DriveTransformer/goalpoint_1gpu/20260226_131720.log

2026-02-26 13:17:32,943 - mmdet - INFO - initialize ResNet with init_cfg {'type': 'Pretrained', 'checkpoint': './ckpts/resnet50-19c8e357.pth'}
2026-02-26 13:17:33,081 - mmdet - INFO - initialize FPN with init_cfg {'type': 'Xavier', 'layer': 'Conv2d', 'distribution': 'uniform'}
Name of parameter - Initialization information

pts_bbox_head.code_weights - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_code_weights - torch.Size([4]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.position_range - torch.Size([6]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.pc_range - torch.Size([6]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_query.weight - torch.Size([900, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_reference_points.weight - torch.Size([900, 3]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.0.weight - torch.Size([512, 9]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_cls_embedding.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.0.weight - torch.Size([512, 768]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.agent_ref_embedding.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.mode_embedding.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_query.weight - torch.Size([100, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_reference_points.weight - torch.Size([100, 2]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.0.weight - torch.Size([512, 6]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_cls_embedding.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.0.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.map_ref_embedding.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.0.weight - torch.Size([512, 141]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_lcf_encoder.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_traj_ref_fix_time_embedding.0.weight - torch.Size([512, 1920]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_time_embedding.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_time_embedding.2.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_time_embedding.2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_time_embedding.4.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_time_embedding.4.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.0.weight - torch.Size([512, 640]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.2.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.2.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.4.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_ref_fix_dist_embedding.4.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.featurized_pe.reduce.0.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.featurized_pe.reduce.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.featurized_pe.gamma.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.featurized_pe.gamma.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.featurized_pe.beta.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.featurized_pe.beta.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.spatial_alignment.reduce.0.weight - torch.Size([512, 8]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.spatial_alignment.reduce.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.spatial_alignment.gamma.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.spatial_alignment.gamma.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.spatial_alignment.beta.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.spatial_alignment.beta.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.time_embedding.0.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.time_embedding.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.time_embedding.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.time_embedding.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.time_embedding.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.time_embedding.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.0.weight - torch.Size([512, 192]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.2.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.2.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.4.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.img_position_encoder.4.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.reduce.0.weight - torch.Size([512, 180]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.reduce.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.gamma.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.gamma.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.beta.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_pe.beta.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_pose_memory.reduce.0.weight - torch.Size([512, 180]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_memory.reduce.0.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_memory.gamma.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_memory.gamma.bias - torch.Size([512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_memory.beta.weight - torch.Size([512, 512]): 
Initialized by user-defined `init_weights` in DriveTransformerlHead  

pts_bbox_head.ego_pose_memory.beta.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.attentions.1.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.agent_prep_decoder.layers.0.ffns.0.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.attentions.1.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_prep_decoder.layers.0.ffns.0.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.task_self_attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.1.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.temporal_cross_attention_layers.2.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.sensor_cross_attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.cross_w_q.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.0.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.1.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.0.ffns.0.2.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.task_self_attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.1.temporal_cross_attention_layers.1.w_q.0.weight - torch.Size([512]2026-02-26 13:17:33,098 - DriveTransformer - INFO - Model:
DriveTransformer(
  (pts_bbox_head): DriveTransformerlHead(
    (loss_cls): FocalLoss()
    (loss_bbox): L1Loss()
    (loss_traj): L1Loss()
    (loss_traj_cls): FocalLoss()
    (loss_map_cls): FocalLoss()
    (loss_map_pts): PtsL1Loss()
    (loss_map_dir): PtsDirCosLoss()
    (loss_plan_reg_fix_dist): L1Loss()
    (loss_plan_reg_fix_time): L1Loss()
    (loss_plan_cls): FocalLoss()
    (agent_query): Embedding(900, 512)
    (agent_reference_points): Embedding(900, 3)
    (agent_cls_embedding): Sequential(
      (0): Linear(in_features=9, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (agent_ref_embedding): Sequential(
      (0): Linear(in_features=768, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (mode_embedding): Embedding(6, 512)
    (map_query): Embedding(100, 512)
    (map_reference_points): Embedding(100, 2)
    (map_cls_embedding): Sequential(
      (0): Linear(in_features=6, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (map_ref_embedding): Sequential(
      (0): Linear(in_features=512, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (ego_lcf_encoder): Sequential(
      (0): Linear(in_features=141, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (ego_traj_ref_fix_time_embedding): Sequential(
      (0): Linear(in_features=1920, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (ego_traj_ref_fix_dist_embedding): Sequential(
      (0): Linear(in_features=640, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (featurized_pe): MLN(
      (reduce): Sequential(
        (0): Linear(in_features=512, out_features=512, bias=True)
        (1): SiLU(inplace=True)
      )
      (gamma): Linear(in_features=512, out_features=512, bias=True)
      (beta): Linear(in_features=512, out_features=512, bias=True)
      (ln): LayerNorm((512,), eps=1e-05, elementwise_affine=False)
    )
    (spatial_alignment): MLN(
      (reduce): Sequential(
        (0): Linear(in_features=8, out_features=512, bias=True)
        (1): SiLU(inplace=True)
      )
      (gamma): Linear(in_features=512, out_features=512, bias=True)
      (beta): Linear(in_features=512, out_features=512, bias=True)
      (ln): LayerNorm((512,), eps=1e-05, elementwise_affine=False)
    )
    (time_embedding): Sequential(
      (0): Linear(in_features=512, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (img_position_encoder): Sequential(
      (0): Linear(in_features=192, out_features=512, bias=True)
      (1): SiLU(inplace=True)
      (2): Linear(in_features=512, out_features=512, bias=True)
      (3): SiLU(inplace=True)
      (4): Linear(in_features=512, out_features=512, bias=True)
    )
    (ego_pose_pe): MLN(
      (reduce): Sequential(
        (0): Linear(in_features=180, out_features=512, bias=True)
        (1): SiLU(inplace=True)
      )
      (gamma): Linear(in_features=512, out_features=512, bias=True)
      (beta): Linear(in_features=512, out_features=512, bias=True)
      (ln): LayerNorm((512,), eps=1e-05, elementwise_affine=False)
    )
    (ego_pose_memory): MLN(
      (reduce): Sequential(
        (0): Linear(in_features=180, out_features=512, bias=True)
        (1): SiLU(inplace=True)
      )
      (gamma): Linear(in_features=512, out_features=512, bias=True)
      (beta): Linear(in_features=512, out_features=512, bias=True)
      (ln): LayerNorm((512,), eps=1e-05, elementwise_affine=False)
    )
    (agent_prep_decoder): DriveTransformerPreDecoder(
      (layers): ModuleList(
        (0): DriveTransformerPreDecoderLayer(
          (attentions): ModuleList(
            (0-1): 2 x AttentionLayer(
              (w_q): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (w_k): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (w_v): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (proj): Linear(in_features=512, out_features=512, bias=True)
              (postnorm): Identity()
            )
          )
          (ffns): ModuleList(
            (0): SwiGLULayer(
              (swiglu): SwiGLU(
                (w12): Linear(in_features=512, out_features=2736, bias=True)
                (w3): Linear(in_features=1368, out_features=512, bias=True)
              )
              (prenorm): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
              (postnorm): Identity()
            )
          )
        )
      )
    )
    (map_prep_decoder): DriveTransformerPreDecoder(
      (layers): ModuleList(
        (0): DriveTransformerPreDecoderLayer(
          (attentions): ModuleList(
            (0-1): 2 x AttentionLayer(
              (w_q): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (w_k): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (w_v): Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
              (proj): Linear(in_features=512, out_features=512, bias=True)
              (postnorm): Identity()
            )
          )
          (ffns): ModuleList(
            (0): SwiGLULayer(
              (swiglu): SwiGLU(
                (w12): Linear(in_features=512, out_features=2736, bias=True)
                (w3): Linear(in_features=1368, out_features=512, bias=True)
              )
              (prenorm): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
              (postnorm): Identity()
            )
          )
        )
      )
    )
    (transformer): DriveTransformerWrapper(
      (decoder): DriveTransformerDecoder(
        (layers): ModuleList(
          (0-5): 6 x DriveTransformerDecoderLayer(
            (task_self_attentions): ModuleList(
              (0): AttentionLayer(
                (w_q): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (w_k): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (w_v): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (proj): Linear(in_features=512, out_features=512, bias=True)
                (postnorm): Identity()
              )
            )
            (temporal_cross_attention_layers): ModuleList(
              (0-2): 3 x AttentionLayer(
                (w_q): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (w_k): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (w_v): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (proj): Linear(in_features=512, out_features=512, bias=True)
                (postnorm): Identity()
              )
            )
            (sensor_cross_attentions): ModuleList(
              (0): AttentionLayer(
                (w_k): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (w_v): Sequential(
                  (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (1): Linear(in_features=512, out_features=512, bias=True)
                )
                (proj): Linear(in_features=512, out_features=512, bias=True)
                (postnorm): Identity()
              )
            )
            (cross_w_q): ModuleList(
              (0-2): 3 x Sequential(
                (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                (1): Linear(in_features=512, out_features=512, bias=True)
              )
            )
            (ffns): ModuleList(
              (0): ModuleList(
                (0-2): 3 x SwiGLULayer(
                  (swiglu): SwiGLU(
                    (w12): Linear(in_features=512, out_features=2736, bias=True)
                    (w3): Linear(in_features=1368, out_features=512, bias=True)
                  )
                  (prenorm): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
                  (dropout): Dropout(p=0.1, inplace=False)
                  (postnorm): Identity()
                )
              )
            )
          )
        )
      )
    )
    (cls_branches): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=9, bias=False)
      )
    )
    (reg_branches): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=10, bias=True)
      )
    )
    (traj_branches): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=12, bias=True)
      )
    )
    (traj_cls_branches): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (6): Linear(in_features=512, out_features=1, bias=False)
      )
    )
    (map_cls_branches): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=6, bias=False)
      )
    )
    (map_reg_branches): ModuleList(
      (0-5): 6 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=2, bias=True)
      )
      (6): Linear(in_features=512, out_features=40, bias=True)
    )
    (ego_traj_branches_fix_dist): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=20, bias=True)
      )
    )
    (ego_traj_branches_fix_time): ModuleList(
      (0-6): 7 x Sequential(
        (0): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
        (1): Linear(in_features=512, out_features=512, bias=True)
        (2): SiLU(inplace=True)
        (3): Linear(in_features=512, out_features=512, bias=True)
        (4): SiLU(inplace=True)
        (5): Linear(in_features=512, out_features=60, bias=True)
      )
    )
  )
  (img_backbone): ResNet(
    (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)
    (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (relu): ReLU(inplace=True)
    (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)
    (layer1): ResLayer(
      (0): Bottleneck(
        (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
        (downsample): Sequential(
          (0): Conv2d(64, 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)
        )
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (1): Bottleneck(
        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (2): Bottleneck(
        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
    )
    (layer2): ResLayer(
      (0): Bottleneck(
        (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
        (downsample): Sequential(
          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)
          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (1): Bottleneck(
        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (2): Bottleneck(
        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (3): Bottleneck(
        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
    )
    (layer3): ResLayer(
      (0): Bottleneck(
        (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
        (downsample): Sequential(
          (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)
          (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (1): Bottleneck(
        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (2): Bottleneck(
        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (3): Bottleneck(
        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (4): Bottleneck(
        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (5): Bottleneck(
        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
    )
    (layer4): ResLayer(
      (0): Bottleneck(
        (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
        (downsample): Sequential(
          (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)
          (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (1): Bottleneck(
        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
      (2): Bottleneck(
        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        (relu): ReLU(inplace=True)
      )
      init_cfg={'type': 'Constant', 'val': 0, 'override': {'name': 'norm3'}}
    )
  )
  init_cfg={'type': 'Pretrained', 'checkpoint': './ckpts/resnet50-19c8e357.pth'}
  (img_neck): FPN(
    (lateral_convs): ModuleList(
      (0): ConvModule(
        (conv): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1))
      )
    )
    (fpn_convs): ModuleList(
      (0): ConvModule(
        (conv): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
      )
    )
  )
  init_cfg={'type': 'Xavier', 'layer': 'Conv2d', 'distribution': 'uniform'}
  (grid_mask): GridMask()
)
2026-02-26 13:17:33,305 - DriveTransformer - INFO - load checkpoint from path: ./work_dirs/drivetransformer_goalpoint_1gpu/epoch_6.pth
2026-02-26 13:17:35,160 - DriveTransformer - INFO - resumed epoch 6, iter 237168
2026-02-26 13:17:35,162 - DriveTransformer - INFO - Start running, host: humble@ysh-4090, work_dir: /home/Humble/Humble/EndtoEnd/Bench2DriveZoo/work_dirs/drivetransformer_goalpoint_1gpu
2026-02-26 13:17:35,162 - DriveTransformer - INFO - Hooks will be executed in the following order:
before_run:
(VERY_HIGH   ) CosineAnnealingLrUpdaterHook       
(ABOVE_NORMAL) GradientCumulativeFp16OptimizerHook
(NORMAL      ) CheckpointHook                     
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
before_train_epoch:
(VERY_HIGH   ) CosineAnnealingLrUpdaterHook       
(LOW         ) IterTimerHook                      
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
before_train_iter:
(VERY_HIGH   ) CosineAnnealingLrUpdaterHook       
(LOW         ) IterTimerHook                      
 -------------------- 
after_train_iter:
(ABOVE_NORMAL) GradientCumulativeFp16OptimizerHook
(NORMAL      ) CheckpointHook                     
(LOW         ) IterTimerHook                      
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
after_train_epoch:
(NORMAL      ) CheckpointHook                     
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
before_val_epoch:
(LOW         ) IterTimerHook                      
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
before_val_iter:
(LOW         ) IterTimerHook                      
 -------------------- 
after_val_iter:
(LOW         ) IterTimerHook                      
 -------------------- 
after_val_epoch:
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
after_run:
(VERY_LOW    ) TextLoggerHook                     
(VERY_LOW    ) TensorboardLoggerHook              
 -------------------- 
2026-02-26 13:17:35,162 - DriveTransformer - INFO - workflow: [('train', 1)], max: 24 epochs
2026-02-26 13:17:35,164 - DriveTransformer - INFO - Checkpoints will be saved to /home/Humble/Humble/EndtoEnd/Bench2DriveZoo/work_dirs/drivetransformer_goalpoint_1gpu by HardDiskBackend.
2026-02-26 13:17:44,251 - DriveTransformer - WARNING - Resume iter number is not divisible by cumulative_iters in GradientCumulativeOptimizerHook, which means the gradient of some iters is lost and the result may be influenced slightly.
2026-02-26 13:17:44,253 - DriveTransformer - WARNING - GradientCumulativeOptimizerHook may slightly decrease performance if the model has BatchNorm layers.
2026-02-26 13:18:22,810 - DriveTransformer - INFO - Epoch [7][50/39528]	lr: 8.550e-05, eta: 7 days, 19:50:10, time: 0.950, data_time: 0.174, memory: 19865, loss_cls: 0.1678, loss_bbox: 0.3657, loss_traj: 0.6574, loss_traj_cls: 0.0009, loss_map_cls: 0.1842, loss_map_pts: 0.7903, loss_map_dir: 0.0102, d0.loss_plan_l1_fix_time: 2.4778, d0.loss_plan_l1_fix_dist: 0.3595, d1.loss_plan_l1_fix_time: 0.9959, d1.loss_plan_l1_fix_dist: 0.1739, d2.loss_plan_l1_fix_time: 0.9809, d2.loss_plan_l1_fix_dist: 0.1825, d3.loss_plan_l1_fix_time: 0.9840, d3.loss_plan_l1_fix_dist: 0.1962, d4.loss_plan_l1_fix_time: 0.9892, d4.loss_plan_l1_fix_dist: 0.2123, d5.loss_plan_l1_fix_time: 0.9949, d5.loss_plan_l1_fix_dist: 0.2279, d6.loss_plan_l1_fix_time: 1.0014, d6.loss_plan_l1_fix_dist: 0.2453, d0.loss_cls_all: 0.2066, d0.loss_bbox_all: 0.4149, d0.loss_cls: 0.2066, d0.loss_bbox: 0.4149, d1.loss_cls: 0.1968, d1.loss_bbox: 0.3817, d2.loss_cls: 0.1843, d2.loss_bbox: 0.3650, d3.loss_cls: 0.1723, d3.loss_bbox: 0.3617, d4.loss_cls: 0.1675, d4.loss_bbox: 0.3614, d5.loss_cls: 0.1665, d5.loss_bbox: 0.3638, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7644, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.6680, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.6567, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6568, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.6553, d5.loss_traj_cls: 0.0013, d5.loss_traj: 0.6555, d0.loss_map_cls_all: 0.2617, d0.loss_map_pts_all: 1.4863, d0.loss_map_dir_all: 0.0094, d0.loss_map_cls: 0.2617, d0.loss_map_pts: 1.4863, d0.loss_map_dir: 0.0094, d1.loss_map_cls: 0.2316, d1.loss_map_pts: 1.0572, d1.loss_map_dir: 0.0094, d2.loss_map_cls: 0.2102, d2.loss_map_pts: 0.8935, d2.loss_map_dir: 0.0095, d3.loss_map_cls: 0.1947, d3.loss_map_pts: 0.8314, d3.loss_map_dir: 0.0099, d4.loss_map_cls: 0.1896, d4.loss_map_pts: 0.8052, d4.loss_map_dir: 0.0101, d5.loss_map_cls: 0.1863, d5.loss_map_pts: 0.7909, d5.loss_map_dir: 0.0102, loss: 29.1798, grad_norm: 130.8433
2026-02-26 13:19:02,279 - DriveTransformer - INFO - Epoch [7][100/39528]	lr: 8.550e-05, eta: 7 days, 3:54:48, time: 0.789, data_time: 0.032, memory: 19865, loss_cls: 0.1646, loss_bbox: 0.3689, loss_traj: 0.5997, loss_traj_cls: 0.0010, loss_map_cls: 0.1801, loss_map_pts: 0.7029, loss_map_dir: 0.0093, d0.loss_plan_l1_fix_time: 2.2596, d0.loss_plan_l1_fix_dist: 0.4237, d1.loss_plan_l1_fix_time: 1.1355, d1.loss_plan_l1_fix_dist: 0.1592, d2.loss_plan_l1_fix_time: 1.1257, d2.loss_plan_l1_fix_dist: 0.1680, d3.loss_plan_l1_fix_time: 1.1269, d3.loss_plan_l1_fix_dist: 0.1789, d4.loss_plan_l1_fix_time: 1.1312, d4.loss_plan_l1_fix_dist: 0.1917, d5.loss_plan_l1_fix_time: 1.1360, d5.loss_plan_l1_fix_dist: 0.2050, d6.loss_plan_l1_fix_time: 1.1412, d6.loss_plan_l1_fix_dist: 0.2185, d0.loss_cls_all: 0.1874, d0.loss_bbox_all: 0.4137, d0.loss_cls: 0.1874, d0.loss_bbox: 0.4137, d1.loss_cls: 0.1904, d1.loss_bbox: 0.3823, d2.loss_cls: 0.1786, d2.loss_bbox: 0.3659, d3.loss_cls: 0.1727, d3.loss_bbox: 0.3639, d4.loss_cls: 0.1665, d4.loss_bbox: 0.3668, d5.loss_cls: 0.1646, d5.loss_bbox: 0.3689, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7518, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.6204, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.5966, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.5903, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.5955, d5.loss_traj_cls: 0.0013, d5.loss_traj: 0.5938, d0.loss_map_cls_all: 0.2506, d0.loss_map_pts_all: 1.3629, d0.loss_map_dir_all: 0.0080, d0.loss_map_cls: 0.2506, d0.loss_map_pts: 1.3629, d0.loss_map_dir: 0.0080, d1.loss_map_cls: 0.2219, d1.loss_map_pts: 0.9581, d1.loss_map_dir: 0.0080, d2.loss_map_cls: 0.2007, d2.loss_map_pts: 0.7981, d2.loss_map_dir: 0.0084, d3.loss_map_cls: 0.1897, d3.loss_map_pts: 0.7304, d3.loss_map_dir: 0.0088, d4.loss_map_cls: 0.1845, d4.loss_map_pts: 0.7076, d4.loss_map_dir: 0.0091, d5.loss_map_cls: 0.1823, d5.loss_map_pts: 0.7039, d5.loss_map_dir: 0.0092, loss: 28.4697, grad_norm: 148.6855
2026-02-26 13:19:41,444 - DriveTransformer - INFO - Epoch [7][150/39528]	lr: 8.550e-05, eta: 6 days, 22:11:24, time: 0.783, data_time: 0.038, memory: 19865, loss_cls: 0.1925, loss_bbox: 0.4012, loss_traj: 0.6996, loss_traj_cls: 0.0009, loss_map_cls: 0.1695, loss_map_pts: 0.8311, loss_map_dir: 0.0125, d0.loss_plan_l1_fix_time: 2.7957, d0.loss_plan_l1_fix_dist: 0.3557, d1.loss_plan_l1_fix_time: 1.3117, d1.loss_plan_l1_fix_dist: 0.1958, d2.loss_plan_l1_fix_time: 1.2946, d2.loss_plan_l1_fix_dist: 0.1981, d3.loss_plan_l1_fix_time: 1.2931, d3.loss_plan_l1_fix_dist: 0.2058, d4.loss_plan_l1_fix_time: 1.2968, d4.loss_plan_l1_fix_dist: 0.2142, d5.loss_plan_l1_fix_time: 1.3015, d5.loss_plan_l1_fix_dist: 0.2208, d6.loss_plan_l1_fix_time: 1.3060, d6.loss_plan_l1_fix_dist: 0.2255, d0.loss_cls_all: 0.2244, d0.loss_bbox_all: 0.4647, d0.loss_cls: 0.2244, d0.loss_bbox: 0.4647, d1.loss_cls: 0.2177, d1.loss_bbox: 0.4265, d2.loss_cls: 0.2057, d2.loss_bbox: 0.4089, d3.loss_cls: 0.1953, d3.loss_bbox: 0.4030, d4.loss_cls: 0.1935, d4.loss_bbox: 0.4005, d5.loss_cls: 0.1923, d5.loss_bbox: 0.4003, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.9264, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.7359, d2.loss_traj_cls: 0.0017, d2.loss_traj: 0.7039, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6909, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.6925, d5.loss_traj_cls: 0.0012, d5.loss_traj: 0.6980, d0.loss_map_cls_all: 0.2448, d0.loss_map_pts_all: 1.6646, d0.loss_map_dir_all: 0.0114, d0.loss_map_cls: 0.2448, d0.loss_map_pts: 1.6646, d0.loss_map_dir: 0.0114, d1.loss_map_cls: 0.2142, d1.loss_map_pts: 1.1453, d1.loss_map_dir: 0.0112, d2.loss_map_cls: 0.1925, d2.loss_map_pts: 0.9484, d2.loss_map_dir: 0.0116, d3.loss_map_cls: 0.1814, d3.loss_map_pts: 0.8693, d3.loss_map_dir: 0.0122, d4.loss_map_cls: 0.1741, d4.loss_map_pts: 0.8396, d4.loss_map_dir: 0.0124, d5.loss_map_cls: 0.1732, d5.loss_map_pts: 0.8313, d5.loss_map_dir: 0.0125, loss: 32.8692, grad_norm: 168.0135
2026-02-26 13:20:19,301 - DriveTransformer - INFO - Epoch [7][200/39528]	lr: 8.550e-05, eta: 6 days, 18:01:59, time: 0.757, data_time: 0.031, memory: 19865, loss_cls: 0.1436, loss_bbox: 0.3453, loss_traj: 0.5037, loss_traj_cls: 0.0010, loss_map_cls: 0.1383, loss_map_pts: 0.6229, loss_map_dir: 0.0111, d0.loss_plan_l1_fix_time: 2.2557, d0.loss_plan_l1_fix_dist: 0.4571, d1.loss_plan_l1_fix_time: 0.9609, d1.loss_plan_l1_fix_dist: 0.1672, d2.loss_plan_l1_fix_time: 0.9445, d2.loss_plan_l1_fix_dist: 0.1680, d3.loss_plan_l1_fix_time: 0.9389, d3.loss_plan_l1_fix_dist: 0.1751, d4.loss_plan_l1_fix_time: 0.9442, d4.loss_plan_l1_fix_dist: 0.1834, d5.loss_plan_l1_fix_time: 0.9509, d5.loss_plan_l1_fix_dist: 0.1924, d6.loss_plan_l1_fix_time: 0.9582, d6.loss_plan_l1_fix_dist: 0.2033, d0.loss_cls_all: 0.1819, d0.loss_bbox_all: 0.3789, d0.loss_cls: 0.1819, d0.loss_bbox: 0.3789, d1.loss_cls: 0.1700, d1.loss_bbox: 0.3503, d2.loss_cls: 0.1543, d2.loss_bbox: 0.3406, d3.loss_cls: 0.1477, d3.loss_bbox: 0.3372, d4.loss_cls: 0.1438, d4.loss_bbox: 0.3381, d5.loss_cls: 0.1420, d5.loss_bbox: 0.3436, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.6438, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.5324, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.5038, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.4938, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.4981, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.4985, d0.loss_map_cls_all: 0.2104, d0.loss_map_pts_all: 1.4049, d0.loss_map_dir_all: 0.0108, d0.loss_map_cls: 0.2104, d0.loss_map_pts: 1.4049, d0.loss_map_dir: 0.0108, d1.loss_map_cls: 0.1785, d1.loss_map_pts: 0.9015, d1.loss_map_dir: 0.0103, d2.loss_map_cls: 0.1582, d2.loss_map_pts: 0.7037, d2.loss_map_dir: 0.0103, d3.loss_map_cls: 0.1488, d3.loss_map_pts: 0.6423, d3.loss_map_dir: 0.0107, d4.loss_map_cls: 0.1418, d4.loss_map_pts: 0.6221, d4.loss_map_dir: 0.0109, d5.loss_map_cls: 0.1388, d5.loss_map_pts: 0.6186, d5.loss_map_dir: 0.0110, loss: 25.5920, grad_norm: 184.9270
2026-02-26 13:20:56,846 - DriveTransformer - INFO - Epoch [7][250/39528]	lr: 8.550e-05, eta: 6 days, 15:17:20, time: 0.751, data_time: 0.034, memory: 19865, loss_cls: 0.1582, loss_bbox: 0.3690, loss_traj: 0.5757, loss_traj_cls: 0.0011, loss_map_cls: 0.1538, loss_map_pts: 0.7075, loss_map_dir: 0.0104, d0.loss_plan_l1_fix_time: 2.3771, d0.loss_plan_l1_fix_dist: 0.4086, d1.loss_plan_l1_fix_time: 0.9549, d1.loss_plan_l1_fix_dist: 0.2080, d2.loss_plan_l1_fix_time: 0.9525, d2.loss_plan_l1_fix_dist: 0.2094, d3.loss_plan_l1_fix_time: 0.9579, d3.loss_plan_l1_fix_dist: 0.2204, d4.loss_plan_l1_fix_time: 0.9645, d4.loss_plan_l1_fix_dist: 0.2342, d5.loss_plan_l1_fix_time: 0.9715, d5.loss_plan_l1_fix_dist: 0.2510, d6.loss_plan_l1_fix_time: 0.9793, d6.loss_plan_l1_fix_dist: 0.2677, d0.loss_cls_all: 0.1998, d0.loss_bbox_all: 0.4092, d0.loss_cls: 0.1998, d0.loss_bbox: 0.4092, d1.loss_cls: 0.1881, d1.loss_bbox: 0.3754, d2.loss_cls: 0.1721, d2.loss_bbox: 0.3650, d3.loss_cls: 0.1601, d3.loss_bbox: 0.3644, d4.loss_cls: 0.1577, d4.loss_bbox: 0.3638, d5.loss_cls: 0.1574, d5.loss_bbox: 0.3667, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7401, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.6159, d2.loss_traj_cls: 0.0019, d2.loss_traj: 0.5736, d3.loss_traj_cls: 0.0013, d3.loss_traj: 0.5656, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.5729, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.5700, d0.loss_map_cls_all: 0.2141, d0.loss_map_pts_all: 1.3798, d0.loss_map_dir_all: 0.0095, d0.loss_map_cls: 0.2141, d0.loss_map_pts: 1.3798, d0.loss_map_dir: 0.0095, d1.loss_map_cls: 0.1949, d1.loss_map_pts: 0.9627, d1.loss_map_dir: 0.0092, d2.loss_map_cls: 0.1765, d2.loss_map_pts: 0.7935, d2.loss_map_dir: 0.0095, d3.loss_map_cls: 0.1661, d3.loss_map_pts: 0.7355, d3.loss_map_dir: 0.0101, d4.loss_map_cls: 0.1584, d4.loss_map_pts: 0.7163, d4.loss_map_dir: 0.0104, d5.loss_map_cls: 0.1547, d5.loss_map_pts: 0.7063, d5.loss_map_dir: 0.0105, loss: 27.4878, grad_norm: 220.8222
2026-02-26 13:21:33,494 - DriveTransformer - INFO - Epoch [7][300/39528]	lr: 8.550e-05, eta: 6 days, 12:51:52, time: 0.733, data_time: 0.030, memory: 19865, loss_cls: 0.1808, loss_bbox: 0.4054, loss_traj: 0.8366, loss_traj_cls: 0.0010, loss_map_cls: 0.1816, loss_map_pts: 0.7916, loss_map_dir: 0.0099, d0.loss_plan_l1_fix_time: 2.5910, d0.loss_plan_l1_fix_dist: 0.4304, d1.loss_plan_l1_fix_time: 1.1923, d1.loss_plan_l1_fix_dist: 0.2111, d2.loss_plan_l1_fix_time: 1.1724, d2.loss_plan_l1_fix_dist: 0.2057, d3.loss_plan_l1_fix_time: 1.1700, d3.loss_plan_l1_fix_dist: 0.2125, d4.loss_plan_l1_fix_time: 1.1736, d4.loss_plan_l1_fix_dist: 0.2228, d5.loss_plan_l1_fix_time: 1.1775, d5.loss_plan_l1_fix_dist: 0.2365, d6.loss_plan_l1_fix_time: 1.1819, d6.loss_plan_l1_fix_dist: 0.2524, d0.loss_cls_all: 0.2186, d0.loss_bbox_all: 0.4629, d0.loss_cls: 0.2186, d0.loss_bbox: 0.4629, d1.loss_cls: 0.2107, d1.loss_bbox: 0.4271, d2.loss_cls: 0.1976, d2.loss_bbox: 0.4123, d3.loss_cls: 0.1847, d3.loss_bbox: 0.4053, d4.loss_cls: 0.1817, d4.loss_bbox: 0.4043, d5.loss_cls: 0.1807, d5.loss_bbox: 0.4038, d0.loss_traj_cls: 0.0001, d0.loss_traj: 1.1118, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.9047, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.8310, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.8124, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.8264, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.8291, d0.loss_map_cls_all: 0.2594, d0.loss_map_pts_all: 1.5148, d0.loss_map_dir_all: 0.0091, d0.loss_map_cls: 0.2594, d0.loss_map_pts: 1.5148, d0.loss_map_dir: 0.0091, d1.loss_map_cls: 0.2272, d1.loss_map_pts: 1.0887, d1.loss_map_dir: 0.0089, d2.loss_map_cls: 0.2056, d2.loss_map_pts: 0.8954, d2.loss_map_dir: 0.0091, d3.loss_map_cls: 0.1950, d3.loss_map_pts: 0.8184, d3.loss_map_dir: 0.0095, d4.loss_map_cls: 0.1872, d4.loss_map_pts: 0.7957, d4.loss_map_dir: 0.0097, d5.loss_map_cls: 0.1831, d5.loss_map_pts: 0.7900, d5.loss_map_dir: 0.0099, loss: 32.5303, grad_norm: 219.1332
2026-02-26 13:22:09,860 - DriveTransformer - INFO - Epoch [7][350/39528]	lr: 8.550e-05, eta: 6 days, 10:58:18, time: 0.727, data_time: 0.035, memory: 19865, loss_cls: 0.1590, loss_bbox: 0.3686, loss_traj: 0.7263, loss_traj_cls: 0.0010, loss_map_cls: 0.1308, loss_map_pts: 0.6419, loss_map_dir: 0.0078, d0.loss_plan_l1_fix_time: 2.5855, d0.loss_plan_l1_fix_dist: 0.4401, d1.loss_plan_l1_fix_time: 1.0806, d1.loss_plan_l1_fix_dist: 0.1810, d2.loss_plan_l1_fix_time: 1.0574, d2.loss_plan_l1_fix_dist: 0.1757, d3.loss_plan_l1_fix_time: 1.0579, d3.loss_plan_l1_fix_dist: 0.1848, d4.loss_plan_l1_fix_time: 1.0612, d4.loss_plan_l1_fix_dist: 0.1971, d5.loss_plan_l1_fix_time: 1.0646, d5.loss_plan_l1_fix_dist: 0.2099, d6.loss_plan_l1_fix_time: 1.0681, d6.loss_plan_l1_fix_dist: 0.2261, d0.loss_cls_all: 0.1989, d0.loss_bbox_all: 0.4250, d0.loss_cls: 0.1989, d0.loss_bbox: 0.4250, d1.loss_cls: 0.1864, d1.loss_bbox: 0.3929, d2.loss_cls: 0.1766, d2.loss_bbox: 0.3722, d3.loss_cls: 0.1666, d3.loss_bbox: 0.3659, d4.loss_cls: 0.1615, d4.loss_bbox: 0.3651, d5.loss_cls: 0.1596, d5.loss_bbox: 0.3677, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.9087, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.7741, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.7304, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.7142, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.7210, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.7236, d0.loss_map_cls_all: 0.1974, d0.loss_map_pts_all: 1.2846, d0.loss_map_dir_all: 0.0071, d0.loss_map_cls: 0.1974, d0.loss_map_pts: 1.2846, d0.loss_map_dir: 0.0071, d1.loss_map_cls: 0.1712, d1.loss_map_pts: 0.9043, d1.loss_map_dir: 0.0069, d2.loss_map_cls: 0.1521, d2.loss_map_pts: 0.7424, d2.loss_map_dir: 0.0070, d3.loss_map_cls: 0.1421, d3.loss_map_pts: 0.6762, d3.loss_map_dir: 0.0074, d4.loss_map_cls: 0.1360, d4.loss_map_pts: 0.6504, d4.loss_map_dir: 0.0077, d5.loss_map_cls: 0.1314, d5.loss_map_pts: 0.6423, d5.loss_map_dir: 0.0078, loss: 28.5301, grad_norm: 205.1435
2026-02-26 13:22:48,660 - DriveTransformer - INFO - Epoch [7][400/39528]	lr: 8.550e-05, eta: 6 days, 10:47:05, time: 0.777, data_time: 0.035, memory: 19865, loss_cls: 0.2099, loss_bbox: 0.4190, loss_traj: 0.6659, loss_traj_cls: 0.0010, loss_map_cls: 0.1963, loss_map_pts: 0.9068, loss_map_dir: 0.0129, d0.loss_plan_l1_fix_time: 2.7034, d0.loss_plan_l1_fix_dist: 0.3820, d1.loss_plan_l1_fix_time: 1.1466, d1.loss_plan_l1_fix_dist: 0.2028, d2.loss_plan_l1_fix_time: 1.1293, d2.loss_plan_l1_fix_dist: 0.2047, d3.loss_plan_l1_fix_time: 1.1280, d3.loss_plan_l1_fix_dist: 0.2147, d4.loss_plan_l1_fix_time: 1.1313, d4.loss_plan_l1_fix_dist: 0.2262, d5.loss_plan_l1_fix_time: 1.1350, d5.loss_plan_l1_fix_dist: 0.2386, d6.loss_plan_l1_fix_time: 1.1395, d6.loss_plan_l1_fix_dist: 0.2524, d0.loss_cls_all: 0.2493, d0.loss_bbox_all: 0.4822, d0.loss_cls: 0.2493, d0.loss_bbox: 0.4822, d1.loss_cls: 0.2423, d1.loss_bbox: 0.4451, d2.loss_cls: 0.2283, d2.loss_bbox: 0.4264, d3.loss_cls: 0.2187, d3.loss_bbox: 0.4201, d4.loss_cls: 0.2108, d4.loss_bbox: 0.4201, d5.loss_cls: 0.2102, d5.loss_bbox: 0.4196, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.9040, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.7624, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.6886, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6729, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.6718, d5.loss_traj_cls: 0.0013, d5.loss_traj: 0.6652, d0.loss_map_cls_all: 0.2693, d0.loss_map_pts_all: 1.8011, d0.loss_map_dir_all: 0.0111, d0.loss_map_cls: 0.2693, d0.loss_map_pts: 1.8011, d0.loss_map_dir: 0.0111, d1.loss_map_cls: 0.2470, d1.loss_map_pts: 1.2614, d1.loss_map_dir: 0.0111, d2.loss_map_cls: 0.2231, d2.loss_map_pts: 1.0276, d2.loss_map_dir: 0.0116, d3.loss_map_cls: 0.2136, d3.loss_map_pts: 0.9451, d3.loss_map_dir: 0.0123, d4.loss_map_cls: 0.2036, d4.loss_map_pts: 0.9163, d4.loss_map_dir: 0.0126, d5.loss_map_cls: 0.1989, d5.loss_map_pts: 0.9076, d5.loss_map_dir: 0.0127, loss: 33.0904, grad_norm: 197.1436
2026-02-26 13:23:26,335 - DriveTransformer - INFO - Epoch [7][450/39528]	lr: 8.550e-05, eta: 6 days, 10:04:49, time: 0.752, data_time: 0.031, memory: 19865, loss_cls: 0.1616, loss_bbox: 0.3416, loss_traj: 0.6352, loss_traj_cls: 0.0010, loss_map_cls: 0.1518, loss_map_pts: 0.6879, loss_map_dir: 0.0113, d0.loss_plan_l1_fix_time: 2.4292, d0.loss_plan_l1_fix_dist: 0.4253, d1.loss_plan_l1_fix_time: 1.0334, d1.loss_plan_l1_fix_dist: 0.1985, d2.loss_plan_l1_fix_time: 1.0212, d2.loss_plan_l1_fix_dist: 0.2002, d3.loss_plan_l1_fix_time: 1.0258, d3.loss_plan_l1_fix_dist: 0.2101, d4.loss_plan_l1_fix_time: 1.0311, d4.loss_plan_l1_fix_dist: 0.2216, d5.loss_plan_l1_fix_time: 1.0371, d5.loss_plan_l1_fix_dist: 0.2347, d6.loss_plan_l1_fix_time: 1.0436, d6.loss_plan_l1_fix_dist: 0.2491, d0.loss_cls_all: 0.1985, d0.loss_bbox_all: 0.3924, d0.loss_cls: 0.1985, d0.loss_bbox: 0.3924, d1.loss_cls: 0.1896, d1.loss_bbox: 0.3639, d2.loss_cls: 0.1748, d2.loss_bbox: 0.3499, d3.loss_cls: 0.1681, d3.loss_bbox: 0.3426, d4.loss_cls: 0.1626, d4.loss_bbox: 0.3418, d5.loss_cls: 0.1610, d5.loss_bbox: 0.3413, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.8350, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.6900, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.6359, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6307, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.6265, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.6264, d0.loss_map_cls_all: 0.2286, d0.loss_map_pts_all: 1.4912, d0.loss_map_dir_all: 0.0112, d0.loss_map_cls: 0.2286, d0.loss_map_pts: 1.4912, d0.loss_map_dir: 0.0112, d1.loss_map_cls: 0.1925, d1.loss_map_pts: 0.9879, d1.loss_map_dir: 0.0105, d2.loss_map_cls: 0.1711, d2.loss_map_pts: 0.7861, d2.loss_map_dir: 0.0107, d3.loss_map_cls: 0.1618, d3.loss_map_pts: 0.7138, d3.loss_map_dir: 0.0109, d4.loss_map_cls: 0.1569, d4.loss_map_pts: 0.6910, d4.loss_map_dir: 0.0112, d5.loss_map_cls: 0.1534, d5.loss_map_pts: 0.6873, d5.loss_map_dir: 0.0112, loss: 28.3984, grad_norm: 191.9594
2026-02-26 13:24:00,308 - DriveTransformer - INFO - Epoch [7][500/39528]	lr: 8.550e-05, eta: 6 days, 8:04:56, time: 0.679, data_time: 0.031, memory: 19865, loss_cls: 0.1766, loss_bbox: 0.3669, loss_traj: 0.7793, loss_traj_cls: 0.0010, loss_map_cls: 0.1660, loss_map_pts: 0.6652, loss_map_dir: 0.0075, d0.loss_plan_l1_fix_time: 3.1668, d0.loss_plan_l1_fix_dist: 0.3692, d1.loss_plan_l1_fix_time: 1.2056, d1.loss_plan_l1_fix_dist: 0.1642, d2.loss_plan_l1_fix_time: 1.1909, d2.loss_plan_l1_fix_dist: 0.1673, d3.loss_plan_l1_fix_time: 1.1899, d3.loss_plan_l1_fix_dist: 0.1770, d4.loss_plan_l1_fix_time: 1.1926, d4.loss_plan_l1_fix_dist: 0.1886, d5.loss_plan_l1_fix_time: 1.1958, d5.loss_plan_l1_fix_dist: 0.2009, d6.loss_plan_l1_fix_time: 1.1994, d6.loss_plan_l1_fix_dist: 0.2137, d0.loss_cls_all: 0.2153, d0.loss_bbox_all: 0.4302, d0.loss_cls: 0.2153, d0.loss_bbox: 0.4302, d1.loss_cls: 0.2074, d1.loss_bbox: 0.3929, d2.loss_cls: 0.1938, d2.loss_bbox: 0.3758, d3.loss_cls: 0.1856, d3.loss_bbox: 0.3688, d4.loss_cls: 0.1789, d4.loss_bbox: 0.3688, d5.loss_cls: 0.1780, d5.loss_bbox: 0.3664, d0.loss_traj_cls: 0.0001, d0.loss_traj: 1.0168, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.8361, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.7865, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.7811, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.7759, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.7731, d0.loss_map_cls_all: 0.2483, d0.loss_map_pts_all: 1.2565, d0.loss_map_dir_all: 0.0062, d0.loss_map_cls: 0.2483, d0.loss_map_pts: 1.2565, d0.loss_map_dir: 0.0062, d1.loss_map_cls: 0.2156, d1.loss_map_pts: 0.9196, d1.loss_map_dir: 0.0065, d2.loss_map_cls: 0.1958, d2.loss_map_pts: 0.7653, d2.loss_map_dir: 0.0068, d3.loss_map_cls: 0.1812, d3.loss_map_pts: 0.6994, d3.loss_map_dir: 0.0072, d4.loss_map_cls: 0.1721, d4.loss_map_pts: 0.6768, d4.loss_map_dir: 0.0073, d5.loss_map_cls: 0.1667, d5.loss_map_pts: 0.6667, d5.loss_map_dir: 0.0075, loss: 30.7847, grad_norm: 182.3264
2026-02-26 13:24:39,122 - DriveTransformer - INFO - Epoch [7][550/39528]	lr: 8.550e-05, eta: 6 days, 8:11:00, time: 0.776, data_time: 0.033, memory: 19865, loss_cls: 0.1649, loss_bbox: 0.3690, loss_traj: 0.6848, loss_traj_cls: 0.0010, loss_map_cls: 0.1544, loss_map_pts: 0.7949, loss_map_dir: 0.0112, d0.loss_plan_l1_fix_time: 2.4982, d0.loss_plan_l1_fix_dist: 0.4267, d1.loss_plan_l1_fix_time: 1.1892, d1.loss_plan_l1_fix_dist: 0.1871, d2.loss_plan_l1_fix_time: 1.1751, d2.loss_plan_l1_fix_dist: 0.1904, d3.loss_plan_l1_fix_time: 1.1681, d3.loss_plan_l1_fix_dist: 0.1985, d4.loss_plan_l1_fix_time: 1.1706, d4.loss_plan_l1_fix_dist: 0.2080, d5.loss_plan_l1_fix_time: 1.1740, d5.loss_plan_l1_fix_dist: 0.2196, d6.loss_plan_l1_fix_time: 1.1784, d6.loss_plan_l1_fix_dist: 0.2327, d0.loss_cls_all: 0.2020, d0.loss_bbox_all: 0.4310, d0.loss_cls: 0.2020, d0.loss_bbox: 0.4310, d1.loss_cls: 0.1951, d1.loss_bbox: 0.3998, d2.loss_cls: 0.1805, d2.loss_bbox: 0.3814, d3.loss_cls: 0.1714, d3.loss_bbox: 0.3728, d4.loss_cls: 0.1669, d4.loss_bbox: 0.3698, d5.loss_cls: 0.1652, d5.loss_bbox: 0.3698, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.8968, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.7239, d2.loss_traj_cls: 0.0017, d2.loss_traj: 0.6738, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6776, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.6762, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.6810, d0.loss_map_cls_all: 0.2341, d0.loss_map_pts_all: 1.6139, d0.loss_map_dir_all: 0.0105, d0.loss_map_cls: 0.2341, d0.loss_map_pts: 1.6139, d0.loss_map_dir: 0.0105, d1.loss_map_cls: 0.2035, d1.loss_map_pts: 1.1359, d1.loss_map_dir: 0.0102, d2.loss_map_cls: 0.1827, d2.loss_map_pts: 0.9203, d2.loss_map_dir: 0.0102, d3.loss_map_cls: 0.1694, d3.loss_map_pts: 0.8402, d3.loss_map_dir: 0.0106, d4.loss_map_cls: 0.1612, d4.loss_map_pts: 0.8105, d4.loss_map_dir: 0.0110, d5.loss_map_cls: 0.1567, d5.loss_map_pts: 0.7970, d5.loss_map_dir: 0.0111, loss: 30.9193, grad_norm: 183.4456
2026-02-26 13:25:16,096 - DriveTransformer - INFO - Epoch [7][600/39528]	lr: 8.550e-05, eta: 6 days, 7:39:39, time: 0.739, data_time: 0.032, memory: 19865, loss_cls: 0.1609, loss_bbox: 0.3604, loss_traj: 0.7898, loss_traj_cls: 0.0010, loss_map_cls: 0.1596, loss_map_pts: 0.6583, loss_map_dir: 0.0106, d0.loss_plan_l1_fix_time: 2.8711, d0.loss_plan_l1_fix_dist: 0.4110, d1.loss_plan_l1_fix_time: 1.1065, d1.loss_plan_l1_fix_dist: 0.1867, d2.loss_plan_l1_fix_time: 1.0984, d2.loss_plan_l1_fix_dist: 0.1904, d3.loss_plan_l1_fix_time: 1.0994, d3.loss_plan_l1_fix_dist: 0.1977, d4.loss_plan_l1_fix_time: 1.1031, d4.loss_plan_l1_fix_dist: 0.2086, d5.loss_plan_l1_fix_time: 1.1073, d5.loss_plan_l1_fix_dist: 0.2205, d6.loss_plan_l1_fix_time: 1.1127, d6.loss_plan_l1_fix_dist: 0.2356, d0.loss_cls_all: 0.2020, d0.loss_bbox_all: 0.4147, d0.loss_cls: 0.2020, d0.loss_bbox: 0.4147, d1.loss_cls: 0.1945, d1.loss_bbox: 0.3821, d2.loss_cls: 0.1828, d2.loss_bbox: 0.3668, d3.loss_cls: 0.1688, d3.loss_bbox: 0.3625, d4.loss_cls: 0.1608, d4.loss_bbox: 0.3629, d5.loss_cls: 0.1600, d5.loss_bbox: 0.3617, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.9900, d1.loss_traj_cls: 0.0011, d1.loss_traj: 0.8165, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.7652, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.7753, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.7771, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.7807, d0.loss_map_cls_all: 0.2384, d0.loss_map_pts_all: 1.4000, d0.loss_map_dir_all: 0.0102, d0.loss_map_cls: 0.2384, d0.loss_map_pts: 1.4000, d0.loss_map_dir: 0.0102, d1.loss_map_cls: 0.2020, d1.loss_map_pts: 0.9463, d1.loss_map_dir: 0.0098, d2.loss_map_cls: 0.1818, d2.loss_map_pts: 0.7555, d2.loss_map_dir: 0.0099, d3.loss_map_cls: 0.1713, d3.loss_map_pts: 0.6881, d3.loss_map_dir: 0.0103, d4.loss_map_cls: 0.1656, d4.loss_map_pts: 0.6648, d4.loss_map_dir: 0.0105, d5.loss_map_cls: 0.1616, d5.loss_map_pts: 0.6577, d5.loss_map_dir: 0.0105, loss: 30.0805, grad_norm: 183.4591
2026-02-26 13:25:53,540 - DriveTransformer - INFO - Epoch [7][650/39528]	lr: 8.550e-05, eta: 6 days, 7:21:30, time: 0.749, data_time: 0.036, memory: 19865, loss_cls: 0.1715, loss_bbox: 0.3758, loss_traj: 0.6941, loss_traj_cls: 0.0011, loss_map_cls: 0.1593, loss_map_pts: 0.7106, loss_map_dir: 0.0095, d0.loss_plan_l1_fix_time: 2.8005, d0.loss_plan_l1_fix_dist: 0.3746, d1.loss_plan_l1_fix_time: 1.1152, d1.loss_plan_l1_fix_dist: 0.2058, d2.loss_plan_l1_fix_time: 1.1073, d2.loss_plan_l1_fix_dist: 0.2003, d3.loss_plan_l1_fix_time: 1.1062, d3.loss_plan_l1_fix_dist: 0.2064, d4.loss_plan_l1_fix_time: 1.1102, d4.loss_plan_l1_fix_dist: 0.2154, d5.loss_plan_l1_fix_time: 1.1147, d5.loss_plan_l1_fix_dist: 0.2254, d6.loss_plan_l1_fix_time: 1.1201, d6.loss_plan_l1_fix_dist: 0.2360, d0.loss_cls_all: 0.2084, d0.loss_bbox_all: 0.4365, d0.loss_cls: 0.2084, d0.loss_bbox: 0.4365, d1.loss_cls: 0.2020, d1.loss_bbox: 0.3983, d2.loss_cls: 0.1882, d2.loss_bbox: 0.3813, d3.loss_cls: 0.1771, d3.loss_bbox: 0.3763, d4.loss_cls: 0.1731, d4.loss_bbox: 0.3755, d5.loss_cls: 0.1719, d5.loss_bbox: 0.3755, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.9684, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.7502, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.6852, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6925, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.6805, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.6829, d0.loss_map_cls_all: 0.2362, d0.loss_map_pts_all: 1.3807, d0.loss_map_dir_all: 0.0093, d0.loss_map_cls: 0.2362, d0.loss_map_pts: 1.3807, d0.loss_map_dir: 0.0093, d1.loss_map_cls: 0.2116, d1.loss_map_pts: 0.9730, d1.loss_map_dir: 0.0086, d2.loss_map_cls: 0.1856, d2.loss_map_pts: 0.8115, d2.loss_map_dir: 0.0087, d3.loss_map_cls: 0.1757, d3.loss_map_pts: 0.7471, d3.loss_map_dir: 0.0092, d4.loss_map_cls: 0.1680, d4.loss_map_pts: 0.7218, d4.loss_map_dir: 0.0093, d5.loss_map_cls: 0.1613, d5.loss_map_pts: 0.7115, d5.loss_map_dir: 0.0095, loss: 30.0003, grad_norm: 181.8209
2026-02-26 13:26:30,647 - DriveTransformer - INFO - Epoch [7][700/39528]	lr: 8.550e-05, eta: 6 days, 7:00:10, time: 0.742, data_time: 0.030, memory: 19865, loss_cls: 0.1715, loss_bbox: 0.3810, loss_traj: 0.8147, loss_traj_cls: 0.0011, loss_map_cls: 0.1364, loss_map_pts: 0.6120, loss_map_dir: 0.0084, d0.loss_plan_l1_fix_time: 2.4365, d0.loss_plan_l1_fix_dist: 0.4542, d1.loss_plan_l1_fix_time: 1.0783, d1.loss_plan_l1_fix_dist: 0.1880, d2.loss_plan_l1_fix_time: 1.0686, d2.loss_plan_l1_fix_dist: 0.1878, d3.loss_plan_l1_fix_time: 1.0702, d3.loss_plan_l1_fix_dist: 0.1983, d4.loss_plan_l1_fix_time: 1.0748, d4.loss_plan_l1_fix_dist: 0.2133, d5.loss_plan_l1_fix_time: 1.0793, d5.loss_plan_l1_fix_dist: 0.2321, d6.loss_plan_l1_fix_time: 1.0840, d6.loss_plan_l1_fix_dist: 0.2519, d0.loss_cls_all: 0.2056, d0.loss_bbox_all: 0.4304, d0.loss_cls: 0.2056, d0.loss_bbox: 0.4304, d1.loss_cls: 0.2009, d1.loss_bbox: 0.4000, d2.loss_cls: 0.1892, d2.loss_bbox: 0.3856, d3.loss_cls: 0.1799, d3.loss_bbox: 0.3797, d4.loss_cls: 0.1747, d4.loss_bbox: 0.3788, d5.loss_cls: 0.1724, d5.loss_bbox: 0.3805, d0.loss_traj_cls: 0.0001, d0.loss_traj: 1.1088, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.9014, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.8371, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.8161, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.8116, d5.loss_traj_cls: 0.0015, d5.loss_traj: 0.8088, d0.loss_map_cls_all: 0.1950, d0.loss_map_pts_all: 1.2376, d0.loss_map_dir_all: 0.0083, d0.loss_map_cls: 0.1950, d0.loss_map_pts: 1.2376, d0.loss_map_dir: 0.0083, d1.loss_map_cls: 0.1744, d1.loss_map_pts: 0.8613, d1.loss_map_dir: 0.0080, d2.loss_map_cls: 0.1569, d2.loss_map_pts: 0.7095, d2.loss_map_dir: 0.0080, d3.loss_map_cls: 0.1492, d3.loss_map_pts: 0.6443, d3.loss_map_dir: 0.0082, d4.loss_map_cls: 0.1426, d4.loss_map_pts: 0.6200, d4.loss_map_dir: 0.0084, d5.loss_map_cls: 0.1372, d5.loss_map_pts: 0.6142, d5.loss_map_dir: 0.0084, loss: 29.2794, grad_norm: 185.4367
2026-02-26 13:27:09,151 - DriveTransformer - INFO - Epoch [7][750/39528]	lr: 8.550e-05, eta: 6 days, 7:03:43, time: 0.770, data_time: 0.037, memory: 19865, loss_cls: 0.1460, loss_bbox: 0.3344, loss_traj: 0.5343, loss_traj_cls: 0.0011, loss_map_cls: 0.1465, loss_map_pts: 0.6612, loss_map_dir: 0.0103, d0.loss_plan_l1_fix_time: 2.3031, d0.loss_plan_l1_fix_dist: 0.3969, d1.loss_plan_l1_fix_time: 0.9815, d1.loss_plan_l1_fix_dist: 0.1567, d2.loss_plan_l1_fix_time: 0.9689, d2.loss_plan_l1_fix_dist: 0.1616, d3.loss_plan_l1_fix_time: 0.9680, d3.loss_plan_l1_fix_dist: 0.1748, d4.loss_plan_l1_fix_time: 0.9731, d4.loss_plan_l1_fix_dist: 0.1875, d5.loss_plan_l1_fix_time: 0.9787, d5.loss_plan_l1_fix_dist: 0.2010, d6.loss_plan_l1_fix_time: 0.9846, d6.loss_plan_l1_fix_dist: 0.2141, d0.loss_cls_all: 0.1728, d0.loss_bbox_all: 0.3867, d0.loss_cls: 0.1728, d0.loss_bbox: 0.3867, d1.loss_cls: 0.1688, d1.loss_bbox: 0.3566, d2.loss_cls: 0.1592, d2.loss_bbox: 0.3428, d3.loss_cls: 0.1523, d3.loss_bbox: 0.3374, d4.loss_cls: 0.1480, d4.loss_bbox: 0.3359, d5.loss_cls: 0.1469, d5.loss_bbox: 0.3361, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.6793, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.5581, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.5369, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.5360, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.5381, d5.loss_traj_cls: 0.0015, d5.loss_traj: 0.5382, d0.loss_map_cls_all: 0.2080, d0.loss_map_pts_all: 1.3798, d0.loss_map_dir_all: 0.0092, d0.loss_map_cls: 0.2080, d0.loss_map_pts: 1.3798, d0.loss_map_dir: 0.0092, d1.loss_map_cls: 0.1823, d1.loss_map_pts: 0.9517, d1.loss_map_dir: 0.0087, d2.loss_map_cls: 0.1673, d2.loss_map_pts: 0.7672, d2.loss_map_dir: 0.0093, d3.loss_map_cls: 0.1584, d3.loss_map_pts: 0.6987, d3.loss_map_dir: 0.0099, d4.loss_map_cls: 0.1541, d4.loss_map_pts: 0.6707, d4.loss_map_dir: 0.0102, d5.loss_map_cls: 0.1476, d5.loss_map_pts: 0.6639, d5.loss_map_dir: 0.0103, loss: 26.2855, grad_norm: 185.2696
2026-02-26 13:27:47,697 - DriveTransformer - INFO - Epoch [7][800/39528]	lr: 8.550e-05, eta: 6 days, 7:07:20, time: 0.771, data_time: 0.034, memory: 19865, loss_cls: 0.1955, loss_bbox: 0.3897, loss_traj: 0.6169, loss_traj_cls: 0.0011, loss_map_cls: 0.1480, loss_map_pts: 0.7030, loss_map_dir: 0.0120, d0.loss_plan_l1_fix_time: 2.8477, d0.loss_plan_l1_fix_dist: 0.4750, d1.loss_plan_l1_fix_time: 1.2186, d1.loss_plan_l1_fix_dist: 0.2078, d2.loss_plan_l1_fix_time: 1.2071, d2.loss_plan_l1_fix_dist: 0.2047, d3.loss_plan_l1_fix_time: 1.2053, d3.loss_plan_l1_fix_dist: 0.2117, d4.loss_plan_l1_fix_time: 1.2092, d4.loss_plan_l1_fix_dist: 0.2230, d5.loss_plan_l1_fix_time: 1.2141, d5.loss_plan_l1_fix_dist: 0.2391, d6.loss_plan_l1_fix_time: 1.2192, d6.loss_plan_l1_fix_dist: 0.2587, d0.loss_cls_all: 0.2376, d0.loss_bbox_all: 0.4577, d0.loss_cls: 0.2376, d0.loss_bbox: 0.4577, d1.loss_cls: 0.2249, d1.loss_bbox: 0.4214, d2.loss_cls: 0.2108, d2.loss_bbox: 0.4016, d3.loss_cls: 0.1997, d3.loss_bbox: 0.3945, d4.loss_cls: 0.1959, d4.loss_bbox: 0.3888, d5.loss_cls: 0.1947, d5.loss_bbox: 0.3887, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7650, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.6602, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.6317, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.6238, d4.loss_traj_cls: 0.0013, d4.loss_traj: 0.6186, d5.loss_traj_cls: 0.0015, d5.loss_traj: 0.6150, d0.loss_map_cls_all: 0.2167, d0.loss_map_pts_all: 1.4702, d0.loss_map_dir_all: 0.0111, d0.loss_map_cls: 0.2167, d0.loss_map_pts: 1.4702, d0.loss_map_dir: 0.0111, d1.loss_map_cls: 0.1841, d1.loss_map_pts: 0.9858, d1.loss_map_dir: 0.0110, d2.loss_map_cls: 0.1673, d2.loss_map_pts: 0.7992, d2.loss_map_dir: 0.0113, d3.loss_map_cls: 0.1608, d3.loss_map_pts: 0.7340, d3.loss_map_dir: 0.0117, d4.loss_map_cls: 0.1534, d4.loss_map_pts: 0.7103, d4.loss_map_dir: 0.0118, d5.loss_map_cls: 0.1499, d5.loss_map_pts: 0.7047, d5.loss_map_dir: 0.0120, loss: 30.5437, grad_norm: 180.3073
2026-02-26 13:28:24,464 - DriveTransformer - INFO - Epoch [7][850/39528]	lr: 8.550e-05, eta: 6 days, 6:45:43, time: 0.735, data_time: 0.034, memory: 19865, loss_cls: 0.1660, loss_bbox: 0.3554, loss_traj: 0.5312, loss_traj_cls: 0.0011, loss_map_cls: 0.1642, loss_map_pts: 0.6827, loss_map_dir: 0.0096, d0.loss_plan_l1_fix_time: 2.4284, d0.loss_plan_l1_fix_dist: 0.3640, d1.loss_plan_l1_fix_time: 1.0342, d1.loss_plan_l1_fix_dist: 0.1689, d2.loss_plan_l1_fix_time: 1.0275, d2.loss_plan_l1_fix_dist: 0.1662, d3.loss_plan_l1_fix_time: 1.0302, d3.loss_plan_l1_fix_dist: 0.1751, d4.loss_plan_l1_fix_time: 1.0348, d4.loss_plan_l1_fix_dist: 0.1856, d5.loss_plan_l1_fix_time: 1.0405, d5.loss_plan_l1_fix_dist: 0.1996, d6.loss_plan_l1_fix_time: 1.0462, d6.loss_plan_l1_fix_dist: 0.2145, d0.loss_cls_all: 0.2038, d0.loss_bbox_all: 0.4129, d0.loss_cls: 0.2038, d0.loss_bbox: 0.4129, d1.loss_cls: 0.1953, d1.loss_bbox: 0.3822, d2.loss_cls: 0.1859, d2.loss_bbox: 0.3646, d3.loss_cls: 0.1724, d3.loss_bbox: 0.3587, d4.loss_cls: 0.1683, d4.loss_bbox: 0.3548, d5.loss_cls: 0.1674, d5.loss_bbox: 0.3541, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7078, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.5766, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.5411, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.5317, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.5264, d5.loss_traj_cls: 0.0015, d5.loss_traj: 0.5251, d0.loss_map_cls_all: 0.2319, d0.loss_map_pts_all: 1.4063, d0.loss_map_dir_all: 0.0098, d0.loss_map_cls: 0.2319, d0.loss_map_pts: 1.4063, d0.loss_map_dir: 0.0098, d1.loss_map_cls: 0.2033, d1.loss_map_pts: 0.9636, d1.loss_map_dir: 0.0092, d2.loss_map_cls: 0.1879, d2.loss_map_pts: 0.7905, d2.loss_map_dir: 0.0092, d3.loss_map_cls: 0.1772, d3.loss_map_pts: 0.7225, d3.loss_map_dir: 0.0093, d4.loss_map_cls: 0.1712, d4.loss_map_pts: 0.6953, d4.loss_map_dir: 0.0095, d5.loss_map_cls: 0.1665, d5.loss_map_pts: 0.6845, d5.loss_map_dir: 0.0096, loss: 27.4844, grad_norm: 176.1752
2026-02-26 13:29:01,849 - DriveTransformer - INFO - Epoch [7][900/39528]	lr: 8.550e-05, eta: 6 days, 6:34:27, time: 0.748, data_time: 0.032, memory: 19865, loss_cls: 0.1604, loss_bbox: 0.3351, loss_traj: 0.5567, loss_traj_cls: 0.0011, loss_map_cls: 0.1461, loss_map_pts: 0.6831, loss_map_dir: 0.0098, d0.loss_plan_l1_fix_time: 2.4950, d0.loss_plan_l1_fix_dist: 0.4068, d1.loss_plan_l1_fix_time: 1.0974, d1.loss_plan_l1_fix_dist: 0.1712, d2.loss_plan_l1_fix_time: 1.0759, d2.loss_plan_l1_fix_dist: 0.1811, d3.loss_plan_l1_fix_time: 1.0743, d3.loss_plan_l1_fix_dist: 0.1971, d4.loss_plan_l1_fix_time: 1.0778, d4.loss_plan_l1_fix_dist: 0.2112, d5.loss_plan_l1_fix_time: 1.0816, d5.loss_plan_l1_fix_dist: 0.2224, d6.loss_plan_l1_fix_time: 1.0863, d6.loss_plan_l1_fix_dist: 0.2343, d0.loss_cls_all: 0.1935, d0.loss_bbox_all: 0.3942, d0.loss_cls: 0.1935, d0.loss_bbox: 0.3942, d1.loss_cls: 0.1854, d1.loss_bbox: 0.3614, d2.loss_cls: 0.1740, d2.loss_bbox: 0.3439, d3.loss_cls: 0.1656, d3.loss_bbox: 0.3375, d4.loss_cls: 0.1601, d4.loss_bbox: 0.3354, d5.loss_cls: 0.1606, d5.loss_bbox: 0.3340, d0.loss_traj_cls: 0.0001, d0.loss_traj: 0.7302, d1.loss_traj_cls: 0.0012, d1.loss_traj: 0.6225, d2.loss_traj_cls: 0.0018, d2.loss_traj: 0.5789, d3.loss_traj_cls: 0.0012, d3.loss_traj: 0.5673, d4.loss_traj_cls: 0.0012, d4.loss_traj: 0.5577, d5.loss_traj_cls: 0.0014, d5.loss_traj: 0.5547, d0.loss_map_cls_all: 0.2125, d0.loss_map_pts_all: 1.3851, d0.loss_map_dir_all: 0.0097, d0.loss_map_cls: 0.2125, d0.loss_map_pts: 1.3851, d0.loss_map_dir: 0.0097, d1.loss_map_cls: 0.1842, d1.loss_map_pts: 0.9476, d1.loss_map_dir: 0.0093, d2.loss_map_cls: 0.1685, d2.loss_map_pts: 0.7762, d2.loss_map_dir: 0.0095, d3.loss_map_cls: 0.1601, d3.loss_map_pts: 0.7148, d3.loss_map_dir: 0.0096, d4.loss_map_cls: 0.1524, d4.loss_map_pts: 0.6901, d4.loss_map_dir: 0.0097, d5.loss_map_cls: 0.1481, d5.loss_map_pts: 0.6831, d5.loss_map_dir: 0.0097, loss: 27.7435, grad_norm: 173.7089
e is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.1.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.4.ffns.0.2.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.task_self_attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.1.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_q.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_q.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_q.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_q.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.temporal_cross_attention_layers.2.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_k.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_k.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_k.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_k.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_v.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_v.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_v.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.w_v.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.proj.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.sensor_cross_attentions.0.proj.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.cross_w_q.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.0.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.1.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.swiglu.w12.weight - torch.Size([2736, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.swiglu.w12.bias - torch.Size([2736]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.swiglu.w3.weight - torch.Size([512, 1368]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.swiglu.w3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.prenorm.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.transformer.decoder.layers.5.ffns.0.2.prenorm.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.0.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.1.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.2.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.3.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.4.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.5.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.cls_branches.6.5.weight - torch.Size([9, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.0.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.1.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.2.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.3.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.4.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.5.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.5.weight - torch.Size([10, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.reg_branches.6.5.bias - torch.Size([10]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.0.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.1.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.2.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.3.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.4.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.5.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.5.weight - torch.Size([12, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_branches.6.5.bias - torch.Size([12]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.0.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.1.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.2.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.3.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.4.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.5.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.5.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.5.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.traj_cls_branches.6.6.weight - torch.Size([1, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.0.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.1.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.2.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.3.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.4.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.5.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_cls_branches.6.5.weight - torch.Size([6, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.0.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.1.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.2.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.3.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.4.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.5.weight - torch.Size([2, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.5.5.bias - torch.Size([2]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.6.weight - torch.Size([40, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.map_reg_branches.6.bias - torch.Size([40]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.0.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.1.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.2.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.3.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.4.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.5.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.5.weight - torch.Size([20, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_dist.6.5.bias - torch.Size([20]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.0.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.1.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.2.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.3.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.4.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.5.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.0.weight - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.0.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.1.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.1.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.3.weight - torch.Size([512, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.3.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.5.weight - torch.Size([60, 512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

pts_bbox_head.ego_traj_branches_fix_time.6.5.bias - torch.Size([60]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

img_backbone.conv1.weight - torch.Size([64, 3, 7, 7]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.bn1.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.bn1.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.conv1.weight - torch.Size([64, 64, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn1.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn1.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.conv2.weight - torch.Size([64, 64, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn2.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn2.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.conv3.weight - torch.Size([256, 64, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn3.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.bn3.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.downsample.0.weight - torch.Size([256, 64, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.downsample.1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.0.downsample.1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.conv1.weight - torch.Size([64, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn1.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn1.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.conv2.weight - torch.Size([64, 64, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn2.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn2.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.conv3.weight - torch.Size([256, 64, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn3.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.1.bn3.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.conv1.weight - torch.Size([64, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn1.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn1.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.conv2.weight - torch.Size([64, 64, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn2.weight - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn2.bias - torch.Size([64]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.conv3.weight - torch.Size([256, 64, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn3.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer1.2.bn3.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.conv1.weight - torch.Size([128, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn1.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn1.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.conv2.weight - torch.Size([128, 128, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn2.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn2.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.conv3.weight - torch.Size([512, 128, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn3.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.bn3.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.downsample.0.weight - torch.Size([512, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.downsample.1.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.0.downsample.1.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.conv1.weight - torch.Size([128, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn1.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn1.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.conv2.weight - torch.Size([128, 128, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn2.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn2.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.conv3.weight - torch.Size([512, 128, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn3.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.1.bn3.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.conv1.weight - torch.Size([128, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn1.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn1.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.conv2.weight - torch.Size([128, 128, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn2.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn2.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.conv3.weight - torch.Size([512, 128, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn3.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.2.bn3.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.conv1.weight - torch.Size([128, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn1.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn1.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.conv2.weight - torch.Size([128, 128, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn2.weight - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn2.bias - torch.Size([128]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.conv3.weight - torch.Size([512, 128, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn3.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer2.3.bn3.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.conv1.weight - torch.Size([256, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.downsample.0.weight - torch.Size([1024, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.downsample.1.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.0.downsample.1.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.conv1.weight - torch.Size([256, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.1.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.conv1.weight - torch.Size([256, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.2.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.conv1.weight - torch.Size([256, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.3.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.conv1.weight - torch.Size([256, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.4.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.conv1.weight - torch.Size([256, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn1.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn1.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.conv2.weight - torch.Size([256, 256, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn2.weight - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn2.bias - torch.Size([256]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.conv3.weight - torch.Size([1024, 256, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn3.weight - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer3.5.bn3.bias - torch.Size([1024]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.conv1.weight - torch.Size([512, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn1.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn1.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.conv2.weight - torch.Size([512, 512, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn2.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn2.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.conv3.weight - torch.Size([2048, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn3.weight - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.bn3.bias - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.downsample.0.weight - torch.Size([2048, 1024, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.downsample.1.weight - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.0.downsample.1.bias - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.conv1.weight - torch.Size([512, 2048, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn1.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn1.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.conv2.weight - torch.Size([512, 512, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn2.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn2.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.conv3.weight - torch.Size([2048, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn3.weight - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.1.bn3.bias - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.conv1.weight - torch.Size([512, 2048, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn1.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn1.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.conv2.weight - torch.Size([512, 512, 3, 3]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn2.weight - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn2.bias - torch.Size([512]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.conv3.weight - torch.Size([2048, 512, 1, 1]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn3.weight - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_backbone.layer4.2.bn3.bias - torch.Size([2048]): 
PretrainedInit: load from ./ckpts/resnet50-19c8e357.pth 

img_neck.lateral_convs.0.conv.weight - torch.Size([512, 2048, 1, 1]): 
XavierInit: gain=1, distribution=uniform, bias=0 

img_neck.lateral_convs.0.conv.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer  

img_neck.fpn_convs.0.conv.weight - torch.Size([512, 512, 3, 3]): 
XavierInit: gain=1, distribution=uniform, bias=0 

img_neck.fpn_convs.0.conv.bias - torch.Size([512]): 
The value is the same before and after calling `init_weights` of DriveTransformer