Enable some ruff checks for performance and readability (#39383)

* Fix inefficient sequence tests

Signed-off-by: cyy <cyyever@outlook.com>

* Enable PERF102

Signed-off-by: cyy <cyyever@outlook.com>

* Enable PLC1802

Signed-off-by: cyy <cyyever@outlook.com>

* Enable PLC0208

Signed-off-by: cyy <cyyever@outlook.com>

---------

Signed-off-by: cyy <cyyever@outlook.com>

examples/pytorch/3d_parallel_checks.py

@@ -754,9 +754,9 @@ def get_parameters(model: nn.Module) -> Iterable[torch.Tensor]:
     Returns:
         Iterable[torch.Tensor]: An iterator over all parameters in the model
     """
-    for name, module in model._modules.items():
+    for module in model._modules.values():
         # Look for parameters in module attributes
-        for attr_name, attr in module.__dict__.items():
+        for attr in module.__dict__.values():
             if isinstance(attr, torch.Tensor) and attr.requires_grad:
                 yield attr
         # Recursively get parameters from submodules

pyproject.toml

@@ -22,7 +22,7 @@ line-length = 119
 ignore = ["C901", "E501", "E741", "F402", "F823"]
 # RUF013: Checks for the use of implicit Optional
 #  in type annotations when the default parameter value is None.
-select = ["C", "E", "F", "I", "W", "RUF013", "UP006"]
+select = ["C", "E", "F", "I", "W", "RUF013", "UP006", "PERF102", "PLC1802", "PLC0208"]
 extend-safe-fixes = ["UP006"]
 
 # Ignore import violations in all `__init__.py` files.

src/transformers/configuration_utils.py

@@ -607,7 +607,7 @@ class PretrainedConfig(PushToHubMixin):
         if "model_type" in config_dict and hasattr(cls, "model_type") and config_dict["model_type"] != cls.model_type:
             # sometimes the config has no `base_config_key` if the config is used in several composite models
             # e.g. LlamaConfig. In that case we try to see if there is match in `model_type` before raising a warning
-            for k, v in config_dict.items():
+            for v in config_dict.values():
                 if isinstance(v, dict) and v.get("model_type") == cls.model_type:
                     config_dict = v
 

src/transformers/modeling_utils.py

@@ -2166,7 +2166,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, PushToHubMixin, PeftAdapterMi
                 self._tp_plan.update({f"{name}.{k}": v for k, v in plan.copy().items()})
 
         if self._tp_plan is not None and is_torch_greater_or_equal("2.5") and _torch_distributed_available:
-            for _, v in self._tp_plan.items():
+            for v in self._tp_plan.values():
                 if v not in ALL_PARALLEL_STYLES:
                     raise ValueError(
                         f"Unsupported tensor parallel style {v}. Supported styles are {ALL_PARALLEL_STYLES}"
@@ -2845,7 +2845,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, PushToHubMixin, PeftAdapterMi
 
                 all_encoder_weights = {module_name + "/" + sub_name for sub_name in encoder_modules.keys()}
                 encoder_layer_pos = 0
-                for name, module in decoder_modules.items():
+                for name in decoder_modules.keys():
                     if name.isdigit():
                         encoder_name = str(int(name) + encoder_layer_pos)
                         decoder_name = name
@@ -5830,7 +5830,7 @@ def caching_allocator_warmup(model: PreTrainedModel, expanded_device_map: dict,
     accelerator_device_map = {
         param: torch.device(device) for param, device in expanded_device_map.items() if is_accelerator_device(device)
     }
-    if not len(accelerator_device_map):
+    if not accelerator_device_map:
         return
 
     tp_plan_regex = (

src/transformers/models/auto/feature_extraction_auto.py

@@ -133,7 +133,7 @@ def feature_extractor_class_from_name(class_name: str):
             except AttributeError:
                 continue
 
-    for _, extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.items():
+    for extractor in FEATURE_EXTRACTOR_MAPPING._extra_content.values():
         if getattr(extractor, "__name__", None) == class_name:
             return extractor
 

src/transformers/models/auto/image_processing_auto.py

@@ -212,7 +212,7 @@ def get_image_processor_class_from_name(class_name: str):
             except AttributeError:
                 continue
 
-    for _, extractors in IMAGE_PROCESSOR_MAPPING._extra_content.items():
+    for extractors in IMAGE_PROCESSOR_MAPPING._extra_content.values():
         for extractor in extractors:
             if getattr(extractor, "__name__", None) == class_name:
                 return extractor
@@ -533,7 +533,7 @@ class AutoImageProcessor:
                 )
                 use_fast = False
             if use_fast:
-                for _, image_processors in IMAGE_PROCESSOR_MAPPING_NAMES.items():
+                for image_processors in IMAGE_PROCESSOR_MAPPING_NAMES.values():
                     if image_processor_type in image_processors:
                         break
                 else:

src/transformers/models/auto/tokenization_auto.py

@@ -744,7 +744,7 @@ def tokenizer_class_from_name(class_name: str) -> Union[type[Any], None]:
             except AttributeError:
                 continue
 
-    for config, tokenizers in TOKENIZER_MAPPING._extra_content.items():
+    for tokenizers in TOKENIZER_MAPPING._extra_content.values():
         for tokenizer in tokenizers:
             if getattr(tokenizer, "__name__", None) == class_name:
                 return tokenizer

src/transformers/models/auto/video_processing_auto.py

@@ -84,7 +84,7 @@ def video_processor_class_from_name(class_name: str):
             except AttributeError:
                 continue
 
-    for _, extractor in VIDEO_PROCESSOR_MAPPING._extra_content.items():
+    for extractor in VIDEO_PROCESSOR_MAPPING._extra_content.values():
         if getattr(extractor, "__name__", None) == class_name:
             return extractor
 

src/transformers/models/bridgetower/modeling_bridgetower.py

@@ -140,7 +140,7 @@ class BridgeTowerResidualAttention(nn.Module):
     def forward(self, hidden_state: torch.Tensor, attention_mask: Optional[torch.Tensor] = None):
         residual_state = hidden_state + self.attention(self.ln_1(hidden_state), attention_mask)
         hidden_state = self.ln_2(residual_state)
-        for _, layer in self.mlp.items():
+        for layer in self.mlp.values():
             hidden_state = layer(hidden_state)
         hidden_state = residual_state + hidden_state
         return hidden_state

src/transformers/models/donut/processing_donut.py

@@ -199,7 +199,7 @@ class DonutProcessor(ProcessorMixin):
                 if tokens[:6] == r"<sep/>":  # non-leaf nodes
                     return [output] + self.token2json(tokens[6:], is_inner_value=True, added_vocab=added_vocab)
 
-        if len(output):
+        if output:
             return [output] if is_inner_value else output
         else:
             return [] if is_inner_value else {"text_sequence": tokens}

src/transformers/models/grounding_dino/convert_grounding_dino_to_hf.py

@@ -239,7 +239,7 @@ def create_rename_keys(state_dict, config):
     ########################################## DECODER - END
 
     ########################################## Additional - START
-    for layer_name, params in state_dict.items():
+    for layer_name in state_dict.keys():
         #### TEXT BACKBONE
         if "bert" in layer_name:
             rename_keys.append((layer_name, layer_name.replace("bert", "model.text_backbone")))

src/transformers/models/llama4/image_processing_llama4_fast.py

@@ -177,7 +177,7 @@ def find_supported_resolutions(max_num_chunks: int, patch_size: SizeDict) -> tor
 
     # get the resolutions multiplied by the patch_size
     possible_resolutions = []
-    for key, value in asp_dict.items():
+    for value in asp_dict.values():
         for height, depth in value:
             possible_resolutions.append((height * patch_size, depth * patch_size))
 

src/transformers/models/mluke/convert_mluke_original_pytorch_checkpoint_to_pytorch.py

@@ -100,7 +100,7 @@ def convert_luke_checkpoint(checkpoint_path, metadata_path, entity_vocab_path, p
     state_dict.pop("lm_head.decoder.weight")
     state_dict.pop("lm_head.decoder.bias")
     state_dict_for_hugging_face = OrderedDict()
-    for key, value in state_dict.items():
+    for key in state_dict.keys():
         if not (key.startswith("lm_head") or key.startswith("entity_predictions")):
             state_dict_for_hugging_face[f"luke.{key}"] = state_dict[key]
         else:

src/transformers/models/omdet_turbo/convert_omdet_turbo_to_hf.py

@@ -100,7 +100,7 @@ def create_rename_keys_vision(state_dict, config):
     ########################################## VISION BACKBONE - END
 
     ########################################## ENCODER - START
-    for layer_name, params in state_dict.items():
+    for layer_name in state_dict.keys():
         if "neck" in layer_name:
             layer_name_replace = layer_name.replace("neck", "encoder")
             layer_name_replace = layer_name_replace.replace("input_proj", "channel_projection_layers")
@@ -117,7 +117,7 @@ def create_rename_keys_vision(state_dict, config):
     ########################################## ENCODER - END
 
     ########################################## DECODER - START
-    for layer_name, params in state_dict.items():
+    for layer_name in state_dict.keys():
         if layer_name.startswith("decoder"):
             layer_name_replace = layer_name.replace("decoder.decoder.layers", "decoder.layers")
             layer_name_replace = layer_name_replace.replace("input_proj", "channel_projection_layers")

src/transformers/models/xmod/convert_xmod_original_pytorch_checkpoint_to_pytorch.py

@@ -144,7 +144,7 @@ def convert_xmod_checkpoint_to_pytorch(
 
         if sorted(bert_output.adapter_modules.keys()) != sorted(xmod_layer.adapter_modules.keys()):
             raise AssertionError("Lists of language adapters do not match.")
-        for lang_code, adapter in xmod_layer.adapter_modules.items():
+        for lang_code in xmod_layer.adapter_modules.keys():
             to_adapter = bert_output.adapter_modules[lang_code]
             from_adapter = xmod_layer.adapter_modules[lang_code]
             to_adapter.dense1.weight = from_adapter.fc1.weight

src/transformers/models/zoedepth/convert_zoedepth_to_hf.py

@@ -266,7 +266,7 @@ def convert_state_dict(orig_state_dict):
 
 
 def remove_ignore_keys(state_dict):
-    for key, _ in state_dict.copy().items():
+    for key in state_dict.copy().keys():
         if (
             "fc_norm" in key
             or "relative_position_index" in key

src/transformers/pipelines/base.py

@@ -1288,14 +1288,14 @@ class Pipeline(_ScikitCompat, PushToHubMixin):
             if self.task in SUPPORTED_PEFT_TASKS:
                 supported_models_names.extend(SUPPORTED_PEFT_TASKS[self.task])
 
-            for _, model_name in supported_models.items():
+            for model_name in supported_models.values():
                 # Mapping can now contain tuples of models for the same configuration.
                 if isinstance(model_name, tuple):
                     supported_models_names.extend(list(model_name))
                 else:
                     supported_models_names.append(model_name)
             if hasattr(supported_models, "_model_mapping"):
-                for _, model in supported_models._model_mapping._extra_content.items():
+                for model in supported_models._model_mapping._extra_content.values():
                     if isinstance(model_name, tuple):
                         supported_models_names.extend([m.__name__ for m in model])
                     else:

src/transformers/tokenization_utils_base.py

@@ -232,7 +232,7 @@ class BatchEncoding(UserDict):
 
         self._encodings = encoding
 
-        if n_sequences is None and encoding is not None and len(encoding):
+        if n_sequences is None and encoding is not None and encoding:
             n_sequences = encoding[0].n_sequences
 
         self._n_sequences = n_sequences

src/transformers/trainer_pt_utils.py

@@ -149,7 +149,7 @@ def find_batch_size(tensors):
             if result is not None:
                 return result
     elif isinstance(tensors, Mapping):
-        for key, value in tensors.items():
+        for value in tensors.values():
             result = find_batch_size(value)
             if result is not None:
                 return result

src/transformers/utils/import_utils.py

@@ -2183,12 +2183,12 @@ class _LazyModule(ModuleType):
                 self._modules = self._modules.union(module_keys)
 
                 for key, values in module.items():
-                    if len(missing_backends):
+                    if missing_backends:
                         self._object_missing_backend[key] = missing_backends
 
                     for value in values:
                         self._class_to_module[value] = key
-                        if len(missing_backends):
+                        if missing_backends:
                             self._object_missing_backend[value] = missing_backends
                     _import_structure.setdefault(key, []).extend(values)
 

src/transformers/utils/quantization_config.py

@@ -1199,7 +1199,7 @@ class VptqConfig(QuantizationConfigMixin):
         r"""
         Safety checker that arguments are correct
         """
-        for layer_name, layer_param in self.config_for_layers.items():
+        for layer_param in self.config_for_layers.values():
             VptqLayerConfig(**layer_param)
         if self.enable_proxy_error is True:
             raise ValueError("enable_proxy_error should always be False until we support training")

tests/models/auto/test_modeling_auto.py

@@ -125,7 +125,7 @@ class AutoModelTest(unittest.TestCase):
         self.assertIsNotNone(model)
         self.assertIsInstance(model, BertForPreTraining)
         # Only one value should not be initialized and in the missing keys.
-        for key, value in loading_info.items():
+        for value in loading_info.values():
             self.assertEqual(len(value), 0)
 
     @slow

tests/models/auto/test_tokenization_auto.py

@@ -70,7 +70,7 @@ class AutoTokenizerTest(unittest.TestCase):
 
     @slow
     def test_tokenizer_from_pretrained(self):
-        for model_name in {"google-bert/bert-base-uncased", "google-bert/bert-base-cased"}:
+        for model_name in ("google-bert/bert-base-uncased", "google-bert/bert-base-cased"):
             tokenizer = AutoTokenizer.from_pretrained(model_name)
             self.assertIsNotNone(tokenizer)
             self.assertIsInstance(tokenizer, (BertTokenizer, BertTokenizerFast))

tests/models/luke/test_modeling_luke.py

@@ -897,7 +897,7 @@ class LukeModelIntegrationTests(unittest.TestCase):
         encoding = tokenizer(text, entity_spans=[span], add_prefix_space=True, return_tensors="pt")
 
         # move all values to device
-        for key, value in encoding.items():
+        for key in encoding.keys():
             encoding[key] = encoding[key].to(torch_device)
 
         outputs = model(**encoding)
@@ -932,7 +932,7 @@ class LukeModelIntegrationTests(unittest.TestCase):
         encoding = tokenizer(text, entity_spans=[span], add_prefix_space=True, return_tensors="pt")
 
         # move all values to device
-        for key, value in encoding.items():
+        for key in encoding.keys():
             encoding[key] = encoding[key].to(torch_device)
 
         outputs = model(**encoding)

tests/peft_integration/test_peft_integration.py

@@ -757,7 +757,7 @@ class PeftIntegrationTester(unittest.TestCase, PeftTesterMixin):
                     model.load_adapter(tmpdirname, is_trainable=True)
 
                     for name, module in model.named_modules():
-                        if len(list(module.children())):
+                        if list(module.children()):
                             # only check leaf modules
                             continue
 

tests/test_modeling_common.py

@@ -2535,7 +2535,7 @@ class ModelTesterMixin:
 
                 shared_ptrs = {k: v for k, v in ptrs.items() if len(v) > 1}
 
-                for _, shared_names in shared_ptrs.items():
+                for shared_names in shared_ptrs.values():
                     reloaded_ptrs = {reloaded_state[k].data_ptr() for k in shared_names}
                     self.assertEqual(
                         len(reloaded_ptrs),

tests/test_pipeline_mixin.py

@@ -139,7 +139,7 @@ task_to_pipeline_and_spec_mapping = {
     "zero-shot-image-classification": (ZeroShotImageClassificationPipeline, ZeroShotImageClassificationInput),
 }
 
-for task, task_info in pipeline_test_mapping.items():
+for task_info in pipeline_test_mapping.values():
     test = task_info["test"]
     task_info["mapping"] = {
         "pt": getattr(test, "model_mapping", None),

tests/utils/test_import_structure.py

@@ -96,7 +96,7 @@ class TestImportStructures(unittest.TestCase):
             with self.subTest(f"Testing arch {architecture}"):
                 import_structure = define_import_structure(self.models_path / architecture)
                 backend_agnostic_import_structure = {}
-                for requirement, module_object_mapping in import_structure.items():
+                for module_object_mapping in import_structure.values():
                     for module, objects in module_object_mapping.items():
                         if module not in backend_agnostic_import_structure:
                             backend_agnostic_import_structure[module] = []

utils/add_pipeline_model_mapping_to_test.py

@@ -37,7 +37,7 @@ from tests.test_pipeline_mixin import pipeline_test_mapping
 
 
 PIPELINE_TEST_MAPPING = {}
-for task, _ in pipeline_test_mapping.items():
+for task in pipeline_test_mapping.keys():
     PIPELINE_TEST_MAPPING[task] = {"pt": None, "tf": None}
 
 

utils/check_repo.py

@@ -790,7 +790,7 @@ def check_all_auto_object_names_being_defined():
         mappings_to_check.update({name: getattr(module, name) for name in mapping_names})
 
     for name, mapping in mappings_to_check.items():
-        for _, class_names in mapping.items():
+        for class_names in mapping.values():
             if not isinstance(class_names, tuple):
                 class_names = (class_names,)
                 for class_name in class_names:

utils/notification_service_doc_tests.py

@@ -332,7 +332,7 @@ if __name__ == "__main__":
 
     doc_test_results = {}
     # `artifact_key` is the artifact path
-    for artifact_key, artifact_obj in available_artifacts.items():
+    for artifact_obj in available_artifacts.values():
         artifact_path = artifact_obj.paths[0]
         if not artifact_path["path"].startswith("doc_tests_gpu_test_reports_"):
             continue