lychee's stats

# the actual value the program is looking for (we pretend that the guess function is unaware of this variable)
TRUE_VALUE=$(($(dd if=/dev/urandom count=1 2> /dev/null | cksum | cut -d' ' -f1) % 100))



# $1 should be the guess
# using echo as return is for little skibidi toilets, hence im using error codes
# (this function is implemented for testing purposes)
submit() {
    echo "guess: $1"
    # Too big
    if (( $1 < TRUE_VALUE )); then
        return -1 # overflows to 255

    elif (( $1 > TRUE_VALUE )); then
        return 1

    elif [[ $1 -eq $TRUE_VALUE ]]; then
        return 0

    else
        while true; do
            printf "crazy? i was crazy once. they put me in a room. a rubber room. a rubber room with rats. the rats made me ..."
        done
    fi

}

guess() {
  # I was initially planning to do a perfectly suboptimal search (binary search, but guessing all values in the wrong direction before proceeding), but then it occurred to me... i couldn't be bothered
  for i in {0..100}; do

    submit $i

    if [[ $? -eq "0" ]]; then
        echo "welp that was fun, im off"

        return 0
    fi
  done
}

guess

import keras
import math
import numpy as np

# create a dataset
data = np.arange(0.001, 100, 0.00025)

np.random.shuffle(data)

# take a sample for testing purposes
split_data = np.array_split(data, 4)

x_train = np.concatenate((split_data[0], split_data[1], split_data[2]))
x_test = split_data[3]

split_data = None # wipe my temp variable

y_train = np.reciprocal(np.sqrt(x_train))
y_test = np.reciprocal(np.sqrt(x_test))

# set up a model / topology or whatever
model = keras.models.Sequential([
  keras.layers.Input(shape=(1,)),
  keras.layers.Dense(32, activation='relu'),
  keras.layers.Dense(16, activation='relu'),
  keras.layers.Dense(16, activation='relu'),
  keras.layers.Dense(1, activation='exponential')
])


# define a loss function
loss_fn = keras.losses.MeanSquaredLogarithmicError()

model.dropout = keras.layers.Dropout(0.0)

# compile the model, with the loss function
model.compile(optimizer='adam',
              loss=loss_fn,
              metrics=['mean_absolute_percentage_error'],
              auto_scale_loss=True)

# adjust the model to minimise loss
model.fit(x_train, y_train, validation_data=(x_test, y_test), epochs=3)


# evaluate the model's performance
# for the purposes of this, I think 1 colour channel is fine
model.evaluate(x_test,  y_test, verbose=2)

while True: # bad bad bad error management, but good enough. Don't want it to hard crash after spending ages training.
  try:
    value = float(input("Enter test value: ").strip())
    prediction = model.predict(np.reshape(value, (1)))
    print("1 / sqrt({0}) ≈ {1}".format(value, prediction))
  except KeyboardInterrupt:
    break # (effectively just exit, but without importing sys)
  except:
    print("Invalid value! Please retry")