RLCard 1.0.7

Support CPU training of DMC

RLCard 1.0.6

• Small fix on testing.
• Support PettingZoo.

RLCard 1.0.4

• Fixed bugs in num_players
• Improved efficiency of Mahjong
• Fixed step_back bug in no-limit Holdem

RLCard 1.0.2

• Support 8 environments.
• Support 4 algorithms with Torch training.
• Support Deep Monte-Carlo (DMC) described in DouZero.
• Clean and light-weighted codebase.