makepad/libs/ggml
Admin 8d7d90a843 ggml/llama: fix non-flat unary metal dispatch truncating rows past 1024 floats
Grid width for the non-flat unary path was ne01 while the kernel decomposes
tgpig.x into (ne0-chunk, row) — rows past ntg were silently dropped, masked
below the 32768-element flat-path threshold; silu on [8192, n] crossed it at
prefill batch 4 and corrupted qwen35 batched prefill. Grid is now
ne01 * ceil(ne0/nth). Adds metal unary/gated-delta-net/ssm_conv regression
tests, the llama-batch-probe bisect harness, prefill batch default 32, and
keeps the graph cache until a reserve-retry actually needs eviction.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-28 13:46:57 +02:00
..
src ggml/llama: fix non-flat unary metal dispatch truncating rows past 1024 floats 2026-07-28 13:46:57 +02:00
build.rs Checkpoint Qwen CUDA exact progress 2026-04-20 13:48:29 +02:00
Cargo.toml Add CUDA NVFP4 backend and GPU decode path 2026-04-11 11:44:04 +02:00