dao027

> Are you using the pretrained model provided by them or your own trained model? And have you tried v.6.0? i have tried the pretrained model provided by them according ...

> I meet the some problem by using yolov5s v6.1 . the same picture but different results. @dao027 , do you solve it? not yet, but result of TRT is...

> I found the reason, but don't know how to fix it. There is no 'model.24.anchor_grid' in trained model with yolov5-6.0, but it do exist in pretrained models, and exist...

> Guys, you need to find out why the trained model doesn't contain anchor info. roger, i will check this

> negative, cant use engine file from yolov5 official convertion, u need to convert pth model to engine by way in this repo

same issue, fixed by https://github.com/WongKinYiu/yolov7/issues/783 move ur `with torch.no_grad():` position

> ```c > dst_drm_fd > ``` 请问你这里的dst_drm_fd是怎么定义的呢，需要怎样保存imcvtcolor转换以后的数据？我的src_fd是从mppframe拿过来的，那我转换以后也是存到mppbuffer里面吗？

> > 请问你这里的dst_drm_fd是怎么定义的呢，需要怎样保存imcvtcolor转换以后的数据？我的src_fd是从mppframe拿过来的，那我转换以后也是存到mppbuffer里面吗？ > > 我并没有存到mppbuffer中，而是自行维护转换后的数据 感谢回复，再请问下你这边有使用到sample中的这部分吗？ ret = dma_buf_alloc(DMA_HEAP_DMA32_UNCACHED_PATH, dst_buf_size, &dst_dma_fd, (void **)&dst_buf); 这里的dst_dma_fd和你代码中的dst_drm_fd有区别吗？

> > 感谢回复，再请问下你这边有使用到sample中的这部分吗？ ret = dma_buf_alloc(DMA_HEAP_DMA32_UNCACHED_PATH, dst_buf_size, &dst_dma_fd, (void **)&dst_buf); 这里的dst_dma_fd和你代码中的dst_drm_fd有区别吗？ > > 最初我使用的为drm_buf，但是后来在alloc过程中遇到了错误，所以我现在也在使用dma32_buf 你好，遇到一点问题，想请教一下： ``` dst_buf = (char *)malloc(dst_buf_size); memset(dst_buf, 0x80, dst_buf_size); dst_handle = importbuffer_virtualaddr(dst_buf, dst_buf_size); ``` 这里importbuffer_virtualaddr返回的值一直是0，然后就报importbuffer failed!这个是驱动版本的问题吗

同问，dst_fd该如何定义呢