【已解决】报错“copying a param with shape torch.Size([1280, 1280]) from checkpoint”

原创 已于 2024-07-19 15:15:07 修改 · 3.0k 阅读 · 2 ·
CC 4.0 BY-SA版权
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
文章标签：

#深度学习 #python #pytorch #人工智能 #stable diffusion

于 2024-07-12 12:28:45 首次发布

在使用SDXL时，报错“copying a param with shape torch.Size([1280, 1280]) from checkpoint, the shape in current model is torch.Size([1280, 1664]).”

出现问题

在做图像生成项目时，使用ip-adapter-plus_sdxl_vit-h.bin 作为ip-adapter，时出现问题

分析问题原因

查阅官方文档点这里会发现

这里提到了一个这个玩意

找了半天找到了这么一条评论

原来是ip-adapter-plus_sdxl_vit-h.bin 必须搭配vit-h版本的图片解码器使用

解决办法

进入网站

这个是官网，国内进不去

这个是我找到的国内镜像

下载我箭头指的这两个文件，然后给它们两放到一个文件夹里面，记得要把json文件的名字改成config，这样才能识别出它是配置文件

之后把它当作图片解码文件使用即可

学习笔记26-解决：载入预训练模型时Pytorch遇到权重不匹配的问题（附+修改后的预训练模型载入和冻结特征权重完整代码）
LZL2020LZL的博客
03-15 2万+
在pytorch微调mobilenetV3模型时遇到的问题 1.KeyError: ‘features.4.block.2.fc1.weight’ 这个是因为模型结构修改了，没有正确修改预训练权重，导致载入权重与模型不同，使用下面说的两种方法适当修改载入权重即可。 2.size mismatch for fc.weight: copying a param with shape torch.Size([1000, 1280]) from checkpoint, the shape in current mo
size mismatch for yolo_head2.1.bias: copying a param with shape torch.Size(【75】) from checkpoint...
justwaityou1314的博客
04-27 3562
凯哥英语视频 今天一个朋友用YOLO4预测图片报错：size mismatch for yolo_head2.1.bias: copying a param with shape torch.Size(【75】) from checkpoint, the shape in current model is torch.Size(【18】). 网上有说，是在dataloader环节出现了问题。在数据读取时一个判断写的越界了，如下所示 def make_dataset(dir, opt): image.
size mismatch for head.weight: copying a param with shape torch.Size([1000, 768]) from checkpoint
Green____hand的博客
08-01 912
我在github上的vit分类模型（）使用了自己的数据集，训练的时候发生报错。报错原因：训练模型时出现FC层不匹配，我的数据集是 3个类的，下载的预权重是ImageNet 1000个类的。
copying a param with shape torch.Size([84, 48, 3, 3]) from checkpoint, the shape in current model is
justwaityou1314的博客
05-13 5763
前言 无 凯哥英语视频 @[TOC](YOLO中copying a param with shape torch.Size([84, 48, 3, 3]) from checkpoint, the shape in current model is torch.Size([24, 48, 3, 3]).相关报错处理) 不用想太多,就是classes错误 看看这两个方面是否对应 “classes_path” : ‘model_data/coco_classes.txt’ 另外就是你的模型, 是否也是
copying a param with shape torch.Size([262, 2048]) ，参数大小不匹配
weixin_46436505的博客
05-28 7437
从检查点复制具有shape torch.size（[262]）的参数，当前模型中的形状为torch.size（[290]）。 fc.weight的参数大小不匹配，只需修改参数即可。
异常错误：([4, 512]) from checkpoint, the shape in current model is torch.Size([5, 512]). ogb包的版本问题
Drug discovery
10-09 346
1.3.6的版本会将原子变为5，而以前的老版本会是4，所以只需要。
RuntimeError: Error(s) in loading state_dict for DataParallel: size mismatch for module.block.upBlocks.0.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.upBlocks.1.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.upBlocks.2.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.upBlocks.3.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.upBlocks.4.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.upBlocks.5.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.0.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.1.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.2.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.3.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.4.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.block.downBlocks.5.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]). size mismatch for module.out.0.weight: copying a param with shape torch.Size([64, 64, 6, 6]) from checkpoint, the shape in current model is torch.Size([64, 64, 8, 8]).
06-07
model.load_state_dict(torch.load('checkpoint.pth'), strict=False) ``` 使用`strict=False`参数可跳过尺寸不匹配的层[^3]，但需确保核心层能正确加载。 4. **验证环境一致性** ```python # 检查PyTorch版本...
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for Model: size mismatch for model.24.m.0.weight: copying a param with shape torch.Size([18, 128, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 128, 1, 1]). size mismatch for model.24.m.0.bias: copying a param with shape torch.Size([18]) from checkpoint, the shape in current model is torch.Size([21]). size mismatch for model.24.m.1.weight: copying a param with shape torch.Size([18, 256, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 256, 1, 1]). size mismatch for model.24.m.1.bias: copying a param with shape torch.Size([18]) from checkpoint, the shape in current model is torch.Size([21]). size mismatch for model.24.m.2.weight: copying a param with shape torch.Size([18, 512, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 512, 1, 1]). size mismatch for model.24.m.2.bias: copying a param with shape torch.Size([18]) from checkpoint, the shape in current model is torch.Size([21]).
最新发布
07-26
具体错误信息是：`RuntimeError: Error(s) in loading state_dict for ...: size mismatch for param model.24.m.0.weight and others`。这通常是因为我们尝试加载的预训练权重的某些层与当前模型定义中的相应层的...
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for YoloBody: size mismatch for yolo_head_P3.weight: copying a param with shape torch.Size([255, 256, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 256, 1, 1]). size mismatch for yolo_head_P3.bias: copying a param with shape torch.Size([255]) from checkpoint, the shape in current model is torch.Size([21]). size mismatch for yolo_head_P4.weight: copying a param with shape torch.Size([255, 512, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 512, 1, 1]). size mismatch for yolo_head_P4.bias: copying a param with shape torch.Size([255]) from checkpoint, the shape in current model is torch.Size([21]). size mismatch for yolo_head_P5.weight: copying a param with shape torch.Size([255, 1024, 1, 1]) from checkpoint, the shape in current model is torch.Size([21, 1024, 1, 1]). size mismatch for yolo_head_P5.bias: copying a param with shape torch.Size([255]) from checkpoint, the shape in current model is torch.Size([21]).
06-07
让我们考虑一个常见的解决方案：选择性加载权重（忽略检测头）代码示例：```pythonmodel=YoloBody(config)#用自己的配置构建模型pretrained_weights="path/to/pretrained.pth"pretrained_dict=torch.load(pretrained...
RuntimeError: Error(s) in loading state_dict for Wav2vec2Model: size mismatch for decoder.model.0.weight: copying a param with shape torch.Size([256, 512]) from checkpoint, the shape in current model is torch.Size([128, 512]). size mismatch for decoder.model.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([128]). size mismatch for decoder.model.3.weight: copying a param with shape torch.Size([64, 256]) from checkpoint, the shape in current model is torch.Size([7, 128]). size mismatch for decoder.model.3.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([7]).如何解决报错请举例
05-25
解决这个问题的方法是调整模型的参数维度，使其与预训练模型的参数维度相匹配。下面是一个简单的示例： ```python import torch class MyModel(torch.nn.Module): def __init__(self): super(MyModel, self).__...
strict=False 但还是size mismatch for []: copying a param with shape [] from checkpoint,the shape in cur
weixin_44966641的博客
09-03 1万+
strict=False 但还是size mismatch for []: copying a param with shape [] from checkpoint,the shape in cur 问题 我们知道通过 model.load_state_dict(state_dict, strict=False) 可以暂且忽略掉模型和参数文件中不匹配的参数，先将正常匹配的参数从文件中载入模型。 笔者在使用时遇到了这样一个报错： RuntimeError: Error(s) in loading stat
strict=False 但还是 size mismatch for xx.weight copying a param with shape [] from checkpoint错误的解决方法
weixin_44001371的博客
12-17 2981
pop掉。
pytorch-维度变幻
weixin_45797937的博客
03-03 117
维度变换
pytorch 训练时错误size mismatch for module_list.88.Conv2d.weight: copying a param with shape torch.Size
07-28 6963
size mismatch for module_list.88.Conv2d.weight: copying a param with shape torch.Size([75, 1024, 1, 1]) from checkpoint, the shape in current model is torch.Size([255, 1024, 1, 1]). size mismatch for module_list.88.Conv2d.bias: copying a param with shape t
size mismatch for fc.weight: copying a param with shape torch.Size([1000, 2048]) from checkpoint, th
weixin_61785507的博客
06-10 1万+
我想在我自己的项目更换其他的模型，下载的预训练模型出现了FC层不匹配的问题，找了好多人都写了这个点，今天总结一下： 首先我们遇到的问题如下： 他的意思是resnet50的fc层是1000分类，而我的是62分类，所以不匹配，那怎么处理呢？ 网传不可靠方法： 应用了pop，p掉全连接的参数， 这样呢，我试了， 根本就不行，别着急，我还有方法： 这样不论你的模型文件和预训练文件放在那，都可以传入......
copying a param of torch.Size([128]) from checkpoint, where the shape is torch.Size([1, 128])解决方式
kongkongqixi的博客
03-05 9429
关于torch的在加载bn_inception、x_inception时出现copying a param of torch.Size([128]) from checkpoint, where the shape is torch.Size([1, 128])类型的错误，观察发现可以用过修改加载模型中权重维度解决问题。 例如：bn_inception中都是两维元组的数据转为一维：如（1,128...
size mismatch for rpn.head.cls_logits.weight
weixin_43537097的博客
10-19 627
对mobileNetv2+faster_RCNN进行训练之后，将权重保存。但是在预测时，导入权重出现错误。但笔者在后续的过程，将该行代码注释后运行，竟然没有报错。反倒是进行冻结后，预测准确率低，时常抽风。根据报错信息，报错的层为 rpn，
size mismatch问题：训练权重不匹配问题
热门推荐
python_Ezreal的博客
04-23 2万+
在测试二阶段和三阶段模型的时候程序一直报错： RuntimeError: Error(s) in loading state_dict for Eff: size mismatch for fc.weight: copying a param with shape torch.Size([18, 1000]) from checkpoint, the shape in current model is torch.Size([14, 1000]). size mismatch for fc...
VGG11使用卷积层代替全连接层进行测试
qq_52062754的博客
11-09 2245
VGG11使用卷积层代替全连接层进行测试
评论 1
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值