TensorFlow waxaa laga yaqaan TensorFlow.

Qalabka Overview Haku Nala soo xiriir Haku Shahaado shahaado Muuqaalka ee DTypes Telefoonka Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Tensors Qalabka Tensile Qalabka Tensile Tensors waa arrayada multi-dimensional oo ka mid ah a Waxaad ka mid ah ka mid ah dhismaha Qalabka . dtype dtypes tf.dtypes Waxaad ka soo xiriir Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka . Haku np.arrays All tensors are immutable like Python numbers and strings: Waxaad ka mid ka mid ah in la soo saarka ah ee tensors, kaliya in la soo saarka cusub. import tensorflow as tf import numpy as np 2024-08-15 03:05:18.327501: 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 03:05:18.348450: 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 03:05:18.354825: 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 Haku Waayo, waxaa loo isticmaali karaa tensors basics. Waxaad ka mid ah "scalar" ama "rank-0" tensor. A scalar waa mid ka mid ah mid ka mid ah "ax". # This will be an int32 tensor by default; see "dtypes" below. rank_0_tensor = tf.constant(4) print(rank_0_tensor) tf.Tensor(4, shape=(), dtype=int32) WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1723691120.932442 176945 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 00:00:1723691120.936343 176945 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 00:00:1723691120.940040 176945 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 00:00:1723691120.943264 176945 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 00:00:1723691120.954872 176945 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 00:00:1723691120.958376 176945 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 00:00:1723691120.961894 176945 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 00:00:1723691120.964843 176945 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 00:00:1723691120.967730 176945 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 00:00:1723691120.971300 176945 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 00:00:1723691120.974711 176945 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 00:00:1723691120.977717 176945 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 00:00:1723691122.208679 176945 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 00:00:1723691122.210786 176945 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 00:00:1723691122.212791 176945 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 00:00:1723691122.214776 176945 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 00:00:1723691122.216798 176945 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 00:00:1723691122.218734 176945 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 00:00:1723691122.220650 176945 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 00:00:1723691122.222554 176945 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 00:00:1723691122.224486 176945 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 00:00:1723691122.226429 176945 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 00:00:1723691122.228329 176945 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 00:00:1723691122.230251 176945 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 00:00:1723691122.269036 176945 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 00:00:1723691122.271069 176945 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 00:00:1723691122.273006 176945 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 00:00:1723691122.274956 176945 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 00:00:1723691122.276917 176945 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 00:00:1723691122.278854 176945 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 00:00:1723691122.280754 176945 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 00:00:1723691122.282664 176945 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 00:00:1723691122.284613 176945 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 00:00:1723691122.287058 176945 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 00:00:1723691122.289508 176945 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 00:00:1723691122.291891 176945 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 Tensor "vector" ama "rank-1" waa sida lambarka ah oo ka mid ah wax soo saarka. # Let's make this a float tensor. rank_1_tensor = tf.constant([2.0, 3.0, 4.0]) print(rank_1_tensor) tf.Tensor([2. 3. 4.], shape=(3,), dtype=float32) A "matrix" or "rank-2" tensor has two axes: # If you want to be specific, you can set the dtype (see below) at creation time rank_2_tensor = tf.constant([[1, 2], [3, 4], [5, 6]], dtype=tf.float16) print(rank_2_tensor) tf.Tensor( [[1. 2.] [3. 4.] [5. 6.]], shape=(3, 2), dtype=float16) A scalar, shape: [] A vector, shape: [3] A matrix, shape: [3, 2] Tensors may have more axes; here is a tensor with three axes: # There can be an arbitrary number of # axes (sometimes called "dimensions") rank_3_tensor = tf.constant([ [[0, 1, 2, 3, 4], [5, 6, 7, 8, 9]], [[10, 11, 12, 13, 14], [15, 16, 17, 18, 19]], [[20, 21, 22, 23, 24], [25, 26, 27, 28, 29]],]) print(rank_3_tensor) tf.Tensor( [[[ 0 1 2 3 4] [ 5 6 7 8 9]] [[10 11 12 13 14] [15 16 17 18 19]] [[20 21 22 23 24] [25 26 27 28 29]]], shape=(3, 2, 5), dtype=int32) Waxaad ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah. A 3-axis tensor, shape: [3, 2, 5] Ma rabtaa in aad u isticmaali karaa macluumaadka macluumaadka macluumaadka. ama la Haku: np.array tensor.numpy np.array(rank_2_tensor) array([[1., 2.], [3., 4.], [5., 6.]], dtype=float16) rank_2_tensor.numpy() array([[1., 2.], [3., 4.], [5., 6.]], dtype=float16) Tensors waa mid ka mid ah floats iyo incts, laakiin waxaa ka mid ah oo badan oo ka mid ah: Qalabka Kompleks Haku Qalabka Tensors waa "rectangular" -eyna waa, ka mid ah cadaadis oo dhan, cadaadis oo dhan waa mid ka mid ah. tf.Tensor Qalabka Tensile Qalabka Tension Waxaad ka heli karaa macluumaadka baska ee tensors, oo ka mid ah add-on, multiplication element-like, iyo matrix multiplication. a = tf.constant([[1, 2], [3, 4]]) b = tf.constant([[1, 1], [1, 1]]) # Could have also said `tf.ones([2,2], dtype=tf.int32)` print(tf.add(a, b), "\n") print(tf.multiply(a, b), "\n") print(tf.matmul(a, b), "\n") tf.Tensor( [[2 3] [4 5]], shape=(2, 2), dtype=int32) tf.Tensor( [[1 2] [3 4]], shape=(2, 2), dtype=int32) tf.Tensor( [[3 3] [7 7]], shape=(2, 2), dtype=int32) print(a + b, "\n") # element-wise addition print(a * b, "\n") # element-wise multiplication print(a @ b, "\n") # matrix multiplication tf.Tensor( [[2 3] [4 5]], shape=(2, 2), dtype=int32) tf.Tensor( [[1 2] [3 4]], shape=(2, 2), dtype=int32) tf.Tensor( [[3 3] [7 7]], shape=(2, 2), dtype=int32) Tensors waxaa loo isticmaali karaa in ka mid ah kala duwan oo loo isticmaali karaa (or "Ops"). c = tf.constant([[4.0, 5.0], [10.0, 1.0]]) # Find the largest value print(tf.reduce_max(c)) # Find the index of the largest value print(tf.math.argmax(c)) # Compute the softmax print(tf.nn.softmax(c)) tf.Tensor(10.0, shape=(), dtype=float32) tf.Tensor([1 0], shape=(2,), dtype=int64) tf.Tensor( [[2.6894143e-01 7.3105854e-01] [9.9987662e-01 1.2339458e-04]], shape=(2, 2), dtype=float32) Shuruudaha ugu horeysay ee loo yaqaan 'TensorFlow' waa in la soo xiriir oo ka mid ah t.convert_to_tensor. Shuruudaha ugu horeysay ee loo yaqaan 'TensorFlow' waa in la soo xiriir oo ka mid ah t.convert_to_tensor. tf.convert_to_tensor([1,2,3]) tf.reduce_max([1,2,3]) tf.reduce_max(np.array([1,2,3])) Nala soo xiriir Tensors waxay ka mid ah dhismaha. Shuruudaha: dhismaha (number of elements) ee caadiga ah ee tensor. Xisaabinta: Xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta: xisaabinta Axis ama Dimension: Dhammaan ah ee tensor. Taageerada: Taageerada ugu badan ee tensor, wax soo saarka ee vector shahaado. Markaad: Markaad ka mid ah aad u aragto reference la "tensor two-dimensional", a rank-2 tensor ma ahaatee ma ahaaday in 2D space. Sida loo yaqaan "tensor two-dimensional" (tensor two-dimensional), tensor rank-2 waxay ka dhigi karaa in ay ku yaalaa 2D. Note: Qalabka iyo Objects waxaa la isticmaali karaa in la isticmaali karaa: tf.TensorShape rank_4_tensor = tf.zeros([3, 2, 4, 5]) A rank-4 tensor, shape: [3, 2, 4, 5] print("Type of every element:", rank_4_tensor.dtype) print("Number of axes:", rank_4_tensor.ndim) print("Shape of tensor:", rank_4_tensor.shape) print("Elements along axis 0 of tensor:", rank_4_tensor.shape[0]) print("Elements along the last axis of tensor:", rank_4_tensor.shape[-1]) print("Total number of elements (3*2*4*5): ", tf.size(rank_4_tensor).numpy()) Type of every element: Number of axes: 4 Shape of tensor: (3, 2, 4, 5) Elements along axis 0 of tensor: 3 Elements along the last axis of tensor: 5 Total number of elements (3*2*4*5): 120 Sidaa waxaa loo yaabaa in la iyo Qalabka Qalabka Qalabka Qalabka wax soo saarka.If you need a Haku Marka Function. wax soo saarka waa caadiga ah, laakiin sidoo kale waxaa laga yaabaa in loo yaqaan 'graphs' (wax yar). Tensor.ndim Tensor.shape Tensor Tensor tf.rank tf.shape tf.rank(rank_4_tensor) tf.shape(rank_4_tensor) Markaad ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid Typical axis order Haku Qalabka Single Axis TensorFlow waxay ku yaqaan 'Python Standard Indexing Rules' (Riixaanka Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka). , iyo raaxada ugu weyn ee indexing NumPy. Qalabka dhismaha, dhismaha, dhismaha, dhismaha, dhismaha, dhismaha Qalabka ugu horeysay ee 0 Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Colons, :, waxaa loo isticmaali karaa in ay qaado: start:stop:step rank_1_tensor = tf.constant([0, 1, 1, 2, 3, 5, 8, 13, 21, 34]) print(rank_1_tensor.numpy()) [ 0 1 1 2 3 5 8 13 21 34] Qalabka ugu soo saarka ah waxaa loo yaqaan 'Scalar Indexing Removes the Axis': print("First:", rank_1_tensor[0].numpy()) print("Second:", rank_1_tensor[1].numpy()) print("Last:", rank_1_tensor[-1].numpy()) First: 0 Second: 1 Last: 34 Qalabka A Shuruudaha ugu horeysay ee Axial: : print("Everything:", rank_1_tensor[:].numpy()) print("Before 4:", rank_1_tensor[:4].numpy()) print("From 4 to the end:", rank_1_tensor[4:].numpy()) print("From 2, before 7:", rank_1_tensor[2:7].numpy()) print("Every other item:", rank_1_tensor[::2].numpy()) print("Reversed:", rank_1_tensor[::-1].numpy()) Everything: [ 0 1 1 2 3 5 8 13 21 34] Before 4: [0 1 1 2] From 4 to the end: [ 3 5 8 13 21 34] From 2, before 7: [1 2 3 5 8] Every other item: [ 0 1 3 8 21] Reversed: [34 21 13 8 5 3 2 1 1 0] Qalabka Multi-axis Tensors heerka ugu horeysay waa indexed by passing multiple indices. Riixaanka ugu fiican oo ku saabsan xisaabinta single-axis waxaa loo isticmaali karaa in cadaadis ah. print(rank_2_tensor.numpy()) [[1. 2.] [3. 4.] [5. 6.]] Waayo, waxa uu ka mid ah in ay ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ka mid ah. # Pull out a single value from a 2-rank tensor print(rank_2_tensor[1, 1].numpy()) You can index using any combination of integers and slices: # Get row and column tensors print("Second row:", rank_2_tensor[1, :].numpy()) print("Second column:", rank_2_tensor[:, 1].numpy()) print("Last row:", rank_2_tensor[-1, :].numpy()) print("First item in last column:", rank_2_tensor[0, -1].numpy()) print("Skip the first row:") print(rank_2_tensor[1:, :].numpy(), "\n") Second row: [3. 4.] Second column: [2. 4. 6.] Last row: [5. 6.] First item in last column: 2.0 Skip the first row: [[3. 4.] [5. 6.]] Here is an example with a 3-axis tensor: print(rank_3_tensor[:, :, 4]) tf.Tensor( [[ 4 9] [14 19] [24 29]], shape=(3, 2), dtype=int32) Selecting the last feature across all locations in each example in the batch QEEBE Sida loo isticmaali karaa indexing si ay u isticmaali karaa elementada kala duwan ee tensors aad. Qalabka Tensor Shahaado shahaado Shuruudaha ugu horeysay ee Tensor waxaa loo isticmaali karaa. # Shape returns a `TensorShape` object that shows the size along each axis x = tf.constant([[1], [2], [3]]) print(x.shape) (3, 1) # You can convert this object into a Python list, too print(x.shape.as_list()) [3, 1] Waxaad ka mid ah ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah. wax soo saarka waa la soo saarka iyo wax soo saarka xawaaraha xawaaraha xawaaraha waaweyn. tf.reshape # You can reshape a tensor to a new shape. # Note that you're passing in a list reshaped = tf.reshape(x, [1, 3]) print(x.shape) print(reshaped.shape) (3, 1) (1, 3) TensorFlow waxaa loo isticmaali karaa nidaamka "Row-major" ah ee ku saabsan muujinada C-style, oo loo isticmaali karaa si ay u isticmaali karaa in ay u isticmaali karaa in ay u isticmaali karaa in ay u isticmaali karaa si ay u isticmaali karaa in ay u isticmaali karaa in ay u isticmaali karaa in ay u isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa in ay isticmaali karaa. print(rank_3_tensor) tf.Tensor( [[[ 0 1 2 3 4] [ 5 6 7 8 9]] [[10 11 12 13 14] [15 16 17 18 19]] [[20 21 22 23 24] [25 26 27 28 29]]], shape=(3, 2, 5), dtype=int32) Sida loo yaqaan Tensor, sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale sidoo kale. # A `-1` passed in the `shape` argument says "Whatever fits". print(tf.reshape(rank_3_tensor, [-1])) tf.Tensor( [ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29], shape=(30,), dtype=int32) Sida loo isticmaali karaa mid ka mid ah Waxaa la soo bandhigiisa in la soo bandhigiis ah oo la soo bandhigiis ah oo la soo bandhigiis. QEEBE tf.reshape 1 3x2x5 tensor, dhismaha ku (3x2)x5 ama 3x(2x5) waa mid ka mid ah wax soo saarka ah in la sameeyo, sida yar oo ka mid ah ka mid ah: print(tf.reshape(rank_3_tensor, [3*2, 5]), "\n") print(tf.reshape(rank_3_tensor, [3, -1])) tf.Tensor( [[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19] [20 21 22 23 24] [25 26 27 28 29]], shape=(6, 5), dtype=int32) tf.Tensor( [[ 0 1 2 3 4 5 6 7 8 9] [10 11 12 13 14 15 16 17 18 19] [20 21 22 23 24 25 26 27 28 29]], shape=(3, 10), dtype=int32) Some good reshapes. Reshaping waa in la "shahaado" oo ka mid ah wax soo saarka cusub oo ka mid ah oo ka mid ah oo ka mid ah wax soo saarka ah, laakiin waxa uu ka mid ah wax soo saarka ah oo ka mid ah wax soo saarka. Qalabka Qalabka Ma rabtaa, waxaan u baahan tahay Waayo, tf.reshape tf.transpose # Bad examples: don't do this # You can't reorder axes with reshape. print(tf.reshape(rank_3_tensor, [2, 3, 5]), "\n") # This is a mess print(tf.reshape(rank_3_tensor, [5, 6]), "\n") # This doesn't work at all try: tf.reshape(rank_3_tensor, [7, -1]) except Exception as e: print(f"{type(e).__name__}: {e}") tf.Tensor( [[[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14]] [[15 16 17 18 19] [20 21 22 23 24] [25 26 27 28 29]]], shape=(2, 3, 5), dtype=int32) tf.Tensor( [[ 0 1 2 3 4 5] [ 6 7 8 9 10 11] [12 13 14 15 16 17] [18 19 20 21 22 23] [24 25 26 27 28 29]], shape=(5, 6), dtype=int32) InvalidArgumentError: { {function_node __wrapped__Reshape_device_/job:localhost/replica:0/task:0/device:GPU:0} } Input to reshape is a tensor with 30 values, but the requested shape requires a multiple of 7 [Op:Reshape] Some bad reshapes. Ma rabtaa in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan yahay. (Axaas-duurka waa ma aha) ama shuruudaha oo dhan waa (Range of the tensor waa in la soo xiriir). None None Riix oo ka mid ah tf.RaggedTensor, dhismaha ka mid ah ka mid ah TensorFlow's symbolic, graph-building APIs: Qalabka Qalabka adeegga ah ee API. Muuqaalka Haku Haku Waayo, a Dhammaan waxaa loo isticmaali karaa Qalabka tf.Tensor Tensor.dtype Sida loo helo a ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ka mid ah. tf.Tensor TensorFlow waxaa loo yaqaan "TensorFlow" oo loo yaqaan "TensorFlow" waxaa loo yaqaan "TensorFlow" oo loo yaqaan "TensorFlow". Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Sidaas, TensorFlow waxay isticmaali karaan NumPy waxay isticmaali karaa in ay u isticmaali karaa arrayada. tf.int32 tf.float32 Waxaad ka mid ah ka mid ah Type. the_f64_tensor = tf.constant([2.2, 3.3, 4.4], dtype=tf.float64) the_f16_tensor = tf.cast(the_f64_tensor, dtype=tf.float16) # Now, cast to an uint8 and lose the decimal precision the_u8_tensor = tf.cast(the_f16_tensor, dtype=tf.uint8) print(the_u8_tensor) tf.Tensor([2 3 4], shape=(3,), dtype=uint8) Telefoonka Shuruudaha waa mid ka mid ah kharashka ah oo ku saabsan Shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan oo ka mid ah shuruudaha ugu badan. Qalabka ugu fiican ee NumPy Waxaa laga yaabaa in ay ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah. x = tf.constant([1, 2, 3]) y = tf.constant(2) z = tf.constant([2, 2, 2]) # All of these are the same computation print(tf.multiply(x, 2)) print(x * y) print(x * z) tf.Tensor([2 4 6], shape=(3,), dtype=int32) tf.Tensor([2 4 6], shape=(3,), dtype=int32) tf.Tensor([2 4 6], shape=(3,), dtype=int32) Sidaas, xayawaanka ka badan 1 waxaa laga yaabaa in la xayawaan in la xayawaan xayawaanka kale. Dhamaan xayawaanka waxaa laga yaabaa in la xayawaan in la xayawaan. 3x1 matrix waa matrix 1x4 si ay u soo saarka matrix 3x4. . [4] # These are the same computations x = tf.reshape(x,[3,1]) y = tf.range(1, 5) print(x, "\n") print(y, "\n") print(tf.multiply(x, y)) tf.Tensor( [[1] [2] [3]], shape=(3, 1), dtype=int32) tf.Tensor([1 2 3 4], shape=(4,), dtype=int32) tf.Tensor( [[ 1 2 3 4] [ 2 4 6 8] [ 3 6 9 12]], shape=(3, 4), dtype=int32) A broadcasted add: a times a gives a [3, 1] [1, 4] [3,4] Sida loo yaabaa, waxaa laga yaabaa in la soo bandhigay: x_stretch = tf.constant([[1, 1, 1, 1], [2, 2, 2, 2], [3, 3, 3, 3]]) y_stretch = tf.constant([[1, 2, 3, 4], [1, 2, 3, 4], [1, 2, 3, 4]]) print(x_stretch * y_stretch) # Again, operator overloading tf.Tensor( [[ 1 2 3 4] [ 2 4 6 8] [ 3 6 9 12]], shape=(3, 4), dtype=int32) Waayo, broadcasting waa mid ka mid ah waqti iyo dhismaha ah, sida wax soo saarka broadcast waxay ka mid ah wax soo saarka tensors in memory. Sida loo isticmaali karaa waxaa loo isticmaali karaa . tf.broadcast_to print(tf.broadcast_to(tf.constant([1, 2, 3]), [3, 3])) tf.Tensor( [[1 2 3] [1 2 3] [1 2 3]], shape=(3, 3), dtype=int32) Markaas ka mid ah macluumaadka, Ma rabtaa in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan tahay in aad u baahan yahay. broadcast_to Waxaa laga yaabaa in ay ku yaalaa in ay ka mid ah. Jake VanderPlas oo ku saabsan waxaa lagu soo bandhigay in ka badan (wax ka mid ah NumPy). Qalabka Qalabka Data Science ee Python Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka ugu badan, sida iyo Qalabka Qalabka Qalabka Sidaas, waxaad ka mid ahay in ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah. tf.matmul tf.reshape tf.Tensor Inta badan, laakiin sidoo kale oo dhan, ops call Qalabka dhismaha iyo dhismaha dhismaha iyo dhismaha dhismaha iyo dhismaha Haku Qalabka Python iyo Waxaad ka heli karaa automatically. convert_to_tensor ndarray TensorShape tf.Variable Haku wax soo saarka, iyo haddii aad u soo saarka ah oo aad u baahan tahay si ay u isticmaali karaa tensor. tf.register_tensor_conversion_function Qalabka Tensors Tensor oo ka mid ah oo ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah. Qalabka dhismaha. tf.ragged.RaggedTensor Sida loo yaabaa, waxaa loo yaqaan "Tensor Regular" (Tensor Regular): A , shape: tf.RaggedTensor [4, None] ragged_list = [ [0, 1, 2, 3], [4, 5], [6, 7, 8], [9]] try: tensor = tf.constant(ragged_list) except Exception as e: print(f"{type(e).__name__}: {e}") ValueError: Can't convert non-rectangular Python sequence to Tensor. Sida loo helo a Haku : tf.RaggedTensor tf.ragged.constant ragged_tensor = tf.ragged.constant(ragged_list) print(ragged_tensor) Qalabka A Waxaad ka mid ah ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ah mid ka mid ah mid ka mid ah: tf.RaggedTensor print(ragged_tensor.shape) (4, None) Qalabka String Waxaa la , oo waa in aad u isticmaali karaa data sida string (variable-length byte array) in tensors. tf.string dtype Shuruudaha waa atomic oo ay ka horumariyaan sida shuruudaha Python waa. Length of the string is not one of the axes of the tensor. Haku waa in la isticmaali karaa. tf.strings Sida loo yaqaan Tensor String Scalar: # Tensors can be strings, too here is a scalar string. scalar_string_tensor = tf.constant("Gray wolf") print(scalar_string_tensor) tf.Tensor(b'Gray wolf', shape=(), dtype=string) Sida loo isticmaali karaa vektor of strings: A vector of strings, shape: [3,] # If you have three string tensors of different lengths, this is OK. tensor_of_strings = tf.constant(["Gray wolf", "Quick brown fox", "Lazy dog"]) # Note that the shape is (3,). The string length is not included. print(tensor_of_strings) tf.Tensor([b'Gray wolf' b'Quick brown fox' b'Lazy dog'], shape=(3,), dtype=string) Sida loo yaqaan 'Printing the above' Prefix waxaa loo yaqaan dtype waa mid ka mid ah dunida Unicode, laakiin mid ka mid ah byte-string. Sida loo isticmaali karaa Unicode in TensorFlow. b tf.string Qalabka Unicode Markaas oo ka mid ah macluumaadka Unicode, waxaa loo isticmaali karaa codka Utf-8. tf.constant("🥳👍") Qalabka dhismaha ugu fiican ee loo isticmaali karaa in Waxaa ku saabsan . tf.strings tf.strings.split # You can use split to split a string into a set of tensors print(tf.strings.split(scalar_string_tensor, sep=" ")) tf.Tensor([b'Gray' b'wolf'], shape=(2,), dtype=string) # ...but it turns into a `RaggedTensor` if you split up a tensor of strings, # as each string might be split into a different number of parts. print(tf.strings.split(tensor_of_strings)) Three strings split, shape: [3, None] iyo : tf.strings.to_number text = tf.constant("1 10 100") print(tf.strings.to_number(tf.strings.split(text, " "))) tf.Tensor([ 1. 10. 100.], shape=(3,), dtype=float32) Marka aad u isticmaali karaa Si loo isticmaali karaa in ay u isticmaali karaa macluumaadka macluumaadka, waxaa loo isticmaali karaa in macluumaadka macluumaadka macluumaadka. tf.cast byte_strings = tf.strings.bytes_split(tf.constant("Duck")) byte_ints = tf.io.decode_raw(tf.constant("Duck"), tf.uint8) print("Byte strings:", byte_strings) print("Bytes:", byte_ints) Byte strings: tf.Tensor([b'D' b'u' b'c' b'k'], shape=(4,), dtype=string) Bytes: tf.Tensor([ 68 117 99 107], shape=(4,), dtype=uint8) # Or split it up as unicode and then decode it unicode_bytes = tf.constant("アヒル 🦆") unicode_char_bytes = tf.strings.unicode_split(unicode_bytes, "UTF-8") unicode_values = tf.strings.unicode_decode(unicode_bytes, "UTF-8") print("\nUnicode bytes:", unicode_bytes) print("\nUnicode chars:", unicode_char_bytes) print("\nUnicode values:", unicode_values) Unicode bytes: tf.Tensor(b'\xe3\x82\xa2\xe3\x83\x92\xe3\x83\xab \xf0\x9f\xa6\x86', shape=(), dtype=string) Unicode chars: tf.Tensor([b'\xe3\x82\xa2' b'\xe3\x83\x92' b'\xe3\x83\xab' b' ' b'\xf0\x9f\xa6\x86'], shape=(5,), dtype=string) Unicode values: tf.Tensor([ 12450 12498 12523 32 129414], shape=(5,), dtype=int32) EE.UU dtype waxaa loo isticmaali karaa waayo-aragnimo dhererka dhererka ah ee TensorFlow. module waxay ku yaalaa funktions for converting data to and from bytes, oo ka mid ah decoding images iyo parsing csv. tf.string tf.io Qalabka Tension Sida loo yaabaa, waxaa loo isticmaali karaa in ay ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah mid ka mid ah. Shuruudaha dhismaha waaweyn ee shuruudaha dhismaha ah. tf.sparse.SparseTensor A , shape: tf.SparseTensor [3, 4] # Sparse tensors store values by index in a memory-efficient manner sparse_tensor = tf.sparse.SparseTensor(indices=[[0, 0], [1, 2]], values=[1, 2], dense_shape=[3, 4]) print(sparse_tensor, "\n") # You can convert sparse tensors to dense print(tf.sparse.to_dense(sparse_tensor)) SparseTensor(indices=tf.Tensor( [[0 0] [1 2]], shape=(2, 2), dtype=int64), values=tf.Tensor([1 2], shape=(2,), dtype=int32), dense_shape=tf.Tensor([3 4], shape=(2,), dtype=int64)) tf.Tensor( [[1 0 0 0] [0 0 2 0] [0 0 0 0]], shape=(3, 4), dtype=int32) Markaas oo ku yaalaa TensorFlow website, this article appears here under a new headline and is licensed under CC BY 4.0. code samples shared under the Apache 2.0 License. this article appears here under a new title and is licensed under Qalabka Qalabka Qalabka Qalabka Qalabka Qalabka Originally published on the , TensorFlow website CC BY 4.0. Apache 2.0 License. TensorFlow ee website