Removes predict_with_metadata() overrides in Model subclasses.

This change leaves lit_nlp.api.model.Model as the only implementer of predict_with_metadata(), which sets up the removal of this method in the future.

PiperOrigin-RevId: 551569578

lit_nlp/api/model.py

@@ -285,17 +285,6 @@ def predict(
   ) -> Iterable[JsonDict]:
     return self.wrapped.predict(inputs, *args, **kw)
 
-  # NOTE: if a subclass modifies predict(), it should also override this to
-  # call the custom predict() method - otherwise this will delegate to the
-  # wrapped class and call /that class's/ predict() method, likely leading to
-  # incorrect results.
-  # b/171513556 will solve this problem by removing the need for any
-  # *_with_metadata() methods.
-  def predict_with_metadata(
-      self, indexed_inputs: Iterable[JsonDict], **kw
-  ) -> Iterable[JsonDict]:
-    return self.wrapped.predict_with_metadata(indexed_inputs, **kw)
-
   def load(self, path: str):
     """Load a new model and wrap it with this class."""
     new_model = self.wrapped.load(path)

lit_nlp/examples/models/t5.py

@@ -436,10 +436,6 @@ def predict(self, inputs):
     outputs = self.wrapped.predict(model_inputs)
     return (utils.remap_dict(mo, self.FIELD_RENAMES) for mo in outputs)
 
-  def predict_with_metadata(self, indexed_inputs):
-    """As predict(), but inputs are IndexedInput."""
-    return self.predict((ex["data"] for ex in indexed_inputs))
-
   def input_spec(self):
     spec = lit_types.remap_spec(self.wrapped.input_spec(), self.FIELD_RENAMES)
     spec["source_language"] = lit_types.CategoryLabel()
@@ -505,10 +501,6 @@ def predict(self, inputs):
       mo["rougeL"] = float(score["rougeL"].fmeasure)
       yield mo
 
-  def predict_with_metadata(self, indexed_inputs):
-    """As predict(), but inputs are IndexedInput."""
-    return self.predict((ex["data"] for ex in indexed_inputs))
-
   def input_spec(self):
     return lit_types.remap_spec(self.wrapped.input_spec(), self.FIELD_RENAMES)
 

lit_nlp/lib/caching.py

@@ -285,13 +285,6 @@ def predict(self,
 
     return cached_results
 
-  # TODO(b/171513556): remove this method once we no longer need to override
-  # ModelWrapper.predict_with_metadata()
-  def predict_with_metadata(self, indexed_inputs: Iterable[JsonDict], **kw):
-    """As predict(), but inputs are IndexedInput."""
-    results = self.predict((ex["data"] for ex in indexed_inputs), **kw)
-    return results
-
   def _get_results_from_cache(self, input_keys: list[CacheKey]):
     with self._cache.lock:
       return [self._cache.get(input_key) for input_key in input_keys]

lit_nlp/lib/caching_test.py

@@ -37,21 +37,21 @@ def test_caching_model_wrapper_no_dataset_skip_cache(self):
     model = testing_utils.IdentityRegressionModelForTesting()
     wrapper = caching.CachingModelWrapper(model, "test")
     examples = [{"data": {"val": 1}, "id": "my_id"}]
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(1, model.count)
     self.assertEqual({"score": 1}, results[0])
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(2, model.count)
     self.assertEqual({"score": 1}, results[0])
 
   def test_caching_model_wrapper_use_cache(self):
     model = testing_utils.IdentityRegressionModelForTesting()
     wrapper = caching.CachingModelWrapper(model, "test")
     examples = [{"data": {"val": 1, "_id": "id_to_cache"}, "id": "id_to_cache"}]
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(1, model.count)
     self.assertEqual({"score": 1}, results[0])
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(1, model.count)
     self.assertEqual({"score": 1}, results[0])
     self.assertEmpty(wrapper._cache._pred_locks)
@@ -60,11 +60,11 @@ def test_caching_model_wrapper_not_cached(self):
     model = testing_utils.IdentityRegressionModelForTesting()
     wrapper = caching.CachingModelWrapper(model, "test")
     examples = [{"data": {"val": 1}, "id": "my_id"}]
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(1, model.count)
     self.assertEqual({"score": 1}, results[0])
     examples = [{"data": {"val": 2}, "id": "other_id"}]
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(2, model.count)
     self.assertEqual({"score": 2}, results[0])
 
@@ -98,14 +98,14 @@ def test_caching_model_wrapper_mixed_list(self):
     subset = examples[:1]
 
     # Run the CachingModelWrapper over a subset of examples
-    results = wrapper.predict_with_metadata(subset)
+    results = list(wrapper.predict_with_metadata(subset))
     self.assertEqual(1, model.count)
     self.assertEqual({"score": 0}, results[0])
 
     # Now, run the CachingModelWrapper over all of the examples. This should
     # only pass the examples that were not in subset to the wrapped model, and
     # the total number of inputs processed by the wrapped model should be 3
-    results = wrapper.predict_with_metadata(examples)
+    results = list(wrapper.predict_with_metadata(examples))
     self.assertEqual(3, model.count)
     self.assertEqual({"score": 0}, results[0])
     self.assertEqual({"score": 1}, results[1])