Python Keras中的自定义损失函数（IoU损失函数）和梯度误差？

我是ML新手，我正在尝试实现我自己的损失函数（IoU损失函数），但是得到了一个关于梯度的错误（“没有为任何变量提供梯度”）。 注意，例如，我试图预测表示矩阵中行数的数字（例如y_pred=[1 5 3 9]）。在损失函数中，我应该计算正确预测行的数量，然后将其除以y_true元素的总数。由于损失应该最小化，我在函数的末尾放了1-IoU

这是我的功能，更清楚地说，因为我不知道如何详细解释我的问题：

**def** loss_IoU(y_true, y_pred):

  intersection = []
  #roundig the predicted values y_pred, since the values could be floats ex. [1.5 2 8.98 ...]
  Roundy_pred = tf.round(y_pred)   # (none, 6) with shape of none, and the size of the outputs 6.
  intersection = tf.math.count_nonzero( Roundy_pred == y_true, axis=1, keepdims=True)   # (none, 1) count the number of equal values for each row (i.e. for each predicted y and a true y)
  union = y_true.shape[1] # which is 6 here
  iou = intersection/union
  returned_iou = 1 - iou
  return returned_iou

下面是我得到的错误：

    ValueError                                Traceback (most recent call last)
<ipython-input-176-631f68b50b34> in <module>()
      9 history = model.fit(AS_Training_Set, Label_Training_Set, 
     10                     steps_per_epoch=8, epochs=600, validation_data=
---> 11           (AS_Validation_Set, Label_Validation_Set))

10 frames
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/func_graph.py in wrapper(*args, **kwargs)
    971           except Exception as e:  # pylint:disable=broad-except
    972             if hasattr(e, "ag_error_metadata"):
--> 973               raise e.ag_error_metadata.to_exception(e)
    974             else:
    975               raise

ValueError: in user code:

    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:806 train_function  *
        return step_function(self, iterator)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:796 step_function  **
        outputs = model.distribute_strategy.run(run_step, args=(data,))
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/distribute/distribute_lib.py:1211 run
        return self._extended.call_for_each_replica(fn, args=args, kwargs=kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/distribute/distribute_lib.py:2585 call_for_each_replica
        return self._call_for_each_replica(fn, args, kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/distribute/distribute_lib.py:2945 _call_for_each_replica
        return fn(*args, **kwargs)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:789 run_step  **
        outputs = model.train_step(data)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:757 train_step
        self.trainable_variables)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:2737 _minimize
        trainable_variables))
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:562 _aggregate_gradients
        filtered_grads_and_vars = _filter_grads(grads_and_vars)
    /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:1271 _filter_grads
        ([v.name for _, v in grads_and_vars],))

    ValueError: No gradients provided for any variable: ['conv2d_16/kernel:0', 'conv2d_16/bias:0', 'conv2d_17/kernel:0', 'conv2d_17/bias:0', 'batch_normalization_8/gamma:0', 'batch_normalization_8/beta:0', 'conv2d_18/kernel:0', 'conv2d_18/bias:0', 'conv2d_19/kernel:0', 'conv2d_19/bias:0', 'batch_normalization_9/gamma:0', 'batch_normalization_9/beta:0', 'conv2d_20/kernel:0', 'conv2d_20/bias:0', 'conv2d_21/kernel:0', 'conv2d_21/bias:0', 'batch_normalization_10/gamma:0', 'batch_normalization_10/beta:0', 'conv2d_22/kernel:0', 'conv2d_22/bias:0', 'conv2d_23/kernel:0', 'conv2d_23/bias:0', 'batch_normalization_11/gamma:0', 'batch_normalization_11/beta:0', 'dense_4/kernel:0', 'dense_4/bias:0', 'dense_5/kernel:0', 'dense_5/bias:0'].

ValueError回溯（最近一次调用）
在（）
9历史=模型拟合（作为训练集、标签训练集、，
10步/epoch=8，epoch=600，验证数据=
--->11（作为验证集、标签验证集）
10帧
/包装器中的usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/func_graph.py（*args，**kwargs）
971例外情况为e:#pylint:disable=broad except
972如果hasattr（e，“AGU错误元数据”）：
-->973将e.ag\u错误\u元数据引发到\u异常（e）
974其他：
975提高
ValueError:在用户代码中：
/usr/local/lib/python3.6/dist包/tensorflow/python/keras/engine/training.py:806 train_函数*
返回步骤_函数（self、迭代器）
/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:796-step_函数**
输出=模型。分配策略。运行（运行步骤，参数=（数据，）
/usr/local/lib/python3.6/dist包/tensorflow/python/distribute/distribute_lib.py:1211运行
返回self.\u扩展。为每个\u副本调用\u（fn，args=args，kwargs=kwargs）
/usr/local/lib/python3.6/dist包/tensorflow/python/distribute/distribute_lib.py:2585调用每个副本
返回自我。为每个副本（fn、ARG、kwargs）调用
/usr/local/lib/python3.6/dist-packages/tensorflow/python/distribute/distribute\u-lib.py:2945\u为每个复制副本调用
返回fn（*args，**kwargs）
/usr/local/lib/python3.6/dist包/tensorflow/python/keras/engine/training.py:789运行步骤**
输出=型号列车步进（数据）
/usr/local/lib/python3.6/dist包/tensorflow/python/keras/engine/training.py:757 train\u步骤
可自我训练的变量）
/usr/local/lib/python3.6/dist包/tensorflow/python/keras/engine/training.py:2737
可训练变量）
/usr/local/lib/python3.6/dist packages/tensorflow/python/keras/optimizer\u v2/optimizer\u v2.py:562\u aggregate\u gradients
过滤的梯度和变量=\u过滤梯度（梯度和变量）
/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/optimizer\u v2/optimizer\u v2.py:1271\u filter\u grads
（[v.name代表u，v在grads和vars中]，）
ValueError：没有为任何变量提供渐变：['conv2d_16/内核：0'、'conv2d_16/偏差：0'、'conv2d_17/内核：0'、'conv2d_17/偏差：0'、'batch_normalization_8/gamma:0'、'batch_normalization_8/beta:0'、'conv2d_18/偏差：0'、'conv2d_19/内核：0'、'conv2d_19/偏差：0'、'batch_normalization_9/伽马：0'、'batch_normalization_9/beta:0'、'2d_normalization_20/内核：0'、'convu 20/偏差：0'、'_21/内核：0'，conv2d\u 21/偏差：0'，批处理标准化\u 10/伽马：0'，批处理标准化\u 10/β：0'，conv2d\u 22/内核：0'，conv2d\u 22/偏差：0'，conv2d\u 23/内核：0'，conv2d\u 23/偏差：0'，批处理标准化\u 11/伽马：0'，批处理标准化\u 11/β：0'，稠密\u 4/内核：0'，稠密\u 4/偏差：0'，稠密\u 5/偏差：0']。

有人能帮我解决这个错误吗？我试图修复它，但没有任何结果，我不知道问题是否出在舍入函数上

---

从现在开始感谢。

对于IoU损失函数，我将此函数用于Pascal VOC数据集

def IoU_loss(y_true, y_pred):
    nb_classes = K.int_shape(y_pred)[-1]
    iou = []
    pred_pixels = K.argmax(y_pred, axis=-1)
    for i in range(0, nb_classes):  # exclude first label (background) and last label (void)
        true_labels = K.equal(y_true[:, :, 0], i)
        pred_labels = K.equal(pred_pixels, i)
        inter = tf.cast(true_labels & pred_labels, dtype=tf.int32)
        union = tf.cast(true_labels | pred_labels, dtype=tf.int32)
        legal_batches = K.sum(tf.cast(true_labels, dtype=tf.int32), axis=1) > 0
        ious = K.sum(inter, axis=1) / K.sum(union, axis=1)
        iou.append(K.mean(ious[legal_batches]))

    iou = tf.stack(iou)
    legal_labels = ~tf.math.is_nan(iou)
    iou = iou[legal_labels]
    return K.mean(iou)

---

它需要一些修改，但也适用于您。

非常感谢您，我会尝试一下，希望它适用于我。