关键词 "token count" 的搜索结果, 共 24 条, 只显示前 480 条
Receipt-AI simplifies expense management by using AI and text messages to upload receipts to accounting software.
mesha is a comprehensive financial solution for businesses, including finance, taxes, accounting, bookkeeping, and payroll.
AI co-pilot for productivity and task management during calls.
AI-powered tool for effortless coupon and promo code finds.
Accountant workflow, client portal, and more in one solution.
Build LinkedIn audience with AI.
Turn your business idea into reality!
NewsDeck is a platform for latest news with extensive filtering and analysis options.
Email as quick and easy as texting.
Free tool to manage multiple social media accounts efficiently.
Effortlessly bypass CAPTCHAs with advanced AI.
Cold email tool for automating outreach with unlimited accounts.
Compare and evaluate various AI models and their specifications.
Token-gate content access with AI writing assistance.
A free AI content detector, ChatGPT plagiarism checker, and word counter.
PolyAI offers customer-led voice assistants to businesses, enabling consistent brand experience and data-driven opportunities.
Nanonets is an AI platform that automates processes and extracts actionable insights from unstructured data.
Rate and discover games with Minimap for personalized gaming recommendations.
Lexica is a fast and accurate search engine powered by Stable Diffusion technology.
TCMLLM由北京交通大学计算机与信息技术学院医学智能团队开发的中医药大语言模型项目,旨在通过大模型方式实现中医临床辅助诊疗(病证诊断、处方推荐等)中医药知识问答等任务,推动中医知识问答、临床辅助诊疗等领域的快速发展。目前针对中医临床智能诊疗问题中的处方推荐任务,发布了中医处方推荐指令微调大模型TCMLLM-PR。研发团队整合了8个数据来源,涵盖4本中医经典教科书《中医内科学》、《中医外科学》、《
DeepWiki :基于 GitHub Repo 源代码生成最新版可对话式文档,由 Devin驱动。 开源项目免费使用,无需注册。 私有项目中使用需在 http://devin.ai 注册账号。 直接访问 https://deepwiki.com,或将 GitHub 链接中的 github 替换为 deepwiki。 即:GitHub 仓库链接中的 github 替换为 deepwiki,
在本研究中,我们推出了 MiMo-7B 系列模型,这一系列模型从零开始训练,专为推理任务而生。我们基于 MiMo-7B-Base 进行的强化学习实验表明,我们的模型拥有非凡的推理潜力,甚至超越了规模更大的 32B 模型。此外,我们还对冷启动的 SFT 模型进行了强化学习训练,最终形成了 MiMo-7B-RL,它在数学和代码推理任务上均表现出色,性能堪比 OpenAI o1-mini。 我们开
我们在 Lean 4 中引入了 DeepSeek-Prover-V2,这是一个专为形式化定理证明而设计的开源大型语言模型,其初始化数据通过 DeepSeek-V3 驱动的递归定理证明流程收集。冷启动训练过程首先促使 DeepSeek-V3 将复杂问题分解为一系列子目标。已解决子目标的证明被合成为一个思路链,并结合 DeepSeek-V3 的逐步推理,为强化学习创建初始冷启动。这一过程使我们能够将非
an mcp server serving tools to manage AWS account
只显示前20页数据,更多请搜索
Showing 217 to 240 of 324 results