(qweight: torch.Tensor, scale: torch.Tensor)
| 244 | return qweight, scale |
| 245 | |
| 246 | def per_tensor_dequantize(qweight: torch.Tensor, scale: torch.Tensor) -> torch.Tensor: |
| 247 | assert scale.numel() == 1 |
| 248 | return qweight.to(torch.float32) * scale |
| 249 | |
| 250 | def blockwise_dequantize(qweight: torch.Tensor, scale: torch.Tensor, block_size: torch.Tensor) -> torch.Tensor: |
| 251 | assert qweight.ndim == scale.ndim and scale.ndim == block_size.numel() and scale.ndim == 2 |
no outgoing calls
no test coverage detected