code：

from paddle import nn
import numpy as np
import paddle

embedding = nn.Embedding(100, 200)

x = np.array([[4, 1, 4, 10000, 3, 9, 900, 1, 1, 5],
              [8, 6, 5, 8000, 2, 7, 3, 1, 1, 6]])
input_paddle = paddle.to_tensor(x, dtype='int64')

output = embedding(input_paddle)
print(output)

console：

Tensor(shape=[2, 10, 200], dtype=float32, place=CUDAPlace(0), stop_gradient=False,
       [[[ 0.07328835, -0.09469348,  0.07219494, ...,  0.01943331,
          -0.12398233, -0.08197384],
         [-0.08139029, -0.10521033,  0.11584321, ..., -0.03910951,
           0.12038675, -0.09803125],
         [ 0.07328835, -0.09469348,  0.07219494, ...,  0.01943331,
          -0.12398233, -0.08197384],
         ...,
         [-0.08139029, -0.10521033,  0.11584321, ..., -0.03910951,
           0.12038675, -0.09803125],
         [-0.08139029, -0.10521033,  0.11584321, ..., -0.03910951,
           0.12038675, -0.09803125],
         [ 0.01028429,  0.04357532, -0.07706941, ...,  0.07706572,
           0.09165546,  0.07484353]],
        [[ 0.08599947, -0.00237131,  0.08609001, ..., -0.00191966,
           0.10842200, -0.08339714],
         [ 0.02185206,  0.10013362,  0.05010634, ...,  0.00987017,
           0.13578133, -0.01342258],
         [ 0.01028429,  0.04357532, -0.07706941, ...,  0.07706572,
           0.09165546,  0.07484353],
         ...,
         [-0.08139029, -0.10521033,  0.11584321, ..., -0.03910951,
           0.12038675, -0.09803125],
         [-0.08139029, -0.10521033,  0.11584321, ..., -0.03910951,
           0.12038675, -0.09803125],
         [ 0.02185206,  0.10013362,  0.05010634, ...,  0.00987017,
           0.13578133, -0.01342258]]])

paddle：2.2.1

我embedding的大小只有100
为什么输入10000， 900， 800000却不会报错