Type | Field |
---|---|
float | Learning rate |
float | Gamma |
int | Number of actions |
float | Epsilon |
int | batch size |
int | input dimensions |
float | epsilon end |
int | memory size |
string | model save name |
Agent(gamma, epsilon, lr, input_dims, n_actions, mem_size, batch_size, epsilon_end)
Type | Field |
---|---|
environment variables (depends on what you are trading the agent on) | observation |
Agent.choose_action(observation)
Type | Field |
---|---|
float | state |
float | action |
float | reward |
float | new state |
bool | done |
Agent.store_transition(state, action, reward, new_state, done)
Description |
---|
Runs the data is has collected through the neural network |
Agent.learn()
Description |
---|
saves the model |
Agent.save_model()
Description |
---|
loads the model |
Agent.loads_model()