You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

471 lines
19 KiB

from __future__ import annotations
import inspect
import logging
import os
from collections.abc import Callable
from pathlib import Path
from typing import TYPE_CHECKING, Any
from sentence_transformers.backend import load_onnx_model, load_openvino_model
try:
from typing import Self
except ImportError:
from typing_extensions import Self
import torch
from transformers import AutoConfig, AutoModel, AutoTokenizer, MT5Config, PretrainedConfig, T5Config
from transformers.utils.import_utils import is_peft_available
from transformers.utils.peft_utils import find_adapter_config_file
from sentence_transformers.models.InputModule import InputModule
logger = logging.getLogger(__name__)
if TYPE_CHECKING and is_peft_available():
from peft import PeftConfig
def _save_pretrained_wrapper(_save_pretrained_fn: Callable, subfolder: str) -> Callable[..., None]:
def wrapper(save_directory: str | Path, **kwargs) -> None:
os.makedirs(Path(save_directory) / subfolder, exist_ok=True)
return _save_pretrained_fn(Path(save_directory) / subfolder, **kwargs)
return wrapper
class Transformer(InputModule):
"""Hugging Face AutoModel to generate token embeddings.
Loads the correct class, e.g. BERT / RoBERTa etc.
Args:
model_name_or_path: Hugging Face models name
(https://huggingface.co/models)
max_seq_length: Truncate any inputs longer than max_seq_length
model_args: Keyword arguments passed to the Hugging Face
Transformers model
tokenizer_args: Keyword arguments passed to the Hugging Face
Transformers tokenizer
config_args: Keyword arguments passed to the Hugging Face
Transformers config
cache_dir: Cache dir for Hugging Face Transformers to store/load
models
do_lower_case: If true, lowercases the input (independent if the
model is cased or not)
tokenizer_name_or_path: Name or path of the tokenizer. When
None, then model_name_or_path is used
backend: Backend used for model inference. Can be `torch`, `onnx`,
or `openvino`. Default is `torch`.
"""
config_file_name: str = "sentence_bert_config.json"
config_keys: list[str] = ["max_seq_length", "do_lower_case"]
save_in_root: bool = True
def __init__(
self,
model_name_or_path: str,
max_seq_length: int | None = None,
model_args: dict[str, Any] | None = None,
tokenizer_args: dict[str, Any] | None = None,
config_args: dict[str, Any] | None = None,
cache_dir: str | None = None,
do_lower_case: bool = False,
tokenizer_name_or_path: str | None = None,
backend: str = "torch",
) -> None:
super().__init__()
self.do_lower_case = do_lower_case
self.backend = backend
if model_args is None:
model_args = {}
if tokenizer_args is None:
tokenizer_args = {}
if config_args is None:
config_args = {}
config, is_peft_model = self._load_config(model_name_or_path, cache_dir, backend, config_args)
self._load_model(model_name_or_path, config, cache_dir, backend, is_peft_model, **model_args)
# Get the signature of the auto_model's forward method to pass only the expected arguments from `features`,
# plus some common values like "input_ids", "attention_mask", etc.
model_forward_params = list(inspect.signature(self.auto_model.forward).parameters)
self.model_forward_params = set(model_forward_params) | {
"input_ids",
"attention_mask",
"token_type_ids",
"inputs_embeds",
}
if max_seq_length is not None and "model_max_length" not in tokenizer_args:
tokenizer_args["model_max_length"] = max_seq_length
self.tokenizer = AutoTokenizer.from_pretrained(
tokenizer_name_or_path if tokenizer_name_or_path is not None else model_name_or_path,
cache_dir=cache_dir,
**tokenizer_args,
)
# No max_seq_length set. Try to infer from model
if max_seq_length is None:
if (
hasattr(self.auto_model, "config")
and hasattr(self.auto_model.config, "max_position_embeddings")
and hasattr(self.tokenizer, "model_max_length")
):
max_seq_length = min(self.auto_model.config.max_position_embeddings, self.tokenizer.model_max_length)
self.max_seq_length = max_seq_length
if tokenizer_name_or_path is not None:
self.auto_model.config.tokenizer_class = self.tokenizer.__class__.__name__
def _load_config(
self, model_name_or_path: str, cache_dir: str | None, backend: str, config_args: dict[str, Any]
) -> tuple[PeftConfig | PretrainedConfig, bool]:
"""Loads the transformers or PEFT configuration
Args:
model_name_or_path (str): The model name on Hugging Face (e.g. 'sentence-transformers/all-MiniLM-L6-v2')
or the path to a local model directory.
cache_dir (str | None): The cache directory to store the model configuration.
backend (str): The backend used for model inference. Can be `torch`, `onnx`, or `openvino`.
config_args (dict[str, Any]): Keyword arguments passed to the Hugging Face Transformers config.
Returns:
tuple[PretrainedConfig, bool]: The model configuration and a boolean indicating whether the model is a PEFT model.
"""
if (
find_adapter_config_file(
model_name_or_path,
cache_dir=cache_dir,
token=config_args.get("token"),
revision=config_args.get("revision"),
local_files_only=config_args.get("local_files_only", False),
)
is not None
):
if not is_peft_available():
raise Exception(
"Loading a PEFT model requires installing the `peft` package. You can install it via `pip install peft`."
)
if backend != "torch":
# TODO: Consider following these steps automatically so we can load PEFT models with other backends
raise ValueError(
"PEFT models can currently only be loaded with the `torch` backend. "
'To use other backends, load the model with `backend="torch"`, call `model.transformers_model.merge_and_unload()`, '
"save that model with `model.save_pretrained()` and then load the model with the desired backend."
)
from peft import PeftConfig
return PeftConfig.from_pretrained(model_name_or_path, **config_args, cache_dir=cache_dir), True
return AutoConfig.from_pretrained(model_name_or_path, **config_args, cache_dir=cache_dir), False
def _load_model(
self,
model_name_or_path: str,
config: PeftConfig | PretrainedConfig,
cache_dir: str,
backend: str,
is_peft_model: bool,
**model_args,
) -> None:
"""Loads the transformers or PEFT model into the `auto_model` attribute
Args:
model_name_or_path (str): The model name on Hugging Face (e.g. 'sentence-transformers/all-MiniLM-L6-v2')
or the path to a local model directory.
config ("PeftConfig" | PretrainedConfig): The model configuration.
cache_dir (str | None): The cache directory to store the model configuration.
backend (str): The backend used for model inference. Can be `torch`, `onnx`, or `openvino`.
is_peft_model (bool): Whether the model is a PEFT model.
model_args (dict[str, Any]): Keyword arguments passed to the Hugging Face Transformers model.
"""
if backend == "torch":
# When loading a PEFT model, we need to load the base model first,
# but some model_args are only for the adapter
if is_peft_model:
for adapter_only_kwarg in ["revision"]:
model_args.pop(adapter_only_kwarg, None)
if isinstance(config, T5Config):
self._load_t5_model(model_name_or_path, config, cache_dir, **model_args)
elif isinstance(config, MT5Config):
self._load_mt5_model(model_name_or_path, config, cache_dir, **model_args)
else:
self.auto_model = AutoModel.from_pretrained(
model_name_or_path, config=config, cache_dir=cache_dir, **model_args
)
elif backend == "onnx":
self.auto_model = load_onnx_model(
model_name_or_path=model_name_or_path,
config=config,
task_name="feature-extraction",
**model_args,
)
elif backend == "openvino":
self.auto_model = load_openvino_model(
model_name_or_path=model_name_or_path,
config=config,
task_name="feature-extraction",
**model_args,
)
else:
raise ValueError(f"Unsupported backend '{backend}'. `backend` should be `torch`, `onnx`, or `openvino`.")
def _load_t5_model(self, model_name_or_path: str, config: PretrainedConfig, cache_dir: str, **model_args) -> None:
"""Loads the encoder model from T5"""
from transformers import T5EncoderModel
T5EncoderModel._keys_to_ignore_on_load_unexpected = ["decoder.*"]
self.auto_model = T5EncoderModel.from_pretrained(
model_name_or_path, config=config, cache_dir=cache_dir, **model_args
)
def _load_mt5_model(self, model_name_or_path: str, config: PretrainedConfig, cache_dir: str, **model_args) -> None:
"""Loads the encoder model from T5"""
from transformers import MT5EncoderModel
MT5EncoderModel._keys_to_ignore_on_load_unexpected = ["decoder.*"]
self.auto_model = MT5EncoderModel.from_pretrained(
model_name_or_path, config=config, cache_dir=cache_dir, **model_args
)
def __repr__(self) -> str:
return f"Transformer({dict(self.get_config_dict(), architecture=self.auto_model.__class__.__name__)})"
def forward(self, features: dict[str, torch.Tensor], **kwargs) -> dict[str, torch.Tensor]:
"""
Forward pass through the transformer model.
This method processes the input features through the underlying transformers model
and returns the token embeddings along with any other relevant outputs.
Notes:
- Only passes arguments that are expected by the underlying transformer model
Args:
features (dict[str, torch.Tensor]): Input features dictionary containing at least
'input_ids' and 'attention_mask'. May also contain other tensors required by
the underlying transformer model.
**kwargs: Additional keyword arguments to pass to the underlying transformer model.
Returns:
dict[str, torch.Tensor]: Updated features dictionary containing the input features, plus:
- 'token_embeddings': Token-level embeddings from the transformer model
- 'attention_mask': Possibly modified attention mask if using PeftModel with prompt learning
- 'all_layer_embeddings': If the model outputs hidden states, contains embeddings from all layers
"""
trans_features = {key: value for key, value in features.items() if key in self.model_forward_params}
outputs = self.auto_model(**trans_features, **kwargs, return_dict=True)
token_embeddings = outputs[0]
features["token_embeddings"] = token_embeddings
# If the AutoModel is wrapped with a PeftModelForFeatureExtraction, then it may have added virtual tokens
# We need to extend the attention mask to include these virtual tokens, or the pooling will fail
if is_peft_available():
from peft import PeftModelForFeatureExtraction
if (
isinstance(self.auto_model, PeftModelForFeatureExtraction)
and self.auto_model.active_peft_config.is_prompt_learning
):
batch_size = token_embeddings.size(0)
attention_mask = features["attention_mask"]
prefix_attention_mask = torch.ones(
batch_size, self.auto_model.active_peft_config.num_virtual_tokens, device=attention_mask.device
)
features["attention_mask"] = torch.cat((prefix_attention_mask, attention_mask), dim=1)
if self.auto_model.config.output_hidden_states and "hidden_states" in outputs:
features["all_layer_embeddings"] = outputs["hidden_states"]
return features
def get_word_embedding_dimension(self) -> int:
return self.auto_model.config.hidden_size
def tokenize(
self, texts: list[str] | list[dict] | list[tuple[str, str]], padding: str | bool = True
) -> dict[str, torch.Tensor]:
"""Tokenizes a text and maps tokens to token-ids"""
output = {}
if isinstance(texts[0], str):
to_tokenize = [texts]
elif isinstance(texts[0], dict):
to_tokenize = []
output["text_keys"] = []
for lookup in texts:
text_key, text = next(iter(lookup.items()))
to_tokenize.append(text)
output["text_keys"].append(text_key)
to_tokenize = [to_tokenize]
else:
batch1, batch2 = [], []
for text_tuple in texts:
batch1.append(text_tuple[0])
batch2.append(text_tuple[1])
to_tokenize = [batch1, batch2]
# strip
to_tokenize = [[str(s).strip() for s in col] for col in to_tokenize]
# Lowercase
if self.do_lower_case:
to_tokenize = [[s.lower() for s in col] for col in to_tokenize]
output.update(
self.tokenizer(
*to_tokenize,
padding=padding,
truncation="longest_first",
return_tensors="pt",
max_length=self.max_seq_length,
)
)
return output
def save(self, output_path: str, safe_serialization: bool = True, **kwargs) -> None:
self.auto_model.save_pretrained(output_path, safe_serialization=safe_serialization)
self.tokenizer.save_pretrained(output_path)
self.save_config(output_path)
@classmethod
def load(
cls,
model_name_or_path: str,
# Loading arguments
subfolder: str = "",
token: bool | str | None = None,
cache_folder: str | None = None,
revision: str | None = None,
local_files_only: bool = False,
# Module-specific arguments
trust_remote_code: bool = False,
model_kwargs: dict[str, Any] | None = None,
tokenizer_kwargs: dict[str, Any] | None = None,
config_kwargs: dict[str, Any] | None = None,
backend: str = "torch",
**kwargs,
) -> Self:
init_kwargs = cls._load_init_kwargs(
model_name_or_path=model_name_or_path,
subfolder=subfolder,
token=token,
cache_folder=cache_folder,
revision=revision,
local_files_only=local_files_only,
trust_remote_code=trust_remote_code,
model_kwargs=model_kwargs,
tokenizer_kwargs=tokenizer_kwargs,
config_kwargs=config_kwargs,
backend=backend,
)
return cls(model_name_or_path=model_name_or_path, **init_kwargs)
@classmethod
def _load_init_kwargs(
cls,
model_name_or_path: str,
# Loading arguments
subfolder: str = "",
token: bool | str | None = None,
cache_folder: str | None = None,
revision: str | None = None,
local_files_only: bool = False,
# Module-specific arguments
trust_remote_code: bool = False,
model_kwargs: dict[str, Any] | None = None,
tokenizer_kwargs: dict[str, Any] | None = None,
config_kwargs: dict[str, Any] | None = None,
backend: str = "torch",
**kwargs,
) -> dict[str, Any]:
config = cls.load_config(
model_name_or_path=model_name_or_path,
subfolder=subfolder,
token=token,
cache_folder=cache_folder,
revision=revision,
local_files_only=local_files_only,
)
hub_kwargs = {
"subfolder": subfolder,
"token": token,
"revision": revision,
"local_files_only": local_files_only,
"trust_remote_code": trust_remote_code,
}
# 3rd priority: config file
if "model_args" not in config:
config["model_args"] = {}
if "tokenizer_args" not in config:
config["tokenizer_args"] = {}
if "config_args" not in config:
config["config_args"] = {}
# 2nd priority: hub_kwargs
config["model_args"].update(hub_kwargs)
config["tokenizer_args"].update(hub_kwargs)
config["config_args"].update(hub_kwargs)
# 1st priority: kwargs passed to SentenceTransformer
if model_kwargs:
config["model_args"].update(model_kwargs)
if tokenizer_kwargs:
config["tokenizer_args"].update(tokenizer_kwargs)
if config_kwargs:
config["config_args"].update(config_kwargs)
return {**config, "cache_dir": cache_folder, "backend": backend}
@classmethod
def load_config(
cls,
model_name_or_path: str,
subfolder: str = "",
config_filename: str | None = None,
token: bool | str | None = None,
cache_folder: str | None = None,
revision: str | None = None,
local_files_only: bool = False,
) -> dict[str, Any]:
config_filenames = (
[config_filename]
if config_filename
else [
"sentence_bert_config.json",
"sentence_roberta_config.json",
"sentence_distilbert_config.json",
"sentence_camembert_config.json",
"sentence_albert_config.json",
"sentence_xlm-roberta_config.json",
"sentence_xlnet_config.json",
]
)
for config_filename in config_filenames:
config = super().load_config(
model_name_or_path=model_name_or_path,
subfolder=subfolder,
config_filename=config_filename,
token=token,
cache_folder=cache_folder,
revision=revision,
local_files_only=local_files_only,
)
if config:
break
# Don't allow configs to set trust_remote_code
if "model_args" in config and "trust_remote_code" in config["model_args"]:
config["model_args"].pop("trust_remote_code")
if "tokenizer_args" in config and "trust_remote_code" in config["tokenizer_args"]:
config["tokenizer_args"].pop("trust_remote_code")
if "config_args" in config and "trust_remote_code" in config["config_args"]:
config["config_args"].pop("trust_remote_code")
return config