跳到主要內容
別錯過 Microsoft Ignite 接下來的精彩內容
別錯過 Microsoft Ignite 接下來的精彩內容直接接觸 Microsoft 技術背後的人員、產品和專業知識,您將有機會提出問題、培養技能,並與社群一起學習。
立即登錄
Microsoft
separator
https://catalogartifact.azureedge.net/publicartifacts/bcloudllc1671615348068.tinygradd-1e97d106-10af-4108-aa39-0d67ddc06e0c/image2_bCloudlog.png

Tinygrad

作者 bCloud LLC

(1 評分)

Version 0.14.0 + Free Support on Ubuntu 26.04

Tinygrad is a lightweight open-source deep learning framework designed for simplicity, transparency, and rapid experimentation in artificial intelligence and machine learning. It provides tensor operations, automatic differentiation, neural network components, and compiler capabilities, enabling developers and researchers to build and test AI models with a minimal and flexible framework.

Features of Tinygrad:
  • Lightweight and minimalist deep learning framework for AI development.
  • Supports tensor operations with automatic differentiation (autograd).
  • Includes neural network layers, optimizers, and training utilities.
  • Provides JIT compilation for optimized model execution.
  • Supports multiple hardware backends including CPU and compatible GPU accelerators.
  • Simple and readable codebase suitable for learning and AI research.
  • Enables rapid prototyping and experimentation with machine learning models.
  • Compatible with Python-based AI and scientific computing workflows.

Usage Instructions :

$sudo su
$cd /opt

For verifying Tinygrad installation:
$ source ~/tinygrad-env/bin/activate

Check Tinygrad version:
$ python -m pip show tinygrad

Disclaimer: Tinygrad is open-source software provided under the MIT License. It is offered "as is," without any warranty, express or implied. Users are responsible for ensuring the ethical and legal use of Tinygrad in their projects.
中文(香港特別行政區)
Your Privacy Choices Opt-Out Icon 您在加州的隱私選擇
Consumer Health Privacy Sitemap Contact Us Privacy & Cookies Terms of Use About our ads Manage cookies