点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 面向现实世界的机器人感知评测任务，五大赛道，全链路挑战，全球征集解决方案！ 为什么需要 RoboSense？ 在机器人系统不断迈向真实世界的进程中，感知系统的稳定性、鲁棒性与泛化能力正成为制约其部署能 力的关键因素。面对动态人群、恶劣天气、传感器故障、跨平台部署等复杂环境条件，传统感知算法往 往面临性能大幅下降的挑战。 为此， RoboSense Challenge 2025 应运而生。该挑战赛旨在系统性评估机器人在真实场景下的感知与理 解能力，推动多模态感知模型的稳健性研究，鼓励跨模态融合与任务泛化方向的创新探索。 | Registration | From June 2025 | | --- | --- | | Competition Server Online | June 15th, 2025 | | Phase One Deadline | August 15th, 2025 | | Phase Two Deadline | September 15th, 2025 | | Award Decisi ...

【导读】科大讯飞及讯飞医疗国际化战略升级，国产大模型吹响国际化号角 中国基金报记者 郭玟君 国产大模型吹响国际化号角，首站登陆香港。 6月24日，科大讯飞股份有限公司（以下简称科大讯飞）及讯飞医疗科技股份有限公司（以下 简称讯飞医疗）技术应用发布会暨香港公司开幕典礼在香港数字科技枢纽、人工智能加速器 及创业培育基地——香港数码港（以下简称数码港）举行。 以香港为"桥头堡" 科大讯飞全球战略部署全面启航 科大讯飞同时发布了基于讯飞星火大模型底座能力研发的医疗、教育、会议、办公等多领域 人工智能产品香港版及国际版，这是2024年年中科大讯飞及讯飞医疗落户香港并设立国际总 部以来在香港的首场重大发布。 香港积极吸引重点企业 共同打造具有全球竞争力的创科枢纽 引进办主任任景信出席开幕典礼时表示："科大讯飞和讯飞医疗在人工智能领域成就卓越，其 技术应用与香港创新科技发展的方向高度契合。科大讯飞及讯飞医疗选择在香港设立国际总 部和研究院，将有助于香港本地人工智能及大模型等产业链的发展，特别是在智慧医疗领 域，进一步提高医疗服务效率，让更多患者受益。引进办将继续积极支持重点企业在港拓 展，共同将香港打造成为具有全球竞争力的 ...

What goes into fine-tuning models like DeepSeek-R1?- Dataset preparation- Defining LoRA config- Defining a Trainer- Fine-tuning the LLM- Exporting to OllamaThe thread below covers this: https://t.co/78A3YXiWSyAvi Chawla (@_avichawla):Let's fine-tune DeepSeek-R1 (distilled Llama) 100% locally: ...

RDDT Reddit When investors spot an unusual amount of buying for a given stock, this typically creates the sort of curiosity that can lead to some of the most profitable rabbit holes. However, as this information is now readily available to all retail investors, the "smart money" in the market has figured out a way to still shift massive amounts of capital in a much quieter manner. Reddit Today $142.41 +8.27 (+6.17%) 52-Week Range $49.13 ▼ $230.41 P/E Ratio 241.26 Price Target $139.35 Add to Watchlist Throug ...

Thank you. So when I was 19, I was in Montreal. I was an undergrad student and I managed to get sick with the flu.So I sat down and I Googled, "Is flu going to kill me?" And Google Google brought up Wikipedia. Wikipedia said no. So, I stayed at home for 3 days eating my typical diet of ramen and water and uh didn't get better.So, I decided to go to the hospital. Now, has anyone here been to a hospital or an emergency room. Raise of hands.So, who thought it was a really good use of their time. Very efficient ...

CoreWeave's (NASDAQ: CRWV ) is an AI-focused Cloud computing start-up that benefits from a massive upsurge in generative AI spending. Strong demand for GPUs that are being used to train and run large language models is why the company’s MarchAnalyst’s Disclosure:I/we have a beneficial long position in the shares of NBIS, NVDA either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Se ...

Generally, we think the cost of innovation is collapsing here. It is going to transform our lives. We're going to look back and see this is a hugely important era.This is obviously the most disruptive technology in the history of mankind. At the center of many of the latest AI breakthroughs are large language models or LMS. The Chat TS and Geminis of this world.But as powerful as they're becoming, we learn more every day about their limitations, how they can be prone to errors, hallucinations and uncertaint ...

If you found it insightful, reshare it with your network.Find me → @_avichawlaEvery day, I share tutorials and insights on DS, ML, LLMs, and RAGs.Avi Chawla (@_avichawla):Let's build an MCP server (100% locally): ...

以下文章来源于AIGC面面观 ，作者欠阿贝尔两块钱 AIGC面面观 . 整理LLM、AIGC的入门笔记 | 论文学习笔记 | 一线大厂面经 | 探索AIGC落地 作者 | 欠阿贝尔两块钱 来源 | AIGC面面观 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 >>点击进入→ 自动驾驶之心 『大模型』技术交流群 主要贡献 1. 高效混合架构设计 ：结合MoE架构与Lightning Attention）的模型MiniMax-M1， 支持百万级上下文窗 口（1M tokens） ，生成长度达80K tokens时FLOPs仅为传统注意力模型的25%。 2. 超越DAPO的算法CISPO ：通过 剪裁重要性采样权重 提升RL效率，相比DAPO实现2倍加速，避免了 传统方法（如PPO/GRPO）对低概率token有更好的采样效果。 3. 可扩展上下文 ：支持从40K到80K Token生成长度的扩展。 本文只做学术分享，如有侵权，联系删文 1.混合注意力架构 Lighting Attention : 采用I/O感知的线性注意力计算，通过分块计算和内存优化 ，将长 ...

Uh hello everyone. So today I will be talking about AI and sustainability. So let me give you a brief introduction about myself.I've been in education and AI industry from past 10 years. One of the major thing that I have noticed in my life is that no matter how fast you go, there is always innovation happening faster than your rate of learning. Now let me explain that in terms.If I suddenly woke up today and start looking at uh what these large language models are doing, I will be shocked. It would appear ...