Embedding

class torch.ao.nn.quantized.Embedding(num_embeddings, embedding_dim, padding_idx=None, max_norm=None, norm_type=2.0, scale_grad_by_freq=False, sparse=False, _weight=None, dtype=torch.quint8)

A quantized Embedding module with quantized packed weights as inputs. We adopt the same interface as torch.nn.Embedding, please see https://pytorch.org/docs/stable/generated/torch.nn.Embedding.html for documentation.

Similar to Embedding, attributes will be randomly initialized at module creation time and will be overwritten later

Variables

weight (Tensor) – the non-learnable quantized weights of the module of shape $(\text{num\_embeddings}, \text{embedding\_dim})$.

Examples::

>>> m = nn.quantized.Embedding(num_embeddings=10, embedding_dim=12)
>>> indices = torch.tensor([9, 6, 5, 7, 8, 8, 9, 2, 8])
>>> output = m(indices)
>>> print(output.size())
torch.Size([9, 12])

classmethod from_float(mod, use_precomputed_fake_quant=False)

Create a quantized embedding module from a float module

Parameters

mod (Module) – a float module, either produced by torch.ao.quantization utilities or provided by user