在 TensorFlow.js 中切片 3d 张量
Slice a 3d tensor in TensorFlow.js
如何将以下 TensorFlow Python 代码翻译成 TensorFlow.js?
New_tensor = A_3D_tensor[:, :seq_len, :]
您可以使用 slice
method (or, for more complex cases, with stridedSlice
):
const t = tf.range(0, 36, 1).as3D(2, 6, 3);
const seq_len = 4;
const t_sliced = t.slice([0, 0, 0], [-1, seq_len, -1]);
t_sliced.print();
/*
"Tensor
[[[0 , 1 , 2 ],
[3 , 4 , 5 ],
[6 , 7 , 8 ],
[9 , 10, 11]],
[[18, 19, 20],
[21, 22, 23],
[24, 25, 26],
[27, 28, 29]]]"
*/
如何将以下 TensorFlow Python 代码翻译成 TensorFlow.js?
New_tensor = A_3D_tensor[:, :seq_len, :]
您可以使用 slice
method (or, for more complex cases, with stridedSlice
):
const t = tf.range(0, 36, 1).as3D(2, 6, 3);
const seq_len = 4;
const t_sliced = t.slice([0, 0, 0], [-1, seq_len, -1]);
t_sliced.print();
/*
"Tensor
[[[0 , 1 , 2 ],
[3 , 4 , 5 ],
[6 , 7 , 8 ],
[9 , 10, 11]],
[[18, 19, 20],
[21, 22, 23],
[24, 25, 26],
[27, 28, 29]]]"
*/