[ReRanker] Add reranker-delimiter, adapt it slightly to work with SeeKeR

parlai/agents/reranker/reranker.py

@@ -58,6 +58,12 @@ def add_cmdline_args(cls, parser: ParlaiParser, partial_opt: Optional[Opt] = Non
             help='Which strategy to use when re-ranking response candidates. '
             f"Choices: {','.join(RERANKER_STRATEGIES)}",
         )
+        reranker.add_argument(
+            '--reranker-delimiter',
+            type=str,
+            default=None,
+            help='delimiter for the retriever',
+        )
         return parser
 
     def __init__(self, opt: Opt, shared=None):
@@ -69,7 +75,9 @@ def __init__(self, opt: Opt, shared=None):
         )
         self.reranker_strategy = opt['reranker_strategy']
         self.normalize_candidates = opt['normalize_candidates']
-        self.delimiter = opt.get('delimiter', '\n')
+        self.delimiter = opt.get('reranker_delimiter', None)
+        if not self.delimiter:
+            self.delimiter = opt.get('delimiter', '\n')
         self.include_context = True
         self.include_label_cand_only = False
         self.init_predictor(opt, shared)
@@ -459,6 +467,12 @@ def add_cmdline_args(
             default=False,
             help='specify to enable certain debugging procedures.',
         )
+        gen_agent.add_argument(
+            '--inference-opt-key',
+            type=str,
+            default='inference',
+            help='specify inference opt key for dialogue response model',
+        )
 
         return parser
 
@@ -468,8 +482,9 @@ def __init__(self, opt: Opt, shared=None):
         """
         super().__init__(opt, shared)
         reranker_class = self.get_reranker_class()
+        self.inference_opt_key = opt.get('inference_opt_key', 'inference')
         self.inference_strategies = (
-            opt['inference_strategies'] or opt['inference']
+            opt['inference_strategies'] or opt[self.inference_opt_key]
         ).split(',')
         self.debug_mode = opt.get('debug_mode', False)
         if not shared:
@@ -490,6 +505,11 @@ def set_rerank_strategy(self, strategy: str):
         assert strategy in RERANKER_STRATEGIES
         self.reranker.reranker_strategy = strategy
 
+    def get_observations_for_reranker(
+        self, observations: List[Message]
+    ) -> List[Message]:
+        return observations
+
     def share(self):
         """
         Share model parameters.
@@ -510,15 +530,18 @@ def batch_act(self, observations: List[Message]) -> List[Message]:
         batch_reply = [Message() for _ in range(len(observations))]
         # 1. get all beam texts to consider
         for strategy in self.inference_strategies:
-            self.opt['inference'] = strategy
+            self.opt[self.inference_opt_key] = strategy
             inference_batch_reply = super().batch_act(observations)
             for i, resp in enumerate(inference_batch_reply):
                 beam_texts = batch_reply[i].get('beam_texts', [])
                 batch_reply[i] = resp  # add metrics, other response items
                 new_beam_texts = [(*b, strategy) for b in resp.get('beam_texts', [])]
                 batch_reply[i].force_set('beam_texts', beam_texts + new_beam_texts)
         # 2. Rerank
-        for observation, generator_response in zip(observations, batch_reply):
+        observations_for_reranker = self.get_observations_for_reranker(observations)
+        for observation, generator_response in zip(
+            observations_for_reranker, batch_reply
+        ):
             if (
                 'beam_texts' not in generator_response
                 or not generator_response['beam_texts']