error:Incompatible shapes:[16，3] vs [48,3]

最近在网上找到了源码准备复现center loss，但是在调用的时候一直报错说：Incompatible shapes:[16，3] vs [48,3].(我的分类类别为3，batch size=16)，请问这是哪里错了？源码改的时候需要注意什么吗？

报错信息如下：Traceback (most recent call last):
File “E:/Inception改进LYJ/11111inception_v2_train_val.py”, line 251, in
snapshot_prefix=snapshot_prefix)
File “E:/Inception改进LYJ/11111inception_v2_train_val.py”, line 229, in train
snapshot_prefix, snapshot)
File “E:/Inception改进LYJ/11111inception_v2_train_val.py”, line 91, in step_train
keep_prob: 0.8, is_training: True})
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\client\session.py”, line 929, in run
run_metadata_ptr)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\client\session.py”, line 1152, in _run
feed_dict_tensor, options, run_metadata)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\client\session.py”, line 1328, in _do_run
run_metadata)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\client\session.py”, line 1348, in _do_call
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: Incompatible shapes: [16,3] vs. [48,3]
[[node sub (defined at E:\Inception改进LYJ\island_loss.py:15) ]]

Caused by op ‘sub’, defined at:
File “E:/11111_v_train_val.py”, line 251, in
snapshot_prefix=snapshot_prefix)
File “E:/11111_v_train_val.py”, line 181, in train
Island_loss, centers_update_op = Islandloss(labels=input_labels,features=out, alpha=1., alpha1=10., num_classes=3)
File “E:\island_loss.py”, line 15, in Islandloss
center_loss = tf.nn.l2_loss(features - centers_batch)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\ops\math_ops.py”, line 812, in binary_op_wrapper
return func(x, y, name=name)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\ops\gen_math_ops.py”, line 10130, in sub
“Sub”, x=x, y=y, name=name)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\framework\op_def_library.py”, line 788, in _apply_op_helper
op_def=op_def)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\util\deprecation.py”, line 507, in new_func
return func(*args, **kwargs)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\framework\ops.py”, line 3300, in create_op
op_def=op_def)
File “C:\Users\lyj\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\framework\ops.py”, line 1801, in __init__self._traceback = tf_stack.extract_stack()

error:Incompatible shapes:[16，3] vs [48,3]相关推荐

error LNK2019: 无法解析的外部符号 _WinMain@16，该符号在函数 ___tmainCR... 2011年08月05日 09:08:15
经过测试,能解决问题 error LNK2019: 无法解析的外部符号 _WinMain@16,该符号在函数 ___tmainCR... 一,问题描述 MSVCRTD.lib(crtexew.obj) ...
VS2015 error LNK2019 无法解析的外部符号 _WinMain@16，该符号在函数 int __cdecl invoke_main(void)
前言上一篇文章中,为了编译OpenCV的示例代码,然后搭建环境, 最后想直接运行示例demo,我在创建新工程的时候,直接创建了一个win32窗口项目,然后环境配置完成后,将opencv 的示例代码k ...
无法解析的外部符号 _WinMain@16，该符号在函数 ___tmainCRTStartup 中被引用
一,问题描述 MSVCRTD.lib(crtexew.obj) : error LNK2019: 无法解析的外部符号 _WinMain@16,该符号在函数 ___tmainCRTStartup 中被引 ...
error: incompatible types when assigning to type ‘const struct wiphy_wowlan_support‘
环境说明系统:Centos7 (3.10.0) 驱动版本:00014996-RTL8192EU_linux_v5.6.3.1_34030_COEX20171113-0047.20190613.zip ...
C: error: incompatible types when initializing type ‘int’ using type ‘atomic_t’ {aka ‘struct ＜anonym
typedef.c:17:10: error: incompatible types when initializing type 'int' using type 'atomic_t' {aka ' ...
Error: 17172, Severity: 16 SNIInitialize() failed with error 0x2.
为了安装SCCM 2012,今天将测试数据库升级到SQL Server 2008 SP3 + CU4, 但是改为端口后发现Servcie无法启动. 从SQL Server errorlog中看到下面的 ...
使用Process运行程序提示error: incompatible types: java.lang.Process cannot be converted to android.os.Proces
一.在android framework sevice里面添加下面的代码,使用Process java中运行第三方程序,编译的时候提示 error: incompatible types: jav ...
TensorFlow中出现Incompatible shapes between op input and calculated input gradient
Tensorflow反卷积操作中Incompatible shapes between op input and calculated input gradient问题解决办法欢迎使用Markdow ...
亚马逊：我们提取了BERT的一个最优子架构，只有Bert-large的16%，CPU推理速度提升7倍...
选自arXiv 作者:Adrian de Wynter.Daniel J. Perry 机器之心编译机器之心编辑部提取 BERT 子架构是一个非常值得探讨的问题,但现有的研究在子架构准确率和选择方 ...

error:Incompatible shapes:[16，3] vs [48,3]

error:Incompatible shapes:[16，3] vs [48,3]相关推荐

最新文章

热门文章