Tiny deep-learning framework powered by python3 and numpy
- examples --complete code examples of Trchime
- documentation --Trchime development documentation and usage documentation
| Name | Version |
|---|---|
| Python | 3.7.0+ |
| Numpy | 1.19.0+ |
Compute $ y=kx +b $ gradient with trchime, and output the gradient of $ x $ .
The whole process is completed by graph calculation and automatic differentiation.
from trchime import Variable, Constant
k = Constant([[2, 3], [1, 1]])
b = Constant([[7], [3]])
x = Variable([[0], [0]])
y = k @ x + b
z = y.sum()
z.backward()
print(x.grad.data)result:
[[3.],
[4.]]- Graph Computing
- Auto Gradient
- Algebraic system
- Gradient Descent
- Neural Network API
- Convolutional Neural Network
| Author | yatorho |
| 3227068950 | |
| 3227068950@qq.com |