0 0 0 Python

fix: dequantize GLM 5 FP8 weights on import (#5851)

Signed-off-by: Hersh Godse <hersh@trajectory.ai>
Co-authored-by: Charlie Truong <chtruong@nvidia.com>
Co-authored-by: gautham-kollu <gkollu@nvidia.com>

Hersh Godse committed 4d ago

44c871b4eab107028933ea1a3aaa42dacc260c1c

Parent: 08fc1c6

Committed by GitHub <noreply@github.com> on 9/2/2026, 3:06:03 PM