sonica1987

RuntimeError: Could not run 'torchvision::nms'

5

Epoch gpu_mem GIoU obj cls total targets img_size 4/99 3.91G 0.08337 0.4457 0 0.5291 258 192: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████| 19/19 [00:07

剪枝失败

1

步骤按照readme.md进行操作：1、正常训练 （yolov5s 模型 57M） 2、稀疏训练（模型14.4M） 执行剪枝： prune.py 378行： percent_limit = (sorted_bn == highest_thre).nonzero()[0, 0].item() / len(bn_weights) 报错，改为： percent_limit = (sorted_bn == highest_thre).nonzero(as_tuple=False)[0, 0].item() / len(bn_weights)后继续运行剪枝报错如下： model.0.conv.bn BatchNorm2d(32, eps=0.001, momentum=0.03, affine=True,...

Unable to support the qwen model well

11

   I tried qwen 0.5b, 1.8b, 4b, and 7b, and only 0.5b worked properly. However, the result was not as good as the official demo. The output included...