[Fix] fix clp potential error and support bs>1 (#439)

* [Fix] fix clp potential error and support bs>1

* [Fix] fix clp potential error and support bs>1

* minor fix

* minor fix

opencompass/openicl/icl_inferencer/icl_clp_inferencer.py

@@ -119,7 +119,7 @@ class CLPInferencer(BaseInferencer):
         if self.single_token:
             index = 0
             prompt_list = []
-            choice_target_ids = []
+            target_pos = []
             # TODO: Hard code temperaily, need to modified here
             choices = retriever.test_ds[0]['choices']
             try:
@@ -142,6 +142,13 @@ class CLPInferencer(BaseInferencer):
 
             get_token_len = self.model.get_token_len
 
+            if hasattr(self.model.tokenizer, 'padding_side'):
+                # get padding_side for huggingface model
+                padding_side = self.model.tokenizer.padding_side
+            else:
+                # defaults to left for internal model
+                padding_side = 'left'
+
             # prepare in context for each example and control the length
             for idx in range(len(ice_idx_list)):
                 prompt = retriever.generate_prompt_for_generate_task(
@@ -149,7 +156,7 @@ class CLPInferencer(BaseInferencer):
                     ice[idx],
                     ice_template=ice_template,
                     prompt_template=prompt_template)
-                prompt = self.model.parse_template(prompt, mode='ppl')
+                prompt = self.model.parse_template(prompt, mode='gen')
                 if self.max_seq_len is not None:
                     prompt_token_num = get_token_len(prompt)
                     # add one because additional token will be added in the end
@@ -165,15 +172,19 @@ class CLPInferencer(BaseInferencer):
                             ice_template=ice_template,
                             prompt_template=prompt_template)
                         prompt_token_num = get_token_len(prompt)
-                # Add single token for prompt, this token can be any token
-                prompt += 'yes'
                 prompt_list.append(prompt)
-                # in case prompt token num reaches
+                # in case prompt token num reaches max
                 if self.max_seq_len is not None and \
                         prompt_token_num + 1 > self.max_seq_len:
                     prompt_token_num = self.max_seq_len - 1
-                # minus the bos token
+
-                choice_target_ids.append(prompt_token_num - 1)
+                # get the target position index
+                if padding_side == 'left':
+                    # always the last position
+                    target_pos.append(-1)
+                else:
+                    # the last position of the original prompt
+                    target_pos.append(prompt_token_num - 1)
 
             # 4.1 Fetch and zip prompt & gold answer if output column exists
             ds_reader = retriever.dataset_reader
@@ -182,19 +193,36 @@ class CLPInferencer(BaseInferencer):
             else:
                 gold_ans = [None] * len(prompt_list)
 
+            if hasattr(self.model, 'batch_padding'):
+                # get batch padding for huggingface model
+                batch_padding = self.model.batch_padding
+            else:
+                # defaults to False for internal model
+                batch_padding = False
+
             logger.info('Calculating conditional log probability for prompts.')
             for idx in trange(0,
                               len(prompt_list),
                               self.batch_size,
                               disable=not self.is_main_process):
+                # get batch data
                 sub_prompt_list = prompt_list[idx:idx + self.batch_size]
                 sub_golds = gold_ans[idx:idx + self.batch_size]
-                sub_choice_target_ids = choice_target_ids[idx:idx +
+                sub_target_pos = target_pos[idx:idx + self.batch_size]
-                                                          self.batch_size]
-                sub_res = self.__get_cond_prob(sub_prompt_list,
-                                               sub_choice_target_ids,
-                                               choice_ids)
 
+                # get probability result
+                if batch_padding and self.batch_size > 1:
+                    sub_res = self._get_cond_prob(sub_prompt_list,
+                                                  sub_target_pos, choice_ids)
+                else:
+                    sub_res = []
+                    for prompt, position in zip(sub_prompt_list,
+                                                sub_target_pos):
+                        sub_res.extend(
+                            self._get_cond_prob([prompt], [position],
+                                                choice_ids))
+
+                # save all the result
                 for res, prompt, gold in zip(sub_res, sub_prompt_list,
                                              sub_golds):
                     example_input = prompt.replace(ice[idx], '')
@@ -217,22 +245,29 @@ class CLPInferencer(BaseInferencer):
             for sample in output_handler.results_dict.values()
         ]
 
-    def __get_cond_prob(self,
+    def _get_cond_prob(self, input_texts: List[str], target_pos: List[int],
-                        input_texts: List[str],
+                       choice_ids: List[int]):
-                        sub_choice_target_ids,
+        """Get the condition probability of next token.
-                        choice_ids,
+
-                        mask_length=None):
+        Args:
-        # TODO: support multiple tokens
+            input_texts (List[str]): All the input prompt to be tested.
+            target_pos (List[int]): Target position of next token.
+            choice_ids (List[int]): Choice ids of target tokens.
+        """
         if hasattr(self.model, 'generator'):
-            outputs, _ = self.model.generator.get_logits(input_texts)
+            get_logits = self.model.generator.get_logits
         else:
-            outputs, _ = self.model.get_logits(input_texts)
+            get_logits = self.model.get_logits
 
-        shift_logits = outputs[..., :-1, :].contiguous().float()
+        outputs, _ = get_logits(input_texts)
 
-        shift_logits = F.log_softmax(shift_logits, dim=-1)
+        # we want get the next token probability
+        # therefore no shift here
+        logits = outputs.contiguous().float()
+
+        logits = F.log_softmax(logits, dim=-1)
         log_probs = []
-        for logits, target_ids in zip(shift_logits, sub_choice_target_ids):
+        for logit, target_ids in zip(logits, target_pos):
             log_probs.append(
-                F.softmax(logits[target_ids, choice_ids], dim=-1).tolist())
+                F.softmax(logit[target_ids, choice_ids], dim=-1).tolist())
         return log_probs