谷歌I/O大会

谷歌I/O大会谷歌I/O大会 Modified May 21, 2024 Audio or video is not supported Please download the file and open it with the appropriate software. 36726 1715707503 raw.mp4 · 34.79MB 36726 1715707503 raw 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Project Astra Our vision for the future of AI assistants.mp4 · 51.08MB Project Astra Our vision for the future of AI assistants 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. tVO8pDoDBiabhuca.mp4 · 3.72MB tVO8pDoDBiabhuca 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Search in the Gemini era Google IO 2024.mp4 · 30.03MB Search in the Gemini era Google IO 2024 00:00 Prompt: “A single comic book panel of a boy and his father on a grassy hill, staring at the sunset. A speech bubble points from the boy's mouth says "The sun will rise again". Muted, late 1990s coloring style.” Prompt: “A long exposure photo of the Milky Way in a starry night sky, centered over an ocean beach at magic hour. The milky way is bright and prominent with many stars visible against a dark blue black atmosphere in light painting photography with vivid and bold colors. Shot on a professional camera with high contrast and a cinematic composition in the style.” 亮点：Veo：强大的生成视频模型，sora的对手？它可以创建超过 60 秒的高质量 1080p 短片。从写实主义到超现实主义和动画，它可以应对各种电影风格。下面是一个完全不必要的宣传视频：Filmmaking with Donald Glover and his creative studio, Gilga | Veo 实际生成画面在 8秒 33秒 53秒 1分01秒 1分22秒 1分27秒 No access Filmmaking with Donald Glover and his creative studio, Gilga Veo 00:00 实际生成效果（还不错）： No access SBTrjO7gJTTQdiC0 00:00 Prompt: “Many spotted jellyfish pulsating under water. Their bodies are transparent and glowing in deep ocean.” 提示："许多斑点水母在水下跳动它们的身体是透明的，在深海中发光"。 No access oWkhPEsasG0y vvA 00:00 Prompt: “Timelapse of a water lily opening, dark background.” 提示："睡莲开放的延时摄影，暗色背景" No access 5u3wPAsSNl3tDk5F 00:00 Prompt: “A lone cowboy rides his horse across an open plain at beautiful sunset, soft light, warm colors.” 提示："在美丽的夕阳下，一个孤独的牛仔骑着马穿过开阔的平原，柔和的光线，温暖的色彩。" No access 2UkqleaoP43 onC 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. 36726 1715707503 raw.mp4 · 34.79MB 36726 1715707503 raw 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. 36726 1715707503 raw.mp4 · 34.79MB 36726 1715707503 raw 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Project Astra Our vision for the future of AI assistants.mp4 · 51.08MB Project Astra Our vision for the future of AI assistants 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Project Astra Our vision for the future of AI assistants.mp4 · 51.08MB Project Astra Our vision for the future of AI assistants 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. tVO8pDoDBiabhuca.mp4 · 3.72MB tVO8pDoDBiabhuca 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. tVO8pDoDBiabhuca.mp4 · 3.72MB tVO8pDoDBiabhuca 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Search in the Gemini era Google IO 2024.mp4 · 30.03MB Search in the Gemini era Google IO 2024 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Search in the Gemini era Google IO 2024.mp4 · 30.03MB Search in the Gemini era Google IO 2024 00:00 Prompt: “A single comic book panel of a boy and his father on a grassy hill, staring at the sunset. A speech bubble points from the boy's mouth says "The sun will rise again". Muted, late 1990s coloring style.” Prompt: “A long exposure photo of the Milky Way in a starry night sky, centered over an ocean beach at magic hour. The milky way is bright and prominent with many stars visible against a dark blue black atmosphere in light painting photography with vivid and bold colors. Shot on a professional camera with high contrast and a cinematic composition in the style.” 亮点：Veo：强大的生成视频模型，sora的对手？它可以创建超过 60 秒的高质量 1080p 短片。从写实主义到超现实主义和动画，它可以应对各种电影风格。下面是一个完全不必要的宣传视频：Filmmaking with Donald Glover and his creative studio, Gilga | Veo 实际生成画面在 8秒 33秒 53秒 1分01秒 1分22秒 1分27秒 No access Filmmaking with Donald Glover and his creative studio, Gilga Veo 00:00 No access Filmmaking with Donald Glover and his creative studio, Gilga Veo 00:00 实际生成效果（还不错）： No access SBTrjO7gJTTQdiC0 00:00 No access SBTrjO7gJTTQdiC0 00:00 Prompt: “Many spotted jellyfish pulsating under water. Their bodies are transparent and glowing in deep ocean.” 提示："许多斑点水母在水下跳动它们的身体是透明的，在深海中发光"。 No access oWkhPEsasG0y vvA 00:00 No access oWkhPEsasG0y vvA 00:00 Prompt: “Timelapse of a water lily opening, dark background.” 提示："睡莲开放的延时摄影，暗色背景" No access 5u3wPAsSNl3tDk5F 00:00 No access 5u3wPAsSNl3tDk5F 00:00 Prompt: “A lone cowboy rides his horse across an open plain at beautiful sunset, soft light, warm colors.” 提示："在美丽的夕阳下，一个孤独的牛仔骑着马穿过开阔的平原，柔和的光线，温暖的色彩。" No access 2UkqleaoP43 onC 00:00 No access 2UkqleaoP43 onC 00:00 Prompt: “A spaceship hurdles through the vastness of space, stars streaking past as it, high speed, sci fi.” 提示："宇宙飞船在浩瀚的太空中疾驰" "恒星从它身边划过" "高速，科幻" No access Lobq15xpthUWva74 00:00 No access Lobq15xpthUWva74 00:00 Prompt: “A woman sitting alone in a dimly lit cafe, a half finished novel open in front of her. Film noir aesthetic, mysterious atmosphere. Black and white.” 提示："一个女人独自坐在灯光昏暗的咖啡馆里，面前放着一本写了一半的小说。黑色电影美学，神秘氛围。黑白"。 No access yY5X0vfYJEgqnxPg 00:00 No access yY5X0vfYJEgqnxPg 00:00 Prompt: “Extreme close up of chicken and green pepper kebabs grilling on a barbeque with flames. Shallow focus and light smoke. vivid colours.” 提示："烧烤架上烤着鸡肉和青椒串的火焰特写。焦距较浅，烟雾较淡。色彩鲜艳。" No access pBrvDw jVM3HhMuw 00:00 No access pBrvDw jVM3HhMuw 00:00 Prompt: “A panning shot of a serene mountain landscape, the camera slowly revealing snow capped peaks, granite rocks and a crystal clear lake reflecting the sky.” 提示："一个宁静山景的平移镜头，镜头慢慢展现出白雪皑皑的山峰、花岗岩石和倒映着天空的清澈湖水"。 No access g4f8mZxTumw4Btmp 00:00 No access g4f8mZxTumw4Btmp 00:00 Prompt: “A golden retriever walks on a winding mountain trail, its tail wagging excitedly as it explores the sights and scents of the wilderness.” 提示："一只金毛猎犬走在蜿蜒的山间小路上，兴奋地摇着尾巴，探索着荒野的景色和气味"。 No access Ofj4U1YN438LIGi2 00:00 No access Ofj4U1YN438LIGi2 00:00 这些视频未经编辑，是 Veo 的原始输出。您现在可以加入 @labs.google 的候补名单，Google 最新图像模型 Imagen 3 和视频模型 Veo 的 waitlist 申请地址：https://aitestkitchen.withgoogle.com/tools/video fx 开场视频 No access Google IO 2024 Opening Film 00:00 No access Google IO 2024 Opening Film 00:00 💡 Gemini 1.5 pro 多模态 + 2M上下文 +Agents 35种语言 35种语言 Gemini 1.5 pro 多模态 + 2M上下文 +Agents 35种语言 35种语言 Gemma 2: Next Gen Performance and Efficiency 下一代 Gemma 型号 Gemma 2 即将面世。Gemma 2采用了全新的架构，旨在实现突破性的性能和效率，其具有以下优势 • Gemma 2 拥有 270 亿个参数，其性能可与 Llama 3 70B 相媲美，而体积却不到 Llama 3 70B 的一半。这一突破性效率为开放模型领域树立了新标准。 • Gemma 2 的高效设计使其只需不到同类机型一半的计算量。27B 机型经过优化，可在英伟达™（NVIDIA®）的 GPU 上运行，也可在 Vertex AI 的单个 TPU 主机上高效运行，从而使更多用户更方便、更经济地进行部署。 • 多功能调整工具链：Gemma 2 将为开发人员提供跨各种平台和工具生态系统的强大调整功能。从 Google Cloud 等基于云的解决方案到 Axolotl 等流行的社区工具，Gemma 2 的微调将比以往更加简单。此外，与 Hugging Face 和 NVIDIA TensorRT LLM 的无缝合作伙伴集成，以及我们自己的 JAX 和 Keras，可确保您优化性能并在各种硬件配置上高效部署。 • 未来几周内 Gemma 2 正式发布 💡 Gemini 1.5 Flash It’s a lighter weight model, optimized for tasks where low latency and cost matter most. Starting today, developers can use it with up to 1 million tokens in Google AI Studio and Vertex AI. GoogleIO It’s a lighter weight model, optimized for tasks where low latency and cost matter most. Starting today, developers can use it with up to 1 million tokens in Google AI Studio and Vertex AI. GoogleIO 这是一个轻量级模型，针对低延迟和低成本任务进行了优化。从今天开始，开发人员可以在 Google AI Studio 和 Vertex AI 中使用它，token代币数量最多可达 100 万枚介绍双子座 1.5 闪存这是一个轻量级模型，针对低延迟和低成本任务进行了优化。从今天开始，开发人员可以在 Google AI Studio 和 Vertex AI 中使用它，代币数量最多可达 100 万枚。 GoogleIO Gemini 1.5 Flash It’s a lighter weight model, optimized for tasks where low latency and cost matter most. Starting today, developers can use it with up to 1 million tokens in Google AI Studio and Vertex AI. GoogleIO It’s a lighter weight model, optimized for tasks where low latency and cost matter most. Starting today, developers can use it with up to 1 million tokens in Google AI Studio and Vertex AI. GoogleIO GoogleIO 这是一个轻量级模型，针对低延迟和低成本任务进行了优化。从今天开始，开发人员可以在 Google AI Studio 和 Vertex AI 中使用它，token代币数量最多可达 100 万枚介绍双子座 1.5 闪存这是一个轻量级模型，针对低延迟和低成本任务进行了优化。从今天开始，开发人员可以在 Google AI Studio 和 Vertex AI 中使用它，代币数量最多可达 100 万枚。亮点：PaliGemma：开放式视觉语言模型 PaliGemma 是一款功能强大的开放式 VLM，其灵感来自 PaLI 3。PaliGemma 基于包括 SigLIP 视觉模型和 Gemma 语言模型在内的开放式组件，可在各种视觉语言任务中实现领先的微调性能。这包括图像和短视频字幕、视觉问题解答、理解图像中的文字、物体检测和物体分割。立即开始使用 PaliGemma。您可以在 GitHub、Hugging Face models、Kaggle、Vertex AI Model Garden 和 ai.nvidia.com 上找到 PaliGemma（使用 TensoRT LLM 加速），并可通过 JAX 和 Hugging Face Transformers 轻松集成。(Keras 集成即将推出）您还可以通过 Hugging Face Space 与模型进行交互。 PaliGemma GitHub Hugging Face models Kaggle Vertex AI Model Garden ai.nvidia.com Hugging Face Space Gemini 1.5回顾 Project Astra 对AI助手未来的愿景。不过那个眼镜是什么情况？！ 💡 google这办公室该说不说真不错，还可以养狗 google这办公室该说不说真不错，还可以养狗你永远想象不到直播中的视频还没其他渠道的视频有意思 Gemini era Take a look at some of the new ways AI in Search can do the hard work so you don’t have to. Just ask. 反正吧就这么回事吧针对印度语进行开发 |杰玛和纳瓦拉萨虽然许多早期的大型语言模型主要使用英语语言数据进行训练，但该领域正在迅速发展。越来越多的新模型在多语言数据集上进行训练，并且越来越关注开发专门针对世界语言的模型。然而，在确保不同语言的公平代表性和性能方面仍然存在挑战，特别是那些可用数据和计算资源较少的语言。 Gemma 是 Google 的开放模型系列，旨在通过支持非日耳曼语言的项目开发来应对这些挑战。它的分词器和大量标语词汇使其特别适合处理各种语言。观看印度的开发人员如何使用 Gemma 创建 Navarasa，这是一个针对印度语言进行微调的 Gemma 模型。 Audio or video is not supported Please download the file and open it with the appropriate software. Developing for Indic languages Gemma and Navarasa.mp4 · 28.44MB Developing for Indic languages Gemma and Navarasa 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Developing for Indic languages Gemma and Navarasa.mp4 · 28.44MB Developing for Indic languages Gemma and Navarasa 00:00 通过LearnLM使每位教师都能接触到每个学生谷歌一直在试行由LearnLM提供支持的Google Classroom的新功能，以帮助减轻教师的工作量。应用生成式人工智能，我们正在探索如何帮助简化课程计划流程，使教师能够根据学生的个人需求定制课程和内容，以便他们能够扩大学习影响力并满足学生的需求。 Audio or video is not supported Please download the file and open it with the appropriate software. Empowering Every Teacher to Reach Every Student with LearnLM.mp4 · 27.47MB Empowering Every Teacher to Reach Every Student with LearnLM 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Empowering Every Teacher to Reach Every Student with LearnLM.mp4 · 27.47MB Empowering Every Teacher to Reach Every Student with LearnLM 00:00 Try the latest generative AI tools at labs.google 🏝️ https://labs.google/ 可以玩到一些“新”技术，比如Imagen 3 https://labs.google/ 可以玩到一些“新”技术，比如Imagen 3 Audio or video is not supported Please download the file and open it with the appropriate software. Try the latest generative AI tools at labs.google.mp4 · 5.76MB Try the latest generative AI tools at labs.google 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Try the latest generative AI tools at labs.google.mp4 · 5.76MB Try the latest generative AI tools at labs.google 00:00 亮点：Imagen 3 文本到图像生成模型 Visual Artists x Imagen | Google Lab Sessions Audio or video is not supported Please download the file and open it with the appropriate software. Visual Artists x Imagen Google Lab Sessions.mp4 · 113.53MB Visual Artists x Imagen Google Lab Sessions 00:00 Audio or video is not supported Please download the file and open it with the appropriate software. Visual Artists x Imagen Google Lab Sessions.mp4 · 113.53MB Visual Artists x Imagen Google Lab Sessions 00:00 Prompt: “A photo of a man with short hair and beard smiling at the camera. The background is blurry and it shows trees and buildings in light colors.” Prompt: “View from above,

在小宇宙note 阅读完整内容