def plot(episode, rewards, value_losses, policy_losses, noise):
clear_output(True)
rewards_x, rewards_smooth = smooth_plot(10, rewards, 500)
value_losses_x, value_losses_smooth = smooth_plot(10, value_losses, 10000)
policy_losses_x, policy_losses_smooth = smooth_plot(
10, policy_losses, 10000)
noise_x, noise_smooth = smooth_plot(10, noise, 100)
plt.figure(figsize=(18, 12))
plt.subplot(411)
plt.title('episode %s. reward: %s' % (episode, rewards_smooth[-1]))
plt.plot(rewards, label="Rewards", color='lightsteelblue', linewidth='1')
plt.plot(rewards_x,
rewards_smooth,
label='Smothed_Rewards',
color='darkorange',
linewidth='3')
plt.legend(loc='best')
plt.subplot(412)
plt.title('Value_Losses')
plt.plot(value_losses,
label="Value_Losses",
color='lightsteelblue',
linewidth='1')
plt.plot(value_losses_x,
value_losses_smooth,
label="Smoothed_Value_Losses",
color='darkorange',
linewidth='3')
plt.legend(loc='best')
plt.subplot(413)
plt.title('Policy_Losses')
plt.plot(policy_losses,
label="Policy_Losses",
color='lightsteelblue',
linewidth='1')
plt.plot(policy_losses_x,
policy_losses_smooth,
label="Smoothed_Policy_Losses",
color='darkorange',
linewidth='3')
plt.legend(loc='best')
plt.subplot(414)
plt.title('Noise')
plt.plot(noise, label="Noise", color='lightsteelblue', linewidth='1')
plt.plot(noise_x,
noise_smooth,
label="Smoothed_Noise",
color='darkorange',
linewidth='3')
plt.legend(loc='best')
plt.show()
def listS3Objects(bucket_name):
s3 = boto3.resource('s3')
bucket = s3.Bucket(bucket_name)
s3_files = []
i = 0
for f in bucket.objects.all():
s3_files.append(f)
i += 1
if i % 100 == 0:
print(f)
print('object #:', i)
clear_output(wait=True)
return s3_files
def plot_durations():
plt.figure(2)
plt.clf()
durations_t = torch.tensor(episode_durations, dtype=torch.float)
plt.title('Training')
plt.xlabel('Episode')
plt.ylabel('Duration')
plt.plot(durations_t.numpy())
if len(durations_t) >=100:
means = duratoins_t.unfold(0,100,1).mean(1).view(-1)
means = torch.cat((torch.zeros(99), means))
plt.plot(means.numpy())
plt.pause(0.001)
if is_ipython:
display.clear_output(wait=True)
display.display(plt.gcf())
def plot_durations():
plt.figure(2)
plt.clf()
durations_t = torch.tensor(episode_durations, dtype=torch.float)
plt.title('Training...')
plt.xlabel('Episode')
plt.ylabel('Duration')
plt.plot(durations_t.numpy())
# Take 100 episode averages and plot them too
if len(durations_t) >= 100:
means = durations_t.unfold(0, 100, 1).mean(1).view(-1)
means = torch.cat((torch.zeros(99), means))
plt.plot(means.numpy())
plt.pause(0.001) # pause a bit so that plots are updated
if is_ipython:
display.clear_output(wait=True)
display.display(plt.gcf())