Home
cs229- files
- img
- mind
- nlp-roadmap
- 图解数学
- 知识梳理
Activation-Function.md- AI-VNext.md
- algorithm.md
- Apple-Core-ML.md
- Audio.md
- awesome.md
- ChatGPT.md
- CNN.md
- CUDA-Install.md
- CV.md
- dataset.md
- deep-learning.md
- DeepLearning.md
- dotnet-ml.md
- Feature-Engineering.md
- GAN.md
- google-JAX.md
- High-reference-paper.md
- I-Know.md
- ICSharpCore.md
- image.md
- JupyterNotebook.md
- loss-function.md
- Model-Selection.md
- multi-media.md
- NET Core with Jupyter Notebooks.md
- nlp.md
- Ollama.md
- person.md
- ppt素材.md
- pytorch.md
- README.md
- RNN.md
- robot.md
- tensorflow.md
- tools.md
- word2vec.md
- 书籍.md
- 有趣的应用.md
- 有趣的机器学习.md
- 模型汇总.md
- 生成模型和判别模型.md
- 线性回归和逻辑回归.md
https://github.com/google/JAX
JAX 论文链接：https://www.sysml.cc/doc/146.pdf
JAX 的创新之处在于，它基于 XLA 在 GPU 和 TPU 上编译和运行 NumPy 程序.
解释：NumPy计算只能在CPU上计算，所以tensorflow需要定义自己tensor