rework doc_to_visual

lm_eval/api/task.py

@@ -1278,14 +1278,21 @@ class ConfigurableTask(Task):
         else:
             raise TypeError
 
-    def doc_to_visual(self, doc: dict) -> Union[int, str, list]:
+    def doc_to_visual(self, doc: Any) -> Union[int, str, list]:
+        if self.config.doc_to_visual is None:
+            eval_logger.error("doc_to_visual was called but not set in config")
+        else:
+            doc_to_visual = self.config.doc_to_visual
+
         if isinstance(self.config.doc_to_visual, str):
-            assert self.config.doc_to_visual in self.features
-            # Single Image. Still return a list for consistency
-            return doc[self.config.doc_to_visual]
+            if doc_to_visual in self.features:
+                return doc[doc_to_visual]
+            else:
+                return ast.literal_eval(utils.apply_template(doc_to_visual, doc))
+        elif callable(doc_to_visual):
+            return doc_to_visual(doc)
         else:
-            assert callable(self.config.doc_to_visual)
-            return self.config.doc_to_visual(doc)
+            return None
 
     def construct_requests(
         self, doc: dict, ctx: str, **kwargs
@@ -1307,16 +1314,6 @@ class ConfigurableTask(Task):
                 # Otherwise they are placed in the continuation
                 arguments = [(ctx, f"{target_delimiter}{cont}") for cont in choices]
 
-            request_list = [
-                Instance(
-                    request_type="loglikelihood",
-                    doc=doc,
-                    arguments=arg,
-                    idx=i,
-                    **kwargs,
-                )
-                for i, arg in enumerate(arguments)
-            ]
             # TODO: we should raise a warning telling users this will at most ~2x runtime.
             if "acc_mutual_info" in self._metric_fn_list.keys():
                 # if we are calculating multiple choice accuracy
@@ -1325,31 +1322,37 @@ class ConfigurableTask(Task):
                 # here mutual info refers to calculating
                 # log(P(choice|ctx) / P(choice)) = log(P(choice|ctx)) - log(P(choice))
                 # in other words normalizing by subtracting the unconditional logprob of each choice.
-                request_list.extend(
-                    [
-                        Instance(
-                            request_type="loglikelihood",
-                            doc=doc,
-                            arguments=("", "{}".format(choice)),
-                            idx=i,
-                            **kwargs,
-                        )
-                        for i, choice in enumerate(choices)
-                    ]
-                )
-            return request_list
+                aux_arguments = [("", f"{choice}") for choice in choices]
+            else:
+                aux_arguments = None
 
         elif self.OUTPUT_TYPE == "generate_until":
-            if self.INPUT_TYPE == "text_image":
-                arguments = (
-                    ctx,
-                    deepcopy(self.config.generation_kwargs),
-                    self.doc_to_visual,
-                    doc,
-                    self.config.task,
-                )
-            elif self.INPUT_TYPE == "text":
-                arguments = (ctx, deepcopy(self.config.generation_kwargs))
+            arguments = (ctx, deepcopy(self.config.generation_kwargs))
+
+        if self.doc_to_visual:
+            if isinstance(arguments, list):
+                arguments = [arg + (self.doc_to_visual(doc),) for arg in arguments]
+            else:
+                arguments = arguments + (self.doc_to_visual(doc),)
+
+        if isinstance(arguments, type):
+            if aux_arguments is not None:
+                all_arg_list = [arguments, arg_list]
+            else:
+                all_arg_list = [arguments]
+            for arg_list in all_arg_list:
+                request_list = [
+                    Instance(
+                        request_type="loglikelihood",
+                        doc=doc,
+                        arguments=arg,
+                        idx=i,
+                        **kwargs,
+                    )
+                    for i, arg in enumerate(arg_list)
+                ]
+
+            return request_list
 
         return Instance(
             request_type=self.OUTPUT_TYPE,