FEA: add XSimGCL

.github/workflows/python-package.yml

@@ -35,9 +35,9 @@ jobs:
       run: |
         python -m pip install --upgrade pip
         pip install pytest
-        pip install dgl
+        pip install dgl==0.9.1
         pip install torch==${{ matrix.torch-version}}+cpu -f https://download.pytorch.org/whl/torch_stable.html
-        pip install torch-scatter torch-sparse torch-cluster torch-spline-conv torch-geometric -f https://data.pyg.org/whl/torch-${{ matrix.torch-version }}+cpu.html
+        pip install torch-scatter==2.0.9 torch-sparse==0.6.15 torch-cluster==1.6.0 torch-spline-conv==1.2.1 torch-geometric -f https://data.pyg.org/whl/torch-${{ matrix.torch-version }}+cpu.html
         pip install recbole==1.1.1
         conda install -c conda-forge faiss-cpu
     # Use "python -m pytest" instead of "pytest" to fix imports

README.md

@@ -57,6 +57,7 @@ We list currently supported models according to category:
 * **[HMLET](recbole_gnn/model/general_recommender/hmlet.py)** from Kong *et al.*: [Linear, or Non-Linear, That is the Question!](https://arxiv.org/abs/2111.07265) (WSDM 2022).
 * **[NCL](recbole_gnn/model/general_recommender/ncl.py)** from Lin *et al.*: [Improving Graph Collaborative Filtering with Neighborhood-enriched Contrastive Learning](https://arxiv.org/abs/2202.06200) (TheWebConf 2022).
 * **[SimGCL](recbole_gnn/model/general_recommender/simgcl.py)** from Yu *et al.*: [Are Graph Augmentations Necessary? Simple Graph Contrastive Learning for Recommendation](https://arxiv.org/abs/2112.08679) (SIGIR 2022).
+* **[XSimGCL](recbole_gnn/model/general_recommender/xsimgcl.py)** from Yu *et al.*: [XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation](https://arxiv.org/abs/2209.02544) (TKDE 2023).
 
 **Sequential Recommendation**:
 

recbole_gnn/config.py

@@ -1,4 +1,5 @@
 import os
+import recbole
 from recbole.config.configurator import Config as RecBole_Config
 from recbole.utils import ModelType as RecBoleModelType
 
@@ -16,7 +17,8 @@ def __init__(self, model=None, dataset=None, config_file_list=None, config_dict=
             config_file_list (list of str): the external config file, it allows multiple config files, default is None.
             config_dict (dict): the external parameter dictionaries, default is None.
         """
-        self.compatibility_settings()
+        if recbole.__version__ == "1.1.1":
+            self.compatibility_settings()
         super(Config, self).__init__(model, dataset, config_file_list, config_dict)
 
     def compatibility_settings(self):
@@ -59,7 +61,7 @@ def _get_model_and_dataset(self, model, dataset):
             final_dataset = dataset
 
         return final_model, final_model_class, final_dataset
-    
+
     def _load_internal_config_dict(self, model, model_class, dataset):
         super()._load_internal_config_dict(model, model_class, dataset)
         current_path = os.path.dirname(os.path.realpath(__file__))
@@ -75,4 +77,4 @@ def _load_internal_config_dict(self, model, model_class, dataset):
         if self.internal_config_dict['MODEL_TYPE'] == RecBoleModelType.SEQUENTIAL:
             self._update_internal_config_dict(sequential_base_init)
         if self.internal_config_dict['MODEL_TYPE'] == ModelType.SOCIAL:
-            self._update_internal_config_dict(social_base_init)
+            self._update_internal_config_dict(social_base_init)

recbole_gnn/model/general_recommender/__init__.py

@@ -3,3 +3,5 @@
 from recbole_gnn.model.general_recommender.ncl import NCL
 from recbole_gnn.model.general_recommender.ngcf import NGCF
 from recbole_gnn.model.general_recommender.sgl import SGL
+from recbole_gnn.model.general_recommender.simgcl import SimGCL
+from recbole_gnn.model.general_recommender.xsimgcl import XSimGCL

recbole_gnn/model/general_recommender/xsimgcl.py

@@ -0,0 +1,90 @@
+# -*- coding: utf-8 -*-
+r"""
+XSimGCL
+################################################
+Reference:
+    Junliang Yu, Xin Xia, Tong Chen, Lizhen Cui, Nguyen Quoc Viet Hung, Hongzhi Yin. "XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation" in TKDE 2023.
+
+Reference code:
+    https://github.com/Coder-Yu/SELFRec/blob/main/model/graph/XSimGCL.py
+"""
+
+
+import torch
+import torch.nn.functional as F
+
+from recbole_gnn.model.general_recommender import LightGCN
+
+
+class XSimGCL(LightGCN):
+    def __init__(self, config, dataset):
+        super(XSimGCL, self).__init__(config, dataset)
+
+        self.cl_rate = config['lambda']
+        self.eps = config['eps']
+        self.temperature = config['temperature']
+        self.layer_cl = config['layer_cl']
+
+    def forward(self, perturbed=False):
+        all_embs = self.get_ego_embeddings()
+        all_embs_cl = all_embs
+        embeddings_list = []
+
+        for layer_idx in range(self.n_layers):
+            all_embs = self.gcn_conv(all_embs, self.edge_index, self.edge_weight)
+            if perturbed:
+                random_noise = torch.rand_like(all_embs, device=all_embs.device)
+                all_embs = all_embs + torch.sign(all_embs) * F.normalize(random_noise, dim=-1) * self.eps
+            embeddings_list.append(all_embs)
+            if layer_idx == self.layer_cl - 1:
+                all_embs_cl = all_embs
+        lightgcn_all_embeddings = torch.stack(embeddings_list, dim=1)
+        lightgcn_all_embeddings = torch.mean(lightgcn_all_embeddings, dim=1)
+
+        user_all_embeddings, item_all_embeddings = torch.split(lightgcn_all_embeddings, [self.n_users, self.n_items])
+        user_all_embeddings_cl, item_all_embeddings_cl = torch.split(all_embs_cl, [self.n_users, self.n_items])
+        if perturbed:
+            return user_all_embeddings, item_all_embeddings, user_all_embeddings_cl, item_all_embeddings_cl
+        return user_all_embeddings, item_all_embeddings
+
+    def calculate_cl_loss(self, x1, x2):
+        x1, x2 = F.normalize(x1, dim=-1), F.normalize(x2, dim=-1)
+        pos_score = (x1 * x2).sum(dim=-1)
+        pos_score = torch.exp(pos_score / self.temperature)
+        ttl_score = torch.matmul(x1, x2.transpose(0, 1))
+        ttl_score = torch.exp(ttl_score / self.temperature).sum(dim=1)
+        return -torch.log(pos_score / ttl_score).mean()
+
+    def calculate_loss(self, interaction):
+        # clear the storage variable when training
+        if self.restore_user_e is not None or self.restore_item_e is not None:
+            self.restore_user_e, self.restore_item_e = None, None
+
+        user = interaction[self.USER_ID]
+        pos_item = interaction[self.ITEM_ID]
+        neg_item = interaction[self.NEG_ITEM_ID]
+
+        user_all_embeddings, item_all_embeddings, user_all_embeddings_cl, item_all_embeddings_cl = self.forward(perturbed=True)
+        u_embeddings = user_all_embeddings[user]
+        pos_embeddings = item_all_embeddings[pos_item]
+        neg_embeddings = item_all_embeddings[neg_item]
+
+        # calculate BPR Loss
+        pos_scores = torch.mul(u_embeddings, pos_embeddings).sum(dim=1)
+        neg_scores = torch.mul(u_embeddings, neg_embeddings).sum(dim=1)
+        mf_loss = self.mf_loss(pos_scores, neg_scores)
+
+        # calculate regularization Loss
+        u_ego_embeddings = self.user_embedding(user)
+        pos_ego_embeddings = self.item_embedding(pos_item)
+        neg_ego_embeddings = self.item_embedding(neg_item)
+        reg_loss = self.reg_loss(u_ego_embeddings, pos_ego_embeddings, neg_ego_embeddings, require_pow=self.require_pow)
+
+        user = torch.unique(interaction[self.USER_ID])
+        pos_item = torch.unique(interaction[self.ITEM_ID])
+
+        # calculate CL Loss
+        user_cl_loss = self.calculate_cl_loss(user_all_embeddings[user], user_all_embeddings_cl[user])
+        item_cl_loss = self.calculate_cl_loss(item_all_embeddings[pos_item], item_all_embeddings_cl[pos_item])
+
+        return mf_loss, self.reg_weight * reg_loss, self.cl_rate * (user_cl_loss + item_cl_loss)

recbole_gnn/properties/model/XSimGCL.yaml

@@ -0,0 +1,9 @@
+embedding_size: 64
+n_layers: 2
+reg_weight: 0.0001
+
+lambda: 0.1
+eps: 0.2
+temperature: 0.2
+layer_cl: 1
+require_pow: True

results/general/ml-1m.md

@@ -46,16 +46,17 @@ embedding_size: 64
 
 # Evaluation Results
 
-| Method               | Recall@10 | MRR@10 | NDCG@10 | Hit@10 | Precision@10 |
-| -------------------- | --------- | ------ | ------- | ------ | ------------ |
-| **BPR**              | 0.1776    | 0.4187 | 0.2401  | 0.7199 | 0.1779       |
-| **NeuMF**            | 0.1651    | 0.4020 | 0.2271  | 0.7029 | 0.1700       |
-| **NGCF**             | 0.1814    | 0.4354 | 0.2508  | 0.7239 | 0.1850       |
-| **LightGCN**         | 0.1861    | 0.4388 | 0.2538  | 0.7330 | 0.1863       |
-| **SGL**              | 0.1889    | 0.4315 | 0.2505  | 0.7392 | 0.1843       |
-| **HMLET**            | 0.1847    | 0.4297 | 0.2490  | 0.7305 | 0.1836       |
-| **NCL**              | 0.2021    | 0.4599 | 0.2702  | 0.7565 | 0.1962       |
-| **SimGCL**           | 0.2029    | 0.4550 | 0.2667  | 0.7640 | 0.1933       |
+| Method       | Recall@10 | MRR@10 | NDCG@10 | Hit@10 | Precision@10 |
+| ------------ | --------- | ------ | ------- | ------ | ------------ |
+| **BPR**      | 0.1776    | 0.4187 | 0.2401  | 0.7199 | 0.1779       |
+| **NeuMF**    | 0.1651    | 0.4020 | 0.2271  | 0.7029 | 0.1700       |
+| **NGCF**     | 0.1814    | 0.4354 | 0.2508  | 0.7239 | 0.1850       |
+| **LightGCN** | 0.1861    | 0.4388 | 0.2538  | 0.7330 | 0.1863       |
+| **SGL**      | 0.1889    | 0.4315 | 0.2505  | 0.7392 | 0.1843       |
+| **HMLET**    | 0.1847    | 0.4297 | 0.2490  | 0.7305 | 0.1836       |
+| **NCL**      | 0.2021    | 0.4599 | 0.2702  | 0.7565 | 0.1962       |
+| **SimGCL**   | 0.2029    | 0.4550 | 0.2667  | 0.7640 | 0.1933       |
+| **XSimGCL**  | 0.2116    | 0.4638 | 0.2750  | 0.7743 | 0.1987       |
 
 # Hyper-parameters
 
@@ -69,3 +70,4 @@ embedding_size: 64
 | **HMLET**    | learning_rate=0.002<br />n_layers=4<br />activation_function=leakyrelu | learning_rate choice [0.002, 0.001, 0.0005]<br/>n_layers choice [3, 4]<br/>activation_function choice ['elu', 'leakyrelu'] |
 | **NCL**      | learning_rate=0.002<br />n_layers=3<br />reg_weight=0.0001<br />ssl_temp=0.1<br />ssl_reg=1e-06<br />hyper_layers=1<br />alpha=1.5 | learning_rate choice [0.002]<br/>n_layers choice [3]<br/>reg_weight choice [1e-4]<br/>ssl_temp choice [0.1, 0.05]<br/>ssl_reg choice [1e-7, 1e-6]<br/>hyper_layers choice [1]<br/>alpha choice [1, 0.8, 1.5] |
 | **SimGCL**      | learning_rate=0.002<br />n_layers=2<br />reg_weight=0.0001<br />temperature=0.05<br />lambda=1e-5<br />eps=0.1 | learning_rate choice [0.002]<br/>n_layers choice [2, 3]<br/>reg_weight choice [1e-4]<br/>temperature choice [0.05, 0.1, 0.2]<br/>lambda choice [1e-5, 1e-6, 1e-7, 0.005, 0.01, 0.05]<br/>eps choice [0.1, 0.2] |
+| **XSimGCL** | learning_rate=0.002<br />n_layers=2<br />reg_weight=0.0001<br />temperature=0.2<br />lambda=0.1<br />eps=0.2<br />layer_cl=1 | learning_rate choice [0.002]<br/>n_layers choice [2, 3]<br/>reg_weight choice [1e-4]<br/>temperature choice [0.05, 0.1, 0.2]<br/>lambda choice [1e-5, 1e-6, 1e-7, 1e-4, 0.005, 0.01, 0.05, 0.1]<br/>eps choice [0.1, 0.2]<br/>layer_cl choice [1] |

tests/test_model.py

@@ -60,6 +60,12 @@ def test_simgcl(self):
             'model': 'SimGCL'
         }
         quick_test(config_dict)
+
+    def test_xsimgcl(self):
+        config_dict = {
+            'model': 'XSimGCL'
+        }
+        quick_test(config_dict)
 
 
 class TestSequentialRecommender(unittest.TestCase):