when i train the model use your code, something happenyed

[sixleaves]

when i train the model user your code, after 7 minutes, the game speend become very very slowly, i don't whatl's the wrong?

[sixleaves]

the fps will become 3 fps，i don't know what the happend

[Ben Lau]

Do you have GPU card? Honestly I don't see the problem that you mention.

…

On Sun, Jan 7, 2018 at 10:49 PM, sixleaves ***@***.***> wrote: the fps will become 3 fps，i don't know what the happend — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#20 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AO1sY9uFLaXPUPWT_GCgKxtmsB-uwsDuks5tINl8gaJpZM4RVpd3> .

[sixleaves]

i try to train on mac and windows, but it's still appear.on windows i have GPU card, in mac no any one gpu card......。my problem is when i traing, the game’s fps will become only three fps。

[sixleaves]

my env is:
python 3.6.3
anaconda3
pygame
keras 2.0.8

[sixleaves]

there are two warning that pycharm give
1)libpng warning: iCCP: known incorrect sRGB profile
2).UserWarning: The default mode, 'constant', will be changed to 'reflect' in skimage 0.15.

[sixleaves]

I have found when i run into if t > OBSERVE: code 。the game will become slowly。。。

[sixleaves]

i find the cause.beacuse there will spend much time in below code
`
for i in range(0, len(minibatch)):
state_t = minibatch[i][0]
action_t = minibatch[i][1] #This is action index
reward_t = minibatch[i][2]
state_t1 = minibatch[i][3]
terminal = minibatch[i][4]

            inputs[i:i + 1] = state_t    #I saved down s_t

            targets[i] = model.predict(state_t)  # Hitting each buttom probability
            Q_sa = model.predict(state_t1)

            if terminal:
                targets[i, action_t] = reward_t
            else:
                targets[i, action_t] = reward_t + GAMMA * np.max(Q_sa)`

[Ben Lau]

Code updated and it should run very fast now even on CPU.