Lidar_AI_Solution

Lidar_AI_Solution copied to clipboard


Xiao-Hu-Z

Failed to build model model/resnet50int8/build/head.bbox.layernormplugin.plan.

9

; [E] Uncaught exception detected: Unable to open library: libcustom_layernorm.so due to libcustom_layernorm.so: cannot open shared object file: No such file or directory

Happynewyear000

How to export the SwinT model from BEVFusion?

2

I have trained camera+lidar based on SwinT, I have got a pth. How can I generate PTQ model or FP16 model? I used the command "python qat/ptq.py --config=configs/nuscenes/det/transfusion/secfpn/camera+lidar/swint_v0p075/configs.yaml --ckpt=model/swint/epoch_30.pth --calibrate_batch...

slayerlpj

How to run the nuscenes dataset with CUDA-BEVFusion?

1

I want to use CUDA-BEVFusion to process the nuScenes dataset. How do I convert the nuScenes dataset into the example-data format? In the example dataset, how can I view the...

onlywl9598

Thanks for the great work. How to deploy the segmentation header in bevfusion to CUDA-BEVFusion.

whbdmu

export ONNX error

1

Thanks for you repo! I'm tring to convert pth to onnx. But When I use "python qat/export-camera.py --ckpt=model/resnet50/bevfusion-det.pth --fp16". I get a error "model = torch.load(args.ckpt).module AttributeError: 'dict' object has...

slayerlpj

Is there some bug in TensorRT?

1

[stdbuf-1] [NVINFER LOG]: 1: [runtime.cpp::parsePlan::314] Error Code 1: Serialization (Serialization assertion plan->header.magicTag == rt::kPLAN_MAGIC_TAG failed.) [stdbuf-1] Failed to deserialize engine: /workspace/test/shan.li/install/radar_img_fusion/share/radar_img_fusion/model/20240116/CPC_FCR02.a/resnet50/build/camera.backbone.plan [stdbuf-1] Failed to create camera backbone. [stdbuf-1] [2024-07-26 15:04:00.553]...

bujianyiwang

Great work! Is there a version of Voxelization for float type point cloud?

nuaawh

export spconv onnx error

1

When I excute "python centerpoint-export/export-scn.py --ckpt=epoch_20.pth --save-onnx=scn.nuscenes.onnx", occured "AttributeError: module 'cumm.tensorview' has no attribute 'gemm'"，how to slove it？Thanks.

aaa1233211

Thanks for your work! I'm trying to quantize Int8 Centerpoint to accelerate my model on Orin. However the qat code provided only support fp16 quantize and trt model transfer. Could...

zxxzxy