tensorflow ValueError: Dimensions must be equal
2017-05-24 22:56
597 查看
在使用tf.matmul进行矩阵相乘的时候报了这么个错
按道理来说21×8的矩阵乘以21行的列向量是可以乘的，乘完的结果的是21一行的列向量，最后发现tf.matmul函数只能进行矩阵间的乘法操作
。。。。 File "/home/batman/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 610, in call_cpp_shape_fn debug_python_shape_fn, require_shape_fn) File "/home/batman/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 676, in _call_cpp_shape_fn_impl raise ValueError(err.message) ValueError: Dimensions must be equal, but are 8 and 21 for 'MatMul' (op: 'MatMul') with input shapes: [21,8], [21,0].].
按道理来说21×8的矩阵乘以21行的列向量是可以乘的，乘完的结果的是21一行的列向量，最后发现tf.matmul函数只能进行矩阵间的乘法操作
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- ValueError: Tensorflow error: “Tensor must be from the same graph as Tensor…”
- Python3 error: initial_value must be str or None
- python中关于Np.hstack()错误的ValueError: all the input arrays must have same number of dimensions
- Python Error1: ValueError: range parameter must be finite.
- ValueError: Shape must be rank 0 but is rank 1 for 'train_data/ReadFile' (op: 'ReadFile') with input
- 手把手教你 python读取文件报错:ValueError encoding must be one of 'utf_8', 'big5', or 'gbk'.
- TensorBoard报错：InvalidArgumentError: You must feed a value for placeholder tensor 'inputs/x_input' wi
- Tensorflow报错之一：InvalidArgumentError: You must feed a value for placeholder tensor 'inputs/x_input'
- UMLOG: (ERROR) Umeng event's keys and values must be string ,key's max length 128 and value's max le
- tensorflow.python.framework.errors.InvalidArgumentError: Input to reshape is a tensor with xxx value
- ValueError: fill value must be in categories.
- Java - Error: Attribute value must be constant
- 使用 python3.6 调用 jieba 库出现的 ValueError: dictionary file dict.txt must be utf-8 错误
- tensorflow报错：ValueError: features should be a dictionary of `Tensor`s. Given type: ……
- C++ Error : initial value of reference to non-const must be an lvalue
- seq2seq_model.py ValueError: Attempt to reuse RNNCell <tensorflow.contrib.rnn.python.ops.core_rnn_ce
- 处理错误TypeError: initial_value must be unicode or None, not str
- Error: initial value of reference to non const must be lvalue 原因以及解决方法
- [异常整理]java.lang.IllegalArgumentException: Attribute value must not be null
- Can't User UserProfileManager.GetUserProfile Get User Userprofile In Web Service,Error:System.ArgumentNullException: Value Cannot Be Null.(不能在Webservice中使用UserProfileManager的GetUserProfile得到用户配置，报错：System.ArgumentNullException: 值不能为空。)