epoch 465 LR 0.000122 train_oa 99.57, val_oa 92.14, best val oa 92.91 +[04/24 20:50:50] ModelNet40Ply2048 INFO: Epoch 466 LR 0.000121 train_oa 99.75, val_oa 91.49, best val oa 92.91 +[04/24 20:51:36] ModelNet40Ply2048 INFO: Epoch 467 LR 0.000119 train_oa 99.70, val_oa 92.26, best val oa 92.91 +[04/24 20:52:22] ModelNet40Ply2048 INFO: Epoch 468 LR 0.000117 train_oa 99.69, val_oa 91.69, best val oa 92.91 +[04/24 20:53:10] ModelNet40Ply2048 INFO: Epoch 469 LR 0.000116 train_oa 99.68, val_oa 91.77, best val oa 92.91 +[04/24 20:53:56] ModelNet40Ply2048 INFO: Epoch 470 LR 0.000114 train_oa 99.70, val_oa 92.30, best val oa 92.91 +[04/24 20:54:48] ModelNet40Ply2048 INFO: Epoch 471 LR 0.000112 train_oa 99.69, val_oa 92.38, best val oa 92.91 +[04/24 20:55:43] ModelNet40Ply2048 INFO: Epoch 472 LR 0.000111 train_oa 99.66, val_oa 91.61, best val oa 92.91 +[04/24 20:56:35] ModelNet40Ply2048 INFO: Epoch 473 LR 0.000109 train_oa 99.73, val_oa 92.14, best val oa 92.91 +[04/24 20:57:23] ModelNet40Ply2048 INFO: Epoch 474 LR 0.000107 train_oa 99.61, val_oa 92.18, best val oa 92.91 +[04/24 20:58:15] ModelNet40Ply2048 INFO: Epoch 475 LR 0.000106 train_oa 99.74, val_oa 92.50, best val oa 92.91 +[04/24 20:59:04] ModelNet40Ply2048 INFO: Epoch 476 LR 0.000104 train_oa 99.67, val_oa 92.02, best val oa 92.91 +[04/24 20:59:54] ModelNet40Ply2048 INFO: Epoch 477 LR 0.000103 train_oa 99.76, val_oa 92.38, best val oa 92.91 +[04/24 21:00:40] ModelNet40Ply2048 INFO: Epoch 478 LR 0.000101 train_oa 99.71, val_oa 91.94, best val oa 92.91 +[04/24 21:01:26] ModelNet40Ply2048 INFO: Epoch 479 LR 0.000099 train_oa 99.75, val_oa 91.57, best val oa 92.91 +[04/24 21:02:17] ModelNet40Ply2048 INFO: Epoch 480 LR 0.000098 train_oa 99.62, val_oa 91.77, best val oa 92.91 +[04/24 21:03:07] ModelNet40Ply2048 INFO: Epoch 481 LR 0.000096 train_oa 99.66, val_oa 91.61, best val oa 92.91 +[04/24 21:03:50] ModelNet40Ply2048 INFO: Epoch 482 LR 0.000095 train_oa 99.73, val_oa 92.42, best val oa 92.91 +[04/24 21:04:38] ModelNet40Ply2048 INFO: Epoch 483 LR 0.000093 train_oa 99.80, val_oa 92.38, best val oa 92.91 +[04/24 21:05:28] ModelNet40Ply2048 INFO: Epoch 484 LR 0.000092 train_oa 99.77, val_oa 92.02, best val oa 92.91 +[04/24 21:06:17] ModelNet40Ply2048 INFO: Epoch 485 LR 0.000090 train_oa 99.69, val_oa 92.26, best val oa 92.91 +[04/24 21:07:06] ModelNet40Ply2048 INFO: Epoch 486 LR 0.000089 train_oa 99.74, val_oa 92.46, best val oa 92.91 +[04/24 21:07:53] ModelNet40Ply2048 INFO: Epoch 487 LR 0.000087 train_oa 99.71, val_oa 92.34, best val oa 92.91 +[04/24 21:08:45] ModelNet40Ply2048 INFO: Epoch 488 LR 0.000086 train_oa 99.81, val_oa 92.59, best val oa 92.91 +[04/24 21:09:37] ModelNet40Ply2048 INFO: Epoch 489 LR 0.000084 train_oa 99.74, val_oa 92.22, best val oa 92.91 +[04/24 21:10:31] ModelNet40Ply2048 INFO: Epoch 490 LR 0.000083 train_oa 99.80, val_oa 92.38, best val oa 92.91 +[04/24 21:11:17] ModelNet40Ply2048 INFO: Epoch 491 LR 0.000082 train_oa 99.76, val_oa 92.14, best val oa 92.91 +[04/24 21:12:02] ModelNet40Ply2048 INFO: Epoch 492 LR 0.000080 train_oa 99.80, val_oa 92.46, best val oa 92.91 +[04/24 21:12:46] ModelNet40Ply2048 INFO: Epoch 493 LR 0.000079 train_oa 99.78, val_oa 91.69, best val oa 92.91 +[04/24 21:13:37] ModelNet40Ply2048 INFO: Epoch 494 LR 0.000077 train_oa 99.88, val_oa 92.18, best val oa 92.91 +[04/24 21:14:24] ModelNet40Ply2048 INFO: Epoch 495 LR 0.000076 train_oa 99.78, val_oa 92.34, best val oa 92.91 +[04/24 21:15:13] ModelNet40Ply2048 INFO: Epoch 496 LR 0.000075 train_oa 99.74, val_oa 92.50, best val oa 92.91 +[04/24 21:15:58] ModelNet40Ply2048 INFO: Epoch 497 LR 0.000073 train_oa 99.80, val_oa 92.18, best val oa 92.91 +[04/24 21:16:45] ModelNet40Ply2048 INFO: Epoch 498 LR 0.000072 train_oa 99.80, val_oa 92.63, best val oa 92.91 +[04/24 21:17:32] ModelNet40Ply2048 INFO: Epoch 499 LR 0.000071 train_oa 99.82, val_oa 92.14, best val oa 92.91 +[04/24 21:18:26] ModelNet40Ply2048 INFO: Epoch 500 LR 0.000069 train_oa 99.79, val_oa 92.06, best val oa 92.91 +[04/24 21:19:13] ModelNet40Ply2048 INFO: Epoch 501 LR 0.000068 train_oa 99.81, val_oa 92.14, best val oa 92.91 +[04/24 21:20:00] ModelNet40Ply2048 INFO: Epoch 502 LR 0.000067 train_oa 99.79, val_oa 92.42, best val oa 92.91 +[04/24 21:20:44] ModelNet40Ply2048 INFO: Epoch 503 LR 0.000065 train_oa 99.84, val_oa 92.22, best val oa 92.91 +[04/24 21:21:31] ModelNet40Ply2048 INFO: Epoch 504 LR 0.000064 train_oa 99.84, val_oa 92.50, best val oa 92.91 +[04/24 21:22:22] ModelNet40Ply2048 INFO: Epoch 505 LR 0.000063 train_oa 99.87, val_oa 92.30, best val oa 92.91 +[04/24 21:23:12] ModelNet40Ply2048 INFO: Epoch 506 LR 0.000062 train_oa 99.86, val_oa 92.18, best val oa 92.91 +[04/24 21:24:01] ModelNet40Ply2048 INFO: Epoch 507 LR 0.000060 train_oa 99.78, val_oa 92.14, best val oa 92.91 +[04/24 21:24:53] ModelNet40Ply2048 INFO: Epoch 508 LR 0.000059 train_oa 99.88, val_oa 91.90, best val oa 92.91 +[04/24 21:25:39] ModelNet40Ply2048 INFO: Epoch 509 LR 0.000058 train_oa 99.82, val_oa 92.22, best val oa 92.91 +[04/24 21:26:25] ModelNet40Ply2048 INFO: Epoch 510 LR 0.000057 train_oa 99.83, val_oa 91.82, best val oa 92.91 +[04/24 21:27:11] ModelNet40Ply2048 INFO: Epoch 511 LR 0.000055 train_oa 99.85, val_oa 92.14, best val oa 92.91 +[04/24 21:27:58] ModelNet40Ply2048 INFO: Epoch 512 LR 0.000054 train_oa 99.80, val_oa 92.34, best val oa 92.91 +[04/24 21:28:44] ModelNet40Ply2048 INFO: Epoch 513 LR 0.000053 train_oa 99.86, val_oa 92.30, best val oa 92.91 +[04/24 21:29:31] ModelNet40Ply2048 INFO: Epoch 514 LR 0.000052 train_oa 99.83, val_oa 92.10, best val oa 92.91 +[04/24 21:30:26] ModelNet40Ply2048 INFO: Epoch 515 LR 0.000051 train_oa 99.83, val_oa 92.26, best val oa 92.91 +[04/24 21:31:21] ModelNet40Ply2048 INFO: Epoch 516 LR 0.000050 train_oa 99.83, val_oa 91.86, best val oa 92.91 +[04/24 21:32:15] ModelNet40Ply2048 INFO: Epoch 517 LR 0.000049 train_oa 99.89, val_oa 91.98, best val oa 92.91 +[04/24 21:33:05] ModelNet40Ply2048 INFO: Epoch 518 LR 0.000047 train_oa 99.85, val_oa 92.14, best val oa 92.91 +[04/24 21:33:56] ModelNet40Ply2048 INFO: Epoch 519 LR 0.000046 train_oa 99.87, val_oa 92.02, best val oa 92.91 +[04/24 21:34:48] ModelNet40Ply2048 INFO: Epoch 520 LR 0.000045 train_oa 99.78, val_oa 92.26, best val oa 92.91 +[04/24 21:35:42] ModelNet40Ply2048 INFO: Epoch 521 LR 0.000044 train_oa 99.87, val_oa 92.50, best val oa 92.91 +[04/24 21:36:29] ModelNet40Ply2048 INFO: Epoch 522 LR 0.000043 train_oa 99.88, val_oa 92.30, best val oa 92.91 +[04/24 21:37:23] ModelNet40Ply2048 INFO: Epoch 523 LR 0.000042 train_oa 99.87, val_oa 92.38, best val oa 92.91 +[04/24 21:38:11] ModelNet40Ply2048 INFO: Epoch 524 LR 0.000041 train_oa 99.88, val_oa 92.38, best val oa 92.91 +[04/24 21:39:03] ModelNet40Ply2048 INFO: Epoch 525 LR 0.000040 train_oa 99.84, val_oa 92.38, best val oa 92.91 +[04/24 21:39:49] ModelNet40Ply2048 INFO: Epoch 526 LR 0.000039 train_oa 99.80, val_oa 92.54, best val oa 92.91 +[04/24 21:40:35] ModelNet40Ply2048 INFO: Epoch 527 LR 0.000038 train_oa 99.89, val_oa 92.38, best val oa 92.91 +[04/24 21:41:26] ModelNet40Ply2048 INFO: Epoch 528 LR 0.000037 train_oa 99.89, val_oa 92.14, best val oa 92.91 +[04/24 21:42:10] ModelNet40Ply2048 INFO: Epoch 529 LR 0.000036 train_oa 99.87, val_oa 92.06, best val oa 92.91 +[04/24 21:42:58] ModelNet40Ply2048 INFO: Epoch 530 LR 0.000035 train_oa 99.86, val_oa 91.98, best val oa 92.91 +[04/24 21:43:45] ModelNet40Ply2048 INFO: Epoch 531 LR 0.000034 train_oa 99.87, val_oa 92.26, best val oa 92.91 +[04/24 21:44:32] ModelNet40Ply2048 INFO: Epoch 532 LR 0.000033 train_oa 99.83, val_oa 92.63, best val oa 92.91 +[04/24 21:45:18] ModelNet40Ply2048 INFO: Epoch 533 LR 0.000032 train_oa 99.90, val_oa 92.38, best val oa 92.91 +[04/24 21:46:04] ModelNet40Ply2048 INFO: Epoch 534 LR 0.000031 train_oa 99.90, val_oa 92.26, best val oa 92.91 +[04/24 21:46:56] ModelNet40Ply2048 INFO: Epoch 535 LR 0.000031 train_oa 99.84, val_oa 92.50, best val oa 92.91 +[04/24 21:47:50] ModelNet40Ply2048 INFO: Epoch 536 LR 0.000030 train_oa 99.83, val_oa 92.54, best val oa 92.91 +[04/24 21:48:43] ModelNet40Ply2048 INFO: Epoch 537 LR 0.000029 train_oa 99.92, val_oa 92.30, best val oa 92.91 +[04/24 21:49:33] ModelNet40Ply2048 INFO: Epoch 538 LR 0.000028 train_oa 99.90, val_oa 92.46, best val oa 92.91 +[04/24 21:50:19] ModelNet40Ply2048 INFO: Epoch 539 LR 0.000027 train_oa 99.89, val_oa 92.10, best val oa 92.91 +[04/24 21:51:05] ModelNet40Ply2048 INFO: Epoch 540 LR 0.000026 train_oa 99.85, val_oa 92.02, best val oa 92.91 +[04/24 21:51:52] ModelNet40Ply2048 INFO: Epoch 541 LR 0.000025 train_oa 99.87, val_oa 92.42, best val oa 92.91 +[04/24 21:52:43] ModelNet40Ply2048 INFO: Epoch 542 LR 0.000025 train_oa 99.86, val_oa 92.14, best val oa 92.91 +[04/24 21:53:28] ModelNet40Ply2048 INFO: Epoch 543 LR 0.000024 train_oa 99.94, val_oa 92.38, best val oa 92.91 +[04/24 21:54:12] ModelNet40Ply2048 INFO: Epoch 544 LR 0.000023 train_oa 99.87, val_oa 92.46, best val oa 92.91 +[04/24 21:55:00] ModelNet40Ply2048 INFO: Epoch 545 LR 0.000022 train_oa 99.84, val_oa 92.38, best val oa 92.91 +[04/24 21:55:46] ModelNet40Ply2048 INFO: Epoch 546 LR 0.000022 train_oa 99.91, val_oa 92.71, best val oa 92.91 +[04/24 21:56:33] ModelNet40Ply2048 INFO: Epoch 547 LR 0.000021 train_oa 99.87, val_oa 92.26, best val oa 92.91 +[04/24 21:57:20] ModelNet40Ply2048 INFO: Epoch 548 LR 0.000020 train_oa 99.91, val_oa 92.46, best val oa 92.91 +[04/24 21:58:11] ModelNet40Ply2048 INFO: Epoch 549 LR 0.000019 train_oa 99.91, val_oa 92.22, best val oa 92.91 +[04/24 21:58:59] ModelNet40Ply2048 INFO: Epoch 550 LR 0.000019 train_oa 99.90, val_oa 92.14, best val oa 92.91 +[04/24 21:59:45] ModelNet40Ply2048 INFO: Epoch 551 LR 0.000018 train_oa 99.89, val_oa 92.54, best val oa 92.91 +[04/24 22:00:32] ModelNet40Ply2048 INFO: Epoch 552 LR 0.000017 train_oa 99.88, val_oa 92.42, best val oa 92.91 +[04/24 22:01:17] ModelNet40Ply2048 INFO: Epoch 553 LR 0.000017 train_oa 99.87, val_oa 92.30, best val oa 92.91 +[04/24 22:02:10] ModelNet40Ply2048 INFO: Epoch 554 LR 0.000016 train_oa 99.91, val_oa 92.18, best val oa 92.91 +[04/24 22:03:02] ModelNet40Ply2048 INFO: Epoch 555 LR 0.000015 train_oa 99.86, val_oa 92.30, best val oa 92.91 +[04/24 22:03:49] ModelNet40Ply2048 INFO: Epoch 556 LR 0.000015 train_oa 99.92, val_oa 92.30, best val oa 92.91 +[04/24 22:04:36] ModelNet40Ply2048 INFO: Epoch 557 LR 0.000014 train_oa 99.86, val_oa 92.42, best val oa 92.91 +[04/24 22:05:22] ModelNet40Ply2048 INFO: Epoch 558 LR 0.000014 train_oa 99.92, val_oa 92.50, best val oa 92.91 +[04/24 22:06:08] ModelNet40Ply2048 INFO: Epoch 559 LR 0.000013 train_oa 99.87, val_oa 92.63, best val oa 92.91 +[04/24 22:06:55] ModelNet40Ply2048 INFO: Epoch 560 LR 0.000012 train_oa 99.92, val_oa 92.42, best val oa 92.91 +[04/24 22:07:48] ModelNet40Ply2048 INFO: Epoch 561 LR 0.000012 train_oa 99.93, val_oa 92.30, best val oa 92.91 +[04/24 22:08:41] ModelNet40Ply2048 INFO: Epoch 562 LR 0.000011 train_oa 99.92, val_oa 92.30, best val oa 92.91 +[04/24 22:09:30] ModelNet40Ply2048 INFO: Epoch 563 LR 0.000011 train_oa 99.93, val_oa 92.46, best val oa 92.91 +[04/24 22:10:22] ModelNet40Ply2048 INFO: Epoch 564 LR 0.000010 train_oa 99.92, val_oa 92.34, best val oa 92.91 +[04/24 22:11:13] ModelNet40Ply2048 INFO: Epoch 565 LR 0.000010 train_oa 99.91, val_oa 92.30, best val oa 92.91 +[04/24 22:11:56] ModelNet40Ply2048 INFO: Epoch 566 LR 0.000009 train_oa 99.92, val_oa 92.34, best val oa 92.91 +[04/24 22:12:42] ModelNet40Ply2048 INFO: Epoch 567 LR 0.000009 train_oa 99.92, val_oa 92.10, best val oa 92.91 +[04/24 22:13:28] ModelNet40Ply2048 INFO: Epoch 568 LR 0.000008 train_oa 99.92, val_oa 92.26, best val oa 92.91 +[04/24 22:14:17] ModelNet40Ply2048 INFO: Epoch 569 LR 0.000008 train_oa 99.92, val_oa 92.30, best val oa 92.91 +[04/24 22:15:06] ModelNet40Ply2048 INFO: Epoch 570 LR 0.000008 train_oa 99.92, val_oa 92.38, best val oa 92.91 +[04/24 22:15:55] ModelNet40Ply2048 INFO: Epoch 571 LR 0.000007 train_oa 99.85, val_oa 92.54, best val oa 92.91 +[04/24 22:16:42] ModelNet40Ply2048 INFO: Epoch 572 LR 0.000007 train_oa 99.91, val_oa 92.38, best val oa 92.91 +[04/24 22:17:28] ModelNet40Ply2048 INFO: Epoch 573 LR 0.000006 train_oa 99.92, val_oa 92.22, best val oa 92.91 +[04/24 22:18:10] ModelNet40Ply2048 INFO: Epoch 574 LR 0.000006 train_oa 99.93, val_oa 92.46, best val oa 92.91 +[04/24 22:18:57] ModelNet40Ply2048 INFO: Epoch 575 LR 0.000006 train_oa 99.93, val_oa 92.30, best val oa 92.91 +[04/24 22:19:41] ModelNet40Ply2048 INFO: Epoch 576 LR 0.000005 train_oa 99.94, val_oa 92.46, best val oa 92.91 +[04/24 22:20:33] ModelNet40Ply2048 INFO: Epoch 577 LR 0.000005 train_oa 99.94, val_oa 92.30, best val oa 92.91 +[04/24 22:21:17] ModelNet40Ply2048 INFO: Epoch 578 LR 0.000005 train_oa 99.96, val_oa 92.46, best val oa 92.91 +[04/24 22:22:01] ModelNet40Ply2048 INFO: Epoch 579 LR 0.000004 train_oa 99.95, val_oa 92.26, best val oa 92.91 +[04/24 22:22:45] ModelNet40Ply2048 INFO: Epoch 580 LR 0.000004 train_oa 99.86, val_oa 92.50, best val oa 92.91 +[04/24 22:23:33] ModelNet40Ply2048 INFO: Epoch 581 LR 0.000004 train_oa 99.95, val_oa 92.38, best val oa 92.91 +[04/24 22:24:22] ModelNet40Ply2048 INFO: Epoch 582 LR 0.000003 train_oa 99.95, val_oa 92.46, best val oa 92.91 +[04/24 22:25:12] ModelNet40Ply2048 INFO: Epoch 583 LR 0.000003 train_oa 99.88, val_oa 92.46, best val oa 92.91 +[04/24 22:26:01] ModelNet40Ply2048 INFO: Epoch 584 LR 0.000003 train_oa 99.95, val_oa 92.38, best val oa 92.91 +[04/24 22:26:55] ModelNet40Ply2048 INFO: Epoch 585 LR 0.000003 train_oa 99.88, val_oa 92.34, best val oa 92.91 +[04/24 22:27:43] ModelNet40Ply2048 INFO: Epoch 586 LR 0.000003 train_oa 99.91, val_oa 92.34, best val oa 92.91 +[04/24 22:28:33] ModelNet40Ply2048 INFO: Epoch 587 LR 0.000002 train_oa 99.90, val_oa 92.50, best val oa 92.91 +[04/24 22:29:17] ModelNet40Ply2048 INFO: Epoch 588 LR 0.000002 train_oa 99.93, val_oa 92.46, best val oa 92.91 +[04/24 22:30:04] ModelNet40Ply2048 INFO: Epoch 589 LR 0.000002 train_oa 99.99, val_oa 92.10, best val oa 92.91 +[04/24 22:30:51] ModelNet40Ply2048 INFO: Epoch 590 LR 0.000002 train_oa 99.93, val_oa 92.22, best val oa 92.91 +[04/24 22:31:43] ModelNet40Ply2048 INFO: Epoch 591 LR 0.000002 train_oa 99.91, val_oa 92.59, best val oa 92.91 +[04/24 22:32:33] ModelNet40Ply2048 INFO: Epoch 592 LR 0.000002 train_oa 99.93, val_oa 92.26, best val oa 92.91 +[04/24 22:33:19] ModelNet40Ply2048 INFO: Epoch 593 LR 0.000001 train_oa 99.89, val_oa 92.67, best val oa 92.91 +[04/24 22:34:06] ModelNet40Ply2048 INFO: Epoch 594 LR 0.000001 train_oa 99.93, val_oa 92.71, best val oa 92.91 +[04/24 22:34:52] ModelNet40Ply2048 INFO: Epoch 595 LR 0.000001 train_oa 99.96, val_oa 92.26, best val oa 92.91 +[04/24 22:35:39] ModelNet40Ply2048 INFO: Epoch 596 LR 0.000001 train_oa 99.92, val_oa 92.22, best val oa 92.91 +[04/24 22:36:27] ModelNet40Ply2048 INFO: Epoch 597 LR 0.000001 train_oa 99.92, val_oa 92.71, best val oa 92.91 +[04/24 22:37:14] ModelNet40Ply2048 INFO: Epoch 598 LR 0.000001 train_oa 99.88, val_oa 92.30, best val oa 92.91 +[04/24 22:38:04] ModelNet40Ply2048 INFO: Epoch 599 LR 0.000001 train_oa 99.95, val_oa 92.50, best val oa 92.91 +[04/24 22:38:51] ModelNet40Ply2048 INFO: Epoch 600 LR 0.000001 train_oa 99.89, val_oa 92.34, best val oa 92.91 +[04/24 22:38:54] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 96.00% +bed : 100.00% +bench : 75.00% +bookshelf : 96.00% +bottle : 98.00% +bowl : 90.00% +car : 99.00% +chair : 99.00% +cone : 95.00% +cup : 70.00% +curtain : 95.00% +desk : 94.19% +door : 100.00% +dresser : 88.37% +flower_pot: 15.00% +glass_box : 96.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 80.00% +laptop : 100.00% +mantel : 98.00% +monitor : 100.00% +night_stand: 79.07% +person : 100.00% +piano : 96.00% +plant : 79.00% +radio : 80.00% +range_hood: 95.00% +sink : 85.00% +sofa : 99.00% +stairs : 90.00% +stool : 85.00% +table : 80.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 86.00% +vase : 78.00% +wardrobe : 90.00% +xbox : 85.00% +E@162 OA: 92.30 mAcc: 89.67 + +[04/24 22:38:54] ModelNet40Ply2048 INFO: Successful Loading the ckpt from log/modelnet40ply2048/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD/checkpoint/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth +[04/24 22:38:54] ModelNet40Ply2048 INFO: ckpts @ 162 epoch( {'best_val': 92.90924072265625} ) +[04/24 22:38:57] ModelNet40Ply2048 INFO: +Classes Acc +airplane : 100.00% +bathtub : 98.00% +bed : 99.00% +bench : 75.00% +bookshelf : 99.00% +bottle : 98.00% +bowl : 95.00% +car : 99.00% +chair : 99.00% +cone : 95.00% +cup : 80.00% +curtain : 95.00% +desk : 91.86% +door : 100.00% +dresser : 81.40% +flower_pot: 15.00% +glass_box : 95.00% +guitar : 100.00% +keyboard : 100.00% +lamp : 85.00% +laptop : 100.00% +mantel : 96.00% +monitor : 99.00% +night_stand: 84.88% +person : 95.00% +piano : 96.00% +plant : 88.00% +radio : 85.00% +range_hood: 97.00% +sink : 90.00% +sofa : 99.00% +stairs : 95.00% +stool : 85.00% +table : 85.00% +tent : 95.00% +toilet : 100.00% +tv_stand : 86.00% +vase : 80.00% +wardrobe : 80.00% +xbox : 95.00% +E@162 OA: 93.11 mAcc: 90.78 + diff --git a/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth b/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth new file mode 100644 index 0000000000000000000000000000000000000000..714219b3b4613f5b0d43a45f17d06cc9ec85be8b --- /dev/null +++ b/checkpoint/modelnet/modelnet40ply2048-train-ppv2-s-ngpus1-seed1234-20250424-143358-QbzzqAa6vYDExr9P9eCVRD_ckpt_best.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:464d37b9b83d45cac060547e4b1be915772b5909e84ff8133416cae4fa710be8 +size 31348318 diff --git a/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk.log b/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk.log new file mode 100644 index 0000000000000000000000000000000000000000..2f0a9637ab60dfb4851142205f3be3bf82f6667f --- /dev/null +++ b/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk.log @@ -0,0 +1,2168 @@ +[03/26 12:21:07] S3DIS INFO: dist_url: tcp://localhost:8888 +dist_backend: nccl +multiprocessing_distributed: False +ngpus_per_node: 1 +world_size: 1 +launcher: mp +local_rank: 0 +use_gpu: True +seed: 1234 +epoch: 0 +epochs: 100 +ignore_index: None +val_fn: validate +deterministic: False +sync_bn: False +criterion_args: + NAME: CrossEntropy + label_smoothing: 0.2 +use_mask: False +grad_norm_clip: 10 +layer_decay: 0 +step_per_update: 1 +start_epoch: 1 +sched_on_epoch: True +wandb: + use_wandb: False + project: PointNeXt-S3DIS + tags: ['s3dis', 'train', 'ppv2-xl', 'ngpus1'] + name: s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk +use_amp: False +use_voting: False +val_freq: 1 +resume: False +test: False +finetune: False +mode: train +logname: None +load_path: None +print_freq: 50 +save_freq: -1 +root_dir: log/s3dis +pretrained_path: None +datatransforms: + train: ['ChromaticAutoContrast', 'PointsToTensor', 'PointCloudScaling', 'PointCloudXYZAlign', 'PointCloudRotation', 'PointCloudJitter', 'ChromaticDropGPU', 'ChromaticNormalize'] + val: ['PointsToTensor', 'PointCloudXYZAlign', 'ChromaticNormalize'] + vote: ['ChromaticDropGPU'] + kwargs: + color_drop: 0.2 + gravity_dim: 2 + scale: [0.9, 1.1] + angle: [0, 0, 1] + jitter_sigma: 0.005 + jitter_clip: 0.02 +feature_keys: x,heights +dataset: + common: + NAME: S3DIS + data_root: data/S3DIS/s3disfull + test_area: 5 + voxel_size: 0.04 + train: + split: train + voxel_max: 24000 + loop: 30 + presample: False + val: + split: val + voxel_max: None + presample: True + test: + split: test + voxel_max: None + presample: False +num_classes: 13 +batch_size: 6 +val_batch_size: 1 +dataloader: + num_workers: 6 +cls_weighed_loss: False +optimizer: + NAME: adamw + weight_decay: 0.0001 +sched: cosine +warmup_epochs: 0 +min_lr: 1e-05 +lr: 0.01 +log_dir: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk +model: + NAME: DpnSeg + encoder_args: + NAME: PPV2Encoder + blocks: [1, 4, 7, 4, 4] + strides: [1, 4, 4, 4, 4] + sa_layers: 1 + sa_use_res: False + width: 64 + in_channels: 4 + expansion: 4 + radius: 0.1 + nsample: 32 + flag: 1 + aggr_args: + feature_type: dp_fj + reduction: max + group_args: + NAME: ballquery + normalize_dp: True + conv_args: + order: conv-norm-act + act_args: + act: relu + norm_args: + norm: bn + decoder_args: + NAME: PPV2Decoder + cls_args: + NAME: SegHead + num_classes: 13 + in_channels: None + norm_args: + norm: bn +rank: 0 +distributed: False +mp: False +task_name: s3dis +cfg_basename: ppv2-xl +opts: +is_training: True +run_name: s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk +run_dir: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk +exp_dir: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk +ckpt_dir: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint +log_path: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk.log +cfg_path: log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/cfg.yaml +[03/26 12:21:07] S3DIS INFO: radius: [[0.1], [0.1, 0.2, 0.2, 0.2], [0.2, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4], [0.4, 0.8, 0.8, 0.8], [0.8, 1.6, 1.6, 1.6]], + nsample: [[32], [32, 32, 32, 32], [32, 32, 32, 32, 32, 32, 32], [32, 32, 32, 32], [32, 32, 32, 32]] +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.1 +nsample: 32 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.1 +nsample: 32 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.35000000000000003 +nsample: 64 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.2 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.2 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.2 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.2 +nsample: 32 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.7000000000000001 +nsample: 64 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.4 +nsample: 32 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.8 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.8 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.8 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 0.8 +nsample: 32 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 1.6 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 1.6 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS WARNING: kwargs: {'dis_02': 3} are not used in LocalAggregation +[03/26 12:21:07] S3DIS INFO: NAME: ballquery +normalize_dp: True +radius: 1.6 +nsample: 8 +return_idx: True +[03/26 12:21:07] S3DIS INFO: DpnSeg( + (encoder): PPV2Encoder( + (grouper0): QueryAndGroup() + (encoder): Sequential( + (0): Sequential( + (0): SetAbstraction( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(12, 64, kernel_size=(1,), stride=(1,)) + ) + ) + ) + ) + (1): Sequential( + (0): SetAbstraction( + (grouper): QueryAndGroup() + (selfattention): SelfAttention( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_v): Sequential( + (0): Conv1d(8, 128, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(128, 128, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 128, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (preconv): Sequential( + (0): Conv1d(64, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (conv_finanal): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + (1): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(384, 128, kernel_size=(1,), stride=(1,), groups=128, bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (2): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(384, 128, kernel_size=(1,), stride=(1,), groups=128, bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (3): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(384, 128, kernel_size=(1,), stride=(1,), groups=128, bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + ) + (2): Sequential( + (0): SetAbstraction( + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(8, 3, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(3, 3, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=24, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_v): Sequential( + (0): Conv1d(24, 256, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(256, 256, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=24, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 256, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (preconv): Sequential( + (0): Conv1d(128, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (conv_finanal): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + (1): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (2): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (3): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (4): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (5): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (6): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), groups=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + ) + (3): Sequential( + (0): SetAbstraction( + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(24, 9, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(9, 9, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=72, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_v): Sequential( + (0): Conv1d(72, 512, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(512, 512, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=72, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(27, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(27, 512, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (preconv): Sequential( + (0): Conv1d(256, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (conv_finanal): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + (1): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(1536, 512, kernel_size=(1,), stride=(1,), groups=512, bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (2): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(1536, 512, kernel_size=(1,), stride=(1,), groups=512, bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (3): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(1536, 512, kernel_size=(1,), stride=(1,), groups=512, bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + ) + (4): Sequential( + (0): SetAbstraction( + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(72, 27, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(27, 27, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=216, out_features=216, bias=True) + (linear_k): Linear(in_features=216, out_features=216, bias=True) + (linear_v): Sequential( + (0): Conv1d(216, 1024, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=216, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(75, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(75, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(75, 1024, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 1024, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (preconv): Sequential( + (0): Conv1d(512, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (conv_finanal): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + (1): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(3072, 1024, kernel_size=(1,), stride=(1,), groups=1024, bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (2): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(3072, 1024, kernel_size=(1,), stride=(1,), groups=1024, bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + (3): InvResMLP( + (convs): LocalAggregation( + (vpsa): VPSA( + (theta_x_alpha): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (theta_x_beta): Sequential( + (0): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (z_x): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (tf_zx): Sequential( + (0): Conv1d(3072, 1024, kernel_size=(1,), stride=(1,), groups=1024, bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (relu): ReLU(inplace=True) + (lrlu): ReLU(inplace=True) + (bn1): Identity() + (result): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pos_x): Sequential( + (0): Conv2d(3, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False) + (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (residual): Sequential( + (0): Conv1d(1024, 1024, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + ) + (grouper): QueryAndGroup() + ) + ) + ) + ) + ) + (decoder): PPV2Decoder( + (decoder): Sequential( + (0): Sequential( + (0): FeaturePropogation( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(192, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + (1): Sequential( + (0): FeaturePropogation( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(384, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + (2): Sequential( + (0): FeaturePropogation( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(768, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + (3): Sequential( + (0): FeaturePropogation( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(1536, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + ) + ) + (head): SegHead( + (head): Sequential( + (0): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Dropout(p=0.5, inplace=False) + (2): Sequential( + (0): Conv1d(64, 13, kernel_size=(1,), stride=(1,)) + ) + ) + ) +) +[03/26 12:21:07] S3DIS INFO: Number of params: 28.7945 M +[03/26 12:21:07] S3DIS INFO: Param groups = { + "decay": { + "weight_decay": 0.0001, + "params": [ + "encoder.encoder.0.0.convs.0.0.weight", + "encoder.encoder.1.0.selfattention.linear_q.weight", + "encoder.encoder.1.0.selfattention.linear_k.weight", + "encoder.encoder.1.0.selfattention.linear_v.0.weight", + "encoder.encoder.1.0.selfattention.linear_v.3.weight", + "encoder.encoder.1.0.pt.linear_q.weight", + "encoder.encoder.1.0.pt.linear_k.weight", + "encoder.encoder.1.0.pt.linear_p.0.weight", + "encoder.encoder.1.0.pt.linear_p.3.weight", + "encoder.encoder.1.0.pt.w.2.weight", + "encoder.encoder.1.0.pt.v.0.weight", + "encoder.encoder.1.0.pt.v.3.weight", + "encoder.encoder.1.0.pt.conv_p.0.weight", + "encoder.encoder.1.0.pt.conv_p.3.weight", + "encoder.encoder.1.0.preconv.0.weight", + "encoder.encoder.1.0.conv_finanal.0.weight", + "encoder.encoder.1.1.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.1.1.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.1.1.convs.vpsa.z_x.weight", + "encoder.encoder.1.1.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.1.1.convs.vpsa.result.0.weight", + "encoder.encoder.1.1.convs.vpsa.pos_x.0.weight", + "encoder.encoder.1.1.convs.vpsa.residual.0.weight", + "encoder.encoder.1.2.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.1.2.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.1.2.convs.vpsa.z_x.weight", + "encoder.encoder.1.2.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.1.2.convs.vpsa.result.0.weight", + "encoder.encoder.1.2.convs.vpsa.pos_x.0.weight", + "encoder.encoder.1.2.convs.vpsa.residual.0.weight", + "encoder.encoder.1.3.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.1.3.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.1.3.convs.vpsa.z_x.weight", + "encoder.encoder.1.3.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.1.3.convs.vpsa.result.0.weight", + "encoder.encoder.1.3.convs.vpsa.pos_x.0.weight", + "encoder.encoder.1.3.convs.vpsa.residual.0.weight", + "encoder.encoder.2.0.scorenet_global.0.weight", + "encoder.encoder.2.0.scorenet_global.3.weight", + "encoder.encoder.2.0.selfattention.linear_q.weight", + "encoder.encoder.2.0.selfattention.linear_k.weight", + "encoder.encoder.2.0.selfattention.linear_v.0.weight", + "encoder.encoder.2.0.selfattention.linear_v.3.weight", + "encoder.encoder.2.0.pt.linear_q.weight", + "encoder.encoder.2.0.pt.linear_k.weight", + "encoder.encoder.2.0.pt.linear_p.0.weight", + "encoder.encoder.2.0.pt.linear_p.3.weight", + "encoder.encoder.2.0.pt.w.2.weight", + "encoder.encoder.2.0.pt.v.0.weight", + "encoder.encoder.2.0.pt.v.3.weight", + "encoder.encoder.2.0.pt.conv_p.0.weight", + "encoder.encoder.2.0.pt.conv_p.3.weight", + "encoder.encoder.2.0.preconv.0.weight", + "encoder.encoder.2.0.conv_finanal.0.weight", + "encoder.encoder.2.1.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.1.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.1.convs.vpsa.z_x.weight", + "encoder.encoder.2.1.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.1.convs.vpsa.result.0.weight", + "encoder.encoder.2.1.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.1.convs.vpsa.residual.0.weight", + "encoder.encoder.2.2.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.2.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.2.convs.vpsa.z_x.weight", + "encoder.encoder.2.2.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.2.convs.vpsa.result.0.weight", + "encoder.encoder.2.2.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.2.convs.vpsa.residual.0.weight", + "encoder.encoder.2.3.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.3.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.3.convs.vpsa.z_x.weight", + "encoder.encoder.2.3.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.3.convs.vpsa.result.0.weight", + "encoder.encoder.2.3.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.3.convs.vpsa.residual.0.weight", + "encoder.encoder.2.4.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.4.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.4.convs.vpsa.z_x.weight", + "encoder.encoder.2.4.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.4.convs.vpsa.result.0.weight", + "encoder.encoder.2.4.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.4.convs.vpsa.residual.0.weight", + "encoder.encoder.2.5.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.5.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.5.convs.vpsa.z_x.weight", + "encoder.encoder.2.5.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.5.convs.vpsa.result.0.weight", + "encoder.encoder.2.5.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.5.convs.vpsa.residual.0.weight", + "encoder.encoder.2.6.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.2.6.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.2.6.convs.vpsa.z_x.weight", + "encoder.encoder.2.6.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.2.6.convs.vpsa.result.0.weight", + "encoder.encoder.2.6.convs.vpsa.pos_x.0.weight", + "encoder.encoder.2.6.convs.vpsa.residual.0.weight", + "encoder.encoder.3.0.scorenet_global.0.weight", + "encoder.encoder.3.0.scorenet_global.3.weight", + "encoder.encoder.3.0.selfattention.linear_q.weight", + "encoder.encoder.3.0.selfattention.linear_k.weight", + "encoder.encoder.3.0.selfattention.linear_v.0.weight", + "encoder.encoder.3.0.selfattention.linear_v.3.weight", + "encoder.encoder.3.0.pt.linear_q.weight", + "encoder.encoder.3.0.pt.linear_k.weight", + "encoder.encoder.3.0.pt.linear_p.0.weight", + "encoder.encoder.3.0.pt.linear_p.3.weight", + "encoder.encoder.3.0.pt.w.2.weight", + "encoder.encoder.3.0.pt.v.0.weight", + "encoder.encoder.3.0.pt.v.3.weight", + "encoder.encoder.3.0.pt.conv_p.0.weight", + "encoder.encoder.3.0.pt.conv_p.3.weight", + "encoder.encoder.3.0.preconv.0.weight", + "encoder.encoder.3.0.conv_finanal.0.weight", + "encoder.encoder.3.1.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.3.1.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.3.1.convs.vpsa.z_x.weight", + "encoder.encoder.3.1.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.3.1.convs.vpsa.result.0.weight", + "encoder.encoder.3.1.convs.vpsa.pos_x.0.weight", + "encoder.encoder.3.1.convs.vpsa.residual.0.weight", + "encoder.encoder.3.2.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.3.2.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.3.2.convs.vpsa.z_x.weight", + "encoder.encoder.3.2.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.3.2.convs.vpsa.result.0.weight", + "encoder.encoder.3.2.convs.vpsa.pos_x.0.weight", + "encoder.encoder.3.2.convs.vpsa.residual.0.weight", + "encoder.encoder.3.3.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.3.3.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.3.3.convs.vpsa.z_x.weight", + "encoder.encoder.3.3.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.3.3.convs.vpsa.result.0.weight", + "encoder.encoder.3.3.convs.vpsa.pos_x.0.weight", + "encoder.encoder.3.3.convs.vpsa.residual.0.weight", + "encoder.encoder.4.0.scorenet_global.0.weight", + "encoder.encoder.4.0.scorenet_global.3.weight", + "encoder.encoder.4.0.selfattention.linear_q.weight", + "encoder.encoder.4.0.selfattention.linear_k.weight", + "encoder.encoder.4.0.selfattention.linear_v.0.weight", + "encoder.encoder.4.0.selfattention.linear_v.3.weight", + "encoder.encoder.4.0.pt.linear_q.weight", + "encoder.encoder.4.0.pt.linear_k.weight", + "encoder.encoder.4.0.pt.linear_p.0.weight", + "encoder.encoder.4.0.pt.linear_p.3.weight", + "encoder.encoder.4.0.pt.w.2.weight", + "encoder.encoder.4.0.pt.v.0.weight", + "encoder.encoder.4.0.pt.v.3.weight", + "encoder.encoder.4.0.pt.conv_p.0.weight", + "encoder.encoder.4.0.pt.conv_p.3.weight", + "encoder.encoder.4.0.preconv.0.weight", + "encoder.encoder.4.0.conv_finanal.0.weight", + "encoder.encoder.4.1.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.4.1.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.4.1.convs.vpsa.z_x.weight", + "encoder.encoder.4.1.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.4.1.convs.vpsa.result.0.weight", + "encoder.encoder.4.1.convs.vpsa.pos_x.0.weight", + "encoder.encoder.4.1.convs.vpsa.residual.0.weight", + "encoder.encoder.4.2.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.4.2.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.4.2.convs.vpsa.z_x.weight", + "encoder.encoder.4.2.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.4.2.convs.vpsa.result.0.weight", + "encoder.encoder.4.2.convs.vpsa.pos_x.0.weight", + "encoder.encoder.4.2.convs.vpsa.residual.0.weight", + "encoder.encoder.4.3.convs.vpsa.theta_x_alpha.0.weight", + "encoder.encoder.4.3.convs.vpsa.theta_x_beta.0.weight", + "encoder.encoder.4.3.convs.vpsa.z_x.weight", + "encoder.encoder.4.3.convs.vpsa.tf_zx.0.weight", + "encoder.encoder.4.3.convs.vpsa.result.0.weight", + "encoder.encoder.4.3.convs.vpsa.pos_x.0.weight", + "encoder.encoder.4.3.convs.vpsa.residual.0.weight", + "decoder.decoder.0.0.convs.0.0.weight", + "decoder.decoder.0.0.convs.1.0.weight", + "decoder.decoder.1.0.convs.0.0.weight", + "decoder.decoder.1.0.convs.1.0.weight", + "decoder.decoder.2.0.convs.0.0.weight", + "decoder.decoder.2.0.convs.1.0.weight", + "decoder.decoder.3.0.convs.0.0.weight", + "decoder.decoder.3.0.convs.1.0.weight", + "head.head.0.0.weight", + "head.head.2.0.weight" + ], + "lr_scale": 1.0 + }, + "no_decay": { + "weight_decay": 0.0, + "params": [ + "encoder.encoder.0.0.convs.0.0.bias", + "encoder.encoder.1.0.beta", + "encoder.encoder.1.0.selfattention.linear_q.bias", + "encoder.encoder.1.0.selfattention.linear_k.bias", + "encoder.encoder.1.0.selfattention.linear_v.0.bias", + "encoder.encoder.1.0.selfattention.linear_v.1.weight", + "encoder.encoder.1.0.selfattention.linear_v.1.bias", + "encoder.encoder.1.0.selfattention.linear_v.3.bias", + "encoder.encoder.1.0.selfattention.linear_v.4.weight", + "encoder.encoder.1.0.selfattention.linear_v.4.bias", + "encoder.encoder.1.0.pt.linear_q.bias", + "encoder.encoder.1.0.pt.linear_k.bias", + "encoder.encoder.1.0.pt.linear_p.0.bias", + "encoder.encoder.1.0.pt.linear_p.1.weight", + "encoder.encoder.1.0.pt.linear_p.1.bias", + "encoder.encoder.1.0.pt.linear_p.3.bias", + "encoder.encoder.1.0.pt.w.0.weight", + "encoder.encoder.1.0.pt.w.0.bias", + "encoder.encoder.1.0.pt.w.2.bias", + "encoder.encoder.1.0.pt.w.3.weight", + "encoder.encoder.1.0.pt.w.3.bias", + "encoder.encoder.1.0.pt.v.0.bias", + "encoder.encoder.1.0.pt.v.1.weight", + "encoder.encoder.1.0.pt.v.1.bias", + "encoder.encoder.1.0.pt.v.3.bias", + "encoder.encoder.1.0.pt.conv_p.0.bias", + "encoder.encoder.1.0.pt.conv_p.1.weight", + "encoder.encoder.1.0.pt.conv_p.1.bias", + "encoder.encoder.1.0.pt.conv_p.3.bias", + "encoder.encoder.1.0.preconv.1.weight", + "encoder.encoder.1.0.preconv.1.bias", + "encoder.encoder.1.0.conv_finanal.1.weight", + "encoder.encoder.1.0.conv_finanal.1.bias", + "encoder.encoder.1.1.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.1.1.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.1.1.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.1.1.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.1.1.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.1.1.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.1.1.convs.vpsa.result.1.weight", + "encoder.encoder.1.1.convs.vpsa.result.1.bias", + "encoder.encoder.1.1.convs.vpsa.pos_x.1.weight", + "encoder.encoder.1.1.convs.vpsa.pos_x.1.bias", + "encoder.encoder.1.1.convs.vpsa.residual.1.weight", + "encoder.encoder.1.1.convs.vpsa.residual.1.bias", + "encoder.encoder.1.2.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.1.2.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.1.2.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.1.2.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.1.2.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.1.2.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.1.2.convs.vpsa.result.1.weight", + "encoder.encoder.1.2.convs.vpsa.result.1.bias", + "encoder.encoder.1.2.convs.vpsa.pos_x.1.weight", + "encoder.encoder.1.2.convs.vpsa.pos_x.1.bias", + "encoder.encoder.1.2.convs.vpsa.residual.1.weight", + "encoder.encoder.1.2.convs.vpsa.residual.1.bias", + "encoder.encoder.1.3.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.1.3.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.1.3.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.1.3.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.1.3.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.1.3.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.1.3.convs.vpsa.result.1.weight", + "encoder.encoder.1.3.convs.vpsa.result.1.bias", + "encoder.encoder.1.3.convs.vpsa.pos_x.1.weight", + "encoder.encoder.1.3.convs.vpsa.pos_x.1.bias", + "encoder.encoder.1.3.convs.vpsa.residual.1.weight", + "encoder.encoder.1.3.convs.vpsa.residual.1.bias", + "encoder.encoder.2.0.beta", + "encoder.encoder.2.0.scorenet_global.0.bias", + "encoder.encoder.2.0.scorenet_global.1.weight", + "encoder.encoder.2.0.scorenet_global.1.bias", + "encoder.encoder.2.0.scorenet_global.3.bias", + "encoder.encoder.2.0.scorenet_global.4.weight", + "encoder.encoder.2.0.scorenet_global.4.bias", + "encoder.encoder.2.0.selfattention.linear_q.bias", + "encoder.encoder.2.0.selfattention.linear_k.bias", + "encoder.encoder.2.0.selfattention.linear_v.0.bias", + "encoder.encoder.2.0.selfattention.linear_v.1.weight", + "encoder.encoder.2.0.selfattention.linear_v.1.bias", + "encoder.encoder.2.0.selfattention.linear_v.3.bias", + "encoder.encoder.2.0.selfattention.linear_v.4.weight", + "encoder.encoder.2.0.selfattention.linear_v.4.bias", + "encoder.encoder.2.0.pt.linear_q.bias", + "encoder.encoder.2.0.pt.linear_k.bias", + "encoder.encoder.2.0.pt.linear_p.0.bias", + "encoder.encoder.2.0.pt.linear_p.1.weight", + "encoder.encoder.2.0.pt.linear_p.1.bias", + "encoder.encoder.2.0.pt.linear_p.3.bias", + "encoder.encoder.2.0.pt.w.0.weight", + "encoder.encoder.2.0.pt.w.0.bias", + "encoder.encoder.2.0.pt.w.2.bias", + "encoder.encoder.2.0.pt.w.3.weight", + "encoder.encoder.2.0.pt.w.3.bias", + "encoder.encoder.2.0.pt.v.0.bias", + "encoder.encoder.2.0.pt.v.1.weight", + "encoder.encoder.2.0.pt.v.1.bias", + "encoder.encoder.2.0.pt.v.3.bias", + "encoder.encoder.2.0.pt.conv_p.0.bias", + "encoder.encoder.2.0.pt.conv_p.1.weight", + "encoder.encoder.2.0.pt.conv_p.1.bias", + "encoder.encoder.2.0.pt.conv_p.3.bias", + "encoder.encoder.2.0.preconv.1.weight", + "encoder.encoder.2.0.preconv.1.bias", + "encoder.encoder.2.0.conv_finanal.1.weight", + "encoder.encoder.2.0.conv_finanal.1.bias", + "encoder.encoder.2.1.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.1.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.1.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.1.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.1.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.1.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.1.convs.vpsa.result.1.weight", + "encoder.encoder.2.1.convs.vpsa.result.1.bias", + "encoder.encoder.2.1.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.1.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.1.convs.vpsa.residual.1.weight", + "encoder.encoder.2.1.convs.vpsa.residual.1.bias", + "encoder.encoder.2.2.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.2.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.2.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.2.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.2.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.2.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.2.convs.vpsa.result.1.weight", + "encoder.encoder.2.2.convs.vpsa.result.1.bias", + "encoder.encoder.2.2.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.2.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.2.convs.vpsa.residual.1.weight", + "encoder.encoder.2.2.convs.vpsa.residual.1.bias", + "encoder.encoder.2.3.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.3.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.3.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.3.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.3.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.3.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.3.convs.vpsa.result.1.weight", + "encoder.encoder.2.3.convs.vpsa.result.1.bias", + "encoder.encoder.2.3.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.3.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.3.convs.vpsa.residual.1.weight", + "encoder.encoder.2.3.convs.vpsa.residual.1.bias", + "encoder.encoder.2.4.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.4.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.4.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.4.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.4.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.4.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.4.convs.vpsa.result.1.weight", + "encoder.encoder.2.4.convs.vpsa.result.1.bias", + "encoder.encoder.2.4.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.4.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.4.convs.vpsa.residual.1.weight", + "encoder.encoder.2.4.convs.vpsa.residual.1.bias", + "encoder.encoder.2.5.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.5.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.5.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.5.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.5.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.5.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.5.convs.vpsa.result.1.weight", + "encoder.encoder.2.5.convs.vpsa.result.1.bias", + "encoder.encoder.2.5.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.5.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.5.convs.vpsa.residual.1.weight", + "encoder.encoder.2.5.convs.vpsa.residual.1.bias", + "encoder.encoder.2.6.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.2.6.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.2.6.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.2.6.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.2.6.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.2.6.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.2.6.convs.vpsa.result.1.weight", + "encoder.encoder.2.6.convs.vpsa.result.1.bias", + "encoder.encoder.2.6.convs.vpsa.pos_x.1.weight", + "encoder.encoder.2.6.convs.vpsa.pos_x.1.bias", + "encoder.encoder.2.6.convs.vpsa.residual.1.weight", + "encoder.encoder.2.6.convs.vpsa.residual.1.bias", + "encoder.encoder.3.0.beta", + "encoder.encoder.3.0.scorenet_global.0.bias", + "encoder.encoder.3.0.scorenet_global.1.weight", + "encoder.encoder.3.0.scorenet_global.1.bias", + "encoder.encoder.3.0.scorenet_global.3.bias", + "encoder.encoder.3.0.scorenet_global.4.weight", + "encoder.encoder.3.0.scorenet_global.4.bias", + "encoder.encoder.3.0.selfattention.linear_q.bias", + "encoder.encoder.3.0.selfattention.linear_k.bias", + "encoder.encoder.3.0.selfattention.linear_v.0.bias", + "encoder.encoder.3.0.selfattention.linear_v.1.weight", + "encoder.encoder.3.0.selfattention.linear_v.1.bias", + "encoder.encoder.3.0.selfattention.linear_v.3.bias", + "encoder.encoder.3.0.selfattention.linear_v.4.weight", + "encoder.encoder.3.0.selfattention.linear_v.4.bias", + "encoder.encoder.3.0.pt.linear_q.bias", + "encoder.encoder.3.0.pt.linear_k.bias", + "encoder.encoder.3.0.pt.linear_p.0.bias", + "encoder.encoder.3.0.pt.linear_p.1.weight", + "encoder.encoder.3.0.pt.linear_p.1.bias", + "encoder.encoder.3.0.pt.linear_p.3.bias", + "encoder.encoder.3.0.pt.w.0.weight", + "encoder.encoder.3.0.pt.w.0.bias", + "encoder.encoder.3.0.pt.w.2.bias", + "encoder.encoder.3.0.pt.w.3.weight", + "encoder.encoder.3.0.pt.w.3.bias", + "encoder.encoder.3.0.pt.v.0.bias", + "encoder.encoder.3.0.pt.v.1.weight", + "encoder.encoder.3.0.pt.v.1.bias", + "encoder.encoder.3.0.pt.v.3.bias", + "encoder.encoder.3.0.pt.conv_p.0.bias", + "encoder.encoder.3.0.pt.conv_p.1.weight", + "encoder.encoder.3.0.pt.conv_p.1.bias", + "encoder.encoder.3.0.pt.conv_p.3.bias", + "encoder.encoder.3.0.preconv.1.weight", + "encoder.encoder.3.0.preconv.1.bias", + "encoder.encoder.3.0.conv_finanal.1.weight", + "encoder.encoder.3.0.conv_finanal.1.bias", + "encoder.encoder.3.1.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.3.1.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.3.1.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.3.1.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.3.1.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.3.1.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.3.1.convs.vpsa.result.1.weight", + "encoder.encoder.3.1.convs.vpsa.result.1.bias", + "encoder.encoder.3.1.convs.vpsa.pos_x.1.weight", + "encoder.encoder.3.1.convs.vpsa.pos_x.1.bias", + "encoder.encoder.3.1.convs.vpsa.residual.1.weight", + "encoder.encoder.3.1.convs.vpsa.residual.1.bias", + "encoder.encoder.3.2.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.3.2.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.3.2.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.3.2.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.3.2.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.3.2.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.3.2.convs.vpsa.result.1.weight", + "encoder.encoder.3.2.convs.vpsa.result.1.bias", + "encoder.encoder.3.2.convs.vpsa.pos_x.1.weight", + "encoder.encoder.3.2.convs.vpsa.pos_x.1.bias", + "encoder.encoder.3.2.convs.vpsa.residual.1.weight", + "encoder.encoder.3.2.convs.vpsa.residual.1.bias", + "encoder.encoder.3.3.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.3.3.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.3.3.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.3.3.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.3.3.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.3.3.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.3.3.convs.vpsa.result.1.weight", + "encoder.encoder.3.3.convs.vpsa.result.1.bias", + "encoder.encoder.3.3.convs.vpsa.pos_x.1.weight", + "encoder.encoder.3.3.convs.vpsa.pos_x.1.bias", + "encoder.encoder.3.3.convs.vpsa.residual.1.weight", + "encoder.encoder.3.3.convs.vpsa.residual.1.bias", + "encoder.encoder.4.0.beta", + "encoder.encoder.4.0.scorenet_global.0.bias", + "encoder.encoder.4.0.scorenet_global.1.weight", + "encoder.encoder.4.0.scorenet_global.1.bias", + "encoder.encoder.4.0.scorenet_global.3.bias", + "encoder.encoder.4.0.scorenet_global.4.weight", + "encoder.encoder.4.0.scorenet_global.4.bias", + "encoder.encoder.4.0.selfattention.linear_q.bias", + "encoder.encoder.4.0.selfattention.linear_k.bias", + "encoder.encoder.4.0.selfattention.linear_v.0.bias", + "encoder.encoder.4.0.selfattention.linear_v.1.weight", + "encoder.encoder.4.0.selfattention.linear_v.1.bias", + "encoder.encoder.4.0.selfattention.linear_v.3.bias", + "encoder.encoder.4.0.selfattention.linear_v.4.weight", + "encoder.encoder.4.0.selfattention.linear_v.4.bias", + "encoder.encoder.4.0.pt.linear_q.bias", + "encoder.encoder.4.0.pt.linear_k.bias", + "encoder.encoder.4.0.pt.linear_p.0.bias", + "encoder.encoder.4.0.pt.linear_p.1.weight", + "encoder.encoder.4.0.pt.linear_p.1.bias", + "encoder.encoder.4.0.pt.linear_p.3.bias", + "encoder.encoder.4.0.pt.w.0.weight", + "encoder.encoder.4.0.pt.w.0.bias", + "encoder.encoder.4.0.pt.w.2.bias", + "encoder.encoder.4.0.pt.w.3.weight", + "encoder.encoder.4.0.pt.w.3.bias", + "encoder.encoder.4.0.pt.v.0.bias", + "encoder.encoder.4.0.pt.v.1.weight", + "encoder.encoder.4.0.pt.v.1.bias", + "encoder.encoder.4.0.pt.v.3.bias", + "encoder.encoder.4.0.pt.conv_p.0.bias", + "encoder.encoder.4.0.pt.conv_p.1.weight", + "encoder.encoder.4.0.pt.conv_p.1.bias", + "encoder.encoder.4.0.pt.conv_p.3.bias", + "encoder.encoder.4.0.preconv.1.weight", + "encoder.encoder.4.0.preconv.1.bias", + "encoder.encoder.4.0.conv_finanal.1.weight", + "encoder.encoder.4.0.conv_finanal.1.bias", + "encoder.encoder.4.1.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.4.1.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.4.1.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.4.1.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.4.1.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.4.1.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.4.1.convs.vpsa.result.1.weight", + "encoder.encoder.4.1.convs.vpsa.result.1.bias", + "encoder.encoder.4.1.convs.vpsa.pos_x.1.weight", + "encoder.encoder.4.1.convs.vpsa.pos_x.1.bias", + "encoder.encoder.4.1.convs.vpsa.residual.1.weight", + "encoder.encoder.4.1.convs.vpsa.residual.1.bias", + "encoder.encoder.4.2.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.4.2.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.4.2.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.4.2.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.4.2.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.4.2.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.4.2.convs.vpsa.result.1.weight", + "encoder.encoder.4.2.convs.vpsa.result.1.bias", + "encoder.encoder.4.2.convs.vpsa.pos_x.1.weight", + "encoder.encoder.4.2.convs.vpsa.pos_x.1.bias", + "encoder.encoder.4.2.convs.vpsa.residual.1.weight", + "encoder.encoder.4.2.convs.vpsa.residual.1.bias", + "encoder.encoder.4.3.convs.vpsa.theta_x_alpha.1.weight", + "encoder.encoder.4.3.convs.vpsa.theta_x_alpha.1.bias", + "encoder.encoder.4.3.convs.vpsa.theta_x_beta.1.weight", + "encoder.encoder.4.3.convs.vpsa.theta_x_beta.1.bias", + "encoder.encoder.4.3.convs.vpsa.tf_zx.1.weight", + "encoder.encoder.4.3.convs.vpsa.tf_zx.1.bias", + "encoder.encoder.4.3.convs.vpsa.result.1.weight", + "encoder.encoder.4.3.convs.vpsa.result.1.bias", + "encoder.encoder.4.3.convs.vpsa.pos_x.1.weight", + "encoder.encoder.4.3.convs.vpsa.pos_x.1.bias", + "encoder.encoder.4.3.convs.vpsa.residual.1.weight", + "encoder.encoder.4.3.convs.vpsa.residual.1.bias", + "decoder.decoder.0.0.convs.0.1.weight", + "decoder.decoder.0.0.convs.0.1.bias", + "decoder.decoder.0.0.convs.1.1.weight", + "decoder.decoder.0.0.convs.1.1.bias", + "decoder.decoder.1.0.convs.0.1.weight", + "decoder.decoder.1.0.convs.0.1.bias", + "decoder.decoder.1.0.convs.1.1.weight", + "decoder.decoder.1.0.convs.1.1.bias", + "decoder.decoder.2.0.convs.0.1.weight", + "decoder.decoder.2.0.convs.0.1.bias", + "decoder.decoder.2.0.convs.1.1.weight", + "decoder.decoder.2.0.convs.1.1.bias", + "decoder.decoder.3.0.convs.0.1.weight", + "decoder.decoder.3.0.convs.0.1.bias", + "decoder.decoder.3.0.convs.1.1.weight", + "decoder.decoder.3.0.convs.1.1.bias", + "head.head.0.1.weight", + "head.head.0.1.bias", + "head.head.2.0.bias" + ], + "lr_scale": 1.0 + } +} +[03/26 12:21:07] S3DIS INFO: +Totally 68 samples in val set +[03/26 12:21:07] S3DIS INFO: length of validation dataset: 68 +[03/26 12:21:07] S3DIS INFO: number of classes of the dataset: 13 +[03/26 12:21:07] S3DIS INFO: Training from scratch +[03/26 12:21:07] S3DIS INFO: +Totally 204 samples in train set +[03/26 12:21:07] S3DIS INFO: length of training dataset: 6120 +[03/26 12:30:24] S3DIS INFO: Find a better ckpt @E1, val_miou 44.93 val_macc 52.33, val_oa 82.53 +mious: [91.7 97.36 73.81 0. 0. 30.64 28.62 68.81 81.88 0. 65.49 0. + 45.81] +[03/26 12:30:24] S3DIS INFO: Epoch 1 LR 0.010000 train_miou 41.91, val_miou 44.93, best val miou 44.93 +[03/26 12:30:25] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 12:39:05] S3DIS INFO: Find a better ckpt @E2, val_miou 48.08 val_macc 55.19, val_oa 82.95 +mious: [90.1 97.39 72.55 0. 0. 39.82 11.02 74.62 81.69 0. 67.31 40.63 + 49.9 ] +[03/26 12:39:05] S3DIS INFO: Epoch 2 LR 0.009998 train_miou 53.69, val_miou 48.08, best val miou 48.08 +[03/26 12:39:06] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 12:47:47] S3DIS INFO: Find a better ckpt @E3, val_miou 57.06 val_macc 64.93, val_oa 85.80 +mious: [91.38 97.39 78.92 0. 23.72 48.84 46.53 74.25 79.67 20.51 67.91 62.96 + 49.73] +[03/26 12:47:47] S3DIS INFO: Epoch 3 LR 0.009990 train_miou 60.76, val_miou 57.06, best val miou 57.06 +[03/26 12:47:47] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 12:56:31] S3DIS INFO: Find a better ckpt @E4, val_miou 59.54 val_macc 66.83, val_oa 86.37 +mious: [92.73 97.52 76.16 0. 29.08 50.59 24.38 77.62 86.87 49.48 71.94 61.83 + 55.8 ] +[03/26 12:56:31] S3DIS INFO: Epoch 4 LR 0.009978 train_miou 67.46, val_miou 59.54, best val miou 59.54 +[03/26 12:56:32] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 13:05:16] S3DIS INFO: Find a better ckpt @E5, val_miou 62.76 val_macc 70.06, val_oa 87.19 +mious: [88.92 97.5 80.22 0. 16.28 53.98 61.54 78.84 87.4 55.94 71.53 70.13 + 53.66] +[03/26 13:05:16] S3DIS INFO: Epoch 5 LR 0.009961 train_miou 72.46, val_miou 62.76, best val miou 62.76 +[03/26 13:05:17] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 13:14:07] S3DIS INFO: Epoch 6 LR 0.009939 train_miou 75.79, val_miou 60.75, best val miou 62.76 +[03/26 13:22:56] S3DIS INFO: Find a better ckpt @E7, val_miou 66.14 val_macc 72.78, val_oa 88.88 +mious: [92.96 97.78 82.32 0. 27.27 56.78 69.31 76.32 86.49 60.71 75.39 75.44 + 59.03] +[03/26 13:22:56] S3DIS INFO: Epoch 7 LR 0.009912 train_miou 78.02, val_miou 66.14, best val miou 66.14 +[03/26 13:22:57] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 13:31:41] S3DIS INFO: Epoch 8 LR 0.009880 train_miou 79.37, val_miou 62.03, best val miou 66.14 +[03/26 13:40:21] S3DIS INFO: Find a better ckpt @E9, val_miou 68.09 val_macc 74.49, val_oa 89.82 +mious: [93.83 97.87 83.12 0. 34.17 55.44 71.4 78.7 89.11 69.41 76.38 72.74 + 63.02] +[03/26 13:40:21] S3DIS INFO: Epoch 9 LR 0.009843 train_miou 80.65, val_miou 68.09, best val miou 68.09 +[03/26 13:40:22] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 13:49:06] S3DIS INFO: Epoch 10 LR 0.009802 train_miou 81.58, val_miou 68.08, best val miou 68.09 +[03/26 13:57:52] S3DIS INFO: Epoch 11 LR 0.009756 train_miou 82.86, val_miou 67.56, best val miou 68.09 +[03/26 14:06:32] S3DIS INFO: Find a better ckpt @E12, val_miou 69.00 val_macc 76.31, val_oa 89.96 +mious: [93.65 97.87 83.6 0. 34.46 57.57 72.32 80.53 90.36 77.84 76.33 70.25 + 62.2 ] +[03/26 14:06:32] S3DIS INFO: Epoch 12 LR 0.009705 train_miou 83.68, val_miou 69.00, best val miou 69.00 +[03/26 14:06:33] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 14:15:14] S3DIS INFO: Epoch 13 LR 0.009649 train_miou 84.50, val_miou 66.66, best val miou 69.00 +[03/26 14:23:53] S3DIS INFO: Epoch 14 LR 0.009589 train_miou 85.21, val_miou 68.58, best val miou 69.00 +[03/26 14:32:33] S3DIS INFO: Epoch 15 LR 0.009525 train_miou 86.15, val_miou 68.26, best val miou 69.00 +[03/26 14:41:07] S3DIS INFO: Epoch 16 LR 0.009456 train_miou 86.15, val_miou 68.03, best val miou 69.00 +[03/26 14:49:48] S3DIS INFO: Find a better ckpt @E17, val_miou 70.15 val_macc 77.10, val_oa 89.72 +mious: [93.01 97.91 82.84 0. 39.53 59.71 78.13 78.92 90.1 79.25 74.03 76.52 + 61.95] +[03/26 14:49:48] S3DIS INFO: Epoch 17 LR 0.009382 train_miou 87.00, val_miou 70.15, best val miou 70.15 +[03/26 14:49:49] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 14:58:31] S3DIS INFO: Epoch 18 LR 0.009304 train_miou 87.74, val_miou 69.52, best val miou 70.15 +[03/26 15:07:22] S3DIS INFO: Epoch 19 LR 0.009222 train_miou 87.74, val_miou 69.40, best val miou 70.15 +[03/26 15:16:06] S3DIS INFO: Epoch 20 LR 0.009136 train_miou 87.91, val_miou 68.85, best val miou 70.15 +[03/26 15:24:50] S3DIS INFO: Find a better ckpt @E21, val_miou 70.19 val_macc 76.47, val_oa 90.19 +mious: [94.42 97.92 82.39 0. 38.29 56.77 74.68 81.48 89.89 77. 76.87 78.2 + 64.54] +[03/26 15:24:50] S3DIS INFO: Epoch 21 LR 0.009046 train_miou 88.68, val_miou 70.19, best val miou 70.19 +[03/26 15:24:51] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 15:33:33] S3DIS INFO: Find a better ckpt @E22, val_miou 70.32 val_macc 77.25, val_oa 90.26 +mious: [93.89 97.86 83.21 0. 43.41 54.89 72.44 81.97 90.83 80.01 78.3 73.76 + 63.59] +[03/26 15:33:33] S3DIS INFO: Epoch 22 LR 0.008952 train_miou 89.13, val_miou 70.32, best val miou 70.32 +[03/26 15:33:34] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 15:42:17] S3DIS INFO: Epoch 23 LR 0.008854 train_miou 89.19, val_miou 68.75, best val miou 70.32 +[03/26 15:51:04] S3DIS INFO: Find a better ckpt @E24, val_miou 70.50 val_macc 77.01, val_oa 90.27 +mious: [94.18 97.92 83.89 0. 47.37 58.44 75.08 80.3 90.05 72.5 75.14 78.46 + 63.15] +[03/26 15:51:04] S3DIS INFO: Epoch 24 LR 0.008752 train_miou 89.72, val_miou 70.50, best val miou 70.50 +[03/26 15:51:05] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 15:59:43] S3DIS INFO: Epoch 25 LR 0.008646 train_miou 89.71, val_miou 69.61, best val miou 70.50 +[03/26 16:08:25] S3DIS INFO: Epoch 26 LR 0.008537 train_miou 89.88, val_miou 70.07, best val miou 70.50 +[03/26 16:17:04] S3DIS INFO: Epoch 27 LR 0.008424 train_miou 90.12, val_miou 69.93, best val miou 70.50 +[03/26 16:25:54] S3DIS INFO: Epoch 28 LR 0.008308 train_miou 90.51, val_miou 69.30, best val miou 70.50 +[03/26 16:34:37] S3DIS INFO: Find a better ckpt @E29, val_miou 70.50 val_macc 77.85, val_oa 90.27 +mious: [94.32 97.95 83.48 0. 46. 56.14 76.87 82.3 90.31 73.24 76.76 75.81 + 63.33] +[03/26 16:34:37] S3DIS INFO: Epoch 29 LR 0.008189 train_miou 90.63, val_miou 70.50, best val miou 70.50 +[03/26 16:34:38] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 16:43:20] S3DIS INFO: Find a better ckpt @E30, val_miou 70.51 val_macc 76.38, val_oa 90.44 +mious: [94.2 97.92 83.36 0. 39.68 55.93 79.02 82.12 91.4 73.26 77.43 77.75 + 64.49] +[03/26 16:43:20] S3DIS INFO: Epoch 30 LR 0.008066 train_miou 90.66, val_miou 70.51, best val miou 70.51 +[03/26 16:43:21] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 16:52:00] S3DIS INFO: Find a better ckpt @E31, val_miou 71.43 val_macc 78.53, val_oa 90.27 +mious: [93.89 97.94 82.99 0. 47.95 56.9 77.06 81.97 91.42 81.48 77.67 76.24 + 63.07] +[03/26 16:52:00] S3DIS INFO: Epoch 31 LR 0.007941 train_miou 90.98, val_miou 71.43, best val miou 71.43 +[03/26 16:52:01] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 17:00:39] S3DIS INFO: Epoch 32 LR 0.007813 train_miou 90.93, val_miou 69.81, best val miou 71.43 +[03/26 17:09:21] S3DIS INFO: Epoch 33 LR 0.007681 train_miou 91.38, val_miou 71.18, best val miou 71.43 +[03/26 17:18:04] S3DIS INFO: Epoch 34 LR 0.007548 train_miou 91.54, val_miou 70.71, best val miou 71.43 +[03/26 17:26:42] S3DIS INFO: Epoch 35 LR 0.007411 train_miou 91.64, val_miou 70.27, best val miou 71.43 +[03/26 17:35:27] S3DIS INFO: Epoch 36 LR 0.007273 train_miou 91.76, val_miou 70.14, best val miou 71.43 +[03/26 17:44:08] S3DIS INFO: Epoch 37 LR 0.007132 train_miou 91.90, val_miou 71.36, best val miou 71.43 +[03/26 17:52:47] S3DIS INFO: Epoch 38 LR 0.006989 train_miou 91.95, val_miou 70.29, best val miou 71.43 +[03/26 18:01:29] S3DIS INFO: Epoch 39 LR 0.006844 train_miou 92.19, val_miou 69.88, best val miou 71.43 +[03/26 18:10:11] S3DIS INFO: Epoch 40 LR 0.006697 train_miou 92.28, val_miou 70.36, best val miou 71.43 +[03/26 18:18:51] S3DIS INFO: Epoch 41 LR 0.006549 train_miou 92.29, val_miou 71.34, best val miou 71.43 +[03/26 18:27:32] S3DIS INFO: Epoch 42 LR 0.006399 train_miou 92.46, val_miou 71.29, best val miou 71.43 +[03/26 18:36:13] S3DIS INFO: Epoch 43 LR 0.006247 train_miou 92.60, val_miou 69.94, best val miou 71.43 +[03/26 18:45:00] S3DIS INFO: Epoch 44 LR 0.006095 train_miou 92.60, val_miou 70.53, best val miou 71.43 +[03/26 18:53:36] S3DIS INFO: Epoch 45 LR 0.005941 train_miou 92.58, val_miou 69.29, best val miou 71.43 +[03/26 19:02:18] S3DIS INFO: Find a better ckpt @E46, val_miou 71.59 val_macc 77.50, val_oa 90.79 +mious: [94.53 98.05 83.78 0. 38.68 57.04 79.55 82.54 91.52 83. 78.92 77.28 + 65.8 ] +[03/26 19:02:18] S3DIS INFO: Epoch 46 LR 0.005786 train_miou 92.79, val_miou 71.59, best val miou 71.59 +[03/26 19:02:19] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 19:10:53] S3DIS INFO: Epoch 47 LR 0.005631 train_miou 93.14, val_miou 70.30, best val miou 71.59 +[03/26 19:19:34] S3DIS INFO: Epoch 48 LR 0.005475 train_miou 93.01, val_miou 70.47, best val miou 71.59 +[03/26 19:28:11] S3DIS INFO: Epoch 49 LR 0.005319 train_miou 93.08, val_miou 70.46, best val miou 71.59 +[03/26 19:36:42] S3DIS INFO: Epoch 50 LR 0.005162 train_miou 93.19, val_miou 71.55, best val miou 71.59 +[03/26 19:45:20] S3DIS INFO: Epoch 51 LR 0.005005 train_miou 93.24, val_miou 70.65, best val miou 71.59 +[03/26 19:54:02] S3DIS INFO: Epoch 52 LR 0.004848 train_miou 93.36, val_miou 71.19, best val miou 71.59 +[03/26 20:02:42] S3DIS INFO: Epoch 53 LR 0.004691 train_miou 93.46, val_miou 71.50, best val miou 71.59 +[03/26 20:11:21] S3DIS INFO: Epoch 54 LR 0.004535 train_miou 93.57, val_miou 70.99, best val miou 71.59 +[03/26 20:20:02] S3DIS INFO: Epoch 55 LR 0.004379 train_miou 93.58, val_miou 71.34, best val miou 71.59 +[03/26 20:28:47] S3DIS INFO: Epoch 56 LR 0.004224 train_miou 93.69, val_miou 70.85, best val miou 71.59 +[03/26 20:37:23] S3DIS INFO: Find a better ckpt @E57, val_miou 71.99 val_macc 78.22, val_oa 90.77 +mious: [94.9 97.99 83.78 0. 41.14 53.86 76.17 83.05 92.19 87.42 78.69 80.76 + 65.87] +[03/26 20:37:23] S3DIS INFO: Epoch 57 LR 0.004069 train_miou 93.82, val_miou 71.99, best val miou 71.99 +[03/26 20:37:24] S3DIS INFO: Found the best model and saved in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/26 20:46:04] S3DIS INFO: Epoch 58 LR 0.003915 train_miou 93.85, val_miou 71.74, best val miou 71.99 +[03/26 20:54:45] S3DIS INFO: Epoch 59 LR 0.003763 train_miou 93.81, val_miou 71.28, best val miou 71.99 +[03/26 21:03:23] S3DIS INFO: Epoch 60 LR 0.003611 train_miou 93.99, val_miou 70.51, best val miou 71.99 +[03/26 21:12:17] S3DIS INFO: Epoch 61 LR 0.003461 train_miou 94.07, val_miou 71.57, best val miou 71.99 +[03/26 21:21:07] S3DIS INFO: Epoch 62 LR 0.003313 train_miou 94.16, val_miou 71.50, best val miou 71.99 +[03/26 21:29:51] S3DIS INFO: Epoch 63 LR 0.003166 train_miou 94.09, val_miou 71.43, best val miou 71.99 +[03/26 21:38:36] S3DIS INFO: Epoch 64 LR 0.003021 train_miou 94.20, val_miou 70.88, best val miou 71.99 +[03/26 21:47:16] S3DIS INFO: Epoch 65 LR 0.002878 train_miou 94.22, val_miou 70.14, best val miou 71.99 +[03/26 21:55:59] S3DIS INFO: Epoch 66 LR 0.002737 train_miou 94.31, val_miou 71.35, best val miou 71.99 +[03/26 22:04:48] S3DIS INFO: Epoch 67 LR 0.002599 train_miou 94.41, val_miou 71.43, best val miou 71.99 +[03/26 22:13:29] S3DIS INFO: Epoch 68 LR 0.002462 train_miou 94.46, val_miou 71.67, best val miou 71.99 +[03/26 22:22:12] S3DIS INFO: Epoch 69 LR 0.002329 train_miou 94.44, val_miou 71.11, best val miou 71.99 +[03/26 22:30:54] S3DIS INFO: Epoch 70 LR 0.002197 train_miou 94.52, val_miou 71.05, best val miou 71.99 +[03/26 22:39:39] S3DIS INFO: Epoch 71 LR 0.002069 train_miou 94.47, val_miou 71.33, best val miou 71.99 +[03/26 22:48:32] S3DIS INFO: Epoch 72 LR 0.001944 train_miou 94.65, val_miou 71.12, best val miou 71.99 +[03/26 22:57:18] S3DIS INFO: Epoch 73 LR 0.001821 train_miou 94.64, val_miou 71.47, best val miou 71.99 +[03/26 23:06:06] S3DIS INFO: Epoch 74 LR 0.001702 train_miou 94.67, val_miou 71.40, best val miou 71.99 +[03/26 23:14:47] S3DIS INFO: Epoch 75 LR 0.001586 train_miou 94.67, val_miou 71.11, best val miou 71.99 +[03/26 23:23:29] S3DIS INFO: Epoch 76 LR 0.001473 train_miou 94.74, val_miou 71.40, best val miou 71.99 +[03/26 23:32:06] S3DIS INFO: Epoch 77 LR 0.001364 train_miou 94.80, val_miou 70.82, best val miou 71.99 +[03/26 23:40:50] S3DIS INFO: Epoch 78 LR 0.001258 train_miou 94.80, val_miou 71.34, best val miou 71.99 +[03/26 23:49:31] S3DIS INFO: Epoch 79 LR 0.001156 train_miou 94.79, val_miou 70.52, best val miou 71.99 +[03/26 23:58:21] S3DIS INFO: Epoch 80 LR 0.001058 train_miou 94.85, val_miou 71.16, best val miou 71.99 +[03/27 00:06:59] S3DIS INFO: Epoch 81 LR 0.000964 train_miou 94.88, val_miou 70.86, best val miou 71.99 +[03/27 00:15:44] S3DIS INFO: Epoch 82 LR 0.000874 train_miou 94.87, val_miou 70.12, best val miou 71.99 +[03/27 00:24:22] S3DIS INFO: Epoch 83 LR 0.000788 train_miou 94.91, val_miou 70.76, best val miou 71.99 +[03/27 00:33:06] S3DIS INFO: Epoch 84 LR 0.000706 train_miou 94.93, val_miou 71.10, best val miou 71.99 +[03/27 00:41:49] S3DIS INFO: Epoch 85 LR 0.000628 train_miou 94.95, val_miou 70.91, best val miou 71.99 +[03/27 00:50:43] S3DIS INFO: Epoch 86 LR 0.000554 train_miou 94.95, val_miou 71.07, best val miou 71.99 +[03/27 00:59:34] S3DIS INFO: Epoch 87 LR 0.000485 train_miou 95.00, val_miou 70.79, best val miou 71.99 +[03/27 01:08:22] S3DIS INFO: Epoch 88 LR 0.000421 train_miou 94.96, val_miou 71.01, best val miou 71.99 +[03/27 01:17:17] S3DIS INFO: Epoch 89 LR 0.000361 train_miou 95.00, val_miou 70.67, best val miou 71.99 +[03/27 01:25:59] S3DIS INFO: Epoch 90 LR 0.000305 train_miou 94.99, val_miou 71.23, best val miou 71.99 +[03/27 01:34:39] S3DIS INFO: Epoch 91 LR 0.000254 train_miou 94.99, val_miou 70.85, best val miou 71.99 +[03/27 01:43:17] S3DIS INFO: Epoch 92 LR 0.000208 train_miou 94.99, val_miou 71.01, best val miou 71.99 +[03/27 01:52:00] S3DIS INFO: Epoch 93 LR 0.000167 train_miou 95.00, val_miou 70.96, best val miou 71.99 +[03/27 02:00:38] S3DIS INFO: Epoch 94 LR 0.000130 train_miou 95.05, val_miou 71.11, best val miou 71.99 +[03/27 02:09:21] S3DIS INFO: Epoch 95 LR 0.000098 train_miou 95.05, val_miou 71.09, best val miou 71.99 +[03/27 02:18:05] S3DIS INFO: Epoch 96 LR 0.000071 train_miou 95.02, val_miou 71.04, best val miou 71.99 +[03/27 02:26:57] S3DIS INFO: Epoch 97 LR 0.000049 train_miou 95.07, val_miou 71.16, best val miou 71.99 +[03/27 02:35:38] S3DIS INFO: Epoch 98 LR 0.000032 train_miou 95.03, val_miou 70.89, best val miou 71.99 +[03/27 02:44:20] S3DIS INFO: Epoch 99 LR 0.000020 train_miou 95.04, val_miou 71.10, best val miou 71.99 +[03/27 02:52:56] S3DIS INFO: Epoch 100 LR 0.000012 train_miou 95.04, val_miou 71.26, best val miou 71.99 +[03/27 02:52:57] S3DIS INFO: Best ckpt @E57, val_oa 90.77, val_macc 78.22, val_miou 71.99, +iou per cls is: [94.9 97.99 83.78 0. 41.14 53.86 76.17 83.05 92.19 87.42 78.69 80.76 + 65.87] +[03/27 02:52:57] S3DIS INFO: Successful Loading the ckpt from log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/checkpoint/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth +[03/27 02:52:57] S3DIS INFO: ckpts @ 57 epoch( {'best_val': 71.9859848022461} ) +[03/27 02:52:57] S3DIS INFO: Test [0]/[68] cloud +[03/27 02:53:12] S3DIS INFO: [0]/[68] cloud, test_oa , test_macc, test_miou: 93.06 97.47 88.18, +iou per cls is: [ 97.04 99.37 82.49 100. 100. 100. 97.7 0. 100. 100. + 100. 100. 69.69] +[03/27 02:53:12] S3DIS INFO: Test [1]/[68] cloud +[03/27 02:53:22] S3DIS INFO: [1]/[68] cloud, test_oa , test_macc, test_miou: 93.47 97.65 80.74, +iou per cls is: [ 95.71 99.3 84.51 100. 100. 100. 98.54 0. 100. 100. + 0. 100. 71.63] +[03/27 02:53:22] S3DIS INFO: Test [2]/[68] cloud +[03/27 02:53:42] S3DIS INFO: [2]/[68] cloud, test_oa , test_macc, test_miou: 89.94 87.96 70.19, +iou per cls is: [ 98.05 98.85 77.5 0. 6.27 66.7 96.44 95.38 95.74 100. + 0. 97.52 80. ] +[03/27 02:53:42] S3DIS INFO: Test [3]/[68] cloud +[03/27 02:55:32] S3DIS INFO: [3]/[68] cloud, test_oa , test_macc, test_miou: 95.45 94.85 83.78, +iou per cls is: [ 94.4 97.83 93.08 0. 91.68 100. 93.49 94.34 94.25 99.83 + 78.17 87.72 64.38] +[03/27 02:55:32] S3DIS INFO: Test [4]/[68] cloud +[03/27 02:56:16] S3DIS INFO: [4]/[68] cloud, test_oa , test_macc, test_miou: 91.60 89.83 70.20, +iou per cls is: [ 96.35 99.56 79.24 0. 77.77 63.67 96.99 71.46 97.24 100. + 0. 53.38 76.9 ] +[03/27 02:56:16] S3DIS INFO: Test [5]/[68] cloud +[03/27 03:00:00] S3DIS INFO: [5]/[68] cloud, test_oa , test_macc, test_miou: 80.40 81.11 37.71, +iou per cls is: [88.87 98.43 46.2 0. 6.1 0.55 25.13 0. 0. 97.22 80.62 0. + 47.16] +[03/27 03:00:00] S3DIS INFO: Test [6]/[68] cloud +[03/27 03:00:31] S3DIS INFO: [6]/[68] cloud, test_oa , test_macc, test_miou: 98.58 98.73 98.11, +iou per cls is: [ 98.38 99.46 97.44 100. 100. 100. 83.95 100. 100. 100. + 100. 100. 96.14] +[03/27 03:00:31] S3DIS INFO: Test [7]/[68] cloud +[03/27 03:00:48] S3DIS INFO: [7]/[68] cloud, test_oa , test_macc, test_miou: 97.68 98.31 90.27, +iou per cls is: [ 96.61 98.98 96.24 100. 100. 100. 81.67 100. 100. 100. + 100. 100. 0. ] +[03/27 03:00:48] S3DIS INFO: Test [8]/[68] cloud +[03/27 03:01:06] S3DIS INFO: [8]/[68] cloud, test_oa , test_macc, test_miou: 98.19 98.46 96.07, +iou per cls is: [ 97.35 99.36 97.17 100. 100. 100. 80.73 100. 100. 100. + 100. 100. 74.36] +[03/27 03:01:06] S3DIS INFO: Test [9]/[68] cloud +[03/27 03:01:41] S3DIS INFO: [9]/[68] cloud, test_oa , test_macc, test_miou: 90.41 91.40 76.96, +iou per cls is: [ 98.31 99.4 90.65 100. 100. 1.39 21.02 100. 0. 100. + 100. 100. 89.75] +[03/27 03:01:41] S3DIS INFO: Test [10]/[68] cloud +[03/27 03:01:53] S3DIS INFO: [10]/[68] cloud, test_oa , test_macc, test_miou: 96.63 96.95 80.98, +iou per cls is: [ 98.65 99.11 94.05 100. 100. 100. 65.58 100. 100. 100. + 0. 0. 95.38] +[03/27 03:01:53] S3DIS INFO: Test [11]/[68] cloud +[03/27 03:02:33] S3DIS INFO: [11]/[68] cloud, test_oa , test_macc, test_miou: 85.24 82.35 44.59, +iou per cls is: [ 98.67 98.79 70.31 100. 0.05 0.14 36.14 0. 0. 100. + 0. 0. 75.52] +[03/27 03:02:33] S3DIS INFO: Test [12]/[68] cloud +[03/27 03:08:16] S3DIS INFO: [12]/[68] cloud, test_oa , test_macc, test_miou: 96.65 96.93 65.50, +iou per cls is: [ 96.92 99.08 94.17 0. 0. 100. 64.74 0. 100. 100. + 0. 100. 96.6 ] +[03/27 03:08:16] S3DIS INFO: Test [13]/[68] cloud +[03/27 03:08:43] S3DIS INFO: [13]/[68] cloud, test_oa , test_macc, test_miou: 97.56 97.00 73.43, +iou per cls is: [ 97.58 98.04 96.74 100. 0. 100. 72.63 100. 100. 100. + 0. 0. 89.67] +[03/27 03:08:43] S3DIS INFO: Test [14]/[68] cloud +[03/27 03:09:02] S3DIS INFO: [14]/[68] cloud, test_oa , test_macc, test_miou: 93.83 83.51 66.16, +iou per cls is: [ 97.96 99.47 94.5 100. 100. 100. 54.2 0. 0. 87.68 + 26.24 100. 0. ] +[03/27 03:09:02] S3DIS INFO: Test [15]/[68] cloud +[03/27 03:11:07] S3DIS INFO: [15]/[68] cloud, test_oa , test_macc, test_miou: 95.45 90.70 78.29, +iou per cls is: [ 98.98 99.82 93.2 100. 100. 100. 71.19 0. 100. 76.14 + 35.32 100. 43.14] +[03/27 03:11:07] S3DIS INFO: Test [16]/[68] cloud +[03/27 03:11:24] S3DIS INFO: [16]/[68] cloud, test_oa , test_macc, test_miou: 96.05 97.29 95.93, +iou per cls is: [ 96.64 99.22 92.97 100. 100. 100. 84.15 100. 100. 100. + 100. 100. 74.05] +[03/27 03:11:24] S3DIS INFO: Test [17]/[68] cloud +[03/27 03:11:54] S3DIS INFO: [17]/[68] cloud, test_oa , test_macc, test_miou: 95.11 97.11 80.91, +iou per cls is: [ 99.03 99.42 91.73 100. 100. 100. 84.33 100. 100. 100. + 0. 0. 77.33] +[03/27 03:11:54] S3DIS INFO: Test [18]/[68] cloud +[03/27 03:12:02] S3DIS INFO: [18]/[68] cloud, test_oa , test_macc, test_miou: 96.57 94.75 92.57, +iou per cls is: [ 97.08 99.59 94.38 100. 100. 100. 68.33 100. 100. 100. + 100. 100. 44.06] +[03/27 03:12:02] S3DIS INFO: Test [19]/[68] cloud +[03/27 03:12:35] S3DIS INFO: [19]/[68] cloud, test_oa , test_macc, test_miou: 95.81 81.51 79.93, +iou per cls is: [ 96.74 98.88 92.65 100. 0. 0. 62.61 100. 100. 100. + 100. 100. 88.22] +[03/27 03:12:35] S3DIS INFO: Test [20]/[68] cloud +[03/27 03:13:01] S3DIS INFO: [20]/[68] cloud, test_oa , test_macc, test_miou: 95.75 94.31 86.01, +iou per cls is: [ 99.77 98.44 96.55 100. 100. 100. 100. 92.57 96.54 100. + 100. 0. 34.25] +[03/27 03:13:01] S3DIS INFO: Test [21]/[68] cloud +[03/27 03:13:16] S3DIS INFO: [21]/[68] cloud, test_oa , test_macc, test_miou: 93.28 92.90 80.94, +iou per cls is: [ 97.4 98.26 72.57 0. 54.64 80.29 97.39 93.28 97.35 100. + 98.09 91.24 71.73] +[03/27 03:13:16] S3DIS INFO: Test [22]/[68] cloud +[03/27 03:13:30] S3DIS INFO: [22]/[68] cloud, test_oa , test_macc, test_miou: 93.89 91.56 78.76, +iou per cls is: [ 98.25 99.04 78.66 0. 30.91 84.28 90.19 79.85 96.16 100. + 97.68 95.56 73.23] +[03/27 03:13:30] S3DIS INFO: Test [23]/[68] cloud +[03/27 03:13:45] S3DIS INFO: [23]/[68] cloud, test_oa , test_macc, test_miou: 95.70 94.30 80.90, +iou per cls is: [ 97.82 98.67 81.78 0. 70.17 69.75 97.29 77.19 92.38 100. + 97.54 94.3 74.84] +[03/27 03:13:45] S3DIS INFO: Test [24]/[68] cloud +[03/27 03:14:01] S3DIS INFO: [24]/[68] cloud, test_oa , test_macc, test_miou: 90.25 91.15 69.97, +iou per cls is: [ 97.54 97.64 74.53 0. 62.07 62.64 94.53 82.43 87.08 100. + 91.25 0. 59.97] +[03/27 03:14:01] S3DIS INFO: Test [25]/[68] cloud +[03/27 03:14:23] S3DIS INFO: [25]/[68] cloud, test_oa , test_macc, test_miou: 92.37 91.48 79.37, +iou per cls is: [ 98.25 98.02 73.99 0. 58.19 85.73 84.79 85.78 94.64 100. + 94.7 91.95 65.78] +[03/27 03:14:23] S3DIS INFO: Test [26]/[68] cloud +[03/27 03:14:56] S3DIS INFO: [26]/[68] cloud, test_oa , test_macc, test_miou: 88.00 86.94 73.24, +iou per cls is: [ 93.32 98.71 74.16 0. 21.9 60.94 95.65 63.38 92.12 100. + 87.36 97.33 67.26] +[03/27 03:14:56] S3DIS INFO: Test [27]/[68] cloud +[03/27 03:15:31] S3DIS INFO: [27]/[68] cloud, test_oa , test_macc, test_miou: 93.27 92.49 80.85, +iou per cls is: [ 95.99 98.66 84.35 0. 55.01 81.2 85.99 87.98 97.06 100. + 91.45 97.2 76.12] +[03/27 03:15:31] S3DIS INFO: Test [28]/[68] cloud +[03/27 03:15:46] S3DIS INFO: [28]/[68] cloud, test_oa , test_macc, test_miou: 92.45 86.42 74.62, +iou per cls is: [ 98.34 97.84 87.83 0. 92. 100. 95.96 89.68 91.75 100. + 29.71 23.2 63.74] +[03/27 03:15:46] S3DIS INFO: Test [29]/[68] cloud +[03/27 03:16:05] S3DIS INFO: [29]/[68] cloud, test_oa , test_macc, test_miou: 90.59 92.69 65.22, +iou per cls is: [ 97.78 98.27 93.8 0. 0. 100. 64.26 60.17 100. 100. + 71.57 0. 62.03] +[03/27 03:16:05] S3DIS INFO: Test [30]/[68] cloud +[03/27 03:16:37] S3DIS INFO: [30]/[68] cloud, test_oa , test_macc, test_miou: 96.84 92.98 82.76, +iou per cls is: [ 99.29 98.71 94.99 0. 53.97 100. 92.49 98.2 94.03 100. + 90.46 77.99 75.8 ] +[03/27 03:16:37] S3DIS INFO: Test [31]/[68] cloud +[03/27 03:17:33] S3DIS INFO: [31]/[68] cloud, test_oa , test_macc, test_miou: 85.77 92.37 63.86, +iou per cls is: [ 94.22 98.89 73.95 0. 0. 0. 91.08 90.34 93.8 100. + 55.34 89.61 42.91] +[03/27 03:17:33] S3DIS INFO: Test [32]/[68] cloud +[03/27 03:17:50] S3DIS INFO: [32]/[68] cloud, test_oa , test_macc, test_miou: 94.84 90.88 79.39, +iou per cls is: [ 97.5 98.47 81.39 0. 25.78 76.5 89.67 89.88 95.16 100. + 96.31 94.66 86.77] +[03/27 03:17:50] S3DIS INFO: Test [33]/[68] cloud +[03/27 03:17:58] S3DIS INFO: [33]/[68] cloud, test_oa , test_macc, test_miou: 91.47 94.08 74.12, +iou per cls is: [ 97.92 98.19 82.9 0. 76.7 100. 82.89 95.53 95.37 100. + 63.18 0. 70.82] +[03/27 03:17:58] S3DIS INFO: Test [34]/[68] cloud +[03/27 03:19:16] S3DIS INFO: [34]/[68] cloud, test_oa , test_macc, test_miou: 93.45 93.42 72.86, +iou per cls is: [ 98.96 99.38 87.48 0. 85.82 0. 97.84 95.28 93.07 100. + 43.91 97.28 48.11] +[03/27 03:19:16] S3DIS INFO: Test [35]/[68] cloud +[03/27 03:19:31] S3DIS INFO: [35]/[68] cloud, test_oa , test_macc, test_miou: 93.60 89.92 77.63, +iou per cls is: [ 96.95 98.76 82.95 0. 25.1 67.47 94.85 85.98 94.8 100. + 97.55 91.15 73.62] +[03/27 03:19:31] S3DIS INFO: Test [36]/[68] cloud +[03/27 03:19:50] S3DIS INFO: [36]/[68] cloud, test_oa , test_macc, test_miou: 95.38 95.06 75.64, +iou per cls is: [97.68 98.29 87.08 0. 67.33 73.78 98.47 90.39 92.21 0. 97.43 93.59 + 87.1 ] +[03/27 03:19:50] S3DIS INFO: Test [37]/[68] cloud +[03/27 03:20:41] S3DIS INFO: [37]/[68] cloud, test_oa , test_macc, test_miou: 84.83 87.42 72.76, +iou per cls is: [97.04 95.22 61.04 0. 69.61 79.56 94.12 81.7 85.75 81.29 53.19 96.18 + 51.19] +[03/27 03:20:41] S3DIS INFO: Test [38]/[68] cloud +[03/27 03:20:52] S3DIS INFO: [38]/[68] cloud, test_oa , test_macc, test_miou: 90.95 92.27 72.75, +iou per cls is: [ 96.7 96.06 82.23 0. 0. 81.42 92.29 71.73 98.14 100. + 91.72 70.8 64.63] +[03/27 03:20:52] S3DIS INFO: Test [39]/[68] cloud +[03/27 03:21:06] S3DIS INFO: [39]/[68] cloud, test_oa , test_macc, test_miou: 87.15 87.17 66.69, +iou per cls is: [95.74 98.02 85.25 0. 70.83 79.15 96.37 81.42 92.2 0. 64.08 57.11 + 46.78] +[03/27 03:21:06] S3DIS INFO: Test [40]/[68] cloud +[03/27 03:21:21] S3DIS INFO: [40]/[68] cloud, test_oa , test_macc, test_miou: 92.51 90.66 78.65, +iou per cls is: [ 96.02 99.48 83.35 0. 42.94 76.07 91.57 88.74 96.37 100. + 91.07 91.66 65.25] +[03/27 03:21:21] S3DIS INFO: Test [41]/[68] cloud +[03/27 03:21:35] S3DIS INFO: [41]/[68] cloud, test_oa , test_macc, test_miou: 94.35 94.55 82.24, +iou per cls is: [ 96.35 99.36 87.74 0. 70.85 88.08 94.24 91.11 94.03 100. + 91.59 93.4 62.43] +[03/27 03:21:35] S3DIS INFO: Test [42]/[68] cloud +[03/27 03:22:32] S3DIS INFO: [42]/[68] cloud, test_oa , test_macc, test_miou: 82.52 86.99 73.07, +iou per cls is: [ 96.45 98.18 61.61 0. 75.93 68.58 97.62 89.42 95.92 100. + 26.69 93.41 46.12] +[03/27 03:22:32] S3DIS INFO: Test [43]/[68] cloud +[03/27 03:22:45] S3DIS INFO: [43]/[68] cloud, test_oa , test_macc, test_miou: 93.35 94.18 67.72, +iou per cls is: [96.91 99.08 74.27 0. 0. 79.97 97.63 86.05 95.01 0. 96.33 96.09 + 59.05] +[03/27 03:22:45] S3DIS INFO: Test [44]/[68] cloud +[03/27 03:22:52] S3DIS INFO: [44]/[68] cloud, test_oa , test_macc, test_miou: 97.33 97.14 95.03, +iou per cls is: [ 94.52 97.86 96.95 100. 100. 100. 96.55 85.62 95.67 100. + 96.64 96.33 75.27] +[03/27 03:22:52] S3DIS INFO: Test [45]/[68] cloud +[03/27 03:23:11] S3DIS INFO: [45]/[68] cloud, test_oa , test_macc, test_miou: 91.18 91.87 79.58, +iou per cls is: [ 96.43 99.57 70.14 0. 41.12 90.73 93.43 85.12 96.22 100. + 91.21 100. 70.54] +[03/27 03:23:11] S3DIS INFO: Test [46]/[68] cloud +[03/27 03:23:25] S3DIS INFO: [46]/[68] cloud, test_oa , test_macc, test_miou: 94.77 93.64 75.72, +iou per cls is: [ 96.05 99.69 87.73 0. 51.91 95.89 96.99 90.13 92.9 100. + 90.3 0. 82.78] +[03/27 03:23:25] S3DIS INFO: Test [47]/[68] cloud +[03/27 03:23:47] S3DIS INFO: [47]/[68] cloud, test_oa , test_macc, test_miou: 94.63 91.63 88.52, +iou per cls is: [ 96.74 99.65 87.81 100. 27.01 91.98 98.35 89.1 95.65 100. + 91.9 98.51 73.99] +[03/27 03:23:47] S3DIS INFO: Test [48]/[68] cloud +[03/27 03:24:08] S3DIS INFO: [48]/[68] cloud, test_oa , test_macc, test_miou: 87.59 88.37 75.10, +iou per cls is: [ 95.71 96.48 71.05 0. 45.57 71.04 96.71 64.43 95.49 100. + 83.58 96.01 60.18] +[03/27 03:24:08] S3DIS INFO: Test [49]/[68] cloud +[03/27 03:24:27] S3DIS INFO: [49]/[68] cloud, test_oa , test_macc, test_miou: 97.17 97.73 88.07, +iou per cls is: [ 98.36 97.26 95.18 0. 100. 100. 97.27 95.61 93.08 100. + 92.99 97.03 78.14] +[03/27 03:24:27] S3DIS INFO: Test [50]/[68] cloud +[03/27 03:24:59] S3DIS INFO: [50]/[68] cloud, test_oa , test_macc, test_miou: 93.62 93.96 81.14, +iou per cls is: [ 98.43 95.69 95.17 100. 0. 100. 95.84 82.14 94.14 100. + 53.15 97.71 42.59] +[03/27 03:24:59] S3DIS INFO: Test [51]/[68] cloud +[03/27 03:27:13] S3DIS INFO: [51]/[68] cloud, test_oa , test_macc, test_miou: 89.57 82.98 55.17, +iou per cls is: [91. 98.64 83.33 0. 2.67 0. 59.98 77.24 91.17 0. 74.82 61.59 + 76.8 ] +[03/27 03:27:13] S3DIS INFO: Test [52]/[68] cloud +[03/27 03:29:55] S3DIS INFO: [52]/[68] cloud, test_oa , test_macc, test_miou: 84.93 89.02 66.55, +iou per cls is: [ 92.31 97.51 71.92 0. 59.18 100. 88.4 70.93 71.97 80.75 + 65.28 0. 66.9 ] +[03/27 03:29:55] S3DIS INFO: Test [53]/[68] cloud +[03/27 03:30:14] S3DIS INFO: [53]/[68] cloud, test_oa , test_macc, test_miou: 92.10 86.07 73.88, +iou per cls is: [ 93.05 98.76 92.03 0. 0. 100. 86.01 87.4 95.11 100. + 57.97 90.41 59.74] +[03/27 03:30:14] S3DIS INFO: Test [54]/[68] cloud +[03/27 03:30:29] S3DIS INFO: [54]/[68] cloud, test_oa , test_macc, test_miou: 94.79 92.92 79.77, +iou per cls is: [ 96.31 98.5 81.4 0. 32.95 79.38 83.91 94.34 96.48 100. + 98.54 97.25 77.95] +[03/27 03:30:29] S3DIS INFO: Test [55]/[68] cloud +[03/27 03:33:02] S3DIS INFO: [55]/[68] cloud, test_oa , test_macc, test_miou: 93.96 93.38 73.38, +iou per cls is: [ 96.91 97.59 91.7 0. 74.4 100. 91.31 83.22 87.48 0. + 85.55 76.4 69.4 ] +[03/27 03:33:02] S3DIS INFO: Test [56]/[68] cloud +[03/27 03:33:50] S3DIS INFO: [56]/[68] cloud, test_oa , test_macc, test_miou: 91.42 94.17 77.28, +iou per cls is: [ 97.26 98.49 69.79 0. 33.24 86.88 97.09 88.91 83.89 100. + 82.43 90.11 76.54] +[03/27 03:33:50] S3DIS INFO: Test [57]/[68] cloud +[03/27 03:34:11] S3DIS INFO: [57]/[68] cloud, test_oa , test_macc, test_miou: 91.21 93.23 78.66, +iou per cls is: [ 97.62 98.41 71.95 0. 42.8 62.15 93.24 89.91 96.12 100. + 89.96 100. 80.43] +[03/27 03:34:11] S3DIS INFO: Test [58]/[68] cloud +[03/27 03:34:23] S3DIS INFO: [58]/[68] cloud, test_oa , test_macc, test_miou: 95.92 96.75 70.99, +iou per cls is: [ 96.98 98.86 90.48 0. 0. 72.57 94.77 91.41 99.4 100. + 96.34 0. 82.01] +[03/27 03:34:23] S3DIS INFO: Test [59]/[68] cloud +[03/27 03:34:37] S3DIS INFO: [59]/[68] cloud, test_oa , test_macc, test_miou: 94.63 94.83 76.32, +iou per cls is: [ 97.74 98.76 79.29 0. 0. 65.14 95.99 88.68 91.28 100. + 96.83 97.31 81.1 ] +[03/27 03:34:37] S3DIS INFO: Test [60]/[68] cloud +[03/27 03:34:56] S3DIS INFO: [60]/[68] cloud, test_oa , test_macc, test_miou: 93.18 87.17 83.61, +iou per cls is: [ 97.66 98.71 74.69 100. 8.5 76.56 85.35 83.23 95.94 100. + 93.94 96.46 75.87] +[03/27 03:34:56] S3DIS INFO: Test [61]/[68] cloud +[03/27 03:35:17] S3DIS INFO: [61]/[68] cloud, test_oa , test_macc, test_miou: 95.18 92.74 87.20, +iou per cls is: [ 97.89 97.24 86.12 100. 61.77 59.33 96.16 92.65 92.92 100. + 97.85 93.75 57.93] +[03/27 03:35:17] S3DIS INFO: Test [62]/[68] cloud +[03/27 03:35:33] S3DIS INFO: [62]/[68] cloud, test_oa , test_macc, test_miou: 95.60 96.23 77.28, +iou per cls is: [ 97.04 97.2 91.65 0. 0. 80.34 98.36 78.72 98.37 85.4 + 97.21 100. 80.34] +[03/27 03:35:33] S3DIS INFO: Test [63]/[68] cloud +[03/27 03:35:43] S3DIS INFO: [63]/[68] cloud, test_oa , test_macc, test_miou: 77.99 94.12 62.90, +iou per cls is: [ 93.78 98.03 95.99 100. 0. 100. 100. 0. 100. 100. + 0. 0. 29.86] +[03/27 03:35:43] S3DIS INFO: Test [64]/[68] cloud +[03/27 03:35:54] S3DIS INFO: [64]/[68] cloud, test_oa , test_macc, test_miou: 74.28 78.93 48.95, +iou per cls is: [ 82.1 96.52 65.56 0. 1.25 0. 88.21 100. 0. 100. + 74.08 0. 28.63] +[03/27 03:35:54] S3DIS INFO: Test [65]/[68] cloud +[03/27 03:36:12] S3DIS INFO: [65]/[68] cloud, test_oa , test_macc, test_miou: 82.49 92.28 72.86, +iou per cls is: [ 63.35 97.3 78.35 100. 0. 100. 84.96 100. 100. 100. + 64.86 0. 58.35] +[03/27 03:36:12] S3DIS INFO: Test [66]/[68] cloud +[03/27 03:36:15] S3DIS INFO: [66]/[68] cloud, test_oa , test_macc, test_miou: 87.47 95.50 78.01, +iou per cls is: [ 69.45 99.07 89.48 100. 100. 100. 96.81 100. 100. 100. + 0. 0. 59.37] +[03/27 03:36:15] S3DIS INFO: Test [67]/[68] cloud +[03/27 03:36:28] S3DIS INFO: [67]/[68] cloud, test_oa , test_macc, test_miou: 87.04 88.63 70.57, +iou per cls is: [ 90.01 98.8 91.6 100. 100. 100. 51.72 59.78 75.5 100. + 0. 0. 50.03] +[03/27 03:36:28] S3DIS INFO: Best ckpt @E57, test_oa 91.77, test_macc 79.15, test_miou 73.41, +iou per cls is: [95.72 98.62 85.71 0. 43.74 57.91 78.43 85.28 92.55 88.03 78.94 83.57 + 65.86] +[03/27 03:36:28] S3DIS INFO: save results in log/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk.csv diff --git a/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth b/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth new file mode 100644 index 0000000000000000000000000000000000000000..bda40f247e66ef623d15a0a0a47f670742375de7 --- /dev/null +++ b/checkpoint/s3dis/s3dis-train-ppv2-xl-ngpus1-20250326-122107-nfhNFTpXV6atsVFWAXCDsk_ckpt_best.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37dddfdca696e615a0d83d8d48a543bcb0d652a7390dbc99293d370d7c3dbc11 +size 346916246 diff --git a/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR.log b/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR.log new file mode 100644 index 0000000000000000000000000000000000000000..a25b5d87199b2ea622b6ea779ad1ee6a3b30b077 --- /dev/null +++ b/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR.log @@ -0,0 +1,1858 @@ +[04/01 16:49:46] ScanObjectNNHardest INFO: dist_url: tcp://localhost:8888 +dist_backend: nccl +multiprocessing_distributed: False +ngpus_per_node: 1 +world_size: 1 +launcher: mp +local_rank: 0 +use_gpu: True +seed: 1234 +epoch: 0 +epochs: 250 +ignore_index: None +val_fn: validate +deterministic: False +sync_bn: False +criterion_args: + NAME: SmoothCrossEntropy + label_smoothing: 0.3 +use_mask: False +grad_norm_clip: 10 +layer_decay: 0 +step_per_update: 1 +start_epoch: 1 +sched_on_epoch: True +wandb: + use_wandb: False + project: PointNeXt-ScanObjectNN + tags: ['scanobjectnn', 'train', 'ppv2-s', 'ngpus1', 'seed1234'] + name: scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR +use_amp: False +use_voting: False +val_freq: 1 +resume: False +test: False +finetune: False +mode: train +logname: None +load_path: None +print_freq: 10 +save_freq: -1 +root_dir: log/scanobjectnn +pretrained_path: None +datatransforms: + train: ['PointsToTensor', 'PointCloudScaling', 'PointCloudCenterAndNormalize', 'PointCloudRotation'] + val: ['PointsToTensor', 'PointCloudCenterAndNormalize'] + vote: ['PointCloudRotation'] + kwargs: + scale: [0.9, 1.1] + angle: [0.0, 1.0, 0.0] + gravity_dim: 1 +feature_keys: pos +dataset: + common: + NAME: ScanObjectNNHardest + data_dir: ./data/ScanObjectNN/h5_files/main_split + train: + split: train + val: + split: val + num_points: 1024 +num_points: 1024 +num_classes: 15 +batch_size: 32 +val_batch_size: 64 +dataloader: + num_workers: 6 +lr: 0.002 +optimizer: + NAME: adamw + weight_decay: 0.05 +sched: cosine +warmup_epochs: 0 +min_lr: 0.0001 +t_max: 200 +log_dir: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR +model: + NAME: DpnCls + encoder_args: + NAME: PPV2Encoder + blocks: [1, 1, 1, 1, 1, 1] + strides: [1, 2, 2, 2, 2, 1] + width: 32 + in_channels: 4 + sa_layers: 2 + sa_use_res: True + radius: 0.15 + flag: 0 + radius_scaling: 1.5 + nsample: 32 + expansion: 4 + aggr_args: + feature_type: dp_fj + reduction: max + group_args: + NAME: ballquery + normalize_dp: True + conv_args: + order: conv-norm-act + act_args: + act: relu + norm_args: + norm: bn + cls_args: + NAME: DpnClsHead + num_classes: 15 + mlps: [512, 256] + norm_args: + norm: bn1d +rank: 0 +distributed: False +mp: False +task_name: scanobjectnn +exp_name: ppv2-s +opts: +run_name: scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR +run_dir: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR +exp_dir: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR +ckpt_dir: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint +log_path: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR.log +cfg_path: log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/cfg.yaml +[04/01 16:49:46] ScanObjectNNHardest INFO: radius: [[0.15], [0.15], [0.22499999999999998], [0.33749999999999997], [0.50625], [0.7593749999999999]], + nsample: [[32], [32], [32], [32], [32], [32]] +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.15 +nsample: 32 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.15 +nsample: 32 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.525 +nsample: 64 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.22499999999999998 +nsample: 32 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.7874999999999999 +nsample: 64 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.33749999999999997 +nsample: 32 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: 0.50625 +nsample: 32 +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: NAME: ballquery +normalize_dp: True +radius: None +nsample: None +return_idx: True +[04/01 16:49:46] ScanObjectNNHardest INFO: DpnCls( + (encoder): PPV2Encoder( + (grouper0): QueryAndGroup() + (encoder): Sequential( + (0): Sequential( + (0): SetAbstractionCls( + (convs): Sequential( + (0): Sequential( + (0): Conv1d(12, 32, kernel_size=(1,), stride=(1,)) + ) + ) + ) + ) + (1): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (preconv): Sequential( + (0): Conv1d(32, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (scorenet_global): Sequential( + (0): Conv1d(8, 1, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(1, 1, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 64, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 64, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(64, 64, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=8, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_v): Sequential( + (0): Conv1d(8, 64, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(64, 64, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + ) + ) + (2): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(8, 3, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(3, 3, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(64, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=24, out_features=8, bias=True) + (linear_k): Linear(in_features=8, out_features=8, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(11, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(11, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(11, 128, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 128, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(128, 128, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=24, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_v): Sequential( + (0): Conv1d(24, 128, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(128, 128, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + (key_grouper): QueryAndGroup() + ) + ) + (3): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(24, 9, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(9, 9, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(9, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(128, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=72, out_features=24, bias=True) + (linear_k): Linear(in_features=24, out_features=24, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(27, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(27, 256, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 256, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(256, 256, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=72, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_v): Sequential( + (0): Conv1d(72, 256, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(256, 256, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + ) + ) + (4): Sequential( + (0): SetAbstractionCls( + (skipconv): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,)) + ) + (act): ReLU(inplace=True) + (grouper): QueryAndGroup() + (scorenet_global): Sequential( + (0): Conv1d(72, 27, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(27, 27, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(27, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + ) + (preconv): Sequential( + (0): Conv1d(256, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (pt): PointTransformerLayer( + (linear_q): Linear(in_features=216, out_features=72, bias=True) + (linear_k): Linear(in_features=72, out_features=72, bias=True) + (linear_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + ) + (w): Sequential( + (0): BatchNorm2d(75, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (1): ReLU(inplace=True) + (2): Conv2d(75, 1, kernel_size=(1, 1), stride=(1, 1)) + (3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (4): ReLU(inplace=True) + ) + (v): Sequential( + (0): Conv2d(75, 512, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + (softmax): Softmax(dim=-1) + (conv_p): Sequential( + (0): Conv2d(3, 3, kernel_size=(1, 1), stride=(1, 1)) + (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv2d(3, 512, kernel_size=(1, 1), stride=(1, 1)) + ) + ) + (conv_finanal): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (selfattention): SelfAttention( + (linear_q): Linear(in_features=216, out_features=216, bias=True) + (linear_k): Linear(in_features=216, out_features=216, bias=True) + (linear_v): Sequential( + (0): Conv1d(216, 512, kernel_size=(1,), stride=(1,)) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + (3): Conv1d(512, 512, kernel_size=(1,), stride=(1,)) + (4): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (5): ReLU(inplace=True) + ) + (softmax): Softmax(dim=-1) + ) + ) + ) + (5): Sequential( + (0): SetAbstractionCls( + (grouper): GroupAll() + (preconv): Sequential( + (0): Conv1d(512, 512, kernel_size=(1,), stride=(1,), bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + ) + ) + ) + ) + (prediction): DpnClsHead( + (head): Sequential( + (0): Sequential( + (0): Linear(in_features=512, out_features=512, bias=False) + (1): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (1): Dropout(p=0.5, inplace=False) + (2): Sequential( + (0): Linear(in_features=512, out_features=256, bias=False) + (1): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) + (2): ReLU(inplace=True) + ) + (3): Dropout(p=0.5, inplace=False) + (4): Sequential( + (0): Linear(in_features=256, out_features=15, bias=True) + ) + ) + ) + (criterion): SmoothCrossEntropy() +) +[04/01 16:49:46] ScanObjectNNHardest INFO: Number of params: 2.5603 M +[04/01 16:49:46] ScanObjectNNHardest INFO: Param groups = { + "decay": { + "weight_decay": 0.05, + "params": [ + "encoder.encoder.0.0.convs.0.0.weight", + "encoder.encoder.1.0.skipconv.0.weight", + "encoder.encoder.1.0.preconv.0.weight", + "encoder.encoder.1.0.scorenet_global.0.weight", + "encoder.encoder.1.0.scorenet_global.3.weight", + "encoder.encoder.1.0.pt.linear_q.weight", + "encoder.encoder.1.0.pt.linear_k.weight", + "encoder.encoder.1.0.pt.linear_p.0.weight", + "encoder.encoder.1.0.pt.linear_p.3.weight", + "encoder.encoder.1.0.pt.w.2.weight", + "encoder.encoder.1.0.pt.v.0.weight", + "encoder.encoder.1.0.pt.v.3.weight", + "encoder.encoder.1.0.pt.conv_p.0.weight", + "encoder.encoder.1.0.pt.conv_p.3.weight", + "encoder.encoder.1.0.conv_finanal.0.weight", + "encoder.encoder.1.0.selfattention.linear_q.weight", + "encoder.encoder.1.0.selfattention.linear_k.weight", + "encoder.encoder.1.0.selfattention.linear_v.0.weight", + "encoder.encoder.1.0.selfattention.linear_v.3.weight", + "encoder.encoder.2.0.skipconv.0.weight", + "encoder.encoder.2.0.scorenet_global.0.weight", + "encoder.encoder.2.0.scorenet_global.3.weight", + "encoder.encoder.2.0.preconv.0.weight", + "encoder.encoder.2.0.pt.linear_q.weight", + "encoder.encoder.2.0.pt.linear_k.weight", + "encoder.encoder.2.0.pt.linear_p.0.weight", + "encoder.encoder.2.0.pt.linear_p.3.weight", + "encoder.encoder.2.0.pt.w.2.weight", + "encoder.encoder.2.0.pt.v.0.weight", + "encoder.encoder.2.0.pt.v.3.weight", + "encoder.encoder.2.0.pt.conv_p.0.weight", + "encoder.encoder.2.0.pt.conv_p.3.weight", + "encoder.encoder.2.0.conv_finanal.0.weight", + "encoder.encoder.2.0.selfattention.linear_q.weight", + "encoder.encoder.2.0.selfattention.linear_k.weight", + "encoder.encoder.2.0.selfattention.linear_v.0.weight", + "encoder.encoder.2.0.selfattention.linear_v.3.weight", + "encoder.encoder.3.0.skipconv.0.weight", + "encoder.encoder.3.0.scorenet_global.0.weight", + "encoder.encoder.3.0.scorenet_global.3.weight", + "encoder.encoder.3.0.preconv.0.weight", + "encoder.encoder.3.0.pt.linear_q.weight", + "encoder.encoder.3.0.pt.linear_k.weight", + "encoder.encoder.3.0.pt.linear_p.0.weight", + "encoder.encoder.3.0.pt.linear_p.3.weight", + "encoder.encoder.3.0.pt.w.2.weight", + "encoder.encoder.3.0.pt.v.0.weight", + "encoder.encoder.3.0.pt.v.3.weight", + "encoder.encoder.3.0.pt.conv_p.0.weight", + "encoder.encoder.3.0.pt.conv_p.3.weight", + "encoder.encoder.3.0.conv_finanal.0.weight", + "encoder.encoder.3.0.selfattention.linear_q.weight", + "encoder.encoder.3.0.selfattention.linear_k.weight", + "encoder.encoder.3.0.selfattention.linear_v.0.weight", + "encoder.encoder.3.0.selfattention.linear_v.3.weight", + "encoder.encoder.4.0.skipconv.0.weight", + "encoder.encoder.4.0.scorenet_global.0.weight", + "encoder.encoder.4.0.scorenet_global.3.weight", + "encoder.encoder.4.0.preconv.0.weight", + "encoder.encoder.4.0.pt.linear_q.weight", + "encoder.encoder.4.0.pt.linear_k.weight", + "encoder.encoder.4.0.pt.linear_p.0.weight", + "encoder.encoder.4.0.pt.linear_p.3.weight", + "encoder.encoder.4.0.pt.w.2.weight", + "encoder.encoder.4.0.pt.v.0.weight", + "encoder.encoder.4.0.pt.v.3.weight", + "encoder.encoder.4.0.pt.conv_p.0.weight", + "encoder.encoder.4.0.pt.conv_p.3.weight", + "encoder.encoder.4.0.conv_finanal.0.weight", + "encoder.encoder.4.0.selfattention.linear_q.weight", + "encoder.encoder.4.0.selfattention.linear_k.weight", + "encoder.encoder.4.0.selfattention.linear_v.0.weight", + "encoder.encoder.4.0.selfattention.linear_v.3.weight", + "encoder.encoder.5.0.preconv.0.weight", + "prediction.head.0.0.weight", + "prediction.head.2.0.weight", + "prediction.head.4.0.weight" + ], + "lr_scale": 1.0 + }, + "no_decay": { + "weight_decay": 0.0, + "params": [ + "encoder.encoder.0.0.convs.0.0.bias", + "encoder.encoder.1.0.beta", + "encoder.encoder.1.0.skipconv.0.bias", + "encoder.encoder.1.0.preconv.1.weight", + "encoder.encoder.1.0.preconv.1.bias", + "encoder.encoder.1.0.scorenet_global.0.bias", + "encoder.encoder.1.0.scorenet_global.1.weight", + "encoder.encoder.1.0.scorenet_global.1.bias", + "encoder.encoder.1.0.scorenet_global.3.bias", + "encoder.encoder.1.0.scorenet_global.4.weight", + "encoder.encoder.1.0.scorenet_global.4.bias", + "encoder.encoder.1.0.pt.linear_q.bias", + "encoder.encoder.1.0.pt.linear_k.bias", + "encoder.encoder.1.0.pt.linear_p.0.bias", + "encoder.encoder.1.0.pt.linear_p.1.weight", + "encoder.encoder.1.0.pt.linear_p.1.bias", + "encoder.encoder.1.0.pt.linear_p.3.bias", + "encoder.encoder.1.0.pt.w.0.weight", + "encoder.encoder.1.0.pt.w.0.bias", + "encoder.encoder.1.0.pt.w.2.bias", + "encoder.encoder.1.0.pt.w.3.weight", + "encoder.encoder.1.0.pt.w.3.bias", + "encoder.encoder.1.0.pt.v.0.bias", + "encoder.encoder.1.0.pt.v.1.weight", + "encoder.encoder.1.0.pt.v.1.bias", + "encoder.encoder.1.0.pt.v.3.bias", + "encoder.encoder.1.0.pt.conv_p.0.bias", + "encoder.encoder.1.0.pt.conv_p.1.weight", + "encoder.encoder.1.0.pt.conv_p.1.bias", + "encoder.encoder.1.0.pt.conv_p.3.bias", + "encoder.encoder.1.0.conv_finanal.1.weight", + "encoder.encoder.1.0.conv_finanal.1.bias", + "encoder.encoder.1.0.selfattention.linear_q.bias", + "encoder.encoder.1.0.selfattention.linear_k.bias", + "encoder.encoder.1.0.selfattention.linear_v.0.bias", + "encoder.encoder.1.0.selfattention.linear_v.1.weight", + "encoder.encoder.1.0.selfattention.linear_v.1.bias", + "encoder.encoder.1.0.selfattention.linear_v.3.bias", + "encoder.encoder.1.0.selfattention.linear_v.4.weight", + "encoder.encoder.1.0.selfattention.linear_v.4.bias", + "encoder.encoder.2.0.beta", + "encoder.encoder.2.0.skipconv.0.bias", + "encoder.encoder.2.0.scorenet_global.0.bias", + "encoder.encoder.2.0.scorenet_global.1.weight", + "encoder.encoder.2.0.scorenet_global.1.bias", + "encoder.encoder.2.0.scorenet_global.3.bias", + "encoder.encoder.2.0.scorenet_global.4.weight", + "encoder.encoder.2.0.scorenet_global.4.bias", + "encoder.encoder.2.0.preconv.1.weight", + "encoder.encoder.2.0.preconv.1.bias", + "encoder.encoder.2.0.pt.linear_q.bias", + "encoder.encoder.2.0.pt.linear_k.bias", + "encoder.encoder.2.0.pt.linear_p.0.bias", + "encoder.encoder.2.0.pt.linear_p.1.weight", + "encoder.encoder.2.0.pt.linear_p.1.bias", + "encoder.encoder.2.0.pt.linear_p.3.bias", + "encoder.encoder.2.0.pt.w.0.weight", + "encoder.encoder.2.0.pt.w.0.bias", + "encoder.encoder.2.0.pt.w.2.bias", + "encoder.encoder.2.0.pt.w.3.weight", + "encoder.encoder.2.0.pt.w.3.bias", + "encoder.encoder.2.0.pt.v.0.bias", + "encoder.encoder.2.0.pt.v.1.weight", + "encoder.encoder.2.0.pt.v.1.bias", + "encoder.encoder.2.0.pt.v.3.bias", + "encoder.encoder.2.0.pt.conv_p.0.bias", + "encoder.encoder.2.0.pt.conv_p.1.weight", + "encoder.encoder.2.0.pt.conv_p.1.bias", + "encoder.encoder.2.0.pt.conv_p.3.bias", + "encoder.encoder.2.0.conv_finanal.1.weight", + "encoder.encoder.2.0.conv_finanal.1.bias", + "encoder.encoder.2.0.selfattention.linear_q.bias", + "encoder.encoder.2.0.selfattention.linear_k.bias", + "encoder.encoder.2.0.selfattention.linear_v.0.bias", + "encoder.encoder.2.0.selfattention.linear_v.1.weight", + "encoder.encoder.2.0.selfattention.linear_v.1.bias", + "encoder.encoder.2.0.selfattention.linear_v.3.bias", + "encoder.encoder.2.0.selfattention.linear_v.4.weight", + "encoder.encoder.2.0.selfattention.linear_v.4.bias", + "encoder.encoder.3.0.beta", + "encoder.encoder.3.0.skipconv.0.bias", + "encoder.encoder.3.0.scorenet_global.0.bias", + "encoder.encoder.3.0.scorenet_global.1.weight", + "encoder.encoder.3.0.scorenet_global.1.bias", + "encoder.encoder.3.0.scorenet_global.3.bias", + "encoder.encoder.3.0.scorenet_global.4.weight", + "encoder.encoder.3.0.scorenet_global.4.bias", + "encoder.encoder.3.0.preconv.1.weight", + "encoder.encoder.3.0.preconv.1.bias", + "encoder.encoder.3.0.pt.linear_q.bias", + "encoder.encoder.3.0.pt.linear_k.bias", + "encoder.encoder.3.0.pt.linear_p.0.bias", + "encoder.encoder.3.0.pt.linear_p.1.weight", + "encoder.encoder.3.0.pt.linear_p.1.bias", + "encoder.encoder.3.0.pt.linear_p.3.bias", + "encoder.encoder.3.0.pt.w.0.weight", + "encoder.encoder.3.0.pt.w.0.bias", + "encoder.encoder.3.0.pt.w.2.bias", + "encoder.encoder.3.0.pt.w.3.weight", + "encoder.encoder.3.0.pt.w.3.bias", + "encoder.encoder.3.0.pt.v.0.bias", + "encoder.encoder.3.0.pt.v.1.weight", + "encoder.encoder.3.0.pt.v.1.bias", + "encoder.encoder.3.0.pt.v.3.bias", + "encoder.encoder.3.0.pt.conv_p.0.bias", + "encoder.encoder.3.0.pt.conv_p.1.weight", + "encoder.encoder.3.0.pt.conv_p.1.bias", + "encoder.encoder.3.0.pt.conv_p.3.bias", + "encoder.encoder.3.0.conv_finanal.1.weight", + "encoder.encoder.3.0.conv_finanal.1.bias", + "encoder.encoder.3.0.selfattention.linear_q.bias", + "encoder.encoder.3.0.selfattention.linear_k.bias", + "encoder.encoder.3.0.selfattention.linear_v.0.bias", + "encoder.encoder.3.0.selfattention.linear_v.1.weight", + "encoder.encoder.3.0.selfattention.linear_v.1.bias", + "encoder.encoder.3.0.selfattention.linear_v.3.bias", + "encoder.encoder.3.0.selfattention.linear_v.4.weight", + "encoder.encoder.3.0.selfattention.linear_v.4.bias", + "encoder.encoder.4.0.beta", + "encoder.encoder.4.0.skipconv.0.bias", + "encoder.encoder.4.0.scorenet_global.0.bias", + "encoder.encoder.4.0.scorenet_global.1.weight", + "encoder.encoder.4.0.scorenet_global.1.bias", + "encoder.encoder.4.0.scorenet_global.3.bias", + "encoder.encoder.4.0.scorenet_global.4.weight", + "encoder.encoder.4.0.scorenet_global.4.bias", + "encoder.encoder.4.0.preconv.1.weight", + "encoder.encoder.4.0.preconv.1.bias", + "encoder.encoder.4.0.pt.linear_q.bias", + "encoder.encoder.4.0.pt.linear_k.bias", + "encoder.encoder.4.0.pt.linear_p.0.bias", + "encoder.encoder.4.0.pt.linear_p.1.weight", + "encoder.encoder.4.0.pt.linear_p.1.bias", + "encoder.encoder.4.0.pt.linear_p.3.bias", + "encoder.encoder.4.0.pt.w.0.weight", + "encoder.encoder.4.0.pt.w.0.bias", + "encoder.encoder.4.0.pt.w.2.bias", + "encoder.encoder.4.0.pt.w.3.weight", + "encoder.encoder.4.0.pt.w.3.bias", + "encoder.encoder.4.0.pt.v.0.bias", + "encoder.encoder.4.0.pt.v.1.weight", + "encoder.encoder.4.0.pt.v.1.bias", + "encoder.encoder.4.0.pt.v.3.bias", + "encoder.encoder.4.0.pt.conv_p.0.bias", + "encoder.encoder.4.0.pt.conv_p.1.weight", + "encoder.encoder.4.0.pt.conv_p.1.bias", + "encoder.encoder.4.0.pt.conv_p.3.bias", + "encoder.encoder.4.0.conv_finanal.1.weight", + "encoder.encoder.4.0.conv_finanal.1.bias", + "encoder.encoder.4.0.selfattention.linear_q.bias", + "encoder.encoder.4.0.selfattention.linear_k.bias", + "encoder.encoder.4.0.selfattention.linear_v.0.bias", + "encoder.encoder.4.0.selfattention.linear_v.1.weight", + "encoder.encoder.4.0.selfattention.linear_v.1.bias", + "encoder.encoder.4.0.selfattention.linear_v.3.bias", + "encoder.encoder.4.0.selfattention.linear_v.4.weight", + "encoder.encoder.4.0.selfattention.linear_v.4.bias", + "encoder.encoder.5.0.preconv.1.weight", + "encoder.encoder.5.0.preconv.1.bias", + "prediction.head.0.1.weight", + "prediction.head.0.1.bias", + "prediction.head.2.1.weight", + "prediction.head.2.1.bias", + "prediction.head.4.0.bias" + ], + "lr_scale": 1.0 + } +} +[04/01 16:49:48] ScanObjectNNHardest INFO: Successfully load ScanObjectNN val size: (2882, 1024, 3), num_classes: 15 +[04/01 16:49:48] ScanObjectNNHardest INFO: length of validation dataset: 2882 +[04/01 16:49:49] ScanObjectNNHardest INFO: Successfully load ScanObjectNN val size: (2882, 1024, 3), num_classes: 15 +[04/01 16:49:49] ScanObjectNNHardest INFO: number of classes of the dataset: 15, number of points sampled from dataset: 1024, number of points as model input: 1024 +[04/01 16:49:49] ScanObjectNNHardest INFO: Training from scratch +[04/01 16:49:52] ScanObjectNNHardest INFO: Successfully load ScanObjectNN train size: (11416, 2048, 3), num_classes: 15 +[04/01 16:49:52] ScanObjectNNHardest INFO: length of training dataset: 11416 +[04/01 16:50:53] ScanObjectNNHardest INFO: Find a better ckpt @E1 +[04/01 16:50:53] ScanObjectNNHardest INFO: +Classes Acc +bag : 0.00% +bin : 16.58% +box : 2.26% +cabinet : 35.48% +chair : 85.90% +desk : 17.33% +display : 64.71% +door : 90.00% +shelf : 58.09% +table : 10.37% +bed : 96.36% +pillow : 16.19% +sink : 3.33% +sofa : 18.10% +toilet : 0.00% +E@1 OA: 41.05 mAcc: 34.31 + +[04/01 16:50:53] ScanObjectNNHardest INFO: Epoch 1 LR 0.002000 train_oa 34.28, val_oa 41.05, best val oa 41.05 +[04/01 16:50:53] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:51:46] ScanObjectNNHardest INFO: Find a better ckpt @E2 +[04/01 16:51:46] ScanObjectNNHardest INFO: +Classes Acc +bag : 0.00% +bin : 55.78% +box : 0.75% +cabinet : 54.30% +chair : 62.31% +desk : 23.33% +display : 64.22% +door : 87.14% +shelf : 81.74% +table : 39.63% +bed : 85.45% +pillow : 1.90% +sink : 24.17% +sofa : 82.86% +toilet : 1.18% +E@2 OA: 52.39 mAcc: 44.32 + +[04/01 16:51:46] ScanObjectNNHardest INFO: Epoch 2 LR 0.002000 train_oa 54.96, val_oa 52.39, best val oa 52.39 +[04/01 16:51:46] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:52:40] ScanObjectNNHardest INFO: Find a better ckpt @E3 +[04/01 16:52:40] ScanObjectNNHardest INFO: +Classes Acc +bag : 39.76% +bin : 67.84% +box : 6.02% +cabinet : 41.40% +chair : 88.46% +desk : 10.67% +display : 61.76% +door : 93.33% +shelf : 75.93% +table : 64.07% +bed : 78.18% +pillow : 56.19% +sink : 16.67% +sofa : 81.90% +toilet : 8.24% +E@3 OA: 59.44 mAcc: 52.69 + +[04/01 16:52:40] ScanObjectNNHardest INFO: Epoch 3 LR 0.002000 train_oa 60.34, val_oa 59.44, best val oa 59.44 +[04/01 16:52:40] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:53:40] ScanObjectNNHardest INFO: Find a better ckpt @E4 +[04/01 16:53:40] ScanObjectNNHardest INFO: +Classes Acc +bag : 9.64% +bin : 84.92% +box : 24.06% +cabinet : 50.81% +chair : 77.95% +desk : 72.67% +display : 77.45% +door : 95.71% +shelf : 72.20% +table : 45.93% +bed : 80.91% +pillow : 52.38% +sink : 55.83% +sofa : 66.67% +toilet : 44.71% +E@4 OA: 64.43 mAcc: 60.79 + +[04/01 16:53:40] ScanObjectNNHardest INFO: Epoch 4 LR 0.001999 train_oa 65.58, val_oa 64.43, best val oa 64.43 +[04/01 16:53:41] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:54:36] ScanObjectNNHardest INFO: Find a better ckpt @E5 +[04/01 16:54:36] ScanObjectNNHardest INFO: +Classes Acc +bag : 45.78% +bin : 54.27% +box : 36.09% +cabinet : 72.58% +chair : 94.36% +desk : 58.67% +display : 78.43% +door : 77.62% +shelf : 80.91% +table : 48.52% +bed : 48.18% +pillow : 57.14% +sink : 39.17% +sofa : 91.43% +toilet : 35.29% +E@5 OA: 67.70 mAcc: 61.23 + +[04/01 16:54:36] ScanObjectNNHardest INFO: Epoch 5 LR 0.001998 train_oa 69.57, val_oa 67.70, best val oa 67.70 +[04/01 16:54:36] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:55:29] ScanObjectNNHardest INFO: Epoch 6 LR 0.001997 train_oa 71.75, val_oa 63.60, best val oa 67.70 +[04/01 16:56:27] ScanObjectNNHardest INFO: Find a better ckpt @E7 +[04/01 16:56:27] ScanObjectNNHardest INFO: +Classes Acc +bag : 38.55% +bin : 83.92% +box : 27.07% +cabinet : 59.95% +chair : 87.44% +desk : 74.00% +display : 80.88% +door : 97.14% +shelf : 81.33% +table : 57.78% +bed : 70.91% +pillow : 70.48% +sink : 61.67% +sofa : 92.86% +toilet : 38.82% +E@7 OA: 72.35 mAcc: 68.19 + +[04/01 16:56:27] ScanObjectNNHardest INFO: Epoch 7 LR 0.001996 train_oa 73.81, val_oa 72.35, best val oa 72.35 +[04/01 16:56:27] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 16:57:22] ScanObjectNNHardest INFO: Epoch 8 LR 0.001994 train_oa 75.90, val_oa 71.65, best val oa 72.35 +[04/01 16:58:18] ScanObjectNNHardest INFO: Epoch 9 LR 0.001993 train_oa 77.22, val_oa 72.00, best val oa 72.35 +[04/01 16:59:16] ScanObjectNNHardest INFO: Find a better ckpt @E10 +[04/01 16:59:16] ScanObjectNNHardest INFO: +Classes Acc +bag : 43.37% +bin : 80.40% +box : 45.11% +cabinet : 76.08% +chair : 79.23% +desk : 78.00% +display : 78.43% +door : 93.81% +shelf : 82.57% +table : 51.11% +bed : 76.36% +pillow : 68.57% +sink : 61.67% +sofa : 95.71% +toilet : 65.88% +E@10 OA: 74.46 mAcc: 71.75 + +[04/01 16:59:16] ScanObjectNNHardest INFO: Epoch 10 LR 0.001991 train_oa 78.49, val_oa 74.46, best val oa 74.46 +[04/01 16:59:16] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:00:11] ScanObjectNNHardest INFO: Find a better ckpt @E11 +[04/01 17:00:11] ScanObjectNNHardest INFO: +Classes Acc +bag : 33.73% +bin : 79.90% +box : 36.09% +cabinet : 76.88% +chair : 94.36% +desk : 72.67% +display : 84.31% +door : 93.33% +shelf : 80.50% +table : 60.37% +bed : 76.36% +pillow : 71.43% +sink : 66.67% +sofa : 93.81% +toilet : 76.47% +E@11 OA: 77.17 mAcc: 73.13 + +[04/01 17:00:11] ScanObjectNNHardest INFO: Epoch 11 LR 0.001988 train_oa 79.71, val_oa 77.17, best val oa 77.17 +[04/01 17:00:11] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:01:06] ScanObjectNNHardest INFO: Epoch 12 LR 0.001986 train_oa 80.06, val_oa 76.16, best val oa 77.17 +[04/01 17:02:03] ScanObjectNNHardest INFO: Epoch 13 LR 0.001983 train_oa 80.84, val_oa 76.72, best val oa 77.17 +[04/01 17:03:05] ScanObjectNNHardest INFO: Find a better ckpt @E14 +[04/01 17:03:05] ScanObjectNNHardest INFO: +Classes Acc +bag : 48.19% +bin : 81.91% +box : 57.89% +cabinet : 83.33% +chair : 78.97% +desk : 65.33% +display : 76.47% +door : 91.43% +shelf : 85.06% +table : 80.37% +bed : 83.64% +pillow : 70.48% +sink : 50.83% +sofa : 88.57% +toilet : 75.29% +E@14 OA: 77.83 mAcc: 74.52 + +[04/01 17:03:05] ScanObjectNNHardest INFO: Epoch 14 LR 0.001980 train_oa 82.26, val_oa 77.83, best val oa 77.83 +[04/01 17:03:05] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:04:03] ScanObjectNNHardest INFO: Epoch 15 LR 0.001977 train_oa 83.02, val_oa 76.79, best val oa 77.83 +[04/01 17:04:59] ScanObjectNNHardest INFO: Epoch 16 LR 0.001974 train_oa 83.96, val_oa 76.06, best val oa 77.83 +[04/01 17:05:58] ScanObjectNNHardest INFO: Find a better ckpt @E17 +[04/01 17:05:58] ScanObjectNNHardest INFO: +Classes Acc +bag : 49.40% +bin : 89.45% +box : 54.14% +cabinet : 76.88% +chair : 94.62% +desk : 57.33% +display : 75.49% +door : 85.71% +shelf : 83.82% +table : 72.22% +bed : 88.18% +pillow : 70.48% +sink : 69.17% +sofa : 85.24% +toilet : 61.18% +E@17 OA: 78.00 mAcc: 74.22 + +[04/01 17:05:58] ScanObjectNNHardest INFO: Epoch 17 LR 0.001970 train_oa 84.21, val_oa 78.00, best val oa 78.00 +[04/01 17:05:58] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:06:55] ScanObjectNNHardest INFO: Find a better ckpt @E18 +[04/01 17:06:55] ScanObjectNNHardest INFO: +Classes Acc +bag : 33.73% +bin : 85.43% +box : 46.62% +cabinet : 80.91% +chair : 93.85% +desk : 78.00% +display : 82.35% +door : 91.43% +shelf : 88.80% +table : 61.85% +bed : 74.55% +pillow : 84.76% +sink : 65.83% +sofa : 94.29% +toilet : 70.59% +E@18 OA: 79.56 mAcc: 75.53 + +[04/01 17:06:55] ScanObjectNNHardest INFO: Epoch 18 LR 0.001966 train_oa 85.17, val_oa 79.56, best val oa 79.56 +[04/01 17:06:55] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:07:51] ScanObjectNNHardest INFO: Epoch 19 LR 0.001962 train_oa 85.67, val_oa 74.77, best val oa 79.56 +[04/01 17:08:45] ScanObjectNNHardest INFO: Find a better ckpt @E20 +[04/01 17:08:45] ScanObjectNNHardest INFO: +Classes Acc +bag : 43.37% +bin : 76.38% +box : 59.40% +cabinet : 80.11% +chair : 91.03% +desk : 76.67% +display : 86.27% +door : 92.38% +shelf : 83.40% +table : 67.04% +bed : 74.55% +pillow : 81.90% +sink : 72.50% +sofa : 95.71% +toilet : 78.82% +E@20 OA: 80.15 mAcc: 77.30 + +[04/01 17:08:45] ScanObjectNNHardest INFO: Epoch 20 LR 0.001958 train_oa 86.52, val_oa 80.15, best val oa 80.15 +[04/01 17:08:45] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:09:39] ScanObjectNNHardest INFO: Find a better ckpt @E21 +[04/01 17:09:39] ScanObjectNNHardest INFO: +Classes Acc +bag : 65.06% +bin : 77.89% +box : 68.42% +cabinet : 83.60% +chair : 89.23% +desk : 74.00% +display : 73.04% +door : 96.19% +shelf : 80.91% +table : 74.81% +bed : 75.45% +pillow : 84.76% +sink : 77.50% +sofa : 94.76% +toilet : 74.12% +E@21 OA: 81.37 mAcc: 79.32 + +[04/01 17:09:39] ScanObjectNNHardest INFO: Epoch 21 LR 0.001954 train_oa 87.10, val_oa 81.37, best val oa 81.37 +[04/01 17:09:39] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:10:40] ScanObjectNNHardest INFO: Find a better ckpt @E22 +[04/01 17:10:40] ScanObjectNNHardest INFO: +Classes Acc +bag : 39.76% +bin : 79.90% +box : 78.95% +cabinet : 79.03% +chair : 89.49% +desk : 82.00% +display : 92.16% +door : 97.14% +shelf : 90.87% +table : 62.96% +bed : 84.55% +pillow : 78.10% +sink : 62.50% +sofa : 93.81% +toilet : 70.59% +E@22 OA: 81.58 mAcc: 78.79 + +[04/01 17:10:40] ScanObjectNNHardest INFO: Epoch 22 LR 0.001949 train_oa 87.21, val_oa 81.58, best val oa 81.58 +[04/01 17:10:40] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:11:37] ScanObjectNNHardest INFO: Epoch 23 LR 0.001944 train_oa 87.70, val_oa 80.57, best val oa 81.58 +[04/01 17:12:34] ScanObjectNNHardest INFO: Epoch 24 LR 0.001939 train_oa 88.47, val_oa 80.95, best val oa 81.58 +[04/01 17:13:32] ScanObjectNNHardest INFO: Epoch 25 LR 0.001933 train_oa 88.60, val_oa 80.40, best val oa 81.58 +[04/01 17:14:26] ScanObjectNNHardest INFO: Epoch 26 LR 0.001928 train_oa 89.15, val_oa 81.05, best val oa 81.58 +[04/01 17:15:21] ScanObjectNNHardest INFO: Epoch 27 LR 0.001922 train_oa 89.88, val_oa 80.01, best val oa 81.58 +[04/01 17:16:12] ScanObjectNNHardest INFO: Epoch 28 LR 0.001916 train_oa 89.87, val_oa 79.15, best val oa 81.58 +[04/01 17:17:11] ScanObjectNNHardest INFO: Epoch 29 LR 0.001910 train_oa 90.19, val_oa 80.60, best val oa 81.58 +[04/01 17:18:10] ScanObjectNNHardest INFO: Epoch 30 LR 0.001903 train_oa 90.85, val_oa 80.92, best val oa 81.58 +[04/01 17:19:05] ScanObjectNNHardest INFO: Epoch 31 LR 0.001896 train_oa 91.25, val_oa 81.47, best val oa 81.58 +[04/01 17:20:04] ScanObjectNNHardest INFO: Find a better ckpt @E32 +[04/01 17:20:04] ScanObjectNNHardest INFO: +Classes Acc +bag : 55.42% +bin : 85.93% +box : 39.10% +cabinet : 86.83% +chair : 95.90% +desk : 60.67% +display : 84.80% +door : 96.67% +shelf : 89.63% +table : 72.59% +bed : 73.64% +pillow : 86.67% +sink : 65.83% +sofa : 90.95% +toilet : 92.94% +E@32 OA: 82.10 mAcc: 78.50 + +[04/01 17:20:04] ScanObjectNNHardest INFO: Epoch 32 LR 0.001890 train_oa 91.56, val_oa 82.10, best val oa 82.10 +[04/01 17:20:05] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:21:07] ScanObjectNNHardest INFO: Epoch 33 LR 0.001882 train_oa 91.90, val_oa 81.78, best val oa 82.10 +[04/01 17:22:03] ScanObjectNNHardest INFO: Find a better ckpt @E34 +[04/01 17:22:03] ScanObjectNNHardest INFO: +Classes Acc +bag : 54.22% +bin : 83.92% +box : 64.66% +cabinet : 85.75% +chair : 86.92% +desk : 70.67% +display : 85.29% +door : 95.71% +shelf : 88.80% +table : 82.59% +bed : 83.64% +pillow : 86.67% +sink : 59.17% +sofa : 92.86% +toilet : 75.29% +E@34 OA: 82.82 mAcc: 79.74 + +[04/01 17:22:03] ScanObjectNNHardest INFO: Epoch 34 LR 0.001875 train_oa 92.25, val_oa 82.82, best val oa 82.82 +[04/01 17:22:03] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:23:02] ScanObjectNNHardest INFO: Epoch 35 LR 0.001868 train_oa 92.51, val_oa 82.34, best val oa 82.82 +[04/01 17:24:01] ScanObjectNNHardest INFO: Epoch 36 LR 0.001860 train_oa 92.84, val_oa 80.78, best val oa 82.82 +[04/01 17:24:59] ScanObjectNNHardest INFO: Epoch 37 LR 0.001852 train_oa 93.11, val_oa 82.10, best val oa 82.82 +[04/01 17:25:56] ScanObjectNNHardest INFO: Find a better ckpt @E38 +[04/01 17:25:56] ScanObjectNNHardest INFO: +Classes Acc +bag : 66.27% +bin : 80.40% +box : 66.17% +cabinet : 84.41% +chair : 93.85% +desk : 80.67% +display : 76.96% +door : 98.57% +shelf : 87.97% +table : 63.70% +bed : 88.18% +pillow : 84.76% +sink : 73.33% +sofa : 91.43% +toilet : 84.71% +E@38 OA: 82.93 mAcc: 81.42 + +[04/01 17:25:56] ScanObjectNNHardest INFO: Epoch 38 LR 0.001844 train_oa 93.75, val_oa 82.93, best val oa 82.93 +[04/01 17:25:57] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:26:54] ScanObjectNNHardest INFO: Epoch 39 LR 0.001836 train_oa 93.66, val_oa 82.62, best val oa 82.93 +[04/01 17:27:49] ScanObjectNNHardest INFO: Epoch 40 LR 0.001827 train_oa 93.26, val_oa 82.55, best val oa 82.93 +[04/01 17:28:44] ScanObjectNNHardest INFO: Epoch 41 LR 0.001819 train_oa 94.15, val_oa 82.72, best val oa 82.93 +[04/01 17:29:48] ScanObjectNNHardest INFO: Epoch 42 LR 0.001810 train_oa 93.75, val_oa 81.16, best val oa 82.93 +[04/01 17:30:44] ScanObjectNNHardest INFO: Epoch 43 LR 0.001801 train_oa 94.18, val_oa 82.34, best val oa 82.93 +[04/01 17:31:41] ScanObjectNNHardest INFO: Epoch 44 LR 0.001791 train_oa 94.46, val_oa 81.12, best val oa 82.93 +[04/01 17:32:37] ScanObjectNNHardest INFO: Find a better ckpt @E45 +[04/01 17:32:37] ScanObjectNNHardest INFO: +Classes Acc +bag : 60.24% +bin : 87.44% +box : 56.39% +cabinet : 87.63% +chair : 85.38% +desk : 67.33% +display : 85.29% +door : 91.43% +shelf : 89.21% +table : 82.96% +bed : 77.27% +pillow : 83.81% +sink : 68.33% +sofa : 96.67% +toilet : 89.41% +E@45 OA: 83.21 mAcc: 80.59 + +[04/01 17:32:37] ScanObjectNNHardest INFO: Epoch 45 LR 0.001782 train_oa 94.85, val_oa 83.21, best val oa 83.21 +[04/01 17:32:37] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:33:32] ScanObjectNNHardest INFO: Find a better ckpt @E46 +[04/01 17:33:32] ScanObjectNNHardest INFO: +Classes Acc +bag : 65.06% +bin : 82.91% +box : 72.18% +cabinet : 81.99% +chair : 94.62% +desk : 88.00% +display : 84.31% +door : 95.71% +shelf : 89.63% +table : 62.96% +bed : 75.45% +pillow : 81.90% +sink : 71.67% +sofa : 97.14% +toilet : 78.82% +E@46 OA: 83.48 mAcc: 81.49 + +[04/01 17:33:32] ScanObjectNNHardest INFO: Epoch 46 LR 0.001772 train_oa 94.63, val_oa 83.48, best val oa 83.48 +[04/01 17:33:32] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:34:33] ScanObjectNNHardest INFO: Epoch 47 LR 0.001763 train_oa 95.06, val_oa 82.06, best val oa 83.48 +[04/01 17:35:27] ScanObjectNNHardest INFO: Epoch 48 LR 0.001753 train_oa 95.17, val_oa 82.96, best val oa 83.48 +[04/01 17:36:23] ScanObjectNNHardest INFO: Epoch 49 LR 0.001743 train_oa 95.73, val_oa 82.10, best val oa 83.48 +[04/01 17:37:18] ScanObjectNNHardest INFO: Find a better ckpt @E50 +[04/01 17:37:18] ScanObjectNNHardest INFO: +Classes Acc +bag : 53.01% +bin : 78.39% +box : 75.19% +cabinet : 80.11% +chair : 91.28% +desk : 82.00% +display : 88.24% +door : 94.29% +shelf : 88.80% +table : 75.19% +bed : 79.09% +pillow : 84.76% +sink : 74.17% +sofa : 95.24% +toilet : 83.53% +E@50 OA: 83.55 mAcc: 81.55 + +[04/01 17:37:18] ScanObjectNNHardest INFO: Epoch 50 LR 0.001732 train_oa 95.29, val_oa 83.55, best val oa 83.55 +[04/01 17:37:18] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:38:14] ScanObjectNNHardest INFO: Epoch 51 LR 0.001722 train_oa 95.41, val_oa 83.48, best val oa 83.55 +[04/01 17:39:11] ScanObjectNNHardest INFO: Epoch 52 LR 0.001711 train_oa 95.67, val_oa 80.50, best val oa 83.55 +[04/01 17:40:05] ScanObjectNNHardest INFO: Epoch 53 LR 0.001700 train_oa 95.51, val_oa 82.82, best val oa 83.55 +[04/01 17:41:00] ScanObjectNNHardest INFO: Epoch 54 LR 0.001689 train_oa 95.93, val_oa 83.03, best val oa 83.55 +[04/01 17:41:57] ScanObjectNNHardest INFO: Epoch 55 LR 0.001678 train_oa 96.25, val_oa 81.30, best val oa 83.55 +[04/01 17:42:53] ScanObjectNNHardest INFO: Epoch 56 LR 0.001667 train_oa 96.10, val_oa 82.34, best val oa 83.55 +[04/01 17:43:48] ScanObjectNNHardest INFO: Epoch 57 LR 0.001656 train_oa 96.38, val_oa 82.69, best val oa 83.55 +[04/01 17:44:47] ScanObjectNNHardest INFO: Epoch 58 LR 0.001644 train_oa 96.72, val_oa 82.65, best val oa 83.55 +[04/01 17:45:43] ScanObjectNNHardest INFO: Epoch 59 LR 0.001632 train_oa 96.44, val_oa 82.48, best val oa 83.55 +[04/01 17:46:38] ScanObjectNNHardest INFO: Find a better ckpt @E60 +[04/01 17:46:38] ScanObjectNNHardest INFO: +Classes Acc +bag : 61.45% +bin : 87.44% +box : 54.14% +cabinet : 88.17% +chair : 95.13% +desk : 79.33% +display : 86.27% +door : 87.62% +shelf : 83.40% +table : 75.93% +bed : 83.64% +pillow : 84.76% +sink : 75.83% +sofa : 87.14% +toilet : 92.94% +E@60 OA: 83.80 mAcc: 81.55 + +[04/01 17:46:38] ScanObjectNNHardest INFO: Epoch 60 LR 0.001620 train_oa 97.09, val_oa 83.80, best val oa 83.80 +[04/01 17:46:38] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:47:37] ScanObjectNNHardest INFO: Epoch 61 LR 0.001608 train_oa 96.50, val_oa 82.89, best val oa 83.80 +[04/01 17:48:35] ScanObjectNNHardest INFO: Epoch 62 LR 0.001596 train_oa 96.48, val_oa 83.31, best val oa 83.80 +[04/01 17:49:34] ScanObjectNNHardest INFO: Epoch 63 LR 0.001584 train_oa 96.39, val_oa 83.59, best val oa 83.80 +[04/01 17:50:29] ScanObjectNNHardest INFO: Epoch 64 LR 0.001572 train_oa 96.74, val_oa 83.21, best val oa 83.80 +[04/01 17:51:24] ScanObjectNNHardest INFO: Find a better ckpt @E65 +[04/01 17:51:24] ScanObjectNNHardest INFO: +Classes Acc +bag : 72.29% +bin : 87.44% +box : 42.11% +cabinet : 84.14% +chair : 89.49% +desk : 73.33% +display : 93.63% +door : 89.52% +shelf : 90.04% +table : 81.85% +bed : 73.64% +pillow : 87.62% +sink : 78.33% +sofa : 92.86% +toilet : 89.41% +E@65 OA: 83.87 mAcc: 81.71 + +[04/01 17:51:24] ScanObjectNNHardest INFO: Epoch 65 LR 0.001559 train_oa 97.06, val_oa 83.87, best val oa 83.87 +[04/01 17:51:25] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:52:19] ScanObjectNNHardest INFO: Epoch 66 LR 0.001546 train_oa 97.38, val_oa 82.03, best val oa 83.87 +[04/01 17:53:17] ScanObjectNNHardest INFO: Epoch 67 LR 0.001534 train_oa 97.03, val_oa 82.34, best val oa 83.87 +[04/01 17:54:12] ScanObjectNNHardest INFO: Find a better ckpt @E68 +[04/01 17:54:12] ScanObjectNNHardest INFO: +Classes Acc +bag : 59.04% +bin : 87.44% +box : 63.91% +cabinet : 88.17% +chair : 92.31% +desk : 76.67% +display : 81.37% +door : 94.76% +shelf : 81.74% +table : 78.15% +bed : 80.00% +pillow : 92.38% +sink : 71.67% +sofa : 93.33% +toilet : 81.18% +E@68 OA: 83.97 mAcc: 81.47 + +[04/01 17:54:12] ScanObjectNNHardest INFO: Epoch 68 LR 0.001521 train_oa 97.20, val_oa 83.97, best val oa 83.97 +[04/01 17:54:12] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:55:04] ScanObjectNNHardest INFO: Epoch 69 LR 0.001508 train_oa 97.73, val_oa 82.93, best val oa 83.97 +[04/01 17:56:00] ScanObjectNNHardest INFO: Epoch 70 LR 0.001495 train_oa 97.32, val_oa 83.52, best val oa 83.97 +[04/01 17:57:00] ScanObjectNNHardest INFO: Find a better ckpt @E71 +[04/01 17:57:00] ScanObjectNNHardest INFO: +Classes Acc +bag : 60.24% +bin : 80.40% +box : 60.15% +cabinet : 92.20% +chair : 93.33% +desk : 78.67% +display : 90.69% +door : 85.71% +shelf : 82.99% +table : 77.78% +bed : 81.82% +pillow : 85.71% +sink : 78.33% +sofa : 90.48% +toilet : 84.71% +E@71 OA: 84.18 mAcc: 81.55 + +[04/01 17:57:00] ScanObjectNNHardest INFO: Epoch 71 LR 0.001481 train_oa 97.33, val_oa 84.18, best val oa 84.18 +[04/01 17:57:01] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:57:58] ScanObjectNNHardest INFO: Epoch 72 LR 0.001468 train_oa 97.37, val_oa 82.93, best val oa 84.18 +[04/01 17:59:00] ScanObjectNNHardest INFO: Find a better ckpt @E73 +[04/01 17:59:00] ScanObjectNNHardest INFO: +Classes Acc +bag : 69.88% +bin : 83.42% +box : 63.91% +cabinet : 89.78% +chair : 95.64% +desk : 82.00% +display : 91.67% +door : 92.86% +shelf : 82.57% +table : 72.22% +bed : 80.91% +pillow : 82.86% +sink : 74.17% +sofa : 93.81% +toilet : 81.18% +E@73 OA: 84.87 mAcc: 82.46 + +[04/01 17:59:00] ScanObjectNNHardest INFO: Epoch 73 LR 0.001454 train_oa 97.73, val_oa 84.87, best val oa 84.87 +[04/01 17:59:00] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 17:59:55] ScanObjectNNHardest INFO: Epoch 74 LR 0.001441 train_oa 97.47, val_oa 84.21, best val oa 84.87 +[04/01 18:00:54] ScanObjectNNHardest INFO: Epoch 75 LR 0.001427 train_oa 97.62, val_oa 83.76, best val oa 84.87 +[04/01 18:01:52] ScanObjectNNHardest INFO: Epoch 76 LR 0.001414 train_oa 97.81, val_oa 84.07, best val oa 84.87 +[04/01 18:02:48] ScanObjectNNHardest INFO: Epoch 77 LR 0.001400 train_oa 98.12, val_oa 82.96, best val oa 84.87 +[04/01 18:03:42] ScanObjectNNHardest INFO: Epoch 78 LR 0.001386 train_oa 97.74, val_oa 84.25, best val oa 84.87 +[04/01 18:04:36] ScanObjectNNHardest INFO: Epoch 79 LR 0.001372 train_oa 97.66, val_oa 84.21, best val oa 84.87 +[04/01 18:05:32] ScanObjectNNHardest INFO: Epoch 80 LR 0.001358 train_oa 98.02, val_oa 82.69, best val oa 84.87 +[04/01 18:06:26] ScanObjectNNHardest INFO: Epoch 81 LR 0.001344 train_oa 97.79, val_oa 84.35, best val oa 84.87 +[04/01 18:07:23] ScanObjectNNHardest INFO: Epoch 82 LR 0.001329 train_oa 98.03, val_oa 84.70, best val oa 84.87 +[04/01 18:08:18] ScanObjectNNHardest INFO: Epoch 83 LR 0.001315 train_oa 98.16, val_oa 83.76, best val oa 84.87 +[04/01 18:09:16] ScanObjectNNHardest INFO: Epoch 84 LR 0.001301 train_oa 98.10, val_oa 84.00, best val oa 84.87 +[04/01 18:10:12] ScanObjectNNHardest INFO: Epoch 85 LR 0.001286 train_oa 97.81, val_oa 84.04, best val oa 84.87 +[04/01 18:11:11] ScanObjectNNHardest INFO: Epoch 86 LR 0.001272 train_oa 98.15, val_oa 84.70, best val oa 84.87 +[04/01 18:12:08] ScanObjectNNHardest INFO: Epoch 87 LR 0.001257 train_oa 98.14, val_oa 84.28, best val oa 84.87 +[04/01 18:13:06] ScanObjectNNHardest INFO: Epoch 88 LR 0.001243 train_oa 98.28, val_oa 84.42, best val oa 84.87 +[04/01 18:14:05] ScanObjectNNHardest INFO: Epoch 89 LR 0.001228 train_oa 98.47, val_oa 84.77, best val oa 84.87 +[04/01 18:15:02] ScanObjectNNHardest INFO: Epoch 90 LR 0.001213 train_oa 98.22, val_oa 82.27, best val oa 84.87 +[04/01 18:15:56] ScanObjectNNHardest INFO: Find a better ckpt @E91 +[04/01 18:15:56] ScanObjectNNHardest INFO: +Classes Acc +bag : 60.24% +bin : 89.95% +box : 59.40% +cabinet : 91.67% +chair : 93.85% +desk : 80.67% +display : 86.27% +door : 91.43% +shelf : 86.31% +table : 72.22% +bed : 87.27% +pillow : 83.81% +sink : 76.67% +sofa : 93.81% +toilet : 87.06% +E@91 OA: 85.15 mAcc: 82.71 + +[04/01 18:15:56] ScanObjectNNHardest INFO: Epoch 91 LR 0.001199 train_oa 98.41, val_oa 85.15, best val oa 85.15 +[04/01 18:15:56] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 18:16:56] ScanObjectNNHardest INFO: Epoch 92 LR 0.001184 train_oa 98.61, val_oa 84.84, best val oa 85.15 +[04/01 18:17:56] ScanObjectNNHardest INFO: Epoch 93 LR 0.001169 train_oa 98.57, val_oa 82.82, best val oa 85.15 +[04/01 18:18:54] ScanObjectNNHardest INFO: Epoch 94 LR 0.001154 train_oa 98.57, val_oa 84.32, best val oa 85.15 +[04/01 18:19:50] ScanObjectNNHardest INFO: Epoch 95 LR 0.001139 train_oa 98.60, val_oa 83.41, best val oa 85.15 +[04/01 18:20:45] ScanObjectNNHardest INFO: Epoch 96 LR 0.001125 train_oa 98.31, val_oa 83.80, best val oa 85.15 +[04/01 18:21:41] ScanObjectNNHardest INFO: Epoch 97 LR 0.001110 train_oa 98.65, val_oa 84.49, best val oa 85.15 +[04/01 18:22:35] ScanObjectNNHardest INFO: Epoch 98 LR 0.001095 train_oa 98.60, val_oa 84.46, best val oa 85.15 +[04/01 18:23:28] ScanObjectNNHardest INFO: Epoch 99 LR 0.001080 train_oa 98.78, val_oa 84.14, best val oa 85.15 +[04/01 18:24:24] ScanObjectNNHardest INFO: Epoch 100 LR 0.001065 train_oa 98.95, val_oa 83.93, best val oa 85.15 +[04/01 18:25:20] ScanObjectNNHardest INFO: Epoch 101 LR 0.001050 train_oa 98.87, val_oa 84.32, best val oa 85.15 +[04/01 18:26:12] ScanObjectNNHardest INFO: Epoch 102 LR 0.001035 train_oa 98.92, val_oa 82.51, best val oa 85.15 +[04/01 18:27:14] ScanObjectNNHardest INFO: Epoch 103 LR 0.001020 train_oa 98.72, val_oa 84.00, best val oa 85.15 +[04/01 18:28:10] ScanObjectNNHardest INFO: Epoch 104 LR 0.001005 train_oa 99.13, val_oa 85.05, best val oa 85.15 +[04/01 18:29:04] ScanObjectNNHardest INFO: Find a better ckpt @E105 +[04/01 18:29:04] ScanObjectNNHardest INFO: +Classes Acc +bag : 60.24% +bin : 86.93% +box : 76.69% +cabinet : 87.37% +chair : 95.90% +desk : 82.00% +display : 90.69% +door : 96.67% +shelf : 84.65% +table : 69.63% +bed : 80.00% +pillow : 89.52% +sink : 77.50% +sofa : 93.33% +toilet : 80.00% +E@105 OA: 85.57 mAcc: 83.41 + +[04/01 18:29:04] ScanObjectNNHardest INFO: Epoch 105 LR 0.000990 train_oa 99.02, val_oa 85.57, best val oa 85.57 +[04/01 18:29:05] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 18:30:01] ScanObjectNNHardest INFO: Epoch 106 LR 0.000975 train_oa 98.76, val_oa 83.45, best val oa 85.57 +[04/01 18:30:56] ScanObjectNNHardest INFO: Epoch 107 LR 0.000961 train_oa 98.53, val_oa 85.01, best val oa 85.57 +[04/01 18:31:50] ScanObjectNNHardest INFO: Epoch 108 LR 0.000946 train_oa 99.17, val_oa 85.46, best val oa 85.57 +[04/01 18:32:44] ScanObjectNNHardest INFO: Epoch 109 LR 0.000931 train_oa 99.13, val_oa 84.07, best val oa 85.57 +[04/01 18:33:43] ScanObjectNNHardest INFO: Epoch 110 LR 0.000916 train_oa 99.32, val_oa 84.98, best val oa 85.57 +[04/01 18:34:37] ScanObjectNNHardest INFO: Epoch 111 LR 0.000901 train_oa 99.19, val_oa 83.83, best val oa 85.57 +[04/01 18:35:33] ScanObjectNNHardest INFO: Epoch 112 LR 0.000887 train_oa 99.08, val_oa 83.48, best val oa 85.57 +[04/01 18:36:29] ScanObjectNNHardest INFO: Epoch 113 LR 0.000872 train_oa 99.18, val_oa 83.55, best val oa 85.57 +[04/01 18:37:29] ScanObjectNNHardest INFO: Epoch 114 LR 0.000857 train_oa 99.37, val_oa 83.87, best val oa 85.57 +[04/01 18:38:23] ScanObjectNNHardest INFO: Epoch 115 LR 0.000843 train_oa 98.91, val_oa 84.14, best val oa 85.57 +[04/01 18:39:20] ScanObjectNNHardest INFO: Epoch 116 LR 0.000828 train_oa 99.25, val_oa 84.21, best val oa 85.57 +[04/01 18:40:13] ScanObjectNNHardest INFO: Epoch 117 LR 0.000814 train_oa 99.30, val_oa 84.04, best val oa 85.57 +[04/01 18:41:11] ScanObjectNNHardest INFO: Epoch 118 LR 0.000799 train_oa 99.32, val_oa 84.80, best val oa 85.57 +[04/01 18:42:06] ScanObjectNNHardest INFO: Epoch 119 LR 0.000785 train_oa 99.46, val_oa 84.66, best val oa 85.57 +[04/01 18:43:01] ScanObjectNNHardest INFO: Epoch 120 LR 0.000771 train_oa 99.25, val_oa 84.66, best val oa 85.57 +[04/01 18:43:56] ScanObjectNNHardest INFO: Epoch 121 LR 0.000756 train_oa 99.28, val_oa 84.52, best val oa 85.57 +[04/01 18:44:53] ScanObjectNNHardest INFO: Epoch 122 LR 0.000742 train_oa 99.37, val_oa 85.46, best val oa 85.57 +[04/01 18:45:47] ScanObjectNNHardest INFO: Epoch 123 LR 0.000728 train_oa 99.36, val_oa 84.59, best val oa 85.57 +[04/01 18:46:47] ScanObjectNNHardest INFO: Find a better ckpt @E124 +[04/01 18:46:47] ScanObjectNNHardest INFO: +Classes Acc +bag : 71.08% +bin : 82.91% +box : 66.92% +cabinet : 87.37% +chair : 95.13% +desk : 84.67% +display : 91.67% +door : 94.29% +shelf : 88.38% +table : 69.26% +bed : 79.09% +pillow : 88.57% +sink : 76.67% +sofa : 95.71% +toilet : 87.06% +E@124 OA: 85.63 mAcc: 83.92 + +[04/01 18:46:47] ScanObjectNNHardest INFO: Epoch 124 LR 0.000714 train_oa 99.61, val_oa 85.63, best val oa 85.63 +[04/01 18:46:48] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 18:47:47] ScanObjectNNHardest INFO: Epoch 125 LR 0.000700 train_oa 99.59, val_oa 84.07, best val oa 85.63 +[04/01 18:48:40] ScanObjectNNHardest INFO: Epoch 126 LR 0.000686 train_oa 99.56, val_oa 85.36, best val oa 85.63 +[04/01 18:49:40] ScanObjectNNHardest INFO: Epoch 127 LR 0.000673 train_oa 99.42, val_oa 85.43, best val oa 85.63 +[04/01 18:50:44] ScanObjectNNHardest INFO: Find a better ckpt @E128 +[04/01 18:50:44] ScanObjectNNHardest INFO: +Classes Acc +bag : 55.42% +bin : 82.91% +box : 70.68% +cabinet : 85.22% +chair : 96.41% +desk : 68.67% +display : 93.14% +door : 95.71% +shelf : 92.12% +table : 78.15% +bed : 80.00% +pillow : 88.57% +sink : 74.17% +sofa : 95.24% +toilet : 87.06% +E@128 OA: 85.67 mAcc: 82.90 + +[04/01 18:50:44] ScanObjectNNHardest INFO: Epoch 128 LR 0.000659 train_oa 99.51, val_oa 85.67, best val oa 85.67 +[04/01 18:50:44] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 18:51:42] ScanObjectNNHardest INFO: Epoch 129 LR 0.000646 train_oa 99.55, val_oa 84.91, best val oa 85.67 +[04/01 18:52:39] ScanObjectNNHardest INFO: Epoch 130 LR 0.000632 train_oa 99.59, val_oa 85.29, best val oa 85.67 +[04/01 18:53:36] ScanObjectNNHardest INFO: Epoch 131 LR 0.000619 train_oa 99.56, val_oa 84.66, best val oa 85.67 +[04/01 18:54:34] ScanObjectNNHardest INFO: Epoch 132 LR 0.000605 train_oa 99.53, val_oa 84.91, best val oa 85.67 +[04/01 18:55:32] ScanObjectNNHardest INFO: Find a better ckpt @E133 +[04/01 18:55:32] ScanObjectNNHardest INFO: +Classes Acc +bag : 63.86% +bin : 88.44% +box : 63.91% +cabinet : 88.71% +chair : 95.38% +desk : 75.33% +display : 90.20% +door : 92.86% +shelf : 88.80% +table : 77.41% +bed : 80.00% +pillow : 86.67% +sink : 80.00% +sofa : 92.86% +toilet : 82.35% +E@133 OA: 85.74 mAcc: 83.12 + +[04/01 18:55:32] ScanObjectNNHardest INFO: Epoch 133 LR 0.000592 train_oa 99.63, val_oa 85.74, best val oa 85.74 +[04/01 18:55:32] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 18:56:29] ScanObjectNNHardest INFO: Epoch 134 LR 0.000579 train_oa 99.53, val_oa 85.32, best val oa 85.74 +[04/01 18:57:23] ScanObjectNNHardest INFO: Epoch 135 LR 0.000566 train_oa 99.52, val_oa 85.11, best val oa 85.74 +[04/01 18:58:18] ScanObjectNNHardest INFO: Epoch 136 LR 0.000554 train_oa 99.62, val_oa 85.60, best val oa 85.74 +[04/01 18:59:19] ScanObjectNNHardest INFO: Epoch 137 LR 0.000541 train_oa 99.69, val_oa 84.80, best val oa 85.74 +[04/01 19:00:19] ScanObjectNNHardest INFO: Epoch 138 LR 0.000528 train_oa 99.74, val_oa 85.74, best val oa 85.74 +[04/01 19:01:16] ScanObjectNNHardest INFO: Epoch 139 LR 0.000516 train_oa 99.69, val_oa 85.50, best val oa 85.74 +[04/01 19:02:13] ScanObjectNNHardest INFO: Epoch 140 LR 0.000504 train_oa 99.75, val_oa 84.77, best val oa 85.74 +[04/01 19:03:13] ScanObjectNNHardest INFO: Epoch 141 LR 0.000492 train_oa 99.72, val_oa 84.98, best val oa 85.74 +[04/01 19:04:06] ScanObjectNNHardest INFO: Epoch 142 LR 0.000480 train_oa 99.72, val_oa 85.36, best val oa 85.74 +[04/01 19:05:04] ScanObjectNNHardest INFO: Epoch 143 LR 0.000468 train_oa 99.70, val_oa 85.70, best val oa 85.74 +[04/01 19:06:00] ScanObjectNNHardest INFO: Find a better ckpt @E144 +[04/01 19:06:00] ScanObjectNNHardest INFO: +Classes Acc +bag : 66.27% +bin : 85.93% +box : 69.17% +cabinet : 90.86% +chair : 95.90% +desk : 85.33% +display : 90.20% +door : 90.00% +shelf : 87.97% +table : 72.96% +bed : 83.64% +pillow : 91.43% +sink : 77.50% +sofa : 93.81% +toilet : 88.24% +E@144 OA: 86.50 mAcc: 84.61 + +[04/01 19:06:00] ScanObjectNNHardest INFO: Epoch 144 LR 0.000456 train_oa 99.82, val_oa 86.50, best val oa 86.50 +[04/01 19:06:01] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:06:58] ScanObjectNNHardest INFO: Epoch 145 LR 0.000444 train_oa 99.76, val_oa 85.50, best val oa 86.50 +[04/01 19:07:57] ScanObjectNNHardest INFO: Epoch 146 LR 0.000433 train_oa 99.75, val_oa 85.74, best val oa 86.50 +[04/01 19:08:55] ScanObjectNNHardest INFO: Epoch 147 LR 0.000422 train_oa 99.81, val_oa 86.26, best val oa 86.50 +[04/01 19:09:55] ScanObjectNNHardest INFO: Find a better ckpt @E148 +[04/01 19:09:55] ScanObjectNNHardest INFO: +Classes Acc +bag : 66.27% +bin : 85.93% +box : 63.16% +cabinet : 92.74% +chair : 95.90% +desk : 76.00% +display : 91.18% +door : 91.90% +shelf : 90.04% +table : 78.52% +bed : 81.82% +pillow : 88.57% +sink : 70.00% +sofa : 94.76% +toilet : 90.59% +E@148 OA: 86.54 mAcc: 83.82 + +[04/01 19:09:55] ScanObjectNNHardest INFO: Epoch 148 LR 0.000411 train_oa 99.70, val_oa 86.54, best val oa 86.54 +[04/01 19:09:55] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:10:53] ScanObjectNNHardest INFO: Find a better ckpt @E149 +[04/01 19:10:53] ScanObjectNNHardest INFO: +Classes Acc +bag : 69.88% +bin : 87.94% +box : 66.17% +cabinet : 88.17% +chair : 95.38% +desk : 76.67% +display : 94.12% +door : 96.67% +shelf : 88.38% +table : 76.30% +bed : 83.64% +pillow : 90.48% +sink : 75.83% +sofa : 94.29% +toilet : 90.59% +E@149 OA: 86.85 mAcc: 84.97 + +[04/01 19:10:53] ScanObjectNNHardest INFO: Epoch 149 LR 0.000400 train_oa 99.82, val_oa 86.85, best val oa 86.85 +[04/01 19:10:54] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:11:53] ScanObjectNNHardest INFO: Epoch 150 LR 0.000389 train_oa 99.79, val_oa 85.91, best val oa 86.85 +[04/01 19:12:53] ScanObjectNNHardest INFO: Epoch 151 LR 0.000378 train_oa 99.86, val_oa 85.88, best val oa 86.85 +[04/01 19:13:51] ScanObjectNNHardest INFO: Epoch 152 LR 0.000368 train_oa 99.85, val_oa 86.33, best val oa 86.85 +[04/01 19:14:46] ScanObjectNNHardest INFO: Epoch 153 LR 0.000357 train_oa 99.85, val_oa 86.61, best val oa 86.85 +[04/01 19:15:42] ScanObjectNNHardest INFO: Epoch 154 LR 0.000347 train_oa 99.90, val_oa 86.29, best val oa 86.85 +[04/01 19:16:38] ScanObjectNNHardest INFO: Epoch 155 LR 0.000337 train_oa 99.87, val_oa 86.26, best val oa 86.85 +[04/01 19:17:34] ScanObjectNNHardest INFO: Epoch 156 LR 0.000328 train_oa 99.89, val_oa 85.46, best val oa 86.85 +[04/01 19:18:33] ScanObjectNNHardest INFO: Epoch 157 LR 0.000318 train_oa 99.90, val_oa 85.81, best val oa 86.85 +[04/01 19:19:29] ScanObjectNNHardest INFO: Epoch 158 LR 0.000309 train_oa 99.88, val_oa 85.77, best val oa 86.85 +[04/01 19:20:27] ScanObjectNNHardest INFO: Epoch 159 LR 0.000299 train_oa 99.95, val_oa 86.33, best val oa 86.85 +[04/01 19:21:27] ScanObjectNNHardest INFO: Epoch 160 LR 0.000290 train_oa 99.87, val_oa 85.88, best val oa 86.85 +[04/01 19:22:26] ScanObjectNNHardest INFO: Epoch 161 LR 0.000281 train_oa 99.89, val_oa 86.26, best val oa 86.85 +[04/01 19:23:23] ScanObjectNNHardest INFO: Epoch 162 LR 0.000273 train_oa 99.89, val_oa 85.95, best val oa 86.85 +[04/01 19:24:21] ScanObjectNNHardest INFO: Epoch 163 LR 0.000264 train_oa 99.82, val_oa 86.26, best val oa 86.85 +[04/01 19:25:18] ScanObjectNNHardest INFO: Epoch 164 LR 0.000256 train_oa 99.94, val_oa 85.81, best val oa 86.85 +[04/01 19:26:16] ScanObjectNNHardest INFO: Epoch 165 LR 0.000248 train_oa 99.96, val_oa 85.81, best val oa 86.85 +[04/01 19:27:12] ScanObjectNNHardest INFO: Epoch 166 LR 0.000240 train_oa 99.97, val_oa 86.40, best val oa 86.85 +[04/01 19:28:11] ScanObjectNNHardest INFO: Epoch 167 LR 0.000232 train_oa 99.96, val_oa 86.78, best val oa 86.85 +[04/01 19:29:11] ScanObjectNNHardest INFO: Epoch 168 LR 0.000225 train_oa 99.93, val_oa 86.22, best val oa 86.85 +[04/01 19:30:10] ScanObjectNNHardest INFO: Epoch 169 LR 0.000218 train_oa 99.96, val_oa 86.47, best val oa 86.85 +[04/01 19:31:08] ScanObjectNNHardest INFO: Epoch 170 LR 0.000210 train_oa 99.97, val_oa 85.95, best val oa 86.85 +[04/01 19:32:07] ScanObjectNNHardest INFO: Find a better ckpt @E171 +[04/01 19:32:07] ScanObjectNNHardest INFO: +Classes Acc +bag : 66.27% +bin : 89.45% +box : 61.65% +cabinet : 91.40% +chair : 96.41% +desk : 82.00% +display : 90.20% +door : 91.90% +shelf : 91.70% +table : 74.44% +bed : 84.55% +pillow : 90.48% +sink : 70.83% +sofa : 94.29% +toilet : 94.12% +E@171 OA: 86.88 mAcc: 84.65 + +[04/01 19:32:07] ScanObjectNNHardest INFO: Epoch 171 LR 0.000204 train_oa 99.96, val_oa 86.88, best val oa 86.88 +[04/01 19:32:07] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:33:00] ScanObjectNNHardest INFO: Epoch 172 LR 0.000197 train_oa 99.98, val_oa 86.68, best val oa 86.88 +[04/01 19:33:56] ScanObjectNNHardest INFO: Epoch 173 LR 0.000190 train_oa 99.94, val_oa 86.19, best val oa 86.88 +[04/01 19:34:51] ScanObjectNNHardest INFO: Epoch 174 LR 0.000184 train_oa 99.95, val_oa 86.50, best val oa 86.88 +[04/01 19:35:47] ScanObjectNNHardest INFO: Find a better ckpt @E175 +[04/01 19:35:47] ScanObjectNNHardest INFO: +Classes Acc +bag : 62.65% +bin : 88.94% +box : 64.66% +cabinet : 90.32% +chair : 96.92% +desk : 84.67% +display : 91.18% +door : 94.29% +shelf : 88.80% +table : 72.59% +bed : 82.73% +pillow : 90.48% +sink : 78.33% +sofa : 95.24% +toilet : 91.76% +E@175 OA: 87.02 mAcc: 84.90 + +[04/01 19:35:47] ScanObjectNNHardest INFO: Epoch 175 LR 0.000178 train_oa 99.94, val_oa 87.02, best val oa 87.02 +[04/01 19:35:47] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:36:42] ScanObjectNNHardest INFO: Epoch 176 LR 0.000172 train_oa 99.96, val_oa 86.85, best val oa 87.02 +[04/01 19:37:38] ScanObjectNNHardest INFO: Epoch 177 LR 0.000167 train_oa 99.95, val_oa 86.75, best val oa 87.02 +[04/01 19:38:33] ScanObjectNNHardest INFO: Epoch 178 LR 0.000161 train_oa 99.96, val_oa 86.88, best val oa 87.02 +[04/01 19:39:28] ScanObjectNNHardest INFO: Epoch 179 LR 0.000156 train_oa 99.96, val_oa 86.33, best val oa 87.02 +[04/01 19:40:24] ScanObjectNNHardest INFO: Epoch 180 LR 0.000151 train_oa 99.96, val_oa 86.05, best val oa 87.02 +[04/01 19:41:23] ScanObjectNNHardest INFO: Epoch 181 LR 0.000146 train_oa 99.98, val_oa 86.95, best val oa 87.02 +[04/01 19:42:21] ScanObjectNNHardest INFO: Epoch 182 LR 0.000142 train_oa 99.99, val_oa 86.43, best val oa 87.02 +[04/01 19:43:20] ScanObjectNNHardest INFO: Epoch 183 LR 0.000138 train_oa 99.93, val_oa 85.98, best val oa 87.02 +[04/01 19:44:18] ScanObjectNNHardest INFO: Epoch 184 LR 0.000134 train_oa 99.96, val_oa 86.54, best val oa 87.02 +[04/01 19:45:14] ScanObjectNNHardest INFO: Epoch 185 LR 0.000130 train_oa 99.99, val_oa 86.88, best val oa 87.02 +[04/01 19:46:10] ScanObjectNNHardest INFO: Epoch 186 LR 0.000126 train_oa 99.97, val_oa 86.92, best val oa 87.02 +[04/01 19:47:06] ScanObjectNNHardest INFO: Epoch 187 LR 0.000123 train_oa 99.99, val_oa 86.43, best val oa 87.02 +[04/01 19:48:03] ScanObjectNNHardest INFO: Epoch 188 LR 0.000120 train_oa 99.98, val_oa 86.43, best val oa 87.02 +[04/01 19:48:58] ScanObjectNNHardest INFO: Epoch 189 LR 0.000117 train_oa 99.94, val_oa 86.19, best val oa 87.02 +[04/01 19:49:54] ScanObjectNNHardest INFO: Epoch 190 LR 0.000114 train_oa 99.98, val_oa 86.50, best val oa 87.02 +[04/01 19:50:48] ScanObjectNNHardest INFO: Epoch 191 LR 0.000112 train_oa 99.97, val_oa 85.81, best val oa 87.02 +[04/01 19:51:47] ScanObjectNNHardest INFO: Epoch 192 LR 0.000109 train_oa 99.98, val_oa 86.61, best val oa 87.02 +[04/01 19:52:46] ScanObjectNNHardest INFO: Epoch 193 LR 0.000107 train_oa 99.99, val_oa 86.78, best val oa 87.02 +[04/01 19:53:42] ScanObjectNNHardest INFO: Epoch 194 LR 0.000106 train_oa 100.00, val_oa 86.85, best val oa 87.02 +[04/01 19:54:38] ScanObjectNNHardest INFO: Epoch 195 LR 0.000104 train_oa 99.98, val_oa 86.16, best val oa 87.02 +[04/01 19:55:35] ScanObjectNNHardest INFO: Epoch 196 LR 0.000103 train_oa 99.99, val_oa 86.57, best val oa 87.02 +[04/01 19:56:33] ScanObjectNNHardest INFO: Epoch 197 LR 0.000102 train_oa 100.00, val_oa 86.95, best val oa 87.02 +[04/01 19:57:33] ScanObjectNNHardest INFO: Epoch 198 LR 0.000101 train_oa 99.99, val_oa 86.47, best val oa 87.02 +[04/01 19:58:27] ScanObjectNNHardest INFO: Find a better ckpt @E199 +[04/01 19:58:27] ScanObjectNNHardest INFO: +Classes Acc +bag : 65.06% +bin : 89.95% +box : 68.42% +cabinet : 89.25% +chair : 96.67% +desk : 85.33% +display : 89.22% +door : 95.24% +shelf : 88.80% +table : 73.70% +bed : 84.55% +pillow : 89.52% +sink : 77.50% +sofa : 94.29% +toilet : 88.24% +E@199 OA: 87.06 mAcc: 85.05 + +[04/01 19:58:27] ScanObjectNNHardest INFO: Epoch 199 LR 0.000100 train_oa 99.98, val_oa 87.06, best val oa 87.06 +[04/01 19:58:27] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 19:59:23] ScanObjectNNHardest INFO: Epoch 200 LR 0.000100 train_oa 99.99, val_oa 86.88, best val oa 87.06 +[04/01 20:00:21] ScanObjectNNHardest INFO: Epoch 201 LR 0.000100 train_oa 100.00, val_oa 86.85, best val oa 87.06 +[04/01 20:01:18] ScanObjectNNHardest INFO: Epoch 202 LR 0.000100 train_oa 99.98, val_oa 86.88, best val oa 87.06 +[04/01 20:02:16] ScanObjectNNHardest INFO: Epoch 203 LR 0.000100 train_oa 99.99, val_oa 86.50, best val oa 87.06 +[04/01 20:03:10] ScanObjectNNHardest INFO: Epoch 204 LR 0.000100 train_oa 100.00, val_oa 86.54, best val oa 87.06 +[04/01 20:04:09] ScanObjectNNHardest INFO: Epoch 205 LR 0.000100 train_oa 99.97, val_oa 86.92, best val oa 87.06 +[04/01 20:05:08] ScanObjectNNHardest INFO: Epoch 206 LR 0.000100 train_oa 99.96, val_oa 86.54, best val oa 87.06 +[04/01 20:06:07] ScanObjectNNHardest INFO: Epoch 207 LR 0.000100 train_oa 99.97, val_oa 86.40, best val oa 87.06 +[04/01 20:07:05] ScanObjectNNHardest INFO: Find a better ckpt @E208 +[04/01 20:07:05] ScanObjectNNHardest INFO: +Classes Acc +bag : 74.70% +bin : 86.93% +box : 60.90% +cabinet : 89.52% +chair : 95.90% +desk : 84.67% +display : 90.20% +door : 94.29% +shelf : 88.38% +table : 75.19% +bed : 86.36% +pillow : 90.48% +sink : 81.67% +sofa : 95.71% +toilet : 88.24% +E@208 OA: 87.16 mAcc: 85.54 + +[04/01 20:07:05] ScanObjectNNHardest INFO: Epoch 208 LR 0.000100 train_oa 100.00, val_oa 87.16, best val oa 87.16 +[04/01 20:07:05] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 20:08:02] ScanObjectNNHardest INFO: Epoch 209 LR 0.000100 train_oa 99.97, val_oa 86.81, best val oa 87.16 +[04/01 20:09:01] ScanObjectNNHardest INFO: Epoch 210 LR 0.000100 train_oa 100.00, val_oa 86.78, best val oa 87.16 +[04/01 20:09:55] ScanObjectNNHardest INFO: Epoch 211 LR 0.000100 train_oa 99.99, val_oa 86.57, best val oa 87.16 +[04/01 20:10:51] ScanObjectNNHardest INFO: Epoch 212 LR 0.000100 train_oa 99.99, val_oa 86.40, best val oa 87.16 +[04/01 20:11:49] ScanObjectNNHardest INFO: Epoch 213 LR 0.000100 train_oa 99.98, val_oa 86.71, best val oa 87.16 +[04/01 20:12:47] ScanObjectNNHardest INFO: Epoch 214 LR 0.000100 train_oa 99.99, val_oa 86.78, best val oa 87.16 +[04/01 20:13:44] ScanObjectNNHardest INFO: Epoch 215 LR 0.000100 train_oa 99.98, val_oa 86.61, best val oa 87.16 +[04/01 20:14:40] ScanObjectNNHardest INFO: Epoch 216 LR 0.000100 train_oa 99.97, val_oa 86.71, best val oa 87.16 +[04/01 20:15:44] ScanObjectNNHardest INFO: Epoch 217 LR 0.000100 train_oa 99.99, val_oa 86.02, best val oa 87.16 +[04/01 20:16:41] ScanObjectNNHardest INFO: Epoch 218 LR 0.000100 train_oa 99.99, val_oa 86.16, best val oa 87.16 +[04/01 20:17:37] ScanObjectNNHardest INFO: Epoch 219 LR 0.000100 train_oa 99.98, val_oa 86.22, best val oa 87.16 +[04/01 20:18:30] ScanObjectNNHardest INFO: Epoch 220 LR 0.000100 train_oa 99.99, val_oa 86.02, best val oa 87.16 +[04/01 20:19:26] ScanObjectNNHardest INFO: Epoch 221 LR 0.000100 train_oa 99.98, val_oa 85.57, best val oa 87.16 +[04/01 20:20:25] ScanObjectNNHardest INFO: Epoch 222 LR 0.000100 train_oa 99.96, val_oa 86.22, best val oa 87.16 +[04/01 20:21:26] ScanObjectNNHardest INFO: Epoch 223 LR 0.000100 train_oa 99.99, val_oa 86.29, best val oa 87.16 +[04/01 20:22:23] ScanObjectNNHardest INFO: Epoch 224 LR 0.000100 train_oa 99.98, val_oa 86.47, best val oa 87.16 +[04/01 20:23:19] ScanObjectNNHardest INFO: Epoch 225 LR 0.000100 train_oa 99.98, val_oa 86.47, best val oa 87.16 +[04/01 20:24:15] ScanObjectNNHardest INFO: Epoch 226 LR 0.000100 train_oa 100.00, val_oa 87.09, best val oa 87.16 +[04/01 20:25:16] ScanObjectNNHardest INFO: Epoch 227 LR 0.000100 train_oa 99.97, val_oa 86.43, best val oa 87.16 +[04/01 20:26:16] ScanObjectNNHardest INFO: Epoch 228 LR 0.000100 train_oa 99.99, val_oa 86.61, best val oa 87.16 +[04/01 20:27:14] ScanObjectNNHardest INFO: Epoch 229 LR 0.000100 train_oa 99.99, val_oa 86.05, best val oa 87.16 +[04/01 20:28:12] ScanObjectNNHardest INFO: Epoch 230 LR 0.000100 train_oa 99.99, val_oa 86.64, best val oa 87.16 +[04/01 20:29:09] ScanObjectNNHardest INFO: Epoch 231 LR 0.000100 train_oa 99.98, val_oa 86.02, best val oa 87.16 +[04/01 20:30:07] ScanObjectNNHardest INFO: Epoch 232 LR 0.000100 train_oa 99.99, val_oa 86.29, best val oa 87.16 +[04/01 20:31:09] ScanObjectNNHardest INFO: Epoch 233 LR 0.000100 train_oa 99.99, val_oa 86.43, best val oa 87.16 +[04/01 20:32:10] ScanObjectNNHardest INFO: Epoch 234 LR 0.000100 train_oa 99.96, val_oa 86.71, best val oa 87.16 +[04/01 20:33:08] ScanObjectNNHardest INFO: Epoch 235 LR 0.000100 train_oa 99.99, val_oa 87.16, best val oa 87.16 +[04/01 20:34:06] ScanObjectNNHardest INFO: Epoch 236 LR 0.000100 train_oa 99.98, val_oa 86.71, best val oa 87.16 +[04/01 20:35:01] ScanObjectNNHardest INFO: Epoch 237 LR 0.000100 train_oa 99.99, val_oa 86.33, best val oa 87.16 +[04/01 20:35:58] ScanObjectNNHardest INFO: Epoch 238 LR 0.000100 train_oa 99.98, val_oa 86.64, best val oa 87.16 +[04/01 20:36:52] ScanObjectNNHardest INFO: Epoch 239 LR 0.000100 train_oa 99.97, val_oa 87.13, best val oa 87.16 +[04/01 20:37:48] ScanObjectNNHardest INFO: Find a better ckpt @E240 +[04/01 20:37:48] ScanObjectNNHardest INFO: +Classes Acc +bag : 67.47% +bin : 88.44% +box : 70.68% +cabinet : 91.40% +chair : 96.15% +desk : 86.00% +display : 90.69% +door : 94.29% +shelf : 87.14% +table : 73.33% +bed : 80.91% +pillow : 90.48% +sink : 79.17% +sofa : 94.29% +toilet : 90.59% +E@240 OA: 87.27 mAcc: 85.40 + +[04/01 20:37:48] ScanObjectNNHardest INFO: Epoch 240 LR 0.000100 train_oa 100.00, val_oa 87.27, best val oa 87.27 +[04/01 20:37:48] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 20:38:44] ScanObjectNNHardest INFO: Epoch 241 LR 0.000100 train_oa 99.98, val_oa 87.06, best val oa 87.27 +[04/01 20:39:39] ScanObjectNNHardest INFO: Epoch 242 LR 0.000100 train_oa 99.99, val_oa 86.92, best val oa 87.27 +[04/01 20:40:36] ScanObjectNNHardest INFO: Find a better ckpt @E243 +[04/01 20:40:36] ScanObjectNNHardest INFO: +Classes Acc +bag : 66.27% +bin : 89.45% +box : 70.68% +cabinet : 91.67% +chair : 95.90% +desk : 84.67% +display : 89.22% +door : 94.29% +shelf : 88.80% +table : 72.22% +bed : 81.82% +pillow : 90.48% +sink : 80.83% +sofa : 94.29% +toilet : 91.76% +E@243 OA: 87.30 mAcc: 85.49 + +[04/01 20:40:36] ScanObjectNNHardest INFO: Epoch 243 LR 0.000100 train_oa 99.99, val_oa 87.30, best val oa 87.30 +[04/01 20:40:36] ScanObjectNNHardest INFO: Found the best model and saved in log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 20:41:33] ScanObjectNNHardest INFO: Epoch 244 LR 0.000100 train_oa 99.98, val_oa 87.16, best val oa 87.30 +[04/01 20:42:27] ScanObjectNNHardest INFO: Epoch 245 LR 0.000100 train_oa 99.99, val_oa 86.78, best val oa 87.30 +[04/01 20:43:22] ScanObjectNNHardest INFO: Epoch 246 LR 0.000100 train_oa 99.98, val_oa 86.81, best val oa 87.30 +[04/01 20:44:22] ScanObjectNNHardest INFO: Epoch 247 LR 0.000100 train_oa 99.99, val_oa 86.78, best val oa 87.30 +[04/01 20:45:21] ScanObjectNNHardest INFO: Epoch 248 LR 0.000100 train_oa 99.99, val_oa 86.81, best val oa 87.30 +[04/01 20:46:17] ScanObjectNNHardest INFO: Epoch 249 LR 0.000100 train_oa 99.98, val_oa 87.23, best val oa 87.30 +[04/01 20:47:12] ScanObjectNNHardest INFO: Epoch 250 LR 0.000100 train_oa 99.99, val_oa 87.02, best val oa 87.30 +[04/01 20:47:15] ScanObjectNNHardest INFO: +Classes Acc +bag : 73.49% +bin : 87.94% +box : 63.91% +cabinet : 90.32% +chair : 96.15% +desk : 84.67% +display : 90.20% +door : 95.71% +shelf : 87.97% +table : 76.67% +bed : 83.64% +pillow : 87.62% +sink : 80.00% +sofa : 93.33% +toilet : 89.41% +E@243 OA: 87.27 mAcc: 85.40 + +[04/01 20:47:15] ScanObjectNNHardest INFO: Successful Loading the ckpt from log/scanobjectnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR/checkpoint/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth +[04/01 20:47:15] ScanObjectNNHardest INFO: ckpts @ 243 epoch( {'best_val': 87.30048370361328} ) +[04/01 20:47:19] ScanObjectNNHardest INFO: +Classes Acc +bag : 67.47% +bin : 88.94% +box : 70.68% +cabinet : 91.67% +chair : 95.90% +desk : 84.00% +display : 88.73% +door : 94.76% +shelf : 88.80% +table : 72.59% +bed : 80.91% +pillow : 91.43% +sink : 80.83% +sofa : 93.81% +toilet : 91.76% +E@243 OA: 87.27 mAcc: 85.49 + diff --git a/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth b/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth new file mode 100644 index 0000000000000000000000000000000000000000..619ff786ea20e295586fffea8a4a658a028d1560 --- /dev/null +++ b/checkpoint/scanobejctnn/scanobjectnn-train-ppv2-s-ngpus1-seed1234-20250401-164946-nvdQKdTEv4B3fR7TMhwGjR_ckpt_best.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c19fa2b293739aee1f714feca27d401b79b3ee51244e3d24fde0d3bd43937ab +size 31266231