新手求解，用pnnx把yolov10从torchscript模型转到ncnn模型时算子不支持

[emptyery]

error log | 日志或报错信息 | ログ

model | 模型 | モデル

1. original model

how to reproduce | 复现步骤 | 再現方法

Tasks

Preview Give feedback

No tasks being tracked yet.

[emptyery]

应该是torch.remainder跟torch.floor_divide两个算子不支持，请问怎么在ncnn模型中替换这两个算子？

[wzyforgit]

https://github.com/Tencent/ncnn/blob/8d2ac57824f1adc224b1e740df6b9a9f49ee9cc1/docs/developer-guide/add-custom-layer.zh.md

[emptyery]

ncnn应该有能替换的算子吧？需要自定义么？ 竹坚·水灵 ***@***.***  

…

------------------ 原始邮件 ------------------ 发件人: "Tencent/ncnn" ***@***.***>; 发送时间: 2024年11月29日(星期五) 上午8:48 ***@***.***>; ***@***.******@***.***>; 主题: Re: [Tencent/ncnn] 新手求解，用pnnx把yolov10从torchscript模型转到ncnn模型时算子不支持 (Issue #5803) https://github.com/Tencent/ncnn/blob/8d2ac57824f1adc224b1e740df6b9a9f49ee9cc1/docs/developer-guide/add-custom-layer.zh.md — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>