07 2024 档案
摘要:语音交互的三驾马车:ASR、NLP、TTS https://www.woshipm.com/ai/2620327.html 01 语音交互的组成 我们以一个智能音箱的例子来开始今天的讨论: 假设我们对智能音箱天猫精灵说“放一首周杰伦的《晴天》”。天猫精灵就会说“好的,马上为你播放周杰伦的《晴天》”,
阅读全文
摘要:Conda https://docs.conda.io/projects/conda/en/latest/index.html Conda Documentation Welcome to conda's documentation! Conda provides package, dependen
阅读全文
摘要:embedding-models https://ollama.com/blog/embedding-models Embedding models April 8, 2024 Ollama supports embedding models, making it possible to build
阅读全文
摘要:talk-to-your-data https://github.com/fanqingsong/talk-to-your-data talk-to-your-data project This project aid you to build a talk-to-your-data chatbot
阅读全文
摘要:text2sql-workshop https://github.com/fanqingsong/text2sql-workshop Text2SQL Workshop Using OpenAI, Langchain and Postgresql to Talk to Your Data Overv
阅读全文
摘要:Postgresql & PgAdmin powered by compose https://github.com/fanqingsong/compose-postgres/blob/master/docker-compose.yml Quick Start Clone or download t
阅读全文
摘要:本地测试 https://www.bingal.com/posts/Qwen-7b-usage/ 通义千问-7B(Qwen-7b)是什么 通义千问(Qwen-7B)是阿里云最新发布的一系列超大规模语言模型,这个牛气十足的大模型令人惊叹。基于Transformer架构,Qwen-7B系列汇聚了70亿参
阅读全文
摘要:LMDeploy https://lmdeploy.readthedocs.io/en/latest/index.html LMDeploy has the following core features: Efficient Inference: LMDeploy delivers up to 1
阅读全文
摘要:LM Studio https://lmstudio.ai/ Discover, download, and run local LLMs With LM Studio, you can ...🤖 - Run LLMs on your laptop, entirely offline👾 - Us
阅读全文
摘要:Text2SQL https://developer.volcengine.com/articles/7370376373124202505 简单的说, Vanna是一个开源的、基于Python的、用于SQL自动生成与相关功能的RAG(检索增强生成)框架 。基本特点: 开放源代码( Github上搜
阅读全文
摘要:LangChain Alternative https://www.chatbees.ai/blog/langchain-alternatives 5 Best LangChain Alternatives in 2024If you're looking for alternatives to L
阅读全文
摘要:LangChain vs LlamaIndex https://www.datacamp.com/blog/langchain-vs-llamaindex While both frameworks support integration with external tools and servic
阅读全文
摘要:Prompt https://prompt-guide.ncuos.com/prompts.zh Prompt Hub 是提示的集合,可用于测试 prompt 在各种基本能力和复杂任务上的能力。我们希望 Prompt Hub 帮助您发现利用、实验和构建提示词的有趣方法。我们鼓励并欢迎人工智能研究和开
阅读全文
摘要:Best practices for prompt engineering with the OpenAI API https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-the-o
阅读全文
摘要:TinyAgent https://github.com/fanqingsong/tiny-universe/tree/main/content/TinyAgent content/TinyAgent/agent_demo.py from tinyAgent.Agent_Ollama import
阅读全文
摘要:What Are Zero-Shot Prompting and Few-Shot Prompting https://machinelearningmastery.com/what-are-zero-shot-prompting-and-few-shot-prompting/ How Do Lar
阅读全文
摘要:Instruction和Question的区别和联系 https://blog.csdn.net/qq_35082030/article/details/130727016 3. Question和Instruction的区别问答(QA)和指令(Instruction)是人机交互中最常见的形式之一。
阅读全文
摘要:Large Language Models: A Survey https://arxiv.org/pdf/2402.06196 导读 https://developer.volcengine.com/articles/7389111961612451866 1、介绍了语言建模的历史背景,从统计语言
阅读全文
摘要:transformer model architecture https://www.datacamp.com/tutorial/how-transformers-work 动手写 https://www.datacamp.com/tutorial/building-a-transformer-wi
阅读全文
摘要:CUDA Refresher: The CUDA Programming Model https://developer.nvidia.com/blog/cuda-refresher-cuda-programming-model/ To execute any CUDA program, there
阅读全文
摘要:vLLM https://github.com/vllm-project/vllm https://docs.vllm.ai/en/latest/ 推理和服务,但是更加偏向推理。 vLLM is a fast and easy-to-use library for LLM inference and
阅读全文
摘要:Open WebUI https://alphasec.io/open-source-chatgpt-ui-alternative-with-open-webui/ https://docs.openwebui.com/getting-started/ Chatbot UI https://alph
阅读全文
摘要:Ludwig Declarative deep learning framework built for scale and efficiency. https://ludwig.ai/latest/ What is Ludwig?¶ Ludwig is a low-code framework f
阅读全文
摘要:词根:ven, vent = to come 来 https://www.dictool.com/root/search?wd=vent 在世界史上有很多有名的帝王将相留下的言语当时看起来没有什么,但是慢慢的就变成了非常有名的名言了,在外国里面其中一位就是恺撒大帝,他在泽拉战役中获胜后写给罗马元老院
阅读全文
摘要:模型本身的格式是什么? 模型本身的格式和存储内容可以因模型类型、框架和用途的不同而有所差异。然而,在一般情况下,我们可以从以下几个方面来探讨模型的文件格式和存储内容: 一、模型文件格式 JSON格式: Hugging Face Transformers等框架中,模型文件常以JSON格式存储,包含了模
阅读全文
摘要:HuggingFace Transformers https://www.hugging-face.org/hugging-face-transformers-2/ Hugging Face Transformers has been built by, with, and for the comm
阅读全文
摘要:modelscope上的模型如何下载? 在ModelScope(魔搭)上下载模型,用户可以通过多种方式进行,主要包括直接通过网站界面下载、使用SDK(软件开发工具包)下载,以及通过Git方式下载。以下是详细的下载步骤:一、通过网站界面下载 访问ModelScope官网: 打开浏览器,访问ModelS
阅读全文
摘要:大模型中的token是指什么? 在大模型(尤其是自然语言处理(NLP)领域中的大模型,如BERT、GPT等)的上下文中,"token"是一个关键概念,它指的是文本数据在被模型处理之前或过程中被分割成的最小单元。这些单元可以是单词、标点符号、子词(subword)等,具体取决于模型的词汇表(vocab
阅读全文
摘要:vLLM: Easy, Fast, and Cheap LLM Serving with PagedAttention https://blog.vllm.ai/2023/06/20/vllm.html LLMs promise to fundamentally change how we use
阅读全文
摘要:微调(Fine-Tuning)和检索增强生成(RAG) https://www.toutiao.com/article/7386094597421810186/?log_from=21b91ee0752_1721118518195 2024-06-30 09:05·架构笔记 微调(Fine-Tuni
阅读全文
摘要:大模型的量化 https://developer.baidu.com/article/details/3211764 一、大模型量化的基本概念 大模型量化,简单来说,就是将深度学习模型的浮点运算转换为整型运算的过程。在默认情况下,模型使用32位浮点数进行所有预测计算,这使得模型变得非常大。通过量化技
阅读全文
摘要:Chain-of-Thought Prompting https://www.promptingguide.ai/zh/techniques/cot#%E9%9B%B6%E6%A0%B7%E6%9C%AC-cot-%E6%8F%90%E7%A4%BA 链式思考(CoT)提示 图片来源:Wei等人(2
阅读全文
摘要:phpMyAdmin https://www.phpmyadmin.net/ phpMyAdmin is a free software tool written in PHP, intended to handle the administration of MySQL over the Web.
阅读全文
摘要:confd+etcd+nginx 服务发现 https://learnku.com/articles/36229 https://www.chenshaowen.com/blog/service-discovery-etcd-confd-nginx.html 在进行应用部署时,服务运行起来后,通过接
阅读全文
摘要:How to Use Terraform and Ansible Together: A Practical Example https://www.cloudbolt.io/terraform-best-practices/terraform-vs-ansible/ Now that you kn
阅读全文
摘要:How to link multiple docker-compose services via network New scenario: Sharing a network between services If you want define services in multiple dock
阅读全文
摘要:Nginx VS Traefik Comparison https://reverseproxy.com/docs/comparison/nginx-vs-traefik/#:~:text=In%20summary%2C%20Nginx%20might%20be%20better%20for%20m
阅读全文
摘要:D - Avoid K Palindrome https://atcoder.jp/contests/abc359/tasks/abc359_d 思路 https://atcoder.jp/contests/abc359/submissions/54822869 状压DP 以 K二进制位表示 K字符
阅读全文
摘要:A - Insert https://atcoder.jp/contests/abc361/tasks/abc361_a https://atcoder.jp/contests/abc361/submissions/55260626 int n, k, x; vector<int> a; int m
阅读全文