Skip to content

fixing reference implementation. #5225

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jjsjann123 merged 1 commit into from
Sep 24, 2025
Merged

fixing reference implementation. #5225

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions tests/python/direct_utils/narrow_precision.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,7 +79,7 @@ def linear_to_swizzled_128_4(a_sf_linear: torch.Tensor):
# details about layout requirement on block-wise scaling factor
# https://docs.nvidia.com/cutlass/media/docs/cpp/blackwell_functionality.html#scale-factor-layouts
tmp = torch.reshape(a_sf_padded, (m_tiles, 4, 32, k_tiles, 4))
return tmp.transpose(1, 3).reshape(mn_padded, k_padded)[:mn, :sf_k]
return tmp.transpose(1, 3).reshape(mn_padded, k_padded)


def dequantize_to_dtype(
Expand Down Expand Up @@ -168,7 +168,7 @@ def activation_scale_to_nvfp4(x, g_sf, offsets, blockscale_offsets, block_size):
else:
r = offsets[i + 1]
l_sf = blockscale_offsets[i]
r_sf = l_sf + r - l
r_sf = l_sf + (r - l + 127) // 128 * 128
v, b_sf = pytorch_nvfp4_quantize(x[l:r], g_sf[i])
v_scaled[l:r] = v
block_scale[l_sf:r_sf] = linear_to_swizzled_128_4(b_sf)
Expand Down