def main(args):
# Initialize the environment
env = magent.GridWorld('battle', map_size=args.map_size)
env.set_render_dir(
os.path.join(BASE_DIR, 'examples/battle_model', 'build/render'))
handles = env.get_handles()
tf_config = tf.ConfigProto(allow_soft_placement=True,
log_device_placement=False)
tf_config.gpu_options.allow_growth = True
log_dir = os.path.join(BASE_DIR, 'data/tmp'.format(args.algo))
model_dir = os.path.join(BASE_DIR, 'data/models/{}'.format(args.algo))
start_from = 0
sess = tf.Session(config=tf_config)
models = [
spawn_ai(args.algo, sess, env, handles[0], args.algo + '-me',
args.max_steps),
spawn_ai(args.algo, sess, env, handles[1], args.algo + '-opponent',
args.max_steps)
]
sess.run(tf.global_variables_initializer())
runner = tools.Runner(sess,
env,
handles,
args.map_size,
args.max_steps,
models,
play,
render_every=args.save_every if args.render else 0,
save_every=args.save_every,
tau=0.01,
log_name=args.algo,
log_dir=log_dir,
model_dir=model_dir,
train=True)
for k in range(start_from, start_from + args.n_round):
eps = linear_decay(k, [0, int(args.n_round * 0.8), args.n_round],
[1, 0.2, 0.1])
runner.run(eps, k)
'data/models/{}-1'.format(args.oppo))
sess = tf.Session(config=tf_config)
models = [
spawn_ai(args.algo, sess, env, handles[0], args.algo + '-me',
args.max_steps),
spawn_ai(args.oppo, sess, env, handles[1], args.oppo + '-opponent',
args.max_steps)
]
sess.run(tf.global_variables_initializer())
models[0].load(main_model_dir, step=args.idx[0])
models[1].load(oppo_model_dir, step=args.idx[1])
runner = tools.Runner(sess,
env,
handles,
args.map_size,
args.max_steps,
models,
battle,
render_every=0)
win_cnt = {'main': 0, 'opponent': 0}
for k in range(0, args.n_round):
runner.run(0.0, k, win_cnt=win_cnt)
print('\n[*] >>> WIN_RATE: [{0}] {1} / [{2}] {3}'.format(
args.algo, win_cnt['main'] / args.n_round, args.oppo,
win_cnt['opponent'] / args.n_round))
args.max_steps),
spawn_ai(args.algo, sess, env, handles[2], args.algo + '-opponent2',
args.max_steps),
spawn_ai(args.algo, sess, env, handles[3], args.algo + '-opponent3',
args.max_steps)
]
sess.run(tf.global_variables_initializer())
if args.algo == 'mtmfq':
runner = tools.Runner(
sess,
env,
handles,
args.map_size,
args.max_steps,
models,
play2,
render_every=args.save_every if args.render else 0,
save_every=args.save_every,
tau=0.01,
log_name=args.algo,
log_dir=log_dir,
model_dir=model_dir,
train=True)
else:
runner = tools.Runner(
sess,
env,
handles,
args.map_size,
args.max_steps,
MsgModels[1].load(oppo_msg_dir, step=args.idx)
if args.crp != 'None':
crp = True
else:
crp = False
runner = tools.Runner(sess,
env,
handles,
args.map_size,
args.max_steps,
models,
MsgModels,
play,
render_every=args.save_every if args.render else 0,
save_every=args.save_every,
tau=0.01,
log_name=args.algo,
log_dir=log_dir,
model_dir=model_dir,
train=True,
len_nei=args.len_nei,
rewardtype=args.rewardtype,
crp=crp,
is_selfplay=True,
is_fix=False)
for k in range(start_from, start_from + args.n_round):
eps = magent.utility.piecewise_decay(k, [0, 700, 1400, 8000],
[1, 0.3, 0.05, 0.01])
runner.run(eps, k)