--- title: Zhihui_LLM_Embedding canonical_url: "https://www.modelscope.cn/models/QwenCollection/Zhihui_LLM_Embedding" md_url: "https://www.modelscope.cn/models/QwenCollection/Zhihui_LLM_Embedding.md" repository: QwenCollection/Zhihui_LLM_Embedding last_updated: 2024-11-21 license: cc-by-nc-4.0 pipeline_tag: text-generation tasks: - text-generation model_type: - qwen2 architectures: - Qwen2Model parameters: 7.1B tensor_type: - F32 library_name: - pytorch - transformer - safetensors frameworks: - pytorch language: - zh inference_backends: - "deploy_task emb/text" - "vllm 0.9.2" downloads: 2245 stars: 1 tags: - mteb --- # Zhihui_LLM_Embedding > Zhihui_LLM_Embedding - QwenCollection 在 ModelScope 开源的模型。ZhihuiLLMEmbedding is an embedding model specifically designed to enhance Chinese text retrieval capabilities. It is built on a 7B LLM and enhanced bidirectional attention mechanism to improved… QwenCollection/Zhihui_LLM_Embedding 是 ModelScope 魔搭社区上的 7.1B 参数text-generation模型，采用 cc-by-nc-4.0 许可，可用 deploy_task emb/text、vllm 0.9.2 部署。 - **Repository**: QwenCollection/Zhihui_LLM_Embedding - **License**: cc-by-nc-4.0 - **Tasks**: text-generation - **Parameters**: 7.1B - **Inference backends**: deploy_task emb/text, vllm 0.9.2 - **Tags**: mteb - **Downloads**: 2245 - **Stars**: 1 - **Last updated**: 2024-11-21 Source: https://www.modelscope.cn/models/QwenCollection/Zhihui_LLM_Embedding --- ## Zhihui_LLM_Embedding ### Model Introduction **Zhihui_LLM_Embedding** is an embedding model specifically designed to enhance Chinese text retrieval capabilities. It is built on a 7B LLM and enhanced bidirectional attention mechanism to improved contextual understanding. The model is trained on an extensive corpus from various fields within an extremely large batch. **Zhihui_LLM_Embedding** excels in retrieval tasks, ranking **1st position** on the C-MTEB leaderboard with a leading performance score of **76.74** as of June 25, 2024. ### Optimization points * Data source enhancement: Leverages the knowledge of LLMs through three types of distillation methods.(GPT3.5 & GPT4) * Data Refinement: LLM scores candidate positive passages to select the most relevant examples. * Query Rewriting: LLM generates queries that can be answered by positive documents but are unrelated to negatives, thus enhancing the query's quality and diversity. * Query Expansion: Queries are expanded based on multiple topics for long documents. * Negative example mining: Use multiple methods and different ranges of negative selection to mine hard negative examples. * Improved Contrastive Loss: Design a novel InfoNCE loss assigns higher weights to the harder negative examples to improve the fine-grained feature representation of the model. * Bidirectional-attention: Remove the causal attention of LLMs during contrastive training of decoder-only LLM to produce rich contextualized representations. * Training efficiency: Using Gradient Cache to scale contrastive learning batches beyond GPU memory constraints allows the model to learn from more challenging negative examples. * Others： Dataset-Homogenous Batching、cross-batch negative sampling ### Model Details * Base Decoder-only LLM: [gte-Qwen2-7B-instruct](https://huggingface.co/Alibaba-NLP/gte-Qwen2-7B-instruct) * Pooling Methods: Last token * Embedding Dimension: 3584 ### Usage ##### Requirements ``` transformers>=4.40.2 flash_attn>=2.5.8 sentence-transformers>=2.7.0 ``` ##### How to use Here is an example of how to encode queries and passages using Huggingface-transformer and Sentence-transformer. ##### Usage (HuggingFace Transformers) ```python import torch import torch.nn.functional as F from torch import Tensor from transformers import AutoTokenizer, AutoModel def last_token_pool(last_hidden_states: Tensor, attention_mask: Tensor) -> Tensor: left_padding = (attention_mask[:, -1].sum() == attention_mask.shape[0]) if left_padding: return last_hidden_states[:, -1] else: sequence_lengths = attention_mask.sum(dim=1) - 1 batch_size = last_hidden_states.shape[0] return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths] def get_detailed_instruct(task_description: str, query: str) -> str: return f'Instruct: {task_description}\nQuery: {query}' task = 'Given a web search query, retrieve relevant passages that answer the query' queries = [ get_detailed_instruct(task, "国家法定节假日共多少天"), get_detailed_instruct(task, "如何查看好友申请") ] documents = [ "一年国家法定节假日为11天。根据公布的国家法定节假日调整方案，调整的主要内容包括：元旦放假1天不变；春节放假3天，放假时间为农历正月初一、初二、初三；“五一”国际劳动节1天不变；“十一”国庆节放假3天；清明节、端午节、中秋节增设为国家法定节假日，各放假1天(农历节日如遇闰月，以第一个月为休假日)。3、允许周末上移下错，与法定节假日形成连休。", "这个直接去我的QQ中心不就好了么那里可以查到 我的好友单向好友好友恢复、 以及好友申请 啊可以是你加别人的 或 别人加你的都可以查得到QQ空间里 这个没注意 要有的话也会在你进空间的时候会提示你的QQ 空间里 上面消息 就可以看见了！望采纳！谢谢这个直接去我的QQ中心不就好了么那里可以查到 我的好友单向好友好友恢复、 以及好友申请 啊可以是你加别人的 或 别人加你的都可以查得到", ] input_texts = queries + documents tokenizer = AutoTokenizer.from_pretrained('Lenovo-Zhihui/Zhihui_LLM_Embedding', trust_remote_code=True) model = AutoModel.from_pretrained('Lenovo-Zhihui/Zhihui_LLM_Embedding', trust_remote_code=True) max_length = 512 # Tokenize the input texts batch_dict = tokenizer(input_texts, max_length=max_length, padding=True, truncation=True, return_tensors='pt') outputs = model(**batch_dict) embeddings = last_token_pool(outputs.last_hidden_state, batch_dict['attention_mask']) # normalize embeddings embeddings = F.normalize(embeddings, p=2, dim=1) scores = (embeddings[:2] @ embeddings[2:].T) print(scores.tolist()) ``` ##### Usage (Sentence-Transformers) ```python from sentence_transformers import SentenceTransformer model = SentenceTransformer("Lenovo-Zhihui/Zhihui_LLM_Embedding", trust_remote_code=True) model.max_seq_length = 512 # 数据来源DuRetrieval https://huggingface.co/datasets/C-MTEB/DuRetrieval queries = [ "国家法定节假日共多少天", "如何查看好友申请", ] documents = [ "一年国家法定节假日为11天。根据公布的国家法定节假日调整方案，调整的主要内容包括：元旦放假1天不变；春节放假3天，放假时间为农历正月初一、初二、初三；“五一”国际劳动节1天不变；“十一”国庆节放假3天；清明节、端午节、中秋节增设为国家法定节假日，各放假1天(农历节日如遇闰月，以第一个月为休假日)。3、允许周末上移下错，与法定节假日形成连休。", "这个直接去我的QQ中心不就好了么那里可以查到 我的好友单向好友好友恢复、 以及好友申请 啊可以是你加别人的 或 别人加你的都可以查得到QQ空间里 这个没注意 要有的话也会在你进空间的时候会提示你的QQ 空间里 上面消息 就可以看见了！望采纳！谢谢这个直接去我的QQ中心不就好了么那里可以查到 我的好友单向好友好友恢复、 以及好友申请 啊可以是你加别人的 或 别人加你的都可以查得到", ] query_embeddings = model.encode(queries, prompt_name="query", normalize_embeddings=True) document_embeddings = model.encode(documents, normalize_embeddings=True) scores = (query_embeddings @ document_embeddings.T) print(scores.tolist()) ``` ### Reproduce our results（C-MTEB）: Check out scripts/eval_mteb.py to reproduce evaluation results on C-MTEB benchmark. | Model | T2Retrieval | MMarcoRetrieval | DuRetrieval | CovidRetrieval | CmedqaRetrieval | EcomRetrieval | MedicalRetrieval | VideoRetrieval | Avg | |:-------------------------------|:--------:|:--------:|:--------:|:--------:|:--------:|:--------:|:--------:|:--------:|:--------:| |**Zhihui_LLM_Embedding** | 88.30 | 84.77 | 91.34 | 84.39 | 48.69 | 71.96 | 65.19 | 79.31 | **76.74** | |zpoint_large_embedding_zh | 83.81 | 82.38 | 89.23 | 89.14 | 47.16 | 70.74 | 68.14 | 80.26 | 76.36 | |gte-Qwen2-7B-instruct | 87.73 | 85.16 | 87.44 | 83.65 | 48.69 | 71.15 | 65.59 | 78.84 | 76.03 | |360Zhinao-search | 87.12 | 83.32 | 87.57 | 85.02 | 46.73 | 68.9 | 63.69 | 78.09 | 75.06 | |AGE_Hybrid | 86.88 | 80.65 | 89.28 | 83.66 | 47.26 | 69.28 | 65.94 | 76.79 | 74.97 |