ZeroPoint v0 research prototype: new from-scratch NumPy framework, memory-slot architecture, audited Transformer comparison
5c6c847 verified | """zerocg: 只用 numpy 实现的迷你深度学习框架。 | |
| import zerocg as zc | |
| x = zc.randn(4, 3, requires_grad=True) | |
| loss = (x * x).sum() | |
| loss.backward() # x.grad == 2 * x | |
| """ | |
| from . import functional, nn, optim | |
| from .tensor import Tensor, concat, count_flops, get_default_dtype, no_grad, ones, randn, set_default_dtype, tensor, zeros | |
| __all__ = ["Tensor", "concat", "count_flops", "no_grad", "set_default_dtype", "get_default_dtype", "tensor", "zeros", "ones", "randn", "nn", "optim", "functional"] | |