jacke121的专栏

transforms mp4 扩展正方形

list index out of range t = bracket[low_pos]

简笔画生成smplx sketch2pose

Set XFORMERS_MORE_DETAILS=1 for more details

聚类算法DBSCAN 改进总结

dets must be a CUDA tensor

目录DUSt3R 三维重建：Windows + Pycharm本地安装使用全流程ViewCrafter 地址：GitHub - Drexubery/ViewCrafter: Official implementation of "ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis"

Could not load library libnvrtc.so.11.2. Error: libnvrtc.so.11.2

【代码】RuntimeError: The shape of the 2D attn_mask is torch.Size([77, 77]), but should be (1, 1).

torch_npu 用法笔记

torchvision.io.write_video 报错替换

RuntimeError: Unrecognized tensor type ID: AutogradCUDA

No module named 'torchvision.transforms.functional_tensor'

图生图学习笔记

NCCL timeout

gpu占用脚本

requests.exceptions.SSLError: (MaxRetryError(“HTTPSConnectionPool(host=‘huggingface.co‘, port=443):

blender 导出bvh脸朝下问题

mmpose 2d姿态转json文件 coco转h36m

yolov5 检测封装

accelerate 多卡训练

lstm 回归

Pytorch用于序数多分类的Pytorch损失函数

【代码】yolov8 snpe Converter does not support ‘expand‘ op type。

华为Atlas 部署YOLOV5

这个问题很明显就是网络内的参数类型不同意；

pytorch 半精度hlaf()

PIL opencv图像互转 翻转

onnx转tensorrt学习笔记

python hardsigmoid实现

pytorch RNN 学习笔记

pytorch Default process group is not initialized

https://www.cnblogs.com/jimchen1218/p/14315008.html

https://blog.youkuaiyun.com/dongjinkun/article/details/115375847

Exporting the operator hardsigmoid to ONNX opset version 11 is not supported

PyTorch中的Sequential、ModuleList和ModuleDict用法总结

pytorch导出 openvino 实战笔记

yolov5.6.1 转openvino c++推理笔记

解决方法1：报错代码：anchor_points = self.anchor_points(samples).repeat(batch_size, 1, 1)output_coord = regression + anchor_points修改后代码：anchor_points = self.anchor_points(samples).repeat(batch_size, 1, 1)# decode the points as prediction# output_coord..

torch distributed.init out of memory设置环境gpu：os.environ["CUDA_VISIBLE_DEVICES"] = "1, 2, 3"local_rank=0torch.cuda.set_device(local_rank)cuda(0)默认是第0块显卡，但是设置CUDA_VISIBLE_DEVICES后：cuda(0)就是CUDA_VISIBLE_DEVICES里面的第一个gpu。...