Shortcuts

TransR-FB15K237-single-gpu || TransR-FB15K237-single-gpu-wandb || TransR-FB15K237-single-gpu-hpo || TransR-FB15K237-multigpu

TransR-FB15K237-single-gpu

这一部分介绍如何用一个 GPU 在 FB15K237 知识图谱上训练 TransR [LLS+15]

导入数据

pybind11-OpenKE 有两个工具用于导入数据: pybind11_ke.data.TrainDataLoaderpybind11_ke.data.TestDataLoader

from pybind11_ke.config import Trainer, Tester
from pybind11_ke.module.model import TransE, TransR
from pybind11_ke.module.loss import MarginLoss
from pybind11_ke.module.strategy import NegativeSampling
from pybind11_ke.data import TrainDataLoader, TestDataLoader

pybind11-KE 提供了很多数据集,它们很多都是 KGE 原论文发表时附带的数据集。 pybind11_ke.data.TrainDataLoader 包含 in_path 用于传递数据集目录。

# dataloader for training
train_dataloader = TrainDataLoader(
    in_path = "../../benchmarks/FB15K237/",
    nbatches = 100,
    threads = 8,
    sampling_mode = "normal",
    bern = True,
    neg_ent = 25,
    neg_rel = 0)

导入模型

pybind11-OpenKE 提供了很多 KGE 模型,它们都是目前最常用的基线模型。我们首先导入 pybind11_ke.module.model.TransE,它是最简单的平移模型, 因为为了避免过拟合,pybind11_ke.module.model.TransR 实体和关系的嵌入向量初始化为 pybind11_ke.module.model.TransE 的结果。

# define the transe
transe = TransE(
    ent_tol = train_dataloader.get_ent_tol(),
    rel_tol = train_dataloader.get_rel_tol(),
    dim = 100,
    p_norm = 1,
    norm_flag = True)

下面导入 pybind11_ke.module.model.TransR 模型, 是一个为实体和关系嵌入向量分别构建了独立的向量空间,将实体向量投影到特定的关系向量空间进行平移操作的模型。

transr = TransR(
    ent_tol = train_dataloader.get_ent_tol(),
    rel_tol = train_dataloader.get_rel_tol(),
    dim_e = 100,
    dim_r = 100,
    p_norm = 1,
    norm_flag = True,
    rand_init = False)

损失函数

我们这里使用了 TransE [BUGD+13] 原论文使用的损失函数:pybind11_ke.module.loss.MarginLosspybind11_ke.module.strategy.NegativeSamplingpybind11_ke.module.loss.MarginLoss 进行了封装,加入权重衰减等额外项。

model_e = NegativeSampling(
    model = transe,
    loss = MarginLoss(margin = 5.0),
    batch_size = train_dataloader.get_batch_size()
)

model_r = NegativeSampling(
    model = transr,
    loss = MarginLoss(margin = 4.0),
    batch_size = train_dataloader.get_batch_size()
)

训练模型

pybind11-OpenKE 将训练循环包装成了 pybind11_ke.config.Trainer, 可以运行它的 pybind11_ke.config.Trainer.run() 函数进行模型学习; 也可以通过传入 pybind11_ke.config.Tester, 使得训练器能够在训练过程中评估模型;pybind11_ke.config.Tester 使用 pybind11_ke.data.TestDataLoader 作为数据采样器。

# pretrain transe
trainer = Trainer(model = model_e, data_loader = train_dataloader,
    epochs = 1, lr = 0.5, opt_method = "sgd", use_gpu = True, device = 'cuda:1')
trainer.run()
parameters = transe.get_parameters()
transe.save_parameters("../../checkpoint/transr_transe.json")

# dataloader for test
test_dataloader = TestDataLoader("../../benchmarks/FB15K237/")

# test the transr
tester = Tester(model = transr, data_loader = test_dataloader, use_gpu = True, device = 'cuda:1')

# train transr
transr.set_parameters(parameters)
trainer = Trainer(model = model_r, data_loader = train_dataloader,
    epochs = 1000, lr = 1.0, opt_method = "sgd", use_gpu = True, device = 'cuda:1',
    tester = tester, test = True, valid_interval = 10,
    log_interval = 10, save_interval = 10, save_path = '../../checkpoint/transr.pth')
trainer.run()

# test the model
transr.load_checkpoint('../../checkpoint/transr.pth')
tester.set_sampling_mode("link_test")
tester.run_link_prediction()

Total running time of the script: ( 0 minutes 0.000 seconds)

Gallery generated by Sphinx-Gallery

Docs

Access comprehensive developer documentation for Pybind11-OpenKE

View Docs