4.2 文本相似度（三）

换个思路，再训练一次。

1 基本框架

试想，如果有一个语句需要从预料库中匹配，每一次匹配都会伴随着大量的耗时：

一次匹配20ms, 1 000 000次呢，1 000 000 *20/ 1000 = 20 000S ～5.56H。效率极其的低：

使用如下策略解决：候选文本与匹配文本分别进入模型（并行）然后输出两个向量，通过cos相似度得到是否匹配结果；

2  使用方法介绍

2.1 CosineSimilarity

   pytorch.nn.CosineSimilarity（余弦相似度）是一种常用的相似度度量方法，用于衡量两个向量之间的相似程度。在自然语言处理和信息检索等领域，余弦相似度常用于比较文本、向量表示或特征之间的相似性。

2.2 CosineEmbeddingLoss

3 代码

from transformers import AutoTokenizer, AutoModelForSequenceClassification, Trainer, TrainingArguments
from datasets import load_dataset,load_from_disk
import traceback
import torch

from sklearn.model_selection import train_test_split

#dataset = load_dataset("json", data_files="../data/train_pair_1w.json", split="train")
dataset = load_dataset("csv", data_files="/Users/user/studyFile/2024/nlp/text_similar/data/Chinese_Text_Similarity.csv", split="train")
datasets = dataset.train_test_split(test_size=0.2,shuffle=True)


tokenizer = AutoTokenizer.from_pretrained("../chinese_macbert_base")
def process_function(examples):
    sentences = []
    labels = []
    for sen1, sen2, label in zip(examples["sentence1"], examples["sentence2"], examples["label"]):
        sentences.append(sen1)
        sentences.append(sen2)
        labels.append(1 if int(label) == 1 else -1)
    # input_ids, attention_mask, token_type_ids
    tokenized_examples = tokenizer(sentences, max_length=128, truncation=True, padding="max_length")
    # tokenized_examples format likes {'input_ids':[[],[],[]], 'token_type_ids':[[],[],[]],'attention_mask':[[],[],[]]}
    # k : input_ids, v: [[],[],[],[]]
    
    tokenized_examples = {k: [v[i: i + 2] for i in range(0, len(v), 2)] for k, v in tokenized_examples.items()}
    tokenized_examples["labels"] = labels
    return tokenized_examples
    
tokenized_datasets = datasets.map(process_function, batched=True, remove_columns=datasets["train"].column_names)
tokenized_datasets


#搭建模型
from transformers import BertForSequenceClassification, BertPreTrainedModel, BertModel
from typing import Optional
from transformers.configuration_utils import PretrainedConfig
from torch.nn import CosineSimilarity, CosineEmbeddingLoss
import torch
class DualModel(BertPreTrainedModel):

    def __init__(self, config: PretrainedConfig, *inputs, **kwargs):
        super().__init__(config, *inputs, **kwargs)
        self.bert = BertModel(config)
        self.post_init()

    def forward(
        self,
        input_ids: Optional[torch.Tensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        token_type_ids: Optional[torch.Tensor] = None,
        position_ids: Optional[torch.Tensor] = None,
        head_mask: Optional[torch.Tensor] = None,
        inputs_embeds: Optional[torch.Tensor] = None,
        labels: Optional[torch.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ):
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        # Step1 分别获取sentenceA 和 sentenceB的输入
        senA_input_ids, senB_input_ids = input_ids[:, 0], input_ids[:, 1]
        senA_attention_mask, senB_attention_mask = attention_mask[:, 0], attention_mask[:, 1]
        senA_token_type_ids, senB_token_type_ids = token_type_ids[:, 0], token_type_ids[:, 1]

        # Step2 分别获取sentenceA 和 sentenceB的向量表示
        senA_outputs = self.bert(
            senA_input_ids,
            attention_mask=senA_attention_mask,
            token_type_ids=senA_token_type_ids,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )

        senA_pooled_output = senA_outputs[1]    # [batch, hidden]

        senB_outputs = self.bert(
            senB_input_ids,
            attention_mask=senB_attention_mask,
            token_type_ids=senB_token_type_ids,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )

        senB_pooled_output = senB_outputs[1]    # [batch, hidden]

        # step3 计算相似度

        cos = CosineSimilarity()(senA_pooled_output, senB_pooled_output)    # [batch, ]

        # step4 计算loss

        loss = None
        if labels is not None:
            loss_fct = CosineEmbeddingLoss(0.3)
            loss = loss_fct(senA_pooled_output, senB_pooled_output, labels)

        output = (cos,)
        return ((loss,) + output) if loss is not None else output
    
model = DualModel.from_pretrained("../chinese_macbert_base")

import evaluate
acc_metric = evaluate.load("./metric_accuracy.py")
f1_metirc = evaluate.load("./metric_f1.py")

def eval_metric(eval_predict):
    predictions, labels = eval_predict
    predictions = [int(p > 0.7) for p in predictions]
    labels = [int(l > 0) for l in labels]
    # predictions = predictions.argmax(axis=-1)
    acc = acc_metric.compute(predictions=predictions, references=labels)
    f1 = f1_metirc.compute(predictions=predictions, references=labels)
    acc.update(f1)
    return acc
# 模型参数
train_args = TrainingArguments(output_dir="./dual_model",      # 输出文件夹
                               per_device_train_batch_size=32,  # 训练时的batch_size
                               per_device_eval_batch_size=32,  # 验证时的batch_size
                               logging_steps=10,                # log 打印的频率
                               evaluation_strategy="epoch",     # 评估策略
                               save_strategy="epoch",           # 保存策略
                               save_total_limit=3,              # 最大保存数
                               learning_rate=2e-5,              # 学习率
                               weight_decay=0.01,               # weight_decay
                               metric_for_best_model="f1",      # 设定评估指标
                               load_best_model_at_end=True)     # 训练完成后加载最优模型

trainer = Trainer(model=model, 
                  args=train_args, 
                  train_dataset=tokenized_datasets["train"], 
                  eval_dataset=tokenized_datasets["test"], 
                  compute_metrics=eval_metric)
trainer.train()

# 预测样本

class SentenceSimilarityPipeline:

    def __init__(self, model, tokenizer) -> None:
        self.model = model.bert
        self.tokenizer = tokenizer
        self.device = model.device

    def preprocess(self, senA, senB):
        return self.tokenizer([senA, senB], max_length=128, truncation=True, return_tensors="pt", padding=True)

    def predict(self, inputs):
        inputs = {k: v.to(self.device) for k, v in inputs.items()}
        #print(inputs)
        return self.model(**inputs)[1]  # [2, 768]

    def postprocess(self, logits):
        cos = CosineSimilarity()(logits[None, 0, :], logits[None,1, :]).squeeze().cpu().item()
        return cos

    def __call__(self, senA, senB, return_vector=False):
        inputs = self.preprocess(senA, senB)
        logits = self.predict(inputs)
        result = self.postprocess(logits)
        if return_vector:
            return result, logits
        else:
            return result

pipe = SentenceSimilarityPipeline(model, tokenizer)

pipe("我喜欢北京", "北京是个真奇怪的地方，但是我喜欢", return_vector=True)

        

太耗时了，没有GPU.....