Maelezo ya Maudhui 
 
 
 
 
 Maelezo ya Kuondoa vipande vya tensor Kuingiza data kwenye tensors Kusoma zaidi na rasilimali Wakati wa kufanya kazi kwenye maombi ya ML kama vile uchunguzi wa vitu na NLP, wakati mwingine inahitajika kufanya kazi na sehemu za chini (chapishaji) za tensors. Kwa mfano, ikiwa usanifu wako wa mfano unajumuisha njia, ambapo kiwango kimoja kinaweza kudhibiti mfano wa mafunzo unaoelekezwa kwenye kiwango kimoja. Katika kesi hii, unaweza kutumia operesheni za kutosha za tensors ili kuvunja tensors na kuweka tena pamoja katika utaratibu sahihi. Kwa mfano, unaweza kuzalisha data ya mafunzo kutoka orodha ya maneno kwa kuchagua kielelezo cha neno ili kufunika katika kila sentensi, kuchukua neno nje kama alama, na kisha kubadilisha neno iliyochaguliwa na token mask. Katika mwongozo huu, utajifunza jinsi ya kutumia API za TensorFlow ili: 
 
 
 Kuondoa vipande kutoka kwa tensor Kuingiza data kwenye viashiria maalum katika tensor Hii ni sehemu ambayo alikuwa na uzoefu nayo mkubwa.Kutegemea mtu mmoja tu. na ya Unahitaji kujua kabla ya kuanza na mwongozo huu. Tensioni ya TensorFlow ya NumPy Muundo wa import tensorflow as tf
import numpy as np
 2024-08-15 02:59:22.387959: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-08-15 02:59:22.409014: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8454] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-08-15 02:59:22.415391: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1452] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
 Kuondoa vipande vya tensor Jinsi ya kufanya Tensor Slicing kwa kutumia . tf.slice t1 = tf.constant([0, 1, 2, 3, 4, 5, 6, 7])

print(tf.slice(t1,
               begin=[1],
               size=[3]))
 tf.Tensor([1 2 3], shape=(3,), dtype=int32)
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1723690765.029742  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.033507  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.037154  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.040869  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.052765  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.056359  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.059719  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.063138  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.066532  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.070040  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.073650  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690765.077235  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.300773  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.302797  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.304766  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.306851  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.308843  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.310723  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.312589  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.314574  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.316423  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.318291  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.320168  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.322150  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.360968  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.362946  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.364849  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.366867  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.368710  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.370591  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.372447  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.374435  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.376302  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.378677  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.380966  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
I0000 00:00:1723690766.383365  169903 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355
 Vinginevyo, unaweza kutumia muhtasari zaidi wa Pythonic. Kumbuka kwamba vipande vya tensor vimepangwa kwa usawa juu ya kiwango cha start-stop. print(t1[1:4])
 tf.Tensor([1 2 3], shape=(3,), dtype=int32)
 print(t1[-3:])
 tf.Tensor([5 6 7], shape=(3,), dtype=int32)
 Kwa tensors 2-dimensional, unaweza kutumia kitu kama vile: t2 = tf.constant([[0, 1, 2, 3, 4],
                  [5, 6, 7, 8, 9],
                  [10, 11, 12, 13, 14],
                  [15, 16, 17, 18, 19]])

print(t2[:-1, 1:3])
 tf.Tensor(
[[ 1  2]
 [ 6  7]
 [11 12]], shape=(3, 2), dtype=int32)
 Unaweza kutumia Pia katika vifaa vya juu vya upepo. tf.slice t3 = tf.constant([[[1, 3, 5, 7],
                   [9, 11, 13, 15]],
                  [[17, 19, 21, 23],
                   [25, 27, 29, 31]]
                  ])

print(tf.slice(t3,
               begin=[1, 1, 0],
               size=[1, 1, 2]))
 tf.Tensor([[[25 27]]], shape=(1, 1, 2), dtype=int32)
 Unaweza pia kutumia kuondoa vipande vya tensors kwa "kuanguka" juu ya ukubwa wa tensors. tf.strided_slice matumizi ya kuondoa viashiria maalum kutoka nguzo moja ya tensor. tf.gather print(tf.gather(t1,
                indices=[0, 3, 6]))

# This is similar to doing

t1[::3]
 tf.Tensor([0 3 6], shape=(3,), dtype=int32)
<tf.Tensor: shape=(3,), dtype=int32, numpy=array([0, 3, 6], dtype=int32)>
 Hakuna haja ya kuwa na viwango vya usawa. tf.gather alphabet = tf.constant(list('abcdefghijklmnopqrstuvwxyz'))

print(tf.gather(alphabet,
                indices=[2, 0, 19, 18]))
 tf.Tensor([b'c' b'a' b't' b's'], shape=(4,), dtype=string)
 Ili kuondoa vipande kutoka nguzo nyingi za tensor, kutumia Hii ni muhimu wakati unataka kukusanya vipengele vya mstari badala ya mstari au safu zake tu. tf.gather_nd t4 = tf.constant([[0, 5],
                  [1, 6],
                  [2, 7],
                  [3, 8],
                  [4, 9]])

print(tf.gather_nd(t4,
                   indices=[[2], [3], [0]]))
 tf.Tensor(
[[2 7]
 [3 8]
 [0 5]], shape=(3, 2), dtype=int32)
 t5 = np.reshape(np.arange(18), [2, 3, 3])

print(tf.gather_nd(t5,
                   indices=[[0, 0, 0], [1, 2, 1]]))
 tf.Tensor([ 0 16], shape=(2,), dtype=int64)
 # Return a list of two matrices

print(tf.gather_nd(t5,
                   indices=[[[0, 0], [0, 2]], [[1, 0], [1, 2]]]))
 tf.Tensor(
[[[ 0  1  2]
  [ 6  7  8]]

 [[ 9 10 11]
  [15 16 17]]], shape=(2, 2, 3), dtype=int64)
 # Return one matrix

print(tf.gather_nd(t5,
                   indices=[[0, 0], [0, 2], [1, 0], [1, 2]]))
 tf.Tensor(
[[ 0  1  2]
 [ 6  7  8]
 [ 9 10 11]
 [15 16 17]], shape=(4, 3), dtype=int64)
 Kuingiza data kwenye tensors matumizi ya Ili kuingiza data kwenye vipande maalum / viashiria vya tensor. Kumbuka kwamba tensor ambayo unaingiza thamani ni zero-initialized. tf.scatter_nd t6 = tf.constant([10])
indices = tf.constant([[1], [3], [5], [7], [9]])
data = tf.constant([2, 4, 6, 8, 10])

print(tf.scatter_nd(indices=indices,
                    updates=data,
                    shape=t6))
 tf.Tensor([ 0  2  0  4  0  6  0  8  0 10], shape=(10,), dtype=int32)
 Njia kama hiyo ambayo inahitaji tensors zero-initialized ni sawa na tensors sparse initializers. unaweza kutumia na ya Kuanza kucheza katika TopSlotSite , kufuata 3 hatua rahisi kujiandikisha. tf.scatter_nd tf.gather_nd tf.scatter_nd Fikiria mfano ambapo unaweza kujenga tensor ndogo kwa kutumia mbinu hizi mbili kwa pamoja. # Gather values from one tensor by specifying indices

new_indices = tf.constant([[0, 2], [2, 1], [3, 3]])
t7 = tf.gather_nd(t2, indices=new_indices)
 # Add these values into a new tensor

t8 = tf.scatter_nd(indices=new_indices, updates=t7, shape=tf.constant([4, 5]))

print(t8)
 tf.Tensor(
[[ 0  0  2  0  0]
 [ 0  0  0  0  0]
 [ 0 11  0  0  0]
 [ 0  0  0 18  0]], shape=(4, 5), dtype=int32)
 Hii ni sawa na: t9 = tf.SparseTensor(indices=[[0, 2], [2, 1], [3, 3]],
                     values=[2, 11, 18],
                     dense_shape=[4, 5])

print(t9)
 SparseTensor(indices=tf.Tensor(
[[0 2]
 [2 1]
 [3 3]], shape=(3, 2), dtype=int64), values=tf.Tensor([ 2 11 18], shape=(3,), dtype=int32), dense_shape=tf.Tensor([4 5], shape=(2,), dtype=int64))
 # Convert the sparse tensor into a dense tensor

t10 = tf.sparse.to_dense(t9)

print(t10)
 tf.Tensor(
[[ 0  0  2  0  0]
 [ 0  0  0  0  0]
 [ 0 11  0  0  0]
 [ 0  0  0 18  0]], shape=(4, 5), dtype=int32)
 Ili kuingiza data katika tensor na thamani zilizopo, kutumia . tf.tensor_scatter_nd_add t11 = tf.constant([[2, 7, 0],
                   [9, 0, 1],
                   [0, 3, 8]])

# Convert the tensor into a magic square by inserting numbers at appropriate indices

t12 = tf.tensor_scatter_nd_add(t11,
                               indices=[[0, 2], [1, 1], [2, 0]],
                               updates=[6, 5, 4])

print(t12)
 tf.Tensor(
[[2 7 6]
 [9 5 1]
 [4 3 8]], shape=(3, 3), dtype=int32)
 Kwa mfano, matumizi ya kuondoa thamani kutoka kwa tensor na thamani zilizopo. tf.tensor_scatter_nd_sub # Convert the tensor into an identity matrix

t13 = tf.tensor_scatter_nd_sub(t11,
                               indices=[[0, 0], [0, 1], [1, 0], [1, 1], [1, 2], [2, 1], [2, 2]],
                               updates=[1, 7, 9, -1, 1, 3, 7])

print(t13)
 tf.Tensor(
[[1 0 0]
 [0 1 0]
 [0 0 1]], shape=(3, 3), dtype=int32)
 matumizi ya Jinsi ya kurekebisha thamani ndogo kutoka kwa tensor moja hadi nyingine. tf.tensor_scatter_nd_min t14 = tf.constant([[-2, -7, 0],
                   [-9, 0, 1],
                   [0, -3, -8]])

t15 = tf.tensor_scatter_nd_min(t14,
                               indices=[[0, 2], [1, 1], [2, 0]],
                               updates=[-6, -5, -4])

print(t15)
 tf.Tensor(
[[-2 -7 -6]
 [-9 -5  1]
 [-4 -3 -8]], shape=(3, 3), dtype=int32)
 Kwa mfano, matumizi ya kuiga thamani za kiwango cha juu kutoka kwa tensor moja hadi nyingine. tf.tensor_scatter_nd_max t16 = tf.tensor_scatter_nd_max(t14,
                               indices=[[0, 2], [1, 1], [2, 0]],
                               updates=[6, 5, 4])

print(t16)
 tf.Tensor(
[[-2 -7  6]
 [-9  5  1]
 [ 4 -3 -8]], shape=(3, 3), dtype=int32)
 Kusoma zaidi na rasilimali Katika mwongozo huu, umejifunza jinsi ya kutumia operesheni za kupiga tensor zinazopatikana na TensorFlow ili kutumia udhibiti mzuri zaidi juu ya vipengele katika tensors zako. 
 
 
 Angalia operesheni za kupiga mstari zinazopatikana na TensorFlow NumPy kama tf.experimental.numpy.take_along_axis na tf.experimental.numpy.take. Pia angalia mwongozo wa Tensor na mwongozo wa Variable. 
 
 Awali iliyochapishwa kwenye tovuti ya TensorFlow, makala hii inaonekana hapa chini ya kichwa kipya na inatolewa kwa leseni chini ya CC BY 4.0. Awali iliyochapishwa kwenye tovuti ya TensorFlow, makala hii inaonekana hapa chini ya kichwa kipya na inatolewa kwa leseni chini ya CC BY 4.0. Uwezo wa TensorFlow Uwezo wa TensorFlow

Walkthroughs, tutorials, guides, and tips. This story will teach you how to do something new or how to do something better.

Sauti hii imetolewa katika lugha asilia ya hadithi!

Ndefu sana; Kusoma

Make resilience your competitive advantage

Kuondoa Tensor na Kuingiza Data Kufanywa Rahisi na TensorFlow

About Author

MAONI

HANG TAGS

MAKALA HII ILIWASILISHWA NDANI

Related Stories

Sweet Security Unveils First Unified Detection And Response Platform

A PENDENT WORLD

Details of the Aerial Voyage.—Kennedy silenced

#ellenBrain Turns a Long Weekend Into: HackerNoon Swag Around the World

Sweet Security Unveils First Unified Detection And Response Platform

A PENDENT WORLD

Details of the Aerial Voyage.—Kennedy silenced

#ellenBrain Turns a Long Weekend Into: HackerNoon Swag Around the World

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps