pyg-nightly 2.7.0.dev20241118__py3-none-any.whl → 2.7.0.dev20241120__py3-none-any.whl

torch_geometric/nn/models/glem.py

@@ -0,0 +1,384 @@
+from typing import List, Optional, Union
+
+import torch
+import torch.nn as nn
+from tqdm import tqdm
+
+from torch_geometric.loader import DataLoader, NeighborLoader
+from torch_geometric.nn.models import GraphSAGE, basic_gnn
+
+
+class GLEM(torch.nn.Module):
+    r"""This GNN+LM co-training model is based on GLEM from the `"Learning on
+    Large-scale Text-attributed Graphs via Variational Inference"
+    <https://arxiv.org/abs/2210.14709>`_ paper.
+
+    Args:
+        lm_to_use (str): A TextEncoder from huggingface model repo
+                with a classifier(default: TinyBERT)
+        gnn_to_use (torch_geometric.nn.models): (default: GraphSAGE)
+        out_channels (int): output channels for LM and GNN, should be same
+        num_gnn_heads Optional[int]: Number of heads for attention, if needed
+        num_gnn_layers (int): number of gnn layers
+        gnn_loss: loss function for gnn, (default: CrossEntropyLoss)
+        lm_loss: loss function for Language Model, (default: CrossEntropyLoss)
+        alpha (float): pseudo label weight of E-step, LM optimization,
+            (default: 0.5)
+        beta (float): pseudo label weight of M-step, GNN optimization,
+            (default: 0.5)
+        lm_dtype (torch.dtype): the data type once you load LM into memory,
+            (default: torch.bfloat16)
+        lm_use_lora (bool): choose if LM use Lora peft for fine tune,
+            (default: True)
+        lora_target_modules: The names of the target modules to apply the lora
+            adapter to, e.g. ['q_proj', 'v_proj'] for LLM , (default: None)
+
+    .. note::
+        See `examples/llm_plus_gnn/glem.py` for example usage.
+    """
+    def __init__(
+            self,
+            lm_to_use: str = 'prajjwal1/bert-tiny',
+            gnn_to_use: basic_gnn = GraphSAGE,
+            out_channels: int = 47,
+            gnn_loss=nn.CrossEntropyLoss(reduction='mean'),
+            lm_loss=nn.CrossEntropyLoss(reduction='mean'),
+            alpha: float = 0.5,
+            beta: float = 0.5,
+            lm_dtype: torch.dtype = torch.bfloat16,
+            lm_use_lora: bool = True,
+            lora_target_modules: Optional[Union[List[str], str]] = None,
+            device: Union[str, torch.device] = torch.device('cpu'),
+    ):
+        super().__init__()
+        self.device = device
+        self.lm_loss = lm_loss
+        self.gnn = gnn_to_use
+        self.gnn_loss = gnn_loss
+        self.alpha = alpha
+        self.beta = beta
+        self.gnn_loss = gnn_loss
+        self.lm = lm_to_use
+        from transformers import AutoModelForSequenceClassification
+        self.lm = AutoModelForSequenceClassification.from_pretrained(
+            lm_to_use, num_labels=out_channels, torch_dtype=lm_dtype,
+            offload_folder="offload", trust_remote_code=True)
+        if lm_use_lora:
+            from peft import (
+                LoraConfig,
+                TaskType,
+                get_peft_model,
+                prepare_model_for_kbit_training,
+            )
+            print("Training LM with LORA!")
+            self.lm = prepare_model_for_kbit_training(self.lm)
+            config = LoraConfig(task_type=TaskType.SEQ_CLS, r=16,
+                                lora_alpha=16, lora_dropout=0.05, bias="none",
+                                target_modules=lora_target_modules)
+            self.lm = get_peft_model(self.lm, config)
+            self.lm.print_trainable_parameters()
+        self.lm.config.pad_token_id = self.lm.config.eos_token_id
+        self.lm_device = self.lm.device
+
+        if self.lm.num_labels != self.gnn.out_channels:
+            raise ValueError('''The output channel of language model \
+                             and gnn should be the same''')
+
+    def pre_train_gnn(self, train_loader: NeighborLoader,
+                      optimizer: torch.optim.Optimizer, num_epochs: int,
+                      patience: int, ext_pseudo_labels: torch.Tensor = None,
+                      is_augmented: bool = False, verbose: bool = True):
+        # Pretrain GNN, optional steps if you do not have pseudo labels.
+        best_acc = 0
+        early_stopping = 0
+        # training only based on gold data
+        for epoch in range(0, num_epochs):
+            acc, loss = self.train_gnn(train_loader, optimizer, epoch,
+                                       ext_pseudo_labels, is_augmented,
+                                       verbose)
+            if acc < best_acc:
+                early_stopping += 1
+                if early_stopping > patience:
+                    print(f'Early stopped by Epoch: {epoch}, '
+                          f'Best acc: {best_acc}')
+                    break
+            best_acc = max(best_acc, acc)
+
+    def pre_train_lm(self, train_loader: DataLoader,
+                     optimizer: torch.optim.Optimizer, num_epochs: int,
+                     patience: int, ext_pseudo_labels: torch.Tensor = None,
+                     is_augmented: bool = False, verbose: bool = True):
+        # Pretrain language model
+        best_acc = 0
+        early_stopping = 0
+        for epoch in range(1, num_epochs + 1):
+            acc, loss = self.train_lm(train_loader, optimizer, epoch,
+                                      ext_pseudo_labels, is_augmented, verbose)
+            if acc < best_acc:
+                early_stopping += 1
+                if early_stopping > patience:
+                    print(f'Early stopped by Epoch: {epoch}, '
+                          f'Best acc: {best_acc}')
+                    break
+            best_acc = max(best_acc, acc)
+
+    def train(self, em_phase: str, train_loader: Union[DataLoader,
+                                                       NeighborLoader],
+              optimizer: torch.optim.Optimizer, pseudo_labels: torch.Tensor,
+              epoch: int, is_augmented: bool = False, verbose: bool = False):
+        r"""GLEM training step, EM steps.
+
+        Args:
+            em_phase(str): 'gnn' or 'lm' choose which phase you are training on
+            train_loader(Union[DataLoader, NeighborLoader]): use DataLoader for
+                lm training, include tokenized data, labels is_gold mask.
+                use NeighborLoader for gnn training, include x, edge_index.
+            optimizer (torch.optim.Optimizer): optimizer for training
+            pseudo_labels(torch.Tensor): the predicted labels used as pseudo
+                labels
+            epoch (int): current epoch
+            is_augmented (bool): will use pseudo_labels or not
+            verbose (bool): print training progress bar or not
+
+        Returns:
+            acc (float): training accuracy
+            loss (float): loss value
+        """
+        pseudo_labels = pseudo_labels.to(self.device)
+        if em_phase == 'gnn':
+            acc, loss = self.train_gnn(train_loader, optimizer, epoch,
+                                       pseudo_labels, is_augmented, verbose)
+        if em_phase == 'lm':
+            acc, loss = self.train_lm(train_loader, optimizer, epoch,
+                                      pseudo_labels, is_augmented, verbose)
+        return acc, loss
+
+    def train_lm(self, train_loader: DataLoader,
+                 optimizer: torch.optim.Optimizer, epoch: int,
+                 pseudo_labels: torch.Tensor = None,
+                 is_augmented: bool = False, verbose: bool = True):
+        r"""Language model Training in every epoch.
+
+        Args:
+            train_loader (loader.dataloader.DataLoader): text token dataloader
+            optimizer (torch.optim.Optimizer): model optimizer
+            epoch (int): current train epoch
+            pseudo_labels (torch.Tensor): 1-D tensor, predictions from gnn
+            is_augmented (bool): train with pseudo labels or not
+            verbose (bool): print training progress bar or not
+
+        Returns:
+            approx_acc (torch.tensor): training accuracy
+            loss (torch.float): loss value
+
+        """
+        all_out = []
+        total_loss = total_correct = 0
+        num_nodes = train_loader.dataset.indices.size(0)
+        self.lm.train()
+        if verbose:
+            pbar = tqdm(total=num_nodes)
+            pbar.set_description(f'Epoch {epoch:02d}')
+        for batch in train_loader:
+            inputs = {k: v.to(self.device) for k, v in batch['input'].items()}
+            out = self.lm(**inputs).logits
+            labels = batch['labels'].to(self.device).squeeze()
+            # training with pseudo labels or not
+            if is_augmented:
+                pl_batch = pseudo_labels[batch['n_id']].to(self.device)
+            else:
+                pl_batch = None
+            loss = self.loss(out, labels, self.lm_loss,
+                             batch['is_gold'].to(self.device), pl_batch,
+                             self.alpha, is_augmented)
+            loss.backward()
+            optimizer.step()
+            optimizer.zero_grad()
+            all_out.append(out)
+            total_correct += int(out.argmax(dim=-1).eq(labels).sum())
+            total_loss += float(loss)
+            if verbose:
+                pbar.update(batch['n_id'].size(0))
+
+        all_out = torch.cat(all_out, dim=0)
+        approx_acc = total_correct / num_nodes
+        loss = total_loss / len(train_loader)
+        if verbose:
+            pbar.close()
+        print(f'Epoch {epoch:02d} Loss: {loss:.4f} '
+              f'Approx. Train: {approx_acc:.4f}')
+        return approx_acc, loss
+
+    def train_gnn(self, train_loader: NeighborLoader,
+                  optimizer: torch.optim.Optimizer, epoch: int,
+                  pseudo_labels: torch.Tensor = None,
+                  is_augmented: bool = False, verbose: bool = True):
+        r"""GNN training step in every epoch.
+
+        Args:
+            train_loader (loader.NeighborLoader): gnn Neighbor node loader
+            optimizer (torch.optim.Optimizer): model optimizer
+            epoch (int): current train epoch
+            pseudo_labels(torch.tensor): 1-D tensor, predictions from lm
+            is_augmented(bool): use pseudo labeled node or not
+            verbose (bool): print training progress or not
+
+        Returns:
+            approx_acc (torch.tensor): training accuracy
+            loss (torch.float): loss value
+        """
+        self.gnn.train()
+        num_nodes = train_loader.input_nodes.size(0)
+        if verbose:
+            pbar = tqdm(total=num_nodes)
+            pbar.set_description(f'Epoch {epoch:02d}')
+        total_loss = total_correct = 0
+        all_out = []
+        for batch in train_loader:
+            batch = batch.to(self.device)
+            out = self.gnn(batch.x, batch.edge_index)[:batch.batch_size]
+            all_out.append(out)
+            labels = batch.y[:batch.batch_size].squeeze()
+            is_gold_batch = batch.is_gold[:batch.batch_size].squeeze()
+            # training with pseudo labels or not
+            if is_augmented and pseudo_labels is not None:
+                pl_batch = pseudo_labels[batch.n_id[:batch.batch_size]]
+            else:
+                pl_batch = None
+            loss = self.loss(out, labels, self.gnn_loss, is_gold_batch,
+                             pl_batch, self.beta, is_augmented)
+            loss.backward()
+            optimizer.step()
+            optimizer.zero_grad()
+            total_loss += float(loss)
+            total_correct += int(out.argmax(dim=-1).eq(labels).sum())
+            if verbose:
+                pbar.update(batch.batch_size)
+
+        all_out = torch.cat(all_out, dim=0)
+        loss = total_loss / len(train_loader)
+        approx_acc = total_correct / num_nodes
+        if verbose:
+            pbar.close()
+        print(f'Epoch: {epoch:02d} Loss: {loss:.4f} '
+              f'Approx. Train: {approx_acc:.4f}')
+        return approx_acc, loss
+
+    @torch.no_grad()
+    def inference(self, em_phase: str, data_loader: Union[NeighborLoader,
+                                                          DataLoader],
+                  verbose: bool = False):
+        r"""GLEM inference step.
+
+        Args:
+            em_phase(str): 'gnn' or 'lm'
+            data_loader(dataloader or Neighborloader):
+                dataloader: for lm training, include tokenized data
+                nodeloader: for gnn training, include x, edge_index
+            verbose(bool): print inference progress or not
+
+        Returns:
+            out (torch.Tensor): n * m tensor, m is number of classes,
+                n is number of nodes
+        """
+        out = None
+        if em_phase == 'gnn':
+            self.gnn.eval()
+            out = self.inference_gnn(data_loader, verbose)
+        elif em_phase == 'lm':
+            self.lm.eval()
+            out = self.inference_lm(data_loader, verbose)
+        return out
+
+    @torch.no_grad()
+    def inference_lm(self, data_loader: DataLoader, verbose: bool = True):
+        r"""LM inference step.
+
+        Args:
+            data_loader (Dataloader): include token, labels, and gold mask
+            verbose (bool): print progress bar or not
+
+        Returns:
+            preds (tensor): prediction from GNN, convert to pseudo labels
+                by preds.argmax(dim=-1).unsqueeze(1)
+        """
+        if verbose:
+            pbar = tqdm(total=data_loader.dataset._data.num_nodes)
+            pbar.set_description('LM inference stage')
+        self.lm.eval()
+        preds = []
+        for batch in data_loader:
+            inputs = {k: v.to(self.device) for k, v in batch['input'].items()}
+            logits = self.lm(**inputs).logits
+            preds.append(logits)
+            if verbose:
+                pbar.update(batch['n_id'].size(0))
+        if verbose:
+            pbar.close()
+        preds = torch.cat(preds)
+        return preds
+
+    @torch.no_grad()
+    def inference_gnn(self, data_loader: NeighborLoader, verbose: bool = True):
+        r"""GNN inference step.
+
+        Args:
+            data_loader(NeighborLoader): include x, edge_index,
+            verbose (bool): print progress bar or not
+
+        Returns:
+            preds (tensor): prediction from GNN,
+                convert to pseudo labels by preds.argmax(dim=-1).unsqueeze(1)
+        """
+        if verbose:
+            pbar = tqdm(total=data_loader.data.num_nodes)
+            pbar.set_description('GNN inference stage')
+        preds = []
+        self.gnn.eval()
+        for batch in data_loader:
+            batch = batch.to(self.device)
+            out = self.gnn(batch.x, batch.edge_index)[:batch.batch_size]
+            preds.append(out)
+            if verbose:
+                pbar.update(batch.batch_size)
+        if verbose:
+            pbar.close()
+        preds = torch.cat(preds, dim=0)
+        return preds
+
+    def loss(self, logits: torch.Tensor, labels: torch.Tensor,
+             loss_func: torch.nn.functional, is_gold: torch.Tensor,
+             pseudo_labels: torch.Tensor = None, pl_weight: float = 0.5,
+             is_augmented: bool = True):
+        r"""Core function of variational EM inference, this function is aming
+        on combining loss value on gold(original train) and loss value on
+        pseudo labels.
+
+        Reference:
+        <https://github.com/AndyJZhao/GLEM/blob/main/src/models/GLEM/GLEM_utils.py> # noqa
+
+        Args:
+            logits(torch.tensor): predict results from LM or GNN
+            labels(torch.tensor): combined node labels from ground truth and
+                pseudo labels(if provided)
+            loss_func(torch.nn.modules.loss): loss function for classification
+            is_gold(tensor): a tensor with bool value that mask ground truth
+                    label and during training, thus ~is_gold mask pseudo labels
+            pseudo_labels(torch.tensor): predictions from other model
+            pl_weight: the pseudo labels used in E-step and M-step optimization
+                        alpha in E-step, beta in M-step respectively
+            is_augmented: use EM or just train GNN and LM with gold data
+
+        """
+        def deal_nan(x):
+            return 0 if torch.isnan(x) else x
+
+        if is_augmented and (sum(~is_gold) > 0):
+            mle_loss = deal_nan(loss_func(logits[is_gold], labels[is_gold]))
+            # all other labels beside from ground truth(gold labels)
+            pseudo_label_loss = deal_nan(
+                loss_func(logits[~is_gold], pseudo_labels[~is_gold]))
+            loss = pl_weight * pseudo_label_loss + (1 - pl_weight) * mle_loss
+        else:
+            loss = loss_func(logits, labels)
+        return loss

torch_geometric/nn/models/molecule_gpt.py

@@ -0,0 +1,222 @@
+from typing import List, Optional
+
+import torch
+from torch import Tensor
+
+from torch_geometric.nn.attention import QFormer
+from torch_geometric.nn.nlp.llm import BOS, LLM, MAX_NEW_TOKENS
+from torch_geometric.utils import to_dense_batch
+
+
+def pad_or_truncate(embeddings: Tensor, max_seq_len: int,
+                    padding_value: int = 0) -> Tensor:
+    batch_size, current_seq_len, d = embeddings.size()
+
+    if current_seq_len > max_seq_len:
+        return embeddings[:, :max_seq_len, :]
+    elif current_seq_len < max_seq_len:
+        pad_tensor = torch.full((batch_size, max_seq_len - current_seq_len, d),
+                                padding_value, dtype=embeddings.dtype,
+                                device=embeddings.device)
+        return torch.cat([embeddings, pad_tensor], dim=1)
+    else:
+        return embeddings
+
+
+class MoleculeGPT(torch.nn.Module):
+    r"""The MoleculeGPT model from the `"MoleculeGPT: Instruction
+    Following Large Language Models for Molecular Property Prediction"
+    <https://ai4d3.github.io/papers/34.pdf>`_ paper.
+
+    Args:
+        llm (LLM): The LLM to use.
+        graph_encoder (torch.nn.Module): Encode 2D molecule graph.
+        smiles_encoder (torch.nn.Module): Encode 1D SMILES.
+        mlp_out_channels (int, optional): The size of each embedding
+            after qformer encoding. (default: :obj:`32`)
+        max_tokens (int, optional): Max output tokens of 1D/2D encoder.
+            (default: :obj:`20`)
+
+    .. warning::
+        This module has been tested with the following HuggingFace models
+
+        * :obj:`llm_to_use="lmsys/vicuna-7b-v1.5"`
+
+        and may not work with other models. See other models at `HuggingFace
+        Models <https://huggingface.co/models>`_ and let us know if you
+        encounter any issues.
+
+    .. note::
+        For an example of using :class:`MoleculeGPT`, see
+        `examples/llm/molecule_gpt.py <https://github.com/pyg-team/
+        pytorch_geometric/blob/master/examples/llm/molecule_gpt.py>`_.
+    """
+    def __init__(
+        self,
+        llm: LLM,
+        graph_encoder: torch.nn.Module,
+        smiles_encoder: torch.nn.Module,
+        mlp_out_channels: int = 32,
+        max_tokens: Optional[int] = 20,
+    ) -> None:
+        super().__init__()
+        self.llm = llm
+        self.graph_encoder = graph_encoder.to(self.llm.device)
+        self.smiles_encoder = smiles_encoder.to(self.llm.device)
+
+        self.graph_qformer = QFormer(
+            input_dim=self.graph_encoder.nn[-1].out_features,
+            hidden_dim=mlp_out_channels,
+            output_dim=mlp_out_channels,
+            num_heads=4,
+            num_layers=2,
+        ).to(self.llm.device)
+
+        self.smiles_qformer = QFormer(
+            input_dim=self.smiles_encoder.model.pooler.dense.out_features,
+            hidden_dim=mlp_out_channels,
+            output_dim=mlp_out_channels,
+            num_heads=4,
+            num_layers=2,
+        ).to(self.llm.device)
+
+        self.max_tokens = max_tokens
+
+        self.word_embedding = self.llm.word_embedding
+        self.llm_generator = self.llm.llm
+
+        # LLMs
+        in_dim = 2 * mlp_out_channels * max_tokens
+        out_dim = self.llm.llm.model.embed_tokens.embedding_dim
+        self.projector = torch.nn.Sequential(
+            torch.nn.Linear(in_dim, in_dim),
+            torch.nn.Sigmoid(),
+            torch.nn.Linear(in_dim, out_dim),
+        ).to(self.llm.device)
+
+    def encode(
+        self,
+        x: Tensor,
+        edge_index: Tensor,
+        batch: Tensor,
+        edge_attr: Optional[Tensor],
+        smiles: List[str],
+    ) -> Tensor:
+        batch_size = len(smiles)
+        # 2D Graph Branch: [bs, node_len, d]
+        x = x.to(self.llm.device)
+        edge_index = edge_index.to(self.llm.device)
+        if edge_attr is not None:
+            edge_attr = edge_attr.to(self.llm.device)
+        batch = batch.to(self.llm.device)
+
+        x_graph = self.graph_encoder(x, edge_index, edge_attr=edge_attr)
+        x_graph = to_dense_batch(x_graph, batch)[0]
+        out_graph = self.graph_qformer(x_graph)
+        out_graph = pad_or_truncate(out_graph, max_seq_len=self.max_tokens,
+                                    padding_value=0)
+        out_graph = out_graph.view(batch_size, -1)
+
+        # 1D SMILES Branch: [bs, seq_len, d]
+        x_smiles = self.smiles_encoder.encode(smiles,
+                                              output_device=self.llm.device)
+        out_smiles = self.smiles_qformer(x_smiles)
+        out_smiles = pad_or_truncate(out_smiles, max_seq_len=self.max_tokens,
+                                     padding_value=0)
+        out_smiles = out_smiles.view(batch_size, -1)
+
+        # Merge into LLMs
+        x_cat = torch.cat([out_graph, out_smiles], dim=1)
+        return x_cat
+
+    def forward(
+        self,
+        x: Tensor,
+        edge_index: Tensor,
+        batch: Tensor,
+        edge_attr: Optional[Tensor],
+        smiles: List[str],
+        instructions: List[str],
+        label: List[str],
+        additional_text_context: Optional[List[str]] = None,
+    ):
+        x = self.encode(x, edge_index, batch, edge_attr, smiles)
+        x = self.projector(x)
+        xs = x.split(1, dim=0)
+
+        batch_unique = batch.unique()
+        batch_size = len(instructions)
+        if len(batch_unique) < batch_size:
+            xs = [
+                xs[i] if i in batch_unique else None for i in range(batch_size)
+            ]
+
+        (
+            inputs_embeds,
+            attention_mask,
+            label_input_ids,
+        ) = self.llm._get_embeds(instructions, additional_text_context, xs,
+                                 label)
+
+        with self.llm.autocast_context:
+            outputs = self.llm_generator(
+                inputs_embeds=inputs_embeds,
+                attention_mask=attention_mask,
+                return_dict=True,
+                labels=label_input_ids,
+            )
+
+        return outputs.loss
+
+    @torch.no_grad()
+    def inference(
+        self,
+        x: Tensor,
+        edge_index: Tensor,
+        batch: Tensor,
+        edge_attr: Optional[Tensor],
+        smiles: List[str],
+        instructions: List[str],
+        additional_text_context: Optional[List[str]] = None,
+        max_out_tokens: Optional[int] = MAX_NEW_TOKENS,
+    ):
+        x = self.encode(x, edge_index, batch, edge_attr, smiles)
+        x = self.projector(x)
+        xs = x.split(1, dim=0)
+
+        # Handle questions without node features:
+        batch_unique = batch.unique()
+        batch_size = len(instructions)
+        if len(batch_unique) < batch_size:
+            xs = [
+                xs[i] if i in batch_unique else None for i in range(batch_size)
+            ]
+
+        inputs_embeds, attention_mask, _ = self.llm._get_embeds(
+            instructions, additional_text_context, xs)
+
+        bos_token = self.llm.tokenizer(
+            BOS,
+            add_special_tokens=False,
+        ).input_ids[0]
+
+        with self.llm.autocast_context:
+            outputs = self.llm_generator.generate(
+                inputs_embeds=inputs_embeds,
+                max_new_tokens=max_out_tokens,
+                attention_mask=attention_mask,
+                bos_token_id=bos_token,
+                use_cache=True  # Important to set!
+            )
+
+        return self.llm.tokenizer.batch_decode(
+            outputs,
+            skip_special_tokens=True,
+        )
+
+    def __repr__(self) -> str:
+        return (f'{self.__class__.__name__}(\n'
+                f'  llm={self.llm},\n'
+                f'  graph={self.graph_encoder.__class__.__name__},\n'
+                f'  smiles={self.smiles_encoder},\n'
+                f')')

torch_geometric/nn/nlp/llm.py

@@ -56,7 +56,7 @@ class LLM(torch.nn.Module):
         allocate the correct number of GPUs needed, given the available GPU
         memory of your GPUs.
     dtype (torch.dtype, optional): The data type to use for the LLM.
-        (default :obj: `torch.bloat16`)
+        (default :obj: `torch.bfloat16`)
     """
     def __init__(
         self,

torch_geometric/nn/nlp/sentence_transformer.py

@@ -10,6 +10,7 @@ class PoolingStrategy(Enum):
     MEAN = 'mean'
     LAST = 'last'
     CLS = 'cls'
+    LAST_HIDDEN_STATE = 'last_hidden_state'
 
 
 class SentenceTransformer(torch.nn.Module):
@@ -38,6 +39,8 @@ class SentenceTransformer(torch.nn.Module):
             emb = mean_pooling(emb, attention_mask)
         elif self.pooling_strategy == PoolingStrategy.LAST:
             emb = last_pooling(emb, attention_mask)
+        elif self.pooling_strategy == PoolingStrategy.LAST_HIDDEN_STATE:
+            emb = out.last_hidden_state
         else:
             assert self.pooling_strategy == PoolingStrategy.CLS
             emb = emb[:, 0, :]