biondizzle 7073daaffa fix: allocate token_indices on CPU, move to GPU AFTER JIT compilation ...

CuTeDSL's cute.compile corrupts GPU memory during JIT compilation.
Tensors allocated on GPU before/during compilation get zeroed.
Fix: create token_indices on CPU, then .to(device) after JIT is done.

2026-05-17 08:22:51 +00:00

..

patches

fix: set _l1_activation_global_scale (with underscore) — attribute name mismatch

2026-05-17 03:35:20 +00:00

nvfp4_cutedsl.py

fix: allocate token_indices on CPU, move to GPU AFTER JIT compilation

2026-05-17 08:22:51 +00:00