Revert number of epochs

convlab/policy/ppo/semanticGenTUS-RuleDST-PPOPolicy.json

@@ -5,7 +5,7 @@
 		"use_pretrained_initialisation": false,
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 500,
+		"epoch": 50,
 		"eval_frequency": 5,
 		"process_num": 1,
 		"num_eval_dialogues": 500,

convlab/policy/ppo/semantic_level_config.json

@@ -5,7 +5,7 @@
 		"pretrained_load_path": "",
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 100,
+		"epoch": 10,
 		"eval_frequency": 5,
 		"process_num": 4,
 		"sys_semantic_to_usr": false,

convlab/policy/ppo/setsumbt_config.json

@@ -5,7 +5,7 @@
 		"use_pretrained_initialisation": false,
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 500,
+		"epoch": 50,
 		"eval_frequency": 5,
 		"process_num": 2,
 		"num_eval_dialogues": 500,

convlab/policy/ppo/setsumbt_unc_config.json

@@ -5,7 +5,7 @@
 		"use_pretrained_initialisation": false,
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 500,
+		"epoch": 50,
 		"eval_frequency": 5,
 		"process_num": 2,
 		"num_eval_dialogues": 500,

convlab/policy/ppo/trippy_config.json

@@ -5,7 +5,7 @@
 		"use_pretrained_initialisation": false,
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 500,
+		"epoch": 50,
 		"eval_frequency": 5,
 		"process_num": 2,
 		"num_eval_dialogues": 500,

convlab/policy/ppo/tus_semantic_level_config.json

@@ -5,7 +5,7 @@
 		"pretrained_load_path": "",
 		"num_train_dialogues": 100,
 		"seed": 0,
-		"epoch": 500,
+		"epoch": 50,
 		"eval_frequency": 5,
 		"process_num": 4,
 		"sys_semantic_to_usr": false,