def __init__(self, t_prof, chief_handle):
assert t_prof.n_seats == 2
EvaluatorMasterBase.__init__(
self,
t_prof=t_prof,
eval_env_bldr=_util.get_env_builder_lbr(t_prof=t_prof),
chief_handle=chief_handle,
eval_type="LBR",
log_conf_interval=True)
self.lbr_args = t_prof.module_args["lbr"]
self.weights_for_eval_agent = None
self.alive_worker_handles = None
def __init__(self, t_prof, chief_handle, eval_agent_cls):
assert t_prof.n_seats == 2
self.t_prof = t_prof
self.lbr_args = t_prof.module_args["lbr"]
self._eval_env_bldr = _util.get_env_builder_lbr(t_prof=t_prof)
self.check_to_round = self.lbr_args.lbr_check_to_round
self.chief_handle = chief_handle
self.agent = _AgentWrapper(t_prof=t_prof, lbr_args=self.lbr_args, eval_agent_cls=eval_agent_cls)
# has different raise sizes than agent's env! This needs to be considered when updating the envs after opp acts
self._env = None
self.agent_range = PokerRange(env_bldr=self._eval_env_bldr)
assert self.check_to_round is None or (self.check_to_round in self._eval_env_bldr.rules.ALL_ROUNDS_LIST)
def __init__(self):
t_prof = TrainingProfile(
name="NLH_EXPLOITABILITY_PLO",
nn_type="feedforward",
DISTRIBUTED=False,
CLUSTER=False,
n_learner_actor_workers=2, # 20 workers
max_buffer_size_adv=1e6,
max_buffer_size_avrg=1e6,
export_each_net=False,
checkpoint_freq=8,
eval_agent_export_freq=4, # produces GBs!
# How many actions out of all legal on current step to branch randomly = action bredth limit
n_actions_traverser_samples=
4, # 3 is the default, 4 is current max for b_2
# number of traversals gives some amount of otcomes to train network on
# mult = 1...4, buffer appends every() step with new data
n_traversals_per_iter=30,
# number of mini_batch fetches and model updates on each step
n_batches_adv_training=801, # 1024
n_batches_avrg_training=2048, # 2048
use_pre_layers_adv=True,
n_cards_state_units_adv=192,
n_merge_and_table_layer_units_adv=64,
n_units_final_adv=64,
# amount of batch to feed to NN at once, fetched from buffer randomly.
mini_batch_size_adv=512, # 256
mini_batch_size_avrg=512, # 512
init_adv_model=
"random", # warm start neural weights with init from last iter
init_avrg_model="random",
# use_pre_layers_avrg=False, # shallower nets
lr_avrg=0.001,
game_cls=DiscretizedNLHoldem, # PLO or DiscretizedNLHoldem
env_bldr_cls=VanillaEnvBuilder,
agent_bet_set=bet_sets.PL_2,
n_seats=2,
start_chips=10000,
# You can specify one or both modes. Choosing both is useful to compare them.
eval_modes_of_algo=(
EvalAgentDeepCFR.EVAL_MODE_SINGLE, # SD-CFR
# EvalAgentDeepCFR.EVAL_MODE_AVRG_NET
),
# enables simplified obs. Default works also for 3+ players
use_simplified_headsup_obs=True,
log_verbose=True,
lbr_args=LBRArgs(
lbr_bet_set=bet_sets.PL_2,
n_lbr_hands_per_seat=100,
lbr_check_to_round=Poker.TURN,
# recommended to set to Poker.TURN for 4-round games.
n_parallel_lbr_workers=1,
use_gpu_for_batch_eval=False,
DISTRIBUTED=True,
),
)
self._eval_env_bldr = _util.get_env_builder_lbr(t_prof=t_prof)
stk = [10000, 10000]
self._env = self._eval_env_bldr.get_new_env(is_evaluating=True,
stack_size=stk)
self.t_prof = t_prof