[WIP] Support bet with any amount. Set min raise to double of previous raise #265
Conversation
| '--cuda', | ||
| type=str, | ||
| default='', | ||
| default='0', |
There was a problem hiding this comment.
Could you make '' as default since some users may not have GPU
| # random_agent = RandomAgent(num_actions=env.num_actions) | ||
|
|
||
| env.set_agents([human_agent, human_agent2]) | ||
| env.set_agents([human_agent, human_agent2, human_agent3, human_agent4]) |
There was a problem hiding this comment.
We you make it 4 players instead of 2?
| for player_id in range(self.num_players): | ||
| state = self.get_state(player_id) | ||
| trajectories[player_id].append(state) | ||
|
|
There was a problem hiding this comment.
I believe we don't want to print trajectories in this method, which will affect all the games.
There was a problem hiding this comment.
Sure. Shld Set this to WIP! WIll resolve the status once the feature is done
| return self.round.get_nolimit_legal_actions(players=self.players) | ||
|
|
||
| def step(self, action): | ||
| def step(self, action_tp): |
There was a problem hiding this comment.
What is the motivation of using action_tp? What would the legal actions be with action_tp?
There was a problem hiding this comment.
action_tp (or action_tuple) [Action, amt: int]
legal actions will only contain FOLD, RAISE, CALL,
In the case of raise, action_tp[1] indicates the amt to raise; In other cases, action_tp[1] is not used
|
@wgzesg You mentioned, "WIll resolve the status once the feature is done". Just check if this feature is done. Sorry, I lose track of this PR> |
Support bet with any amount. Set min raise to double of previous raise