-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path2024-10-19.txt
28 lines (25 loc) · 2.73 KB
/
2024-10-19.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
2024-10-19
Python
1. [TheAlgorithms / Python]:Python 中所有算法的实现(https://github.com/TheAlgorithms/Python)
2. [phidatahq / phidata]:构建具有记忆、知识、工具和推理能力的 AI 代理(https://github.com/phidatahq/phidata)
3. [mandiant / capa]:FLARE 团队的开源工具,用于在可执行文件中识别功能(https://github.com/mandiant/capa)
4. [Sinaptik-AI / pandas-ai]:与您的数据库聊天(SQL、CSV、pandas、polars、mongodb、noSQL 等)。PandasAI 通过 LLMs(GPT 3.5 / 4、Anthropic、VertexAI)和 RAG 实现数据分析对话化(https://github.com/Sinaptik-AI/pandas-ai)
5. [frdel / agent-zero]:Agent Zero AI 框架(https://github.com/frdel/agent-zero)
6. [serengil / deepface]:Python 的轻量级面部识别和面部属性分析(年龄、性别、情绪和种族)库(https://github.com/serengil/deepface)
7. [pytorch / pytorch]:具有强大 GPU 加速的 Python 中的张量动态神经网络(https://github.com/pytorch/pytorch)
8. [Rudrabha / Wav2Lip]:包含“ACM Multimedia 2020”上发表的“一个唇同步专家足以实现语音到唇同步生成”的代码。对于高清商业模型,请尝试 Sync Labs(https://github.com/Rudrabha/Wav2Lip)
9. [spotDL / spotify-downloader]:下载您的 Spotify 播放列表和歌曲,包括专辑封面和元数据(如果找到匹配项则从 YouTube 下载)(https://github.com/spotDL/spotify-downloader)
10. [NVIDIA / NeMo]:为研究者和开发者构建的可扩展生成 AI 框架,用于工作于大型语言模型、多模态和语音 AI(自动语音识别和文本到语音)(https://github.com/NVIDIA/NeMo)
11. [Z4nzu / hackingtool]:黑客的 ALL IN ONE 工具(https://github.com/Z4nzu/hackingtool)
12. [scikit-learn / scikit-learn]:scikit-learn:Python 中的机器学习(https://github.com/scikit-learn/scikit-learn)
13. [mlflow / mlflow]:机器学习生命周期的开源平台(https://github.com/mlflow/mlflow)
14. [VikParuchuri / surya]:90 多种语言的 OCR、布局分析、阅读顺序、表格识别(https://github.com/VikParuchuri/surya)
15. [mlfoundations / open_clip]:CLIP 的开源实现(https://github.com/mlfoundations/open_clip)
惊艳项目推荐:
[PandasAI](https://github.com/Sinaptik-AI/pandas-ai):通过 LLMs 和 RAG 实现数据分析对话化的创新工具。
今天趋势项目关注的领域和特点:
- AI 和机器学习:包括算法实现、AI 代理框架、数据分析和处理工具。
- 语音和音频处理:如语音到唇同步生成、音乐下载工具等。
- 计算机视觉:面部识别和属性分析库。
- 多模态 AI:结合文本、图像和语音的 AI 框架。
- 开源平台和工具:如机器学习生命周期管理和机器学习库。