Files
foxhunt/crates
jgrusewski fee5a7cb3a fix: guard diagnostic buffer overflow — d_adv needed 2304 blocks, partials had 1137
debug_buffer_norm_f32 reuses grad_norm_partials (sized for total_params)
as scratch. d_adv_logits with batch_size*tba=589K elements needs 2304
blocks but partials only has 1137 — GPU memory overflow corrupting
adjacent allocations. On H100 this likely trashed grad_norm_buf,
explaining the persistent grad_norm=0.000000. Guard added to skip
oversized norms.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-13 23:23:03 +02:00
..