my sequential NN model loss is decreasing well but accuracy of model is constant from the first step

Question

for solving this problem i check my_data or metric accuracy but nothing change

even i try precision and recall metric but nothing change :

from sklearn.model_selection import train_test_split
#X, y = make_classification(n_samples=100, random_state=1)
x_train, x_test, y_train, y_test = train_test_split(df, lables, stratify=lables,
                                                    random_state=42,)
...
model = keras.models.Sequential([
   # keras.layers.Flatten(input_shape=[28, 28]),
    keras.layers.Dense(10, activation="relu"),
    keras.layers.Dense(6, activation="relu"),
    keras.layers.Dense(6, activation="relu"),
    keras.layers.Dense(1, activation="softmax")
])
model.compile(loss="mse",
              optimizer="sgd",
             metrics=[tf.keras.metrics.Recall()],
              
               )