I want to learn only 1 and 2 data from MNIST data of handwritten numbers.

 from keras.models import Sequential
from keras.layers import Activation, Dense
from keras.layers import Conv2D, MaxPooling2D, Flatten
import time

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import numpy as np

from keras.datasets import mnist

(x_train, t_train), (x_test, t_test) = mnist.load_data()

# Take out only handwritten images of 1 and 2
x_train1, x_train2 = x_train [np.where(t_train==1)], x_train [np.where(t_train==2)]
t_train1, t_train2 = t_train [np.where(t_train==1)], t_train [np.where(t_train==2)]

Convert an array of #1 and 2 handwritten numbers into a list
x_trainone=[]
x_traintwo=[ ]

for n1 in range (x_train1.shape[0]):
    x_trainone.append(x_train1[n1])

for n2 in range (x_train2.shape[0]):
    x_trainwo.append(x_train2[n2])
    
t_trainone=[ ]
t_trainwo=[ ]

for n3 in range(t_train1.shape[0]):
    t_trainone.append(t_train1[n3])

for n4 in range(t_train2.shape[0]):
    t_trainwo.append(t_train2[n4])
    
add a list of #1 and a list of 2
x_trainall2 = x_trainone + x_trainwo
t_trainall2 = t_trainone + t_trainwo

Return the list of learning data mixed with #1 and 2 to the array

x_trainall, t_trainall = np.array(x_trainall2), np.array(t_trainall2)

#Do the same for test data
x_test1, x_test2 = x_test [np.where(t_test==1)], x_test [np.where(t_test==2)]
t_test1, t_test2 = t_test [np.where(t_test==1)], t_test [np.where(t_test==2)]

Convert an array of #1 and 2 handwritten numbers into a list
x_testone=[ ]
x_testtwo=[ ]

for n1 in range (x_test1.shape[0]):
    x_testone.append(x_test1[n1])

for n2 in range (x_test2.shape[0]):
    x_testtwo.append(x_test2[n2])
    
t_testone=[ ]
t_testtwo=[ ]

for n3 in range (t_test1.shape[0]):
    t_testone.append(t_test1[n3])

for n4 in range (t_test2.shape[0]):
    t_testtwo.append(t_test2[n4])
    
add a list of #1 and a list of 2
x_testall2 = x_testone + x_testtwo
t_testall2 = t_testone + t_testtwo

Return the list of learning data mixed with #1 and 2 to the array

x_testall, t_testall = np.array(x_testall2), np.array(t_testall2)

model=keras.sequential()
model.add(Conv2D(28,(3,3), strides=1, activation='sigmoid', input_shape=(28,28,1)))) 
model.add(Activation("relu"))
model.add(Flatten())
model.add (Dense(28)) 
Activation ("sigmoid")
model.add (Dense(28)) 
Activation ("relu")
model.add(Dense(2))
Activation ("softmax")
model.compile(
    optimizer="adam",
    loss=tf.keras.losses.SparseCategoryCrossentropy(from_logits=False),
    metrics=["accuracy"])
model.summary()

model.fit(history=model.fit(x_trainall, t_trainall, epochs=10, batch_size=100))

Epoch 1/10
2023-03-1101:34:38:Wtensorflow/core/framework/op_kernel] OP_REQUIRES failed at sparse_xent_op: INVALID_ARGUMENT: Received a label value of 2 which is outside the valid range of [0, 2]. Label values: 1 2 2 1 1 1 2 1 2 2 2 2 2 1 2 2 1 2 2 1 2 2 1 2 1 2 2 2 2 1 2 2 1 2 2 2 1 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
---------------------------------------------------------------------------
InvalidArgumentError Traceback (most recent call last)
Cell In [15], line 1
---->1model.fit(history=model.fit(x_trainall, t_trainall, epochs=10, 
      2 validation_data=(x_testall, t_testall)))

File...<locals>.error_handler(*args,**kwargs)
     67 filtered_tb =_process_traceback_frames(e.__traceback__)
     68# To get the full stack trace, call:
     69#`tf.debugging.disable_traceback_filtering()`
--- >70 raise e.with_traceback (filtered_tb) from None
     71finally:
     72delfiltered_tb

File....in quick_execute(op_name, num_outputs, inputs, attrs, ctx, name)
     50try:
     51ctx.ensure_initialized()
--- >52 sensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,
     53 inputs, attrs, num_outputs)
     54 except core._NotOkStatusException as:
     55 if name is not None:

InvalidArgumentError: Graph execution error: