update requirements.txt & setup.py, move tests/ to agent_examples/, move deploy to convlab/deploy

tests/results/BERTNLU-RuleDST-LaRL/res.txt

-complete number of dialogs/tot: 0.406
-success number of dialogs/tot: 0.34
-average precision: 0.4775414345324553
-average recall: 0.5413267072145577
-average f1: 0.4764173612317243
-average book rate: 0.4560897435897436
-average turn (succ): 15.0
-average turn (all): 28.6

tests/results/BERTNLU-RuleDST-MDRG/res.txt

-complete number of dialogs/tot: 0.27
-success number of dialogs/tot: 0.252
-average precision: 0.46578760260101465
-average recall: 0.4306643425802305
-average f1: 0.4197515091550275
-average book rate: 0.4897435897435897
-average turn (succ): 13.634920634920634
-average turn (all): 33.58

tests/results/BERTNLU-RuleDST-MLEPolicy-TemplateNLG/res.txt

-complete number of dialogs/tot: 0.526
-success number of dialogs/tot: 0.484
-average precision: 0.6628294459925087
-average recall: 0.7267286423049664
-average f1: 0.6602935206785587
-average book rate: 0.3551923076923077
-average turn (succ): 12.454545454545455
-average turn (all): 26.27

tests/results/BERTNLU-RuleDST-PGPolicy-TemplateNLG/res.txt

-complete number of dialogs/tot: 0.429
-success number of dialogs/tot: 0.433
-average precision: 0.6187398586642582
-average recall: 0.6681233414099458
-average f1: 0.6039577413844464
-average book rate: 0.30991452991452995
-average turn (succ): 14.69284064665127
-average turn (all): 29.068