add 12-07 ai daily

Freelander · Dec 7, 2024 · dfe45fc · dfe45fc
1 parent 1b64f6b
commit dfe45fc
Show file tree

Hide file tree

Showing 3 changed files with 153 additions and 0 deletions.
diff --git a/docs/.vitepress/utils/createSideBar.ts b/docs/.vitepress/utils/createSideBar.ts
@@ -4,6 +4,7 @@ export function createSideBarZH() {
       text: '2024 年',
       collapsed: false,
       items: [
+        { text: '12 月 07 日', link: '/posts/2024/12-07' },
         { text: '12 月 04 日', link: '/posts/2024/12-04' },
         { text: '12 月 03 日', link: '/posts/2024/12-03' },
         { text: '12 月 02 日', link: '/posts/2024/12-02' },
@@ -171,6 +172,7 @@ export function createSideBarEN() {
       text: 'The Year 2024',
       collapsed: true,
       items: [
+        { text: 'December 07', link: '/en/posts/2024/12-07' },
         { text: 'December 04', link: '/en/posts/2024/12-04' },
         { text: 'December 03', link: '/en/posts/2024/12-03' },
         { text: 'December 02', link: '/en/posts/2024/12-02' },

diff --git a/docs/en/posts/2024/12-07.md b/docs/en/posts/2024/12-07.md
@@ -0,0 +1,75 @@
+---
+title: December 07
+date: 2024-12-07
+cover: https://oss.justin3go.com/blogs/fav0-001.jpg
+---
+
+Today's featured cutting-edge AI news, welcome to read 👇
+
+**💡 OpenAI launches reinforcement fine-tuning feature, requiring only a small amount of data to create domain expert AI, currently in closed beta for institutions only, expected to officially release in early 2025.**
+
+**🤖 X (Twitter) platform's Grok AI is now available to all users, free users can ask 10 questions every 2 hours, Premium members get more usage.**
+
+**🚀 Meta open-sources Llama 3.3 70B model, performance rivals 405B parameter large models, surpassing Gemini 1.5 Pro and GPT-4 in multiple tests.**
+
+**⚡️ Google quietly launches Gemini-exp-1206 experimental model, featuring 200 million context length, improved coding capabilities, available for free use in AI Studio.**
+
+**👁️ Google open-sources PaliGemma 2 vision model, offering multiple sizes, excelling at generating detailed image descriptions, performing exceptionally in multiple professional domains.**
+
+
+
+### Latest News
+
+**1. OpenAI Day 2/12**
+
+Launched reinforcement fine-tuning feature, requiring only a small amount of training data in specific domains to create domain expert AI models.
+
+During today's livestream, OpenAI researchers demonstrated that after reinforcement fine-tuning the o1 mini model, its reasoning capabilities in specific domains significantly improved, even surpassing the official o1 model.
+
+Official introduction: https://openai.com/form/rft-research-program/
+
+Currently, this feature is only open for beta testing applications from research institutions, universities, and enterprises, with plans for official launch in early 2025.
+
+![image-20241207102603357](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207102603357.png)
+
+
+
+**2. Grok AI is now available to all users.**
+
+Starting yesterday, you can use Grok AI on X (Twitter) without requiring a subscription.
+
+Access link: https://x.com/i/grok
+
+However, there are question limits: 10 questions every 2 hours. To get more usage, you'll need to subscribe to Premium membership.
+
+![image-20241207110513324](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207110513324.png)
+
+**3. Meta open-sources new Llama 3.3 70B model.**
+
+Achieves performance levels comparable to Llama's largest 3.1 405B model, surpassing Gemini 1.5 Pro and GPT-4o in multiple benchmark tests.
+
+Model download: https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
+
+![image-20241207111312231](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207111312231.png)
+
+**4. Google quietly releases new experimental model Gemini-exp-1206.**
+
+Features 200 million context length, improved model coding capabilities, currently available for free use on Google AI Studio and Gemini API.
+
+Try it online: https://aistudio.google.com
+
+Official release came without detailed introduction, directly launched, reclaiming the top spot in the arena rankings. Worth trying out.
+
+![image-20241207202824102](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207202824102.png)
+
+
+
+**5. Google open-sources powerful vision model PaliGemma 2.**
+
+Offers multiple model sizes with 3B, 10B, 28B parameters, showing stronger performance than PaliGemma, capable of generating detailed image descriptions, not just identifying objects but also describing actions, emotions, and overall scene narratives.
+
+Official introduction: https://developers.googleblog.com/en/introducing-paligemma-2-powerful-vision-language-models-simple-fine-tuning/
+
+Additionally, it shows excellent performance in chemical formula recognition, music notation recognition, chest X-ray reporting, and other domains.
+
+![img](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/PaliGemma2__BlogHeader.original.png)
diff --git a/docs/posts/2024/12-07.md b/docs/posts/2024/12-07.md
@@ -0,0 +1,76 @@
+---
+title: 12 月 07 日
+date: 2024-12-07
+cover: https://oss.justin3go.com/blogs/fav0-001.jpg
+---
+
+今日精选的前沿 AI 信息，欢迎阅读 👇
+
+**💡 OpenAI 推出强化微调功能，只需少量数据即可打造领域专家 AI，目前仅对机构开放内测，预计 2025 年初正式发布。**
+
+**🤖 X（Twitter）平台的 Grok AI 对所有用户开放使用，免费用户每 2 小时可提问 10 次，Premium 会员可获更多使用次数。**
+
+**🚀 Meta 开源 Llama 3.3 70B 模型，性能媲美 405B 参数的大模型，在多项测试中超越 Gemini 1.5 Pro 和 GPT-4。**
+
+**⚡️ Google 低调上线 Gemini-exp-1206 实验模型，具备 200 万上下文长度，提升编码能力，可在 AI Studio 免费使用。**
+
+**👁️ Google 开源 PaliGemma 2 视觉模型，提供多种尺寸，擅长生成详细图像描述，在多个专业领域表现出色。**
+
+
+
+### 前沿资讯
+
+**1、OpenAI Day 2/12**
+
+推出强化微调功能，在特定领域中只需少量的数据训练，即可打造领域专家 AI 模型。
+
+在今天直播上 OpenAI 相关研究人员展示了，对 o1 mini 模型进行强化微调后，在特定领域中模型的推理能力大幅提升，甚至超过正式版 o1 模型。
+
+官方介绍：https://openai.com/form/rft-research-program/
+
+目前该功能只对研究机构、大学和企业开放申请内测，计划在 2025 年初才正式推出。
+
+![image-20241207102603357](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207102603357.png)
+
+
+
+**2、Grok AI 对所有用户开放使用。**
+
+从昨天开始，在 X（Twitter）上不需要订阅会员，也能使用 Grok AI。
+
+使用地址：https://x.com/i/grok
+
+不过有提问次数限制，每 2 小时提问 10 次，想获得更多使用次数还得开 Premium 会员。
+
+![image-20241207110513324](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207110513324.png)
+
+**3、Meta 新开源 Llama 3.3 70B 模型。**
+
+与 Llama 最大模型 3.1 405B 相当的性能水平，在多项基准测试中，超越 Gemini 1.5 Pro 和 GPT-4o。
+
+模型下载：https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
+
+![image-20241207111312231](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207111312231.png)
+
+**4、Google 低调发布新实验模型 Gemini-exp-1206。**
+
+具有 200 万上下文长度，提升模型的编码能力，目前可以在 Google AI Studio 和 Gemini API 上免费使用。
+
+在线体验：https://aistudio.google.com
+
+官方没有发布详细的介绍，直接上线，夺回了竞技场榜首，可以去体验下。
+
+![image-20241207202824102](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/image-20241207202824102.png)
+
+
+
+**5、Google 开源强大的视觉模型 PaliGemma 2。**
+
+提供了多种模型尺寸，有 3B、10B、28B 参数 ，相对 PaliGemma 性能更强，能够生成详细图像描述，不仅能识别物体，还能描述动作、情感和场景的整体叙事。
+
+官方介绍：https://developers.googleblog.com/en/introducing-paligemma-2-powerful-vision-language-models-simple-fine-tuning/
+
+除此之外，在化学公式识别、音乐词谱识别、胸部 X 光报告等领域，同样表现得非常出色。
+
+![img](https://cdn.jsdelivr.net/gh/freelander/oss@master/ai-daily/2024-12-07/PaliGemma2__BlogHeader.original.png)
+