yang-0201 / yolov6_pro Goto Github PK
View Code? Open in Web Editor NEWMake it easier for yolov6 to change the network structure
License: GNU General Public License v3.0
Make it easier for yolov6 to change the network structure
License: GNU General Public License v3.0
openvino int8 conversion
请问yolov6n的yaml文件什么时候出?
在repghostC3.yaml中好像只用了RepGhostC3,并没有用到repghost模块吧,RepVGGBlock只是实现了Repconv吧?
backbone:
[[-1, 1, RepVGGBlock, [64, 3, 2]], # 0-P1/2 channel: 24
[-1, 1, RepVGGBlock, [128, 3, 2]], # 1-P2/4 channel: 48
[-1, 1, RepGhostC3, [128, 192, 2, 3]],
[-1, 1, RepVGGBlock, [256, 3, 2]], # 3-P3/8 channel: 96
[-1, 1, RepGhostC3, [256, 384, 2, 5, 0.25]],
[-1, 1, RepVGGBlock, [512, 3, 2]], # 5-P4/16
[-1, 1, RepGhostC3, [512, 768, 6, 3]],
[-1, 1, RepVGGBlock, [1024, 3, 2]], # 7-P5/32
[-1, 1, RepGhostC3, [1024, 2560, 5, 5, 0.25]],
[-1, 1, SimSPPF, [1024, 5]]]
报错:
return torch.cat(x, self.d)
RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 2 but got size 1 for tensor number 1 in the list.
输入图片尺寸设置的是1280,六倍下采样也是20,为什么cat的对象是11和22尺寸,实在是没找到问题所在,我看了数据输入是没有问题的
在v5 exportde.py 的基础上要修改什么吗
报错:RuntimeError: The size of tensor a (8) must match the size of tensor b (4) at non-singleton dimension 2
训练命令为:python tools/train.py --conf-file configs/model_yaml/yolov6t_GiraffeNeckV2_yaml.py --data data/data.yaml --device 0 --workers 16 --img-size 640 --batch-size 32
Model Summary: 550 layers, 13156006 parameters, 13155972 gradients
Loading state_dict from weights/yolov6t_yaml_new.pt for fine-tuning...
backbone.10.conv.weightcan not change
backbone.10.bn.weightcan not change
backbone.10.bn.biascan not change
backbone.10.bn.running_meancan not change
backbone.10.bn.running_varcan not change
backbone.14.conv.weightcan not change
backbone.14.bn.weightcan not change
backbone.14.bn.biascan not change
backbone.14.bn.running_meancan not change
backbone.14.bn.running_varcan not change
******************
transform_weight: 314/668 from weights/yolov6t_yaml_new.pt
transform_model_weight: 314/768 from weights/yolov6t_yaml_new.pt
******************
Training start...
Epoch iou_loss dfl_loss cls_loss
0%| | 0/186 [00:00<?, ?it/s] /root/miniconda3/envs/py38/lib/python3.8/site-packages/torch/functional.py:445: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2157.)
return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
0%| | 0/186 [00:00<?, ?it/s]
ERROR in training steps.
ERROR in training loop or eval/save model.
Traceback (most recent call last):
File "tools/train.py", line 127, in <module>
main(args)
File "tools/train.py", line 117, in main
trainer.train()
File "/root/autodl-nas/YOLOv6_pro-main/yolov6/core/engine.py", line 99, in train
self.train_in_loop(self.epoch)
File "/root/autodl-nas/YOLOv6_pro-main/yolov6/core/engine.py", line 113, in train_in_loop
self.train_in_steps(epoch_num, self.step)
File "/root/autodl-nas/YOLOv6_pro-main/yolov6/core/engine.py", line 142, in train_in_steps
total_loss, loss_items = self.compute_loss(preds, targets, epoch_num, step_num)
File "/root/autodl-nas/YOLOv6_pro-main/yolov6/models/loss.py", line 155, in __call__
loss_cls = self.varifocal_loss(pred_scores, target_scores, one_hot_label)
File "/root/miniconda3/envs/py38/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/root/autodl-nas/YOLOv6_pro-main/yolov6/models/loss.py", line 196, in forward
weight = alpha * pred_score.pow(gamma) * (1 - label) + gt_score * label
RuntimeError: The size of tensor a (8) must match the size of tensor b (4) at non-singleton dimension 2
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.