add support of 3UTR-BERT

multimolecule/__init__.py

@@ -24,6 +24,12 @@
     SpliceBertForSequenceClassification,
     SpliceBertForTokenClassification,
     SpliceBertModel,
+    UtrBertConfig,
+    UtrBertForMaskedLM,
+    UtrBertForPretraining,
+    UtrBertForSequenceClassification,
+    UtrBertForTokenClassification,
+    UtrBertModel,
     UtrLmConfig,
     UtrLmForMaskedLM,
     UtrLmForPretraining,
@@ -61,6 +67,12 @@
     "SpliceBertForPretraining",
     "SpliceBertForSequenceClassification",
     "SpliceBertForTokenClassification",
+    "UtrBertConfig",
+    "UtrBertModel",
+    "UtrBertForMaskedLM",
+    "UtrBertForPretraining",
+    "UtrBertForSequenceClassification",
+    "UtrBertForTokenClassification",
     "UtrLmConfig",
     "UtrLmModel",
     "UtrLmForMaskedLM",

multimolecule/models/__init__.py

@@ -31,6 +31,14 @@
     SpliceBertForTokenClassification,
     SpliceBertModel,
 )
+from .utrbert import (
+    UtrBertConfig,
+    UtrBertForMaskedLM,
+    UtrBertForPretraining,
+    UtrBertForSequenceClassification,
+    UtrBertForTokenClassification,
+    UtrBertModel,
+)
 from .utrlm import (
     UtrLmConfig,
     UtrLmForMaskedLM,
@@ -66,6 +74,12 @@
     "SpliceBertForPretraining",
     "SpliceBertForSequenceClassification",
     "SpliceBertForTokenClassification",
+    "UtrBertConfig",
+    "UtrBertModel",
+    "UtrBertForMaskedLM",
+    "UtrBertForPretraining",
+    "UtrBertForSequenceClassification",
+    "UtrBertForTokenClassification",
     "UtrLmConfig",
     "UtrLmForMaskedLM",
     "UtrLmForPretraining",

multimolecule/models/utrbert/__init__.py

@@ -0,0 +1,39 @@
+from transformers import (
+    AutoConfig,
+    AutoModel,
+    AutoModelForMaskedLM,
+    AutoModelForPreTraining,
+    AutoModelForSequenceClassification,
+    AutoModelForTokenClassification,
+    AutoModelWithLMHead,
+    AutoTokenizer,
+)
+
+from multimolecule.tokenizers.rna import RnaTokenizer
+
+from .configuration_utrbert import UtrBertConfig
+from .modeling_utrbert import (
+    UtrBertForMaskedLM,
+    UtrBertForPretraining,
+    UtrBertForSequenceClassification,
+    UtrBertForTokenClassification,
+    UtrBertModel,
+)
+
+__all__ = [
+    "UtrBertConfig",
+    "UtrBertModel",
+    "UtrBertForMaskedLM",
+    "UtrBertForPretraining",
+    "UtrBertForSequenceClassification",
+    "UtrBertForTokenClassification",
+]
+
+AutoConfig.register("utrbert", UtrBertConfig)
+AutoModel.register(UtrBertConfig, UtrBertModel)
+AutoModelForMaskedLM.register(UtrBertConfig, UtrBertForMaskedLM)
+AutoModelForPreTraining.register(UtrBertConfig, UtrBertForPretraining)
+AutoModelForSequenceClassification.register(UtrBertConfig, UtrBertForSequenceClassification)
+AutoModelForTokenClassification.register(UtrBertConfig, UtrBertForTokenClassification)
+AutoModelWithLMHead.register(UtrBertConfig, UtrBertForTokenClassification)
+AutoTokenizer.register(UtrBertConfig, RnaTokenizer)

multimolecule/models/utrbert/configuration_utrbert.py

@@ -0,0 +1,113 @@
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+logger = logging.get_logger(__name__)
+
+
+class UtrBertConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`UtrBertModel`]. It is used to instantiate a
+    3UTRBERT model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of the 3UTRBERT
+    [yangyn533/3UTRBERT](https://github.com/yangyn533/3UTRBERT) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*):
+            Vocabulary size of the UTRBERT model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`BertModel`].
+        nmers (`int`, *optional*):
+            kmer size of the UTRBERT model. Defines the vocabulary size of the model.
+        hidden_size (`int`, *optional*, defaults to 768):
+            Dimensionality of the encoder layers and the pooler layer.
+        num_hidden_layers (`int`, *optional*, defaults to 12):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 12):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        intermediate_size (`int`, *optional*, defaults to 3072):
+            Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
+        hidden_act (`str` or `Callable`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"silu"` and `"gelu_new"` are supported.
+        hidden_dropout (`float`, *optional*, defaults to 0.1):
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
+        attention_dropout (`float`, *optional*, defaults to 0.1):
+            The dropout ratio for the attention probabilities.
+        max_position_embeddings (`int`, *optional*, defaults to 512):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        type_vocab_size (`int`, *optional*, defaults to 2):
+            The vocabulary size of the `token_type_ids` passed when calling [`BertModel`] or [`TFBertModel`].
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-12):
+            The epsilon used by the layer normalization layers.
+        position_embedding_type (`str`, *optional*, defaults to `"absolute"`):
+            Type of position embedding. Choose one of `"absolute"`, `"relative_key"`, `"relative_key_query"`. For
+            positional embeddings use `"absolute"`. For more information on `"relative_key"`, please refer to
+            [Self-Attention with Relative Position Representations (Shaw et al.)](https://arxiv.org/abs/1803.02155).
+            For more information on `"relative_key_query"`, please refer to *Method 4* in [Improve Transformer Models
+            with Better Relative Position Embeddings (Huang et al.)](https://arxiv.org/abs/2009.13658).
+        is_decoder (`bool`, *optional*, defaults to `False`):
+            Whether the model is used as a decoder or not. If `False`, the model is used as an encoder.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`.
+
+    Examples:
+        >>> from multimolecule import UtrBertConfig, UtrBertModel
+
+        >>> # Initializing a UtrBERT multimolecule/utrbert style configuration
+        >>> configuration = UtrBertConfig()
+
+        >>> # Initializing a model (with random weights) from the multimolecule/utrbert style configuration
+        >>> model = UtrBertModel(configuration)
+
+        >>> # Accessing the model configuration
+        >>> configuration = model.config
+    """
+
+    model_type = "utrbert"
+
+    def __init__(
+        self,
+        vocab_size=None,
+        nmers=None,
+        hidden_size=768,
+        num_hidden_layers=12,
+        num_attention_heads=12,
+        intermediate_size=3072,
+        hidden_act="gelu",
+        hidden_dropout=0.1,
+        attention_dropout=0.1,
+        max_position_embeddings=512,
+        type_vocab_size=2,
+        initializer_range=0.02,
+        layer_norm_eps=1e-12,
+        pad_token_id=0,
+        position_embedding_type="absolute",
+        use_cache=True,
+        classifier_dropout=None,
+        **kwargs,
+    ):
+        super().__init__(pad_token_id=pad_token_id, **kwargs)
+
+        self.vocab_size = vocab_size
+        self.nmers = nmers
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.hidden_act = hidden_act
+        self.intermediate_size = intermediate_size
+        self.hidden_dropout = hidden_dropout
+        self.attention_dropout = attention_dropout
+        self.max_position_embeddings = max_position_embeddings
+        self.type_vocab_size = type_vocab_size
+        self.initializer_range = initializer_range
+        self.layer_norm_eps = layer_norm_eps
+        self.position_embedding_type = position_embedding_type
+        self.use_cache = use_cache
+        self.classifier_dropout = classifier_dropout

multimolecule/models/utrbert/convert_checkpoint.py

@@ -0,0 +1,124 @@
+import os
+from typing import Optional
+
+import chanfig
+import torch
+from torch import nn
+
+from multimolecule.models import SpliceBertConfig as Config
+from multimolecule.models import SpliceBertForPretraining as Model
+from multimolecule.tokenizers.rna.utils import get_special_tokens_map, get_tokenizer_config, get_vocab_list
+
+try:
+    from huggingface_hub import HfApi
+except ImportError:
+    HfApi = None
+
+torch.manual_seed(1013)
+
+
+def _convert_checkpoint(config, original_state_dict, vocab_list, original_vocab_list):
+    state_dict = {}
+    for key, value in original_state_dict.items():
+        key = key.replace("LayerNorm", "layer_norm")
+        key = key.replace("gamma", "weight")
+        key = key.replace("beta", "bias")
+        if key.startswith("bert"):
+            state_dict["splice" + key] = value
+            continue
+        if key.startswith("cls"):
+            key = "lm_head" + key[15:]
+            state_dict[key] = value
+            continue
+        state_dict[key] = value
+
+    state_vocab_size = state_dict["splicebert.embeddings.word_embeddings.weight"].size(0)
+    original_vocab_size = len(original_vocab_list)
+    if state_vocab_size != original_vocab_size:
+        raise ValueError(
+            f"Vocabulary size do not match. Expected to have {original_vocab_size}, but got {state_vocab_size}."
+        )
+    word_embed = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=config.pad_token_id)
+    word_embed_weight = word_embed.weight.data
+    predictions_decoder_weight = torch.zeros((config.vocab_size, config.hidden_size))
+    predictions_bias = torch.zeros(config.vocab_size)
+    # nn.init.normal_(pos_embed.weight, std=0.02)
+    for original_index, original_token in enumerate(original_vocab_list):
+        new_index = vocab_list.index(original_token)
+        word_embed_weight[new_index] = state_dict["splicebert.embeddings.word_embeddings.weight"][original_index]
+        predictions_decoder_weight[new_index] = state_dict["lm_head.decoder.weight"][original_index]
+        predictions_bias[new_index] = state_dict["lm_head.decoder.bias"][original_index]
+    state_dict["splicebert.embeddings.word_embeddings.weight"] = word_embed_weight
+    state_dict["lm_head.decoder.weight"] = predictions_decoder_weight
+    state_dict["lm_head.decoder.bias"] = state_dict["lm_head.bias"] = predictions_bias
+    return state_dict
+
+
+def convert_checkpoint(convert_config):
+    config = chanfig.load(os.path.join(convert_config.checkpoint_path, "config.json"))
+    config.hidden_dropout = config.pop("hidden_dropout_prob", 0.1)
+    config.attention_dropout = config.pop("attention_probs_dropout_prob", 0.1)
+    config.nmers = int(convert_config.checkpoint_path.split("/")[-1][0])
+    vocab_list = get_vocab_list(config.nmers)
+    config = Config.from_dict(config)
+    del config._name_or_path
+    config.architectures = ["SpliceBertModel"]
+    config.vocab_size = len(vocab_list)
+
+    model = Model(config)
+
+    ckpt = torch.load(
+        os.path.join(convert_config.checkpoint_path, "pytorch_model.bin"), map_location=torch.device("cpu")
+    )
+    original_vocab_list = []
+    for char in open(os.path.join(convert_config.checkpoint_path, "vocab.txt")).read().splitlines():  # noqa: SIM115
+        if char.startswith("["):
+            char = char.lower().replace("[", "<").replace("]", ">")
+        if char == "T":
+            char = "U"
+        if char == "<sep>":
+            char = "<eos>"
+        original_vocab_list.append(char)
+    state_dict = _convert_checkpoint(config, ckpt, vocab_list, original_vocab_list)
+
+    model.load_state_dict(state_dict)
+    model.save_pretrained(convert_config.output_path, safe_serialization=True)
+    model.save_pretrained(convert_config.output_path, safe_serialization=False)
+    chanfig.NestedDict(get_special_tokens_map()).json(
+        os.path.join(convert_config.output_path, "special_tokens_map.json")
+    )
+    chanfig.NestedDict(get_tokenizer_config()).json(os.path.join(convert_config.output_path, "tokenizer_config.json"))
+
+    if convert_config.push_to_hub:
+        if HfApi is None:
+            raise ImportError("Please install huggingface_hub to push to the hub.")
+        api = HfApi()
+        api.create_repo(
+            convert_config.repo_id,
+            token=convert_config.token,
+            exist_ok=True,
+        )
+        api.upload_folder(
+            repo_id=convert_config.repo_id, folder_path=convert_config.output_path, token=convert_config.token
+        )
+
+
+@chanfig.configclass
+class ConvertConfig:
+    checkpoint_path: str
+    output_path: Optional[str] = None
+    push_to_hub: bool = False
+    repo_id: Optional[str] = output_path
+    token: Optional[str] = None
+
+    def post(self):
+        if self.output_path is None:
+            self.output_path = self.checkpoint_path.split("/")[-1].lower()
+        if self.repo_id is None:
+            self.repo_id = f"multimolecule/{self.output_path}"
+
+
+if __name__ == "__main__":
+    config = ConvertConfig()
+    config.parse()  # type: ignore[attr-defined]
+    convert_checkpoint(config)