figure_file = 'plots/fl.png'
best_score = env.reward_range[0]
score_history = []
learn_iters = 0
avg_score = 0
n_steps = 0
for i in range(n_games):
observation = env.reset()
done = False
score = 0
while not done:
action, prob, val = agent.choose_action(observation)
observation_, reward, done, info = env.step(action)
n_steps += 1
score += reward
agent.remember(observation, action, prob, val, reward, done)
if n_steps % N == 0:
agent.learn()
learn_iters += 1
observation = observation_
score_history.append(score)
avg_score = np.mean(score_history[-100:])
if avg_score > best_score:
best_score = avg_score
agent.save_models()
avg_score = 0
n_steps = 0
if load_checkpoint:
agent.load_models()
for i in tqdm(range(n_episodes)):
observation = env.reset()
done = False
score = 0
while not done:
if render:
env.render(mode='human')
action, prob, val = agent.choose_action(observation)
observation_, reward, done, info = env.step(action)
n_steps += 1
score += reward
agent.remember(observation, action, prob, val, reward, done)
if not load_checkpoint:
if n_steps % N == 0:
loss.append(agent.learn())
learn_iters += 1
observation = observation_
if not load_checkpoint:
avg_loss = np.mean(loss, axis=0)
actor_loss.append(avg_loss[0])