yolov7如何不使用预训练模型

根据引用的描述，YOLOV7是通过从零开始进行训练的，没有使用预训练模型。在实验过程中，使用了微软COCO数据集，并且使用了Train 2017 Set进行训练，然后使用Val 2017 Set进行验证和选择超参数。因此，YOLOV7是一种从头开始训练的目标检测模型，没有使用预训练模型来初始化权重。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* [yolov3 yolov3-tiny yolov4 yolov-tiny预训练模型下载](https://download.csdn.net/download/m0_37794494/12880821)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_2"}}] [.reference_item style="max-width: 50%"]
- *2* *3* [YOLOv7：无需预训练模型的最新最实时的目标检测框架](https://blog.csdn.net/weixin_43509698/article/details/125686861)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_2"}}] [.reference_item style="max-width: 50%"]
[ .reference_list ]