SFR-Embedding-Mistral是目前MTEB榜单非常靠前的模型,之前考虑到相关介绍也只有一个偏技术报告的博客,没有太多模型结构或者训练方法上的革新,所以不打算展开细说,但是最近MTEB出现了一个新的模型Linq-Embed-Mistral,跟SFR-Embedding-Mistral在训练技巧上存在诸多一致性,所以决定在这里展开介绍下SFR-Embedding-Mistral的技术...
from sentence_transformers import SentenceTransformer, util model = SentenceTransformer("Salesforce/SFR-Embedding-Mistral") def get_detailed_instruct(task_description: str, query: str) -> str: return f'Instruct: {task_description}\nQuery: {query}' # Each query must come with a one-sentence ins...
SFR-Embedding-Mistral / model.safetensors.index.json model.safetensors.index.json 21.63 KB 一键复制 编辑 原始数据 按行查看 历史 yeliu918 提交于 1年前 . model v1.0 12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667...
# load model and tokenizer tokenizer = AutoTokenizer.from_pretrained('Salesforce/SFR-Embedding-Mistral') # bnb_config = BitsAndBytesConfig( # load_in_4bit=True, # bnb_4bit_use_double_quant=True, # bnb_4bit_quant_type="nf4", # bnb_4bit_compute_dtype=torch.bfloat16 # ) # model =...
就在本周六 AI 开发者集结!和模力方舟相约贵州首届 AI 开发者大会,点击立即报名 Watch 1Star0Fork0 Hugging Face 模型镜像/SFR-Embedding-Mistral 代码Issues0Pull Requests0Wiki统计流水线 服务 Gitee Pages JavaDoc PHPDoc 质量分析 Jenkins for Gitee
Kaggle is the world’s largest data science community with powerful tools and resources to help you achieve your data science goals.