选择模型报错提示配置文件问题

[zhangzhiz]

在UI页面选择模型路径的时候，出现error,日志有一段提示，这个配置文件需要怎么改？

The config attributes {'decay': 0.9999, 'inv_gamma': 1.0, 'min_decay': 0.0, 'optimization_step': 4400, 'power': 0.6666666666666666, 'update_after_step': 0, 'use_ema_warmup': False} were passed to Transformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.

[yunkchen]

没有修改配置文件的操作；根据readme再看一下每一步的操作流程有没有问题

[zhangzhiz]

是按照操作流程来的，通过ui界面那个步骤下载对应权重放入models文件夹。这个报错的文件指向的就是models/Diffusion_Transformer/EasyAnimateV3-XL-2-InP-512x512/transformer这个里面的config.json,没有找到修改这个配置文件中参数的说明，麻烦指点一下。

[yunkchen]

这个配置文件不需要修改

[yunkchen]

贴一下完整的日志

[zhangzhiz]

el=False, elem_id="i2v_end", sources="upload", type="filepath") Running on local URL: http://0.0.0.0:7860

To create a public link, set share=True in launch(). IMPORTANT: You are using gradio version 3.41.2, however version 4.29.0 is available, please upgrade.

Update diffusion transformer

missing keys: 0;

unexpected keys: 96;

[] ['loss.discriminator.main.0.bias', 'loss.discriminator.main.0.weight', 'loss.discriminator.main.11.bias', 'loss.discriminator.main.11.weight', 'loss.discriminator.main.2.weight', 'loss.discriminator.main.3.bias', 'loss.discriminator.main.3.num_batches_tracked', 'loss.discriminator.main.3.running_mean', 'loss.discriminator.main.3.running_var', 'loss.discriminator.main.3.weight', 'loss.discriminator.main.5.weight', 'loss.discriminator.main.6.bias', 'loss.discriminator.main.6.num_batches_tracked', 'loss.discriminator.main.6.running_mean', 'loss.discriminator.main.6.running_var', 'loss.discriminator.main.6.weight', 'loss.discriminator.main.8.weight', 'loss.discriminator.main.9.bias', 'loss.discriminator.main.9.num_batches_tracked', 'loss.discriminator.main.9.running_mean', 'loss.discriminator.main.9.running_var', 'loss.discriminator.main.9.weight', 'loss.discriminator3d.blocks.0.conv1.bias', 'loss.discriminator3d.blocks.0.conv1.weight', 'loss.discriminator3d.blocks.0.conv2.bias', 'loss.discriminator3d.blocks.0.conv2.weight', 'loss.discriminator3d.blocks.0.downsampler.filt', 'loss.discriminator3d.blocks.0.norm1.bias', 'loss.discriminator3d.blocks.0.norm1.weight', 'loss.discriminator3d.blocks.0.norm2.bias', 'loss.discriminator3d.blocks.0.norm2.weight', 'loss.discriminator3d.blocks.0.shortcut.0.filt', 'loss.discriminator3d.blocks.0.shortcut.1.bias', 'loss.discriminator3d.blocks.0.shortcut.1.weight', 'loss.discriminator3d.blocks.1.conv1.bias', 'loss.discriminator3d.blocks.1.conv1.weight', 'loss.discriminator3d.blocks.1.conv2.bias', 'loss.discriminator3d.blocks.1.conv2.weight', 'loss.discriminator3d.blocks.1.downsampler.filt', 'loss.discriminator3d.blocks.1.norm1.bias', 'loss.discriminator3d.blocks.1.norm1.weight', 'loss.discriminator3d.blocks.1.norm2.bias', 'loss.discriminator3d.blocks.1.norm2.weight', 'loss.discriminator3d.blocks.1.shortcut.0.filt', 'loss.discriminator3d.blocks.1.shortcut.1.bias', 'loss.discriminator3d.blocks.1.shortcut.1.weight', 'loss.discriminator3d.blocks.2.conv1.bias', 'loss.discriminator3d.blocks.2.conv1.weight', 'loss.discriminator3d.blocks.2.conv2.bias', 'loss.discriminator3d.blocks.2.conv2.weight', 'loss.discriminator3d.blocks.2.norm1.bias', 'loss.discriminator3d.blocks.2.norm1.weight', 'loss.discriminator3d.blocks.2.norm2.bias', 'loss.discriminator3d.blocks.2.norm2.weight', 'loss.discriminator3d.blocks.2.shortcut.0.bias', 'loss.discriminator3d.blocks.2.shortcut.0.weight', 'loss.discriminator3d.conv_in.bias', 'loss.discriminator3d.conv_in.weight', 'loss.discriminator3d.conv_norm_out.bias', 'loss.discriminator3d.conv_norm_out.weight', 'loss.discriminator3d.conv_out.bias', 'loss.discriminator3d.conv_out.weight', 'loss.logvar', 'loss.perceptual_loss.lin0.model.1.weight', 'loss.perceptual_loss.lin1.model.1.weight', 'loss.perceptual_loss.lin2.model.1.weight', 'loss.perceptual_loss.lin3.model.1.weight', 'loss.perceptual_loss.lin4.model.1.weight', 'loss.perceptual_loss.net.slice1.0.bias', 'loss.perceptual_loss.net.slice1.0.weight', 'loss.perceptual_loss.net.slice1.2.bias', 'loss.perceptual_loss.net.slice1.2.weight', 'loss.perceptual_loss.net.slice2.5.bias', 'loss.perceptual_loss.net.slice2.5.weight', 'loss.perceptual_loss.net.slice2.7.bias', 'loss.perceptual_loss.net.slice2.7.weight', 'loss.perceptual_loss.net.slice3.10.bias', 'loss.perceptual_loss.net.slice3.10.weight', 'loss.perceptual_loss.net.slice3.12.bias', 'loss.perceptual_loss.net.slice3.12.weight', 'loss.perceptual_loss.net.slice3.14.bias', 'loss.perceptual_loss.net.slice3.14.weight', 'loss.perceptual_loss.net.slice4.17.bias', 'loss.perceptual_loss.net.slice4.17.weight', 'loss.perceptual_loss.net.slice4.19.bias', 'loss.perceptual_loss.net.slice4.19.weight', 'loss.perceptual_loss.net.slice4.21.bias', 'loss.perceptual_loss.net.slice4.21.weight', 'loss.perceptual_loss.net.slice5.24.bias', 'loss.perceptual_loss.net.slice5.24.weight', 'loss.perceptual_loss.net.slice5.26.bias', 'loss.perceptual_loss.net.slice5.26.weight', 'loss.perceptual_loss.net.slice5.28.bias', 'loss.perceptual_loss.net.slice5.28.weight', 'loss.perceptual_loss.scaling_layer.scale', 'loss.perceptual_loss.scaling_layer.shift'] loaded 3D transformer's pretrained weights from F:\clip\EasyAnimate-main\models\Diffusion_Transformer\EasyAnimateV3-XL-2-InP-512x512\transformer ... The config attributes {'decay': 0.9999, 'inv_gamma': 1.0, 'min_decay': 0.0, 'optimization_step': 4400, 'power': 0.6666666666666666, 'update_after_step': 0, 'use_ema_warmup': False} were passed to Transformer3DModel, but are not expected and will be ignored. Please verify your config.json configuration file.

提示上面的问题之后，页面加载模型就显示error，服务就停了

[yunkchen]

停止原因跟这个最后一句日志没有关系 你关注下显存是不是满了

[zhangzhiz]

显存没有满，加载模型路径的时候只用了5g左右，我是16g的一个rtx4060的卡，不确定是不是卡的问题

[yunkchen]

代码是不是最新的？可以git pull一下

[zhangzhiz]

是新代码，用了main和v4分支，都是一样的问题

[yunkchen]

EasyAnimateV3-XL-2-512x512/目录下的文件确认下是不是都是完整的，可以跟huggingface上对比下

[yunkchen]

或者把模型文件删除，重新下载解压一遍

[willows888]

正方形的可以继续下去，虽然也报错

[willows888]

整合包报错后不会继续，comfyui报错后还能跳过错误继续下去，生成视频，但只能生成正方形的

[willows888]

我是2080ti 22G的，不支持bf16，选择的fp16，估计也有这个的原因

[willows888]

在报错与与继续运行之间，等待了3分钟左右，gpu始终满负荷运行