| import torch | |
| from safetensors.torch import load_file | |
| def load_model(path='model.safetensors'): | |
| return load_file(path) | |
| def reverse8(a7, a6, a5, a4, a3, a2, a1, a0, weights): | |
| """8-bit bit reversal.""" | |
| inp = torch.tensor([float(a7), float(a6), float(a5), float(a4), | |
| float(a3), float(a2), float(a1), float(a0)]) | |
| outputs = [] | |
| for i in range(8): | |
| y = int((inp @ weights[f'y{i}.weight'].T + weights[f'y{i}.bias'] >= 0).item()) | |
| outputs.append(y) | |
| return outputs | |
| if __name__ == '__main__': | |
| w = load_model() | |
| print('reverse8 examples:') | |
| test_cases = [ | |
| (1, 0, 0, 0, 0, 0, 0, 0), | |
| (0, 0, 0, 0, 0, 0, 0, 1), | |
| (1, 0, 1, 0, 0, 1, 0, 1), | |
| ] | |
| for bits in test_cases: | |
| result = reverse8(*bits, w) | |
| print(f' {list(bits)} -> {result}') | |