ENH Typing: fix library interface (#2265)

Improve typing (re-export) in __init__.py files.
huggingface · Dec 11, 2024 · 3c61b3e · 3c61b3e
1 parent b516cee
commit 3c61b3e
Show file tree

Hide file tree

Showing 4 changed files with 186 additions and 25 deletions.
diff --git a/src/peft/__init__.py b/src/peft/__init__.py
@@ -1,8 +1,3 @@
-# flake8: noqa
-# There's no way to ignore "F401 '...' imported but unused" warnings in this
-# module, but to preserve other warnings. So, don't check this module at all.
-
-# coding=utf-8
 # Copyright 2023-present the HuggingFace Inc. team.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -19,7 +14,7 @@
 
 __version__ = "0.14.1.dev0"
 
-from .auto import (
+from .auto import (  # noqa: I001
     AutoPeftModel,
     AutoPeftModelForCausalLM,
     AutoPeftModelForSequenceClassification,
@@ -91,7 +86,6 @@
     XLoraModel,
     HRAConfig,
     HRAModel,
-    VBLoRAConfig,
     get_eva_state_dict,
     initialize_lora_eva_weights,
     CPTEmbedding,
@@ -113,3 +107,91 @@
     cast_mixed_precision_params,
 )
 from .config import PeftConfig, PromptLearningConfig
+
+__all__ = [
+    "MODEL_TYPE_TO_PEFT_MODEL_MAPPING",
+    "PEFT_TYPE_TO_CONFIG_MAPPING",
+    "TRANSFORMERS_MODELS_TO_PREFIX_TUNING_POSTPROCESS_MAPPING",
+    "AdaLoraConfig",
+    "AdaLoraModel",
+    "AdaptionPromptConfig",
+    "AdaptionPromptModel",
+    "AutoPeftModel",
+    "AutoPeftModelForCausalLM",
+    "AutoPeftModelForFeatureExtraction",
+    "AutoPeftModelForQuestionAnswering",
+    "AutoPeftModelForSeq2SeqLM",
+    "AutoPeftModelForSequenceClassification",
+    "AutoPeftModelForTokenClassification",
+    "BOFTConfig",
+    "BOFTModel",
+    "BoneConfig",
+    "BoneModel",
+    "CPTConfig",
+    "CPTEmbedding",
+    "EvaConfig",
+    "FourierFTConfig",
+    "FourierFTModel",
+    "HRAConfig",
+    "HRAModel",
+    "IA3Config",
+    "IA3Model",
+    "LNTuningConfig",
+    "LNTuningModel",
+    "LoHaConfig",
+    "LoHaModel",
+    "LoKrConfig",
+    "LoKrModel",
+    "LoftQConfig",
+    "LoraConfig",
+    "LoraModel",
+    "LoraRuntimeConfig",
+    "MultitaskPromptTuningConfig",
+    "MultitaskPromptTuningInit",
+    "OFTConfig",
+    "OFTModel",
+    "PeftConfig",
+    "PeftMixedModel",
+    "PeftModel",
+    "PeftModelForCausalLM",
+    "PeftModelForFeatureExtraction",
+    "PeftModelForQuestionAnswering",
+    "PeftModelForSeq2SeqLM",
+    "PeftModelForSequenceClassification",
+    "PeftModelForTokenClassification",
+    "PeftType",
+    "PolyConfig",
+    "PolyModel",
+    "PrefixEncoder",
+    "PrefixTuningConfig",
+    "PromptEmbedding",
+    "PromptEncoder",
+    "PromptEncoderConfig",
+    "PromptEncoderReparameterizationType",
+    "PromptLearningConfig",
+    "PromptTuningConfig",
+    "PromptTuningInit",
+    "TaskType",
+    "VBLoRAConfig",
+    "VBLoRAConfig",
+    "VBLoRAModel",
+    "VeraConfig",
+    "VeraModel",
+    "XLoraConfig",
+    "XLoraModel",
+    "bloom_model_postprocess_past_key_value",
+    "cast_mixed_precision_params",
+    "get_eva_state_dict",
+    "get_layer_status",
+    "get_model_status",
+    "get_peft_config",
+    "get_peft_model",
+    "get_peft_model_state_dict",
+    "initialize_lora_eva_weights",
+    "inject_adapter_in_model",
+    "load_peft_weights",
+    "prepare_model_for_kbit_training",
+    "replace_lora_weights_loftq",
+    "set_peft_model_state_dict",
+    "shift_tokens_right",
+]
diff --git a/src/peft/tuners/__init__.py b/src/peft/tuners/__init__.py
@@ -1,8 +1,3 @@
-# flake8: noqa
-# There's no way to ignore "F401 '...' imported but unused" warnings in this
-# module, but to preserve other warnings. So, don't check this module at all
-
-# coding=utf-8
 # Copyright 2023-present the HuggingFace Inc. team.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -17,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from .adaption_prompt import AdaptionPromptConfig, AdaptionPromptModel
+from .adaption_prompt import AdaptionPromptConfig, AdaptionPromptModel  # noqa: I001
 from .lora import (
     LoraConfig,
     LoraModel,
@@ -47,3 +42,57 @@
 from .vblora import VBLoRAConfig, VBLoRAModel
 from .cpt import CPTConfig, CPTEmbedding
 from .bone import BoneConfig, BoneModel
+
+__all__ = [
+    "AdaLoraConfig",
+    "AdaLoraModel",
+    "AdaptionPromptConfig",
+    "AdaptionPromptModel",
+    "BOFTConfig",
+    "BOFTModel",
+    "BoneConfig",
+    "BoneModel",
+    "CPTConfig",
+    "CPTEmbedding",
+    "EvaConfig",
+    "FourierFTConfig",
+    "FourierFTModel",
+    "HRAConfig",
+    "HRAModel",
+    "IA3Config",
+    "IA3Model",
+    "LNTuningConfig",
+    "LNTuningModel",
+    "LoHaConfig",
+    "LoHaModel",
+    "LoKrConfig",
+    "LoKrModel",
+    "LoftQConfig",
+    "LoraConfig",
+    "LoraModel",
+    "LoraRuntimeConfig",
+    "MixedModel",
+    "MultitaskPromptEmbedding",
+    "MultitaskPromptTuningConfig",
+    "MultitaskPromptTuningInit",
+    "OFTConfig",
+    "OFTModel",
+    "PolyConfig",
+    "PolyModel",
+    "PrefixEncoder",
+    "PrefixTuningConfig",
+    "PromptEmbedding",
+    "PromptEncoder",
+    "PromptEncoderConfig",
+    "PromptEncoderReparameterizationType",
+    "PromptTuningConfig",
+    "PromptTuningInit",
+    "VBLoRAConfig",
+    "VBLoRAModel",
+    "VeraConfig",
+    "VeraModel",
+    "XLoraConfig",
+    "XLoraModel",
+    "get_eva_state_dict",
+    "initialize_lora_eva_weights",
+]
diff --git a/src/peft/utils/__init__.py b/src/peft/utils/__init__.py
@@ -1,8 +1,3 @@
-# flake8: noqa
-# There's no way to ignore "F401 '...' imported but unused" warnings in this
-# module, but to preserve other warnings. So, don't check this module at all
-
-# coding=utf-8
 # Copyright 2023-present the HuggingFace Inc. team.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -17,8 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-# from .config import PeftConfig, PeftType, PromptLearningConfig, TaskType
-from .integrations import map_cache_to_layer_device_map
+from .integrations import map_cache_to_layer_device_map  # noqa: I001
 from .loftq_utils import replace_lora_weights_loftq
 from .peft_types import PeftType, TaskType
 from .other import (
@@ -54,3 +48,43 @@
     cast_mixed_precision_params,
 )
 from .save_and_load import get_peft_model_state_dict, set_peft_model_state_dict, load_peft_weights
+
+__all__ = [
+    "CONFIG_NAME",
+    "INCLUDE_LINEAR_LAYERS_SHORTHAND",
+    "SAFETENSORS_WEIGHTS_NAME",
+    "TRANSFORMERS_MODELS_TO_ADALORA_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_FOURIERFT_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_IA3_FEEDFORWARD_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_IA3_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_LNTUNING_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_LORA_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_PREFIX_TUNING_POSTPROCESS_MAPPING",
+    "TRANSFORMERS_MODELS_TO_VBLORA_TARGET_MODULES_MAPPING",
+    "TRANSFORMERS_MODELS_TO_VERA_TARGET_MODULES_MAPPING",
+    "WEIGHTS_NAME",
+    "ModulesToSaveWrapper",
+    "PeftType",
+    "TaskType",
+    "_freeze_adapter",
+    "_get_batch_size",
+    "_get_submodules",
+    "_is_valid_match",
+    "_prepare_prompt_learning_config",
+    "_set_adapter",
+    "_set_trainable",
+    "bloom_model_postprocess_past_key_value",
+    "cast_mixed_precision_params",
+    "get_auto_gptq_quant_linear",
+    "get_peft_model_state_dict",
+    "get_quantization_config",
+    "id_tensor_storage",
+    "infer_device",
+    "load_peft_weights",
+    "map_cache_to_layer_device_map",
+    "prepare_model_for_kbit_training",
+    "replace_lora_weights_loftq",
+    "set_peft_model_state_dict",
+    "shift_tokens_right",
+    "transpose",
+]
diff --git a/src/peft/utils/peft_types.py b/src/peft/utils/peft_types.py
@@ -1,8 +1,3 @@
-# flake8: noqa
-# There's no way to ignore "F401 '...' imported but unused" warnings in this
-# module, but to preserve other warnings. So, don't check this module at all
-
-# coding=utf-8
 # Copyright 2023-present the HuggingFace Inc. team.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -16,6 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+
 import enum