RMSnow's picture
add backend inference and inferface output
0883aa1
raw
history blame
1.78 kB
import torch
def _sampler(
pdf: torch.Tensor, num_samples: int, device=torch.device("cpu")
) -> torch.Tensor:
size = pdf.size()
z = -torch.log(torch.rand(size, device=device))
_, indices = torch.topk(pdf + z, num_samples)
return indices
def compute_mask_indices(
size: torch.Size,
mask_prob: float,
mask_length: int,
min_masks: int = 0,
device=torch.device("cpu"),
) -> torch.Tensor:
assert len(size) == 2
batch_size, seq_length = size
# compute number of masked span in batch
num_masked_spans = (
mask_prob * float(seq_length) / float(mask_length) + torch.rand(1)[0]
)
num_masked_spans = int(num_masked_spans)
num_masked_spans = max(num_masked_spans, min_masks)
# num_masked <= seq_length
if num_masked_spans * mask_length > seq_length:
num_masked_spans = seq_length // mask_length
pdf = torch.ones(batch_size, seq_length - (mask_length - 1), device=device)
mask_idxs = _sampler(pdf, num_masked_spans, device=device)
mask_idxs = (
mask_idxs.unsqueeze(-1)
.repeat(1, 1, mask_length)
.view(batch_size, num_masked_spans * mask_length)
) # [B,num_masked_spans*mask_length]
offset = (
torch.arange(mask_length, device=device)
.view(1, 1, -1)
.repeat(1, num_masked_spans, 1)
) # [1,num_masked_spans,mask_length]
offset = offset.view(1, num_masked_spans * mask_length)
mask_idxs = mask_idxs + offset # [B,num_masked_spans, mask_length]
ones = torch.ones(batch_size, seq_length, dtype=torch.bool, device=mask_idxs.device)
# masks to fill
full_mask = torch.zeros_like(ones, dtype=torch.bool, device=mask_idxs.device)
return torch.scatter(full_mask, dim=1, index=mask_idxs, src=ones)