Ws,bs = Q.get_weights()
Q_target.assign(sess, Ws,bs)
ann_fric = (1-EPSILON)/ANNEALING
EXP_PROB = 1
# initialize mdp state structure
mdp = MDP_state(STATE_SIZE_POST, FRAMES)
# initialize replay buffer
R = ReplayBuffer(MDP_STATE_SIZE, 1, BUFFER_SIZE)
buf = R.LoadBuffer(OUT_DIR+BUFFER_FILE)
if buf:
OBSERVATION_PHASE = 0
EXP_PROB = EPSILON
populated = R.GetOccupency()
print("Replay buffer loaded from disk, occupied: " + str(populated))
else:
print("Creating new replay buffer")
# initialize logger
L = Logger()
log_not_empty = L.Load(OUT_DIR+LOG_FILE)
if log_not_empty:
print ("Log file loaded")
else:
("Creating new log file")
L.AddNewLog('network_left')
L.AddNewLog('network_middle')
L.AddNewLog('network_right')
L.AddNewLog('policy_left')
sess.run(init)
Ws, bs = Q.get_weights()
Q_target.assign(sess, Ws, bs)
ann_fric = (1 - EPSILON) / ANNEALING
EXP_PROB = 1
# initialize mdp state structure
mdp = MDP_state(STATE_SIZE_POST, FRAMES)
# initialize replay buffer
R = ReplayBuffer(MDP_STATE_SIZE, 1, BUFFER_SIZE)
buf = R.LoadBuffer(OUT_DIR + BUFFER_FILE)
if buf:
EXP_PROB = EPSILON
populated = R.GetOccupency()
print("Replay buffer loaded from disk, occupied: " + str(populated))
else:
print("Creating new replay buffer")
R_val = ReplayBuffer(MDP_STATE_SIZE, 1, BUFFER_SIZE)
buf = R_val.LoadBuffer(VALIDATION_FILE)
if buf:
populated = R_val.GetOccupency()
print("Validation buffer loaded from disk, size: " + str(populated))
else:
print("No validation buffer found!")
# initialize logger
L = Logger()
log_not_empty = L.Load(OUT_DIR + LOG_FILE)
if log_not_empty:
