YOLOv3v4-ModelCompression-MultidatasetTraining-Multibackbone

YOLOv3v4-ModelCompression-MultidatasetTraining-Multibackbone copied to clipboard

AttributeError: module 'torch.distributed' has no attribute 'init_process_group'

2

1179993113

 您好，请问一下将Yolov4的骨干网络换成ghostnet后能进行稀疏化训练，通道剪枝时好像在生成模型时报错： Traceback (most recent call last): File "normal_prune.py", line 159, in pruned_model = prune_model_keep_size(model, prune_idx, CBL_idx, CBLidx2mask) File "/home/yhk/YOLOv4-GhostNet/utils/prune_utils.py", line 287, in prune_model_keep_size update_activation(i, pruned_model, activation, CBL_idx) File...

kust-yhk

加载yolov3-mobilenet模型

8

如果我对yolov3-mobilenet网络结构做了一点改动，想使用你的yolov3-mobilenet预训练模型，命令行需要改动吗？还是说改动train.py文件？

Note-Liu

 Hello; I try to use your pretrained model, but i got nan value in the result. Can you show me the way to fix it ? Big thank to...

hoangkhoiLE

NameError: name 'Darknet' is not defined

1

``` python3 normal_prune.py --data data/coco2017.data --weights yolov3_SPP_best.pt --cfg yolov3-spp3.cfg --percent 0.2 Namespace(batch_size=8, cfg='yolov3-spp3.cfg', data='data/coco2017.data', img_size=416, percent=0.2, weights='yolov3_SPP_best.pt') Traceback (most recent call last): File "normal_prune.py", line 104, in model = Darknet(opt.cfg).to(device)...

MohamedElsaeidy

训练除yolov3外的cfg文件时，出现AssertionError，加载预训练模型失败

3

Model Summary: 45 layers, 9.02466e+06 parameters, 9.02466e+06 gradients Optimizer groups: 16 .bias, 16 Conv2d.weight, 13 other Traceback (most recent call last): File "train.py", line 632, in train(hyp) # train normally...

1246886181

Hi Is it possible to introduce articles related to pruning methods. for example, Which article is related to Regular pruning ? Thanks

saeedkhanehgir

裁剪得到的模型在test.py里读取模型没有报错，但是在train.py里读取模型函数里就会报错，为什么呢？

Biubiubiu12

多GPU训练

3

你好, 我在使用多GPU训练的时候, 每次都会遇到这个问题 ``` Namespace(BN_Fold=False, FPGA=False, KDstr=-1, a_bit=8, adam=False, batch_size=16, bucket='', cache_images=False, cfg='./cfg/yolov4/yolov4.cfg', data='data/coco2017.data', device='0,1,2,4', ema=False, epochs=300, evolve=False, img_size=[320, 640], multi_scale=False, name='', nosave=False, notest=False, prune=0, pt=False, quantized=0, rect=False, resume=False, s=0.0001,...

chenxyyy

量化训练报错：TypeError: min(): argument 'input' (position 1) must be Tensor, not NoneType

3

a-bit和w-bit都设置为8，在量化进行到 i == 2，a_bit = 16，w_bit = 16时，出现错误：TypeError: min(): argument 'input' (position 1) must be Tensor, not NoneType，这一次的WarmupForQ中构建模型时的GFLOPS是空，请问是哪里出了问题？

tomorrow-rain