🚀 bert-base-dutch-cased-snli
bert-base-dutch-cased-snli是一个句子转换器模型,它可以将句子和段落映射到768维的密集向量空间,可用于聚类或语义搜索等任务。
🚀 快速开始
安装依赖
使用该模型前,你需要安装sentence-transformers
库:
pip install -U sentence-transformers
基本使用
以下是使用该模型的基本示例:
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('bert-base-dutch-cased-snli')
embeddings = model.encode(sentences)
print(embeddings)
高级使用
如果你没有安装sentence-transformers
库,也可以使用以下代码示例:
from transformers import AutoTokenizer, AutoModel
import torch
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0]
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
sentences = ['This is an example sentence', 'Each sentence is converted']
tokenizer = AutoTokenizer.from_pretrained('bert-base-dutch-cased-snli')
model = AutoModel.from_pretrained('bert-base-dutch-cased-snli')
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
with torch.no_grad():
model_output = model(**encoded_input)
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
📚 详细文档
评估结果
要对该模型进行自动评估,请参考句子嵌入基准测试:https://seb.sbert.net
训练参数
该模型的训练参数如下:
数据加载器
sentence_transformers.datasets.NoDuplicatesDataLoader.NoDuplicatesDataLoader
,长度为4807,参数如下:
{'batch_size': 64}
损失函数
sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss
,参数如下:
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
训练方法参数
{
"callback": null,
"epochs": 1,
"evaluation_steps": 0,
"evaluator": "utils.CombEvaluator",
"max_grad_norm": 1,
"optimizer_class": "<class 'transformers.optimization.AdamW'>",
"optimizer_params": {
"lr": 1e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 722,
"weight_decay": 0.01
}
完整模型架构
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
引用与作者
如需了解更多信息,请参考相关文档。