wuxianjun666

I found out what the problem is, in lora.hpp, comment out（注释） this line of code and it works! 

 can not delete sd_ctx->sd; when delete sd_ctx->sd, it should delete GGMLModule many times, because there are many struct succeed GGMLModule,  but always gets stuck here, 

> Can you provide more information, such as the operating system you're using, the compiler version, the command-line arguments you're using, and the corresponding output? I run in windows10, and...

> Can you provide more information, such as the operating system you're using, the compiler version, the command-line arguments you're using, and the corresponding output? So I had to eliminate...

yes，when i do not use lora, it is normal

I found out what the problem is, in lora.hpp, comment out（注释） this line of code and it works! 

can you finish multimodal example

> Thanks for your amazing work. > > **Error** > > ``` > [[{{node ParseSingleSequenceExample/ParseSequenceExample/ParseSequenceExampleV2}}]] [Op:IteratorGetNext] > 2022-04-15 13:19:55.258721: W tensorflow/core/framework/op_kernel.cc:1745] OP_REQUIRES failed at example_parsing_ops.cc:480 : INVALID_ARGUMENT: Name: , Key:...

if i use Visual Studio 16 2019 is ok ,when i use cmake -G "MinGW Makefiles" it broken -- Building for: Visual Studio 16 2019 -- Selecting Windows SDK version...

 我在main函数中训练，在main函数中会存在model = build_model(config["Architecture"])，然后我训练完成后会直接调用paddle2engine，在paddle2engine中会调用main_export，在main_export中，也会创建这个模型，model = build_model(config["Architecture"])，就是在这报错的，报错是在head部分报错的： self.fc = nn.Linear( in_channels, class_dim, weight_attr=ParamAttr( name="fc_0.w_0", initializer=nn.initializer.Uniform(-stdv, stdv) ), bias_attr=ParamAttr(name="fc_0.b_0"), ) 网上搜了一下说没清空缓存，这个我不是很理解，创建model是两个不同的实例，为什么会冲突呢，这个key为什么会被占用