CTC loss incorrect computation

Question

Python's input:

sequence1 = [0, 1] 
output1 = [[1, 0, 0],[0, 1, 0]]
loss = tf.compat.v1.nn.ctc_loss(
                               labels=tf.sparse.from_dense([sequence1]),
                               inputs=np.array([output1]).astype('float32'),
                               sequence_length=[2],
                               time_major = False
                               )
print(loss.numpy())

Python's output:

array([1.2408944], dtype=float32)

Mathematica's input:

CTCLossLayer[<|
   "Input" -> {{1, 0, 0}, {0, 1, 0}}, 
   "Target" -> {1, 2} (* Index starts from 1 and not 0 *)
|>]

Mathematica's output:

0.

Since my sequence is [0,1] and it's one hot encoding (when number of characters is 2 + 1(for CTC)) is [[1,0,0],[0,1,0]] - shouldn't the nn.ctc_loss be 0? As in Mathematica? Thanks!