b10532 b10531
ggml-org/llama.cppAug 21, 2026 → Aug 21, 20260 days between releases
Generating AI comparison
Analyzing what changed between b10532 and b10531. This page updates automatically — usually within a minute.
The Two Releases
b10532
b10532
Aug 21, 2026· 2 months ago
Optimizes Flash Attention on the Metal backend by dequantizing KV cache to F16 in a preprocessing pass, improving accuracy and performance for various quantized KV cache types.
Full release notesb10531
b10531
Aug 21, 2026· 2 months ago
Reverts a previous fix for the tensor-split meta backend to address issues introduced by the change, restoring the previous behavior.
Full release notes