Method forward

policy_gradient/policy_gradient_basic.py:28–32

(self, inputs)

        self.rewards = []

    def forward(self, inputs):
        x = F.relu(F.dropout(self.fc1(inputs), 0.6))
        x = self.fc2(x)

        return F.softmax(x, -1)

    def select_action(self, inputs):
        x = paddle.to_tensor(inputs).astype('float32').unsqueeze(0)