Sequential Decision Making

[erdemath]

I am generating a sequential decision making model. I am using tensorflow.keras.sequantial.
But any time I train my model, I have the following error

Error:
assert_input_compatibility
        str(tuple(shape)))

    ValueError: Input 0 of layer sequential is incompatible with the layer: expected ndim=3, found ndim=2. Full shape received: (None, 30)

Here is the part of the code creating the problem,

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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

import gym import numpy as np import gym_anytrading from gym_anytrading.datasets import FOREX_EURUSD_1H_ASK, STOCKS_GOOGL import tensorflow as tf from tensorflow.keras import layers import pandas as pd import matplotlib.pyplot as plt     class QuantitativeTrading():     def __init__(self, market):         self.market = str(market)         def Environment(self):     # Set up the environment         env = gym.make(self.market)         return env   if __name__ == '__main__':       # Call market     market = 'stocks-v0'       # Call the trader     trader = QuantitativeTrading(market)       # Define the model     env = trader.Environment()     units = env.action_space.n     model = tf.keras.Sequential()     model.add(tf.keras.layers.LSTM(128, input_shape=(2, env.observation_space.shape[0])))     model.add(tf.keras.layers.Dense(units, activation='softmax'))     model.compile(optimizer='adam', loss="mean_squared_error")       # [layers.Dense(2, activation="relu", name="layer1"),     #     layers.Dense(3, activation="relu", name="layer2")]       # Train the model     def ModelTraining(num_epochs, env, model):           for epoch in range(num_epochs):             state = env.reset()             action = model.predict(state.transpose())             next_state, reward, done, _ = env.step(action)             model.fit(state, reward, epochs=num_epochs, verbose=0)             state = next_state           return action, state         num_epochs = 100     action, state = ModelTraining(num_epochs, env, model)

[Larz60+]

Is that all that was included in the error traceback?

[erdemath]

Yes, this part creates the error. If you compile it on your interpreter, you must see the same error.