From 3fef1a029a7038d0ff076a5340e4a26e267b13bc Mon Sep 17 00:00:00 2001 From: sunshun-yuan Date: Wed, 4 Jun 2025 11:14:19 +0800 Subject: [PATCH] optimze BEVFusion model --- model_examples/BEVFusion/README.md | 4 +- model_examples/BEVFusion/bevfusion.patch | 571 ++++++++++++++++++++--- model_examples/BEVFusion/test/env_npu.sh | 2 +- 3 files changed, 510 insertions(+), 67 deletions(-) diff --git a/model_examples/BEVFusion/README.md b/model_examples/BEVFusion/README.md index 555b4793..d4658beb 100644 --- a/model_examples/BEVFusion/README.md +++ b/model_examples/BEVFusion/README.md @@ -138,7 +138,7 @@ cd model_examples/BEVFusion 单机8卡 | NAME | Modality | Voxel type (voxel size) | 训练方式 | Epoch | global batch size | NDS | mAP | FPS | |------------------|-----------|-------------------------|------|-------|-------|-------|-------|-------| -| 8p-Atlas 800T A2 | lidar-cam | 0.075 | FP32 | 6 | 32 | 69.48 | 66.6 | 17.53 | +| 8p-Atlas 800T A2 | lidar-cam | 0.075 | FP32 | 6 | 32 | 69.33 | 66.26 | 19.97 | | 8p-竞品A | lidar-cam | 0.075 | FP32 | 6 | 32 | 69.78 | 67.36 | 22.54 | 双机16卡 @@ -149,6 +149,8 @@ cd model_examples/BEVFusion # 版本说明 ## 变更 +2025.5.20：更新单机性能。 + 2025.5.20：支持双机，更新单机性能。 2024.12.5：首次发布。 diff --git a/model_examples/BEVFusion/bevfusion.patch b/model_examples/BEVFusion/bevfusion.patch index 39ba0135..8a3ebbd9 100644 --- a/model_examples/BEVFusion/bevfusion.patch +++ b/model_examples/BEVFusion/bevfusion.patch @@ -12,42 +12,263 @@ index 56e8440b..b3a6382a 100644 if ground_plane is not None: xyz = sampled_gt_bboxes[:, :3] diff --git a/mmdet3d/models/layers/sparse_block.py b/mmdet3d/models/layers/sparse_block.py -index 6ed7c8f4..13f69b0d 100644 +index 6ed7c8f4..86e707d8 100644 --- a/mmdet3d/models/layers/sparse_block.py +++ b/mmdet3d/models/layers/sparse_block.py -@@ -2,17 +2,22 @@ - from typing import Optional, Tuple, Union - - from mmcv.cnn import build_conv_layer, build_norm_layer +@@ -1,224 +1,348 @@ +-# Copyright (c) OpenMMLab. All rights reserved. +-from typing import Optional, Tuple, Union +- +-from mmcv.cnn import build_conv_layer, build_norm_layer -from mmdet.models.backbones.resnet import BasicBlock, Bottleneck +-from torch import nn +- +-from mmdet3d.utils import OptConfigType +-from .spconv import IS_SPCONV2_AVAILABLE +- +-if IS_SPCONV2_AVAILABLE: +- from spconv.pytorch import SparseConvTensor, SparseModule, SparseSequential +-else: +- from mmcv.ops import SparseConvTensor, SparseModule, SparseSequential +- +- +-def replace_feature(out: SparseConvTensor, +- new_features: SparseConvTensor) -> SparseConvTensor: +- if 'replace_feature' in out.__dir__(): +- # spconv 2.x behaviour +- return out.replace_feature(new_features) +- else: +- out.features = new_features +- return out +- +- +-class SparseBottleneck(Bottleneck, SparseModule): +- """Sparse bottleneck block for PartA^2. +- +- Bottleneck block implemented with submanifold sparse convolution. +- +- Args: +- inplanes (int): Inplanes of block. +- planes (int): Planes of block. +- stride (int or Tuple[int]): Stride of the first block. Defaults to 1. +- downsample (Module, optional): Down sample module for block. +- Defaults to None. +- indice_key (str): Indice key for spconv. Default to None. +- conv_cfg (:obj:`ConfigDict` or dict, optional): Config dict for +- convolution layer. Defaults to None. +- norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for +- normalization layer. Defaults to None. +- """ +- +- expansion = 4 +- +- def __init__(self, +- inplanes: int, +- planes: int, +- stride: Union[int, Tuple[int]] = 1, +- downsample: nn.Module = None, +- indice_key=None, +- conv_cfg: OptConfigType = None, +- norm_cfg: OptConfigType = None) -> None: +- +- SparseModule.__init__(self) +- if conv_cfg is None: +- conv_cfg = dict(type='SubMConv3d') +- conv_cfg.setdefault('indice_key', indice_key) +- if norm_cfg is None: +- norm_cfg = dict(type='BN1d') +- Bottleneck.__init__( +- self, +- inplanes, +- planes, +- stride=stride, +- downsample=downsample, +- conv_cfg=conv_cfg, +- norm_cfg=norm_cfg) +- +- def forward(self, x: SparseConvTensor) -> SparseConvTensor: +- identity = x.features +- +- out = self.conv1(x) +- out = replace_feature(out, self.bn1(out.features)) +- out = replace_feature(out, self.relu(out.features)) +- +- out = self.conv2(out) +- out = replace_feature(out, self.bn2(out.features)) +- out = replace_feature(out, self.relu(out.features)) +- +- out = self.conv3(out) +- out = replace_feature(out, self.bn3(out.features)) +- +- if self.downsample is not None: +- identity = self.downsample(x).features +- +- out = replace_feature(out, out.features + identity) +- out = replace_feature(out, self.relu(out.features)) +- +- return out +- +- +-class SparseBasicBlock(BasicBlock, SparseModule): +- """Sparse basic block for PartA^2. +- +- Sparse basic block implemented with submanifold sparse convolution. +- +- Args: +- inplanes (int): Inplanes of block. +- planes (int): Planes of block. +- stride (int or Tuple[int]): Stride of the first block. Defaults to 1. +- downsample (Module, optional): Down sample module for block. +- Defaults to None. +- indice_key (str): Indice key for spconv. Default to None. +- conv_cfg (:obj:`ConfigDict` or dict, optional): Config dict for +- convolution layer. Defaults to None. +- norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for +- normalization layer. Defaults to None. +- """ +- +- expansion = 1 +- +- def __init__(self, +- inplanes: int, +- planes: int, +- stride: Union[int, Tuple[int]] = 1, +- downsample: nn.Module = None, +- indice_key: Optional[str] = None, +- conv_cfg: OptConfigType = None, +- norm_cfg: OptConfigType = None) -> None: +- SparseModule.__init__(self) +- if conv_cfg is None: +- conv_cfg = dict(type='SubMConv3d') +- conv_cfg.setdefault('indice_key', indice_key) +- if norm_cfg is None: +- norm_cfg = dict(type='BN1d') +- BasicBlock.__init__( +- self, +- inplanes, +- planes, +- stride=stride, +- downsample=downsample, +- conv_cfg=conv_cfg, +- norm_cfg=norm_cfg) +- +- def forward(self, x: SparseConvTensor) -> SparseConvTensor: +- identity = x.features +- +- assert x.features.dim() == 2, f'x.features.dim()={x.features.dim()}' +- out = self.conv1(x) +- out = replace_feature(out, self.norm1(out.features)) +- out = replace_feature(out, self.relu(out.features)) +- +- out = self.conv2(out) +- out = replace_feature(out, self.norm2(out.features)) +- +- if self.downsample is not None: +- identity = self.downsample(x).features +- +- out = replace_feature(out, out.features + identity) +- out = replace_feature(out, self.relu(out.features)) +- +- return out +- +- +-def make_sparse_convmodule(in_channels: int, +- out_channels: int, +- kernel_size: Union[int, Tuple[int]], +- indice_key: Optional[str] = None, +- stride: Union[int, Tuple[int]] = 1, +- padding: Union[int, Tuple[int]] = 0, +- conv_type: str = 'SubMConv3d', +- norm_cfg: OptConfigType = None, +- order: Tuple[str] = ('conv', 'norm', 'act'), +- **kwargs) -> SparseSequential: +- """Make sparse convolution module. +- +- Args: +- in_channels (int): The number of input channels. +- out_channels (int): The number of out channels. +- kernel_size (int | Tuple[int]): Kernel size of convolution. +- indice_key (str): The indice key used for sparse tensor. +- stride (int or tuple[int]): The stride of convolution. +- padding (int or tuple[int]): The padding number of input. +- conv_type (str): Sparse conv type in spconv. Defaults to 'SubMConv3d'. +- norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for +- normalization layer. Defaults to None. +- order (Tuple[str]): The order of conv/norm/activation layers. It is a +- sequence of "conv", "norm" and "act". Common examples are +- ("conv", "norm", "act") and ("act", "conv", "norm"). +- Defaults to ('conv', 'norm', 'act'). +- +- Returns: +- spconv.SparseSequential: sparse convolution module. +- """ +- assert isinstance(order, tuple) and len(order) <= 3 +- assert set(order) | {'conv', 'norm', 'act'} == {'conv', 'norm', 'act'} +- +- conv_cfg = dict(type=conv_type, indice_key=indice_key) +- if norm_cfg is None: +- norm_cfg = dict(type='BN1d') +- +- layers = list() +- for layer in order: +- if layer == 'conv': +- if conv_type not in [ +- 'SparseInverseConv3d', 'SparseInverseConv2d', +- 'SparseInverseConv1d' +- ]: +- layers.append( +- build_conv_layer( +- conv_cfg, +- in_channels, +- out_channels, +- kernel_size, +- stride=stride, +- padding=padding, +- bias=False)) +- else: +- layers.append( +- build_conv_layer( +- conv_cfg, +- in_channels, +- out_channels, +- kernel_size, +- bias=False)) +- elif layer == 'norm': +- layers.append(build_norm_layer(norm_cfg, out_channels)[1]) +- elif layer == 'act': +- layers.append(nn.ReLU(inplace=True)) +- +- layers = SparseSequential(*layers) +- return layers ++# Copyright (c) OpenMMLab. All rights reserved. ++from typing import Optional, Tuple, Union ++ ++from mmcv.cnn import build_conv_layer, build_norm_layer +import inspect +from mmengine.model import BaseModule +import torch - from torch import nn ++from torch import nn +import torch.utils.checkpoint as cp - - from mmdet3d.utils import OptConfigType --from .spconv import IS_SPCONV2_AVAILABLE -+from mx_driving.spconv import SparseSequential, SubMConv3d, SparseConv3d, SparseModule ++ ++from mmdet3d.utils import OptConfigType ++from mx_driving.spconv import SparseSequential, SubMConv3d, SparseConv3d, SparseModule, SparseConvTensor +from mmdet.models.backbones.resnet import BasicBlock, Bottleneck +from mmengine.registry import Registry -+from mx_driving.spconv import SparseConvTensor - --if IS_SPCONV2_AVAILABLE: -- from spconv.pytorch import SparseConvTensor, SparseModule, SparseSequential --else: -- from mmcv.ops import SparseConvTensor, SparseModule, SparseSequential - ++ ++ +MODELS = Registry('Sparse conv layer') +MODELS.register_module('SubMConv3d', module=SubMConv3d) +MODELS.register_module('SparseConv3d', module=SparseConv3d) - - def replace_feature(out: SparseConvTensor, - new_features: SparseConvTensor) -> SparseConvTensor: -@@ -23,6 +28,87 @@ def replace_feature(out: SparseConvTensor, - out.features = new_features - return out - ++ ++def replace_feature(out: SparseConvTensor, ++ new_features: SparseConvTensor) -> SparseConvTensor: ++ if 'replace_feature' in out.__dir__(): ++ # spconv 2.x behaviour ++ return out.replace_feature(new_features) ++ else: ++ out.features = new_features ++ return out ++ +class BasicBlock(BaseModule): + expansion = 1 + @@ -129,31 +350,205 @@ index 6ed7c8f4..13f69b0d 100644 + out = self.relu(out) + + return out - - class SparseBottleneck(Bottleneck, SparseModule): - """Sparse bottleneck block for PartA^2. -@@ -199,7 +285,7 @@ def make_sparse_convmodule(in_channels: int, - 'SparseInverseConv1d' - ]: - layers.append( -- build_conv_layer( ++ ++class SparseBottleneck(Bottleneck, SparseModule): ++ """Sparse bottleneck block for PartA^2. ++ ++ Bottleneck block implemented with submanifold sparse convolution. ++ ++ Args: ++ inplanes (int): Inplanes of block. ++ planes (int): Planes of block. ++ stride (int or Tuple[int]): Stride of the first block. Defaults to 1. ++ downsample (Module, optional): Down sample module for block. ++ Defaults to None. ++ indice_key (str): Indice key for spconv. Default to None. ++ conv_cfg (:obj:`ConfigDict` or dict, optional): Config dict for ++ convolution layer. Defaults to None. ++ norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for ++ normalization layer. Defaults to None. ++ """ ++ ++ expansion = 4 ++ ++ def __init__(self, ++ inplanes: int, ++ planes: int, ++ stride: Union[int, Tuple[int]] = 1, ++ downsample: nn.Module = None, ++ indice_key=None, ++ conv_cfg: OptConfigType = None, ++ norm_cfg: OptConfigType = None) -> None: ++ ++ SparseModule.__init__(self) ++ if conv_cfg is None: ++ conv_cfg = dict(type='SubMConv3d') ++ conv_cfg.setdefault('indice_key', indice_key) ++ if norm_cfg is None: ++ norm_cfg = dict(type='BN1d') ++ Bottleneck.__init__( ++ self, ++ inplanes, ++ planes, ++ stride=stride, ++ downsample=downsample, ++ conv_cfg=conv_cfg, ++ norm_cfg=norm_cfg) ++ ++ def forward(self, x: SparseConvTensor) -> SparseConvTensor: ++ identity = x.features ++ ++ out = self.conv1(x) ++ out = replace_feature(out, self.bn1(out.features)) ++ out = replace_feature(out, self.relu(out.features)) ++ ++ out = self.conv2(out) ++ out = replace_feature(out, self.bn2(out.features)) ++ out = replace_feature(out, self.relu(out.features)) ++ ++ out = self.conv3(out) ++ out = replace_feature(out, self.bn3(out.features)) ++ ++ if self.downsample is not None: ++ identity = self.downsample(x).features ++ ++ out = replace_feature(out, out.features + identity) ++ out = replace_feature(out, self.relu(out.features)) ++ ++ return out ++ ++ ++class SparseBasicBlock(BasicBlock, SparseModule): ++ """Sparse basic block for PartA^2. ++ ++ Sparse basic block implemented with submanifold sparse convolution. ++ ++ Args: ++ inplanes (int): Inplanes of block. ++ planes (int): Planes of block. ++ stride (int or Tuple[int]): Stride of the first block. Defaults to 1. ++ downsample (Module, optional): Down sample module for block. ++ Defaults to None. ++ indice_key (str): Indice key for spconv. Default to None. ++ conv_cfg (:obj:`ConfigDict` or dict, optional): Config dict for ++ convolution layer. Defaults to None. ++ norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for ++ normalization layer. Defaults to None. ++ """ ++ ++ expansion = 1 ++ ++ def __init__(self, ++ inplanes: int, ++ planes: int, ++ stride: Union[int, Tuple[int]] = 1, ++ downsample: nn.Module = None, ++ indice_key: Optional[str] = None, ++ conv_cfg: OptConfigType = None, ++ norm_cfg: OptConfigType = None) -> None: ++ SparseModule.__init__(self) ++ if conv_cfg is None: ++ conv_cfg = dict(type='SubMConv3d') ++ conv_cfg.setdefault('indice_key', indice_key) ++ if norm_cfg is None: ++ norm_cfg = dict(type='BN1d') ++ BasicBlock.__init__( ++ self, ++ inplanes, ++ planes, ++ stride=stride, ++ downsample=downsample, ++ conv_cfg=conv_cfg, ++ norm_cfg=norm_cfg) ++ ++ def forward(self, x: SparseConvTensor) -> SparseConvTensor: ++ identity = x.features ++ ++ assert x.features.dim() == 2, f'x.features.dim()={x.features.dim()}' ++ out = self.conv1(x) ++ out = replace_feature(out, self.norm1(out.features)) ++ out = replace_feature(out, self.relu(out.features)) ++ ++ out = self.conv2(out) ++ out = replace_feature(out, self.norm2(out.features)) ++ ++ if self.downsample is not None: ++ identity = self.downsample(x).features ++ ++ out = replace_feature(out, out.features + identity) ++ out = replace_feature(out, self.relu(out.features)) ++ ++ return out ++ ++ ++def make_sparse_convmodule(in_channels: int, ++ out_channels: int, ++ kernel_size: Union[int, Tuple[int]], ++ indice_key: Optional[str] = None, ++ stride: Union[int, Tuple[int]] = 1, ++ padding: Union[int, Tuple[int]] = 0, ++ conv_type: str = 'SubMConv3d', ++ norm_cfg: OptConfigType = None, ++ order: Tuple[str] = ('conv', 'norm', 'act'), ++ **kwargs) -> SparseSequential: ++ """Make sparse convolution module. ++ ++ Args: ++ in_channels (int): The number of input channels. ++ out_channels (int): The number of out channels. ++ kernel_size (int | Tuple[int]): Kernel size of convolution. ++ indice_key (str): The indice key used for sparse tensor. ++ stride (int or tuple[int]): The stride of convolution. ++ padding (int or tuple[int]): The padding number of input. ++ conv_type (str): Sparse conv type in spconv. Defaults to 'SubMConv3d'. ++ norm_cfg (:obj:`ConfigDict` or dict, optional): Config dict for ++ normalization layer. Defaults to None. ++ order (Tuple[str]): The order of conv/norm/activation layers. It is a ++ sequence of "conv", "norm" and "act". Common examples are ++ ("conv", "norm", "act") and ("act", "conv", "norm"). ++ Defaults to ('conv', 'norm', 'act'). ++ ++ Returns: ++ spconv.SparseSequential: sparse convolution module. ++ """ ++ assert isinstance(order, tuple) and len(order) <= 3 ++ assert set(order) | {'conv', 'norm', 'act'} == {'conv', 'norm', 'act'} ++ ++ conv_cfg = dict(type=conv_type, indice_key=indice_key) ++ if norm_cfg is None: ++ norm_cfg = dict(type='BN1d') ++ ++ layers = list() ++ for layer in order: ++ if layer == 'conv': ++ if conv_type not in [ ++ 'SparseInverseConv3d', 'SparseInverseConv2d', ++ 'SparseInverseConv1d' ++ ]: ++ layers.append( + build_sparse_conv_layer( - conv_cfg, - in_channels, - out_channels, -@@ -209,7 +295,7 @@ def make_sparse_convmodule(in_channels: int, - bias=False)) - else: - layers.append( -- build_conv_layer( ++ conv_cfg, ++ in_channels, ++ out_channels, ++ kernel_size, ++ stride=stride, ++ padding=padding, ++ bias=False)) ++ else: ++ layers.append( + build_sparse_conv_layer( - conv_cfg, - in_channels, - out_channels, -@@ -222,3 +308,42 @@ def make_sparse_convmodule(in_channels: int, - - layers = SparseSequential(*layers) - return layers ++ conv_cfg, ++ in_channels, ++ out_channels, ++ kernel_size, ++ bias=False)) ++ elif layer == 'norm': ++ layers.append(build_norm_layer(norm_cfg, out_channels)[1]) ++ elif layer == 'act': ++ layers.append(nn.ReLU(inplace=True)) ++ ++ layers = SparseSequential(*layers) ++ return layers + + +def build_sparse_conv_layer(cfg, *args, **kwargs): @@ -522,10 +917,24 @@ index 68bf2bce..090363a9 100644 + self.encoder_layers.add_module(stage_name, stage_layers) + return out_channels diff --git a/projects/BEVFusion/bevfusion/transfusion_head.py b/projects/BEVFusion/bevfusion/transfusion_head.py -index 8a3e1750..5dd9ab17 100644 +index 8a3e1750..8414de81 100644 --- a/projects/BEVFusion/bevfusion/transfusion_head.py +++ b/projects/BEVFusion/bevfusion/transfusion_head.py -@@ -696,15 +696,16 @@ class TransFusionHead(nn.Module): +@@ -13,11 +13,12 @@ from mmdet.models.utils import multi_apply + from mmengine.structures import InstanceData + from torch import nn + +-from mmdet3d.models import circle_nms, draw_heatmap_gaussian, gaussian_radius ++from mmdet3d.models import circle_nms + from mmdet3d.models.dense_heads.centerpoint_head import SeparateHead + from mmdet3d.models.layers import nms_bev + from mmdet3d.registry import MODELS + from mmdet3d.structures import xywhr2xyxyr ++from mx_driving import npu_assign_target_of_single_head + + + def clip_sigmoid(x, eps=1e-4): +@@ -696,39 +697,30 @@ class TransFusionHead(nn.Module): [gt_bboxes_3d.gravity_center, gt_bboxes_3d.tensor[:, 3:]], dim=1).to(device) grid_size = torch.tensor(self.train_cfg['grid_size']) @@ -535,26 +944,58 @@ index 8a3e1750..5dd9ab17 100644 + voxel_size = self.train_cfg['voxel_size'] feature_map_size = (grid_size[:2] // self.train_cfg['out_size_factor'] ) # [x_len, y_len] - heatmap = gt_bboxes_3d.new_zeros(self.num_classes, feature_map_size[1], - feature_map_size[0]) - for idx in range(len(gt_bboxes_3d)): +- heatmap = gt_bboxes_3d.new_zeros(self.num_classes, feature_map_size[1], +- feature_map_size[0]) +- for idx in range(len(gt_bboxes_3d)): - width = gt_bboxes_3d[idx][3] - length = gt_bboxes_3d[idx][4] -+ gt_bboxes_3d_digital = gt_bboxes_3d[idx].tolist() -+ width = gt_bboxes_3d_digital[3] -+ length = gt_bboxes_3d_digital[4] - width = width / voxel_size[0] / self.train_cfg['out_size_factor'] - length = length / voxel_size[1] / self.train_cfg['out_size_factor'] - if width > 0 and length > 0: -@@ -712,7 +713,7 @@ class TransFusionHead(nn.Module): - (length, width), - min_overlap=self.train_cfg['gaussian_overlap']) - radius = max(self.train_cfg['min_radius'], int(radius)) +- width = width / voxel_size[0] / self.train_cfg['out_size_factor'] +- length = length / voxel_size[1] / self.train_cfg['out_size_factor'] +- if width > 0 and length > 0: +- radius = gaussian_radius( +- (length, width), +- min_overlap=self.train_cfg['gaussian_overlap']) +- radius = max(self.train_cfg['min_radius'], int(radius)) - x, y = gt_bboxes_3d[idx][0], gt_bboxes_3d[idx][1] -+ x, y = gt_bboxes_3d_digital[0], gt_bboxes_3d_digital[1] +- +- coor_x = ((x - pc_range[0]) / voxel_size[0] / +- self.train_cfg['out_size_factor']) +- coor_y = ((y - pc_range[1]) / voxel_size[1] / +- self.train_cfg['out_size_factor']) +- +- center = torch.tensor([coor_x, coor_y], +- dtype=torch.float32, +- device=device) +- center_int = center.to(torch.int32) +- +- # original +- # draw_heatmap_gaussian(heatmap[gt_labels_3d[idx]], center_int, radius) # noqa: E501 +- # NOTE: fix +- draw_heatmap_gaussian(heatmap[gt_labels_3d[idx]], +- center_int[[1, 0]], radius) ++ ++ if (torch.numel(gt_bboxes_3d[:,:-1]) == 0): ++ heatmap = gt_bboxes_3d.new_zeros(self.num_classes, feature_map_size[1], feature_map_size[0]) ++ ++ gt_bboxes_3d[:, [0, 1]] = gt_bboxes_3d[:, [1, 0]] ++ cur_class_id = gt_labels_3d.int() + 1 ++ feature_map_size_list = list(feature_map_size) ++ heatmap, _, _, _ = npu_assign_target_of_single_head(gt_bboxes_3d, ++ cur_class_id, ++ self.num_classes, ++ self.train_cfg['out_size_factor'], ++ self.train_cfg['gaussian_overlap'], ++ self.train_cfg['min_radius'], ++ voxel_size, ++ pc_range, ++ feature_map_size_list, ++ False, ++ False, ++ False, ++ gt_bboxes_3d.shape[0]) - coor_x = ((x - pc_range[0]) / voxel_size[0] / - self.train_cfg['out_size_factor']) + mean_iou = ious[pos_inds].sum() / max(len(pos_inds), 1) + return ( diff --git a/projects/BEVFusion/configs/bevfusion_lidar-cam_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py b/projects/BEVFusion/configs/bevfusion_lidar-cam_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py index a08bb66a..04b6c0a2 100644 --- a/projects/BEVFusion/configs/bevfusion_lidar-cam_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py diff --git a/model_examples/BEVFusion/test/env_npu.sh b/model_examples/BEVFusion/test/env_npu.sh index 401be673..19ede4c2 100644 --- a/model_examples/BEVFusion/test/env_npu.sh +++ b/model_examples/BEVFusion/test/env_npu.sh @@ -23,7 +23,7 @@ export TASK_QUEUE_ENABLE=2 #设置是否开启combined标志,0-关闭/1-开启 export COMBINED_ENABLE=1 #设置是否开启均匀绑核,0-关闭/1-开启粗粒度绑核/2-开启细粒度绑核 -export CPU_AFFINITY_CONF=0 +export CPU_AFFINITY_CONF=2 #HCCL白名单开关,1-关闭/0-开启 export HCCL_WHITELIST_DISABLE=1 #配置HCCL的初始化root通信网卡IP -- Gitee