@@ -570,6 +572,78 @@ def cross_entropy(input, target, weight=None, size_average=None, ignore_index=-1
     result = ms.ops.cross_entropy(input, target, weight, ignore_index, reduction, label_smoothing)
     return cast_to_adapter_tensor(result)
 
+def ctc_loss(log_probs, targets, input_lengths, target_lengths, blank=0, reduction='mean', zero_infinity=False):

@@ -573,0 +634,4 @@
+    rlt = ms.ops.multilabel_soft_margin_loss(input, target, weight, reduction)
+    return cast_to_adapter_tensor(rlt)
+
+# def multi_margin_loss(input, target, p=1, margin=1, weight=None, size_average=None, reduce=None, reduction='mean'):

@@ -1938,0 +2120,4 @@
+
+def unfold(input, kernel_size, dilation=1, padding=0, stride=1):
+    input_ms = cast_to_ms_tensor(input)
+    output = ms.ops.unfold(input_ms, kernel_size, dilation, padding, stride)

@@ -1938,0 +2136,4 @@
+    else:
+        shape_tmp = kernel_size[0] * kernel_size[1]
+    input_ms = input_ms.reshape(shape[0], -1, shape_tmp, shape[2])
+    output = ms.ops.fold(input_ms, ms.Tensor(output_size), kernel_size, dilation, padding, stride)

@@ -0,0 +31,4 @@
+    assert np.allclose(result_ms.asnumpy(), result_torch.numpy())
+    assert result_ms.shape == result_torch.shape
+
+def gaussian_nll_loss():