nextrec 0.1.10__py3-none-any.whl → 0.2.1__py3-none-any.whl

nextrec/loss/listwise.py

@@ -0,0 +1,6 @@
+"""
+Loss functions for listwise tasks
+
+Date: create on 22/11/2025
+Author: Yang Zhou,zyaztec@gmail.com
+"""

nextrec/loss/loss_utils.py

@@ -2,8 +2,7 @@
 Loss utilities for NextRec
 
 Date: create on 09/11/2025
-Author:
-    Yang Zhou,zyaztec@gmail.com
+Author: Yang Zhou,zyaztec@gmail.com
 """
 import torch
 import torch.nn as nn

nextrec/loss/match_losses.py

@@ -2,8 +2,7 @@
 Loss functions for matching tasks
 
 Date: create on 13/11/2025
-Author:
-    Yang Zhou,zyaztec@gmail.com
+Author: Yang Zhou,zyaztec@gmail.com
 """
 
 import torch
@@ -222,9 +221,9 @@ class ListMLELoss(nn.Module):
 
 class ApproxNDCGLoss(nn.Module):
     """
-    Approximate NDCG loss for learning to rank
-    Reference: Qin et al. A General Approximation Framework for Direct Optimization of 
-               Information Retrieval Measures (Information Retrieval 2010)
+    Approximate NDCG loss for learning to rank.
+    Reference: Qin et al. "A General Approximation Framework for Direct Optimization of
+    Information Retrieval Measures" (Information Retrieval 2010).
     """
     def __init__(self, temperature: float = 1.0, reduction: str = 'mean'):
         super(ApproxNDCGLoss, self).__init__()

nextrec/loss/pairwise.py

@@ -0,0 +1,6 @@
+"""
+Loss functions for pairwise tasks
+
+Date: create on 22/11/2025
+Author: Yang Zhou,zyaztec@gmail.com
+"""

nextrec/loss/pointwise.py

@@ -0,0 +1,6 @@
+"""
+Loss functions for pointwise tasks
+
+Date: create on 22/11/2025
+Author: Yang Zhou,zyaztec@gmail.com
+"""

nextrec/models/match/dssm.py

@@ -48,7 +48,7 @@ class DSSM(BaseMatchModel):
                  embedding_l2_reg: float = 0.0,
                  dense_l2_reg: float = 0.0,
                  early_stop_patience: int = 20,
-                 model_id: str = 'dssm'):
+                 **kwargs):
         
         super(DSSM, self).__init__(
             user_dense_features=user_dense_features,
@@ -67,7 +67,7 @@ class DSSM(BaseMatchModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=early_stop_patience,
-            model_id=model_id
+            **kwargs
         )
         
         self.embedding_dim = embedding_dim

nextrec/models/match/dssm_v2.py

@@ -44,7 +44,7 @@ class DSSM_v2(BaseMatchModel):
                  embedding_l2_reg: float = 0.0,
                  dense_l2_reg: float = 0.0,
                  early_stop_patience: int = 20,
-                 model_id: str = 'dssm_v2'):
+                 **kwargs):
         
         super(DSSM_v2, self).__init__(
             user_dense_features=user_dense_features,
@@ -63,7 +63,7 @@ class DSSM_v2(BaseMatchModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=early_stop_patience,
-            model_id=model_id
+            **kwargs
         )
         
         self.embedding_dim = embedding_dim

nextrec/models/match/mind.py

@@ -51,7 +51,7 @@ class MIND(BaseMatchModel):
                  embedding_l2_reg: float = 0.0,
                  dense_l2_reg: float = 0.0,
                  early_stop_patience: int = 20,
-                 model_id: str = 'mind'):
+                 **kwargs):
         
         super(MIND, self).__init__(
             user_dense_features=user_dense_features,
@@ -70,7 +70,7 @@ class MIND(BaseMatchModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=early_stop_patience,
-            model_id=model_id
+            **kwargs
         )
         
         self.embedding_dim = embedding_dim

nextrec/models/match/sdm.py

@@ -52,7 +52,7 @@ class SDM(BaseMatchModel):
                  embedding_l2_reg: float = 0.0,
                  dense_l2_reg: float = 0.0,
                  early_stop_patience: int = 20,
-                 model_id: str = 'sdm'):
+                 **kwargs):
         
         super(SDM, self).__init__(
             user_dense_features=user_dense_features,
@@ -71,7 +71,7 @@ class SDM(BaseMatchModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=early_stop_patience,
-            model_id=model_id
+            **kwargs
         )
         
         self.embedding_dim = embedding_dim

nextrec/models/match/youtube_dnn.py

@@ -50,7 +50,7 @@ class YoutubeDNN(BaseMatchModel):
                  embedding_l2_reg: float = 0.0,
                  dense_l2_reg: float = 0.0,
                  early_stop_patience: int = 20,
-                 model_id: str = 'youtube_dnn'):
+                 **kwargs):
         
         super(YoutubeDNN, self).__init__(
             user_dense_features=user_dense_features,
@@ -69,7 +69,7 @@ class YoutubeDNN(BaseMatchModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=early_stop_patience,
-            model_id=model_id
+            **kwargs
         )
         
         self.embedding_dim = embedding_dim

nextrec/models/multi_task/esmm.py

@@ -47,11 +47,11 @@ class ESMM(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | list[str | nn.Module] | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         # ESMM requires exactly 2 targets: ctr and ctcvr
         if len(target) != 2:
@@ -69,7 +69,7 @@ class ESMM(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/multi_task/mmoe.py

@@ -45,11 +45,11 @@ class MMOE(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | list[str | nn.Module] | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(MMOE, self).__init__(
             dense_features=dense_features,
@@ -63,7 +63,7 @@ class MMOE(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/multi_task/ple.py

@@ -48,11 +48,11 @@ class PLE(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | list[str | nn.Module] | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(PLE, self).__init__(
             dense_features=dense_features,
@@ -66,7 +66,7 @@ class PLE(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/multi_task/share_bottom.py

@@ -36,11 +36,11 @@ class ShareBottom(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | list[str | nn.Module] | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(ShareBottom, self).__init__(
             dense_features=dense_features,
@@ -54,7 +54,7 @@ class ShareBottom(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/afm.py

@@ -35,11 +35,10 @@ class AFM(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4, **kwargs):
         
         super(AFM, self).__init__(
             dense_features=dense_features,
@@ -53,7 +52,7 @@ class AFM(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/autoint.py

@@ -40,11 +40,11 @@ class AutoInt(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(AutoInt, self).__init__(
             dense_features=dense_features,
@@ -58,7 +58,7 @@ class AutoInt(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/dcn.py

@@ -36,11 +36,11 @@ class DCN(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(DCN, self).__init__(
             dense_features=dense_features,
@@ -54,7 +54,7 @@ class DCN(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/deepfm.py

@@ -32,11 +32,10 @@ class DeepFM(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4, **kwargs):
         
         super(DeepFM, self).__init__(
             dense_features=dense_features,
@@ -50,7 +49,7 @@ class DeepFM(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/dien.py

@@ -39,11 +39,11 @@ class DIEN(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(DIEN, self).__init__(
             dense_features=dense_features,
@@ -57,7 +57,7 @@ class DIEN(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/din.py

@@ -38,11 +38,11 @@ class DIN(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(DIN, self).__init__(
             dense_features=dense_features,
@@ -56,7 +56,7 @@ class DIN(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/fibinet.py

@@ -43,11 +43,11 @@ class FiBiNET(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(FiBiNET, self).__init__(
             dense_features=dense_features,
@@ -61,7 +61,7 @@ class FiBiNET(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/fm.py

@@ -31,11 +31,11 @@ class FM(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(FM, self).__init__(
             dense_features=dense_features,
@@ -49,7 +49,7 @@ class FM(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/masknet.py

@@ -37,11 +37,11 @@ class MaskNet(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(MaskNet, self).__init__(
             dense_features=dense_features,
@@ -55,7 +55,7 @@ class MaskNet(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/pnn.py

@@ -35,11 +35,11 @@ class PNN(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(PNN, self).__init__(
             dense_features=dense_features,
@@ -53,7 +53,7 @@ class PNN(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/widedeep.py

@@ -35,11 +35,11 @@ class WideDeep(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(WideDeep, self).__init__(
             dense_features=dense_features,
@@ -53,7 +53,7 @@ class WideDeep(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/models/ranking/xdeepfm.py

@@ -38,11 +38,11 @@ class xDeepFM(BaseModel):
                  optimizer_params: dict = {},
                  loss: str | nn.Module | None = "bce",
                  device: str = 'cpu',
-                 model_id: str = "baseline",
                  embedding_l1_reg=1e-6,
                  dense_l1_reg=1e-5,
                  embedding_l2_reg=1e-5,
-                 dense_l2_reg=1e-4):
+                 dense_l2_reg=1e-4,
+                 **kwargs):
         
         super(xDeepFM, self).__init__(
             dense_features=dense_features,
@@ -56,7 +56,7 @@ class xDeepFM(BaseModel):
             embedding_l2_reg=embedding_l2_reg,
             dense_l2_reg=dense_l2_reg,
             early_stop_patience=20,
-            model_id=model_id
+            **kwargs
         )
 
         self.loss = loss

nextrec/utils/__init__.py

@@ -1,18 +1,14 @@
-from nextrec.utils.optimizer import get_optimizer_fn, get_scheduler_fn
-from nextrec.utils.initializer import get_initializer_fn
-from nextrec.utils.embedding import get_auto_embedding_dim
-from nextrec.utils.common import get_task_type
-
-from nextrec.utils import optimizer, initializer, embedding, common
+from .optimizer import get_optimizer_fn, get_scheduler_fn
+from .initializer import get_initializer_fn
+from .embedding import get_auto_embedding_dim
+from . import optimizer, initializer, embedding
 
 __all__ = [
     'get_optimizer_fn',
     'get_scheduler_fn',
     'get_initializer_fn',
     'get_auto_embedding_dim',
-    'get_task_type',
     'optimizer',
     'initializer',
     'embedding',
-    'common',
 ]

nextrec/utils/embedding.py

@@ -2,8 +2,7 @@
 Embedding utilities for NextRec
 
 Date: create on 13/11/2025
-Author:
-    Yang Zhou, zyaztec@gmail.com
+Author: Yang Zhou, zyaztec@gmail.com
 """
 
 import numpy as np
@@ -13,7 +12,6 @@ def get_auto_embedding_dim(num_classes: int) -> int:
     """
     Calculate the dim of embedding vector according to number of classes in the category.
     Formula: emb_dim = [6 * (num_classes)^(1/4)]
-    Reference: 
-        Deep & Cross Network for Ad Click Predictions.(ADKDD'17)
+    Reference: Deep & Cross Network for Ad Click Predictions.(ADKDD'17)
     """
     return int(np.floor(6 * np.power(num_classes, 0.25)))

nextrec/utils/initializer.py

@@ -2,8 +2,7 @@
 Initialization utilities for NextRec
 
 Date: create on 13/11/2025
-Author:
-    Yang Zhou, zyaztec@gmail.com
+Author: Yang Zhou, zyaztec@gmail.com
 """
 
 import torch.nn as nn

nextrec/utils/optimizer.py

@@ -2,8 +2,7 @@
 Optimizer and Scheduler utilities for NextRec
 
 Date: create on 13/11/2025
-Author:
-    Yang Zhou, zyaztec@gmail.com
+Author: Yang Zhou, zyaztec@gmail.com
 """
 
 import torch

{nextrec-0.1.10.dist-info → nextrec-0.2.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nextrec
-Version: 0.1.10
+Version: 0.2.1
 Summary: A comprehensive recommendation library with match, ranking, and multi-task learning models
 Project-URL: Homepage, https://github.com/zerolovesea/NextRec
 Project-URL: Repository, https://github.com/zerolovesea/NextRec
@@ -61,9 +61,9 @@ Description-Content-Type: text/markdown
 ![Python](https://img.shields.io/badge/Python-3.10+-blue.svg)
 ![PyTorch](https://img.shields.io/badge/PyTorch-1.10+-ee4c2c.svg)
 ![License](https://img.shields.io/badge/License-Apache%202.0-green.svg)
-![Version](https://img.shields.io/badge/Version-0.1.0-orange.svg)
+![Version](https://img.shields.io/badge/Version-0.2.1-orange.svg)
 
-[中文版](README_zh.md)
+English | [中文版](README_zh.md)
 
 **A Unified, Efficient, and Scalable Recommendation System Framework**
 
@@ -123,7 +123,7 @@ model = DeepFM(
     mlp_params={"dims": [256, 128], "activation": "relu", "dropout": 0.5},
     target=target,
     device='cpu',
-    model_id="deepfm_with_processor",
+    session_id="deepfm_with_processor",
     embedding_l1_reg=1e-6,
     dense_l1_reg=1e-5,
     embedding_l2_reg=1e-5,
@@ -269,7 +269,6 @@ NextRec is inspired by the following great open-source projects:
 - **torch-rechub** - A Lighting Pytorch Framework for Recommendation Models, Easy-to-use and Easy-to-extend.
 - **FuxiCTR** — Configurable and reproducible CTR prediction library  
 - **RecBole** — Unified and efficient recommendation library  
-- **PaddleRec** — Large-scale recommendation algorithm library  
 
 Special thanks to all open-source contributors!