perf: optimize H3 inference and segmented weight prefetch (#2240)

This commit is contained in:
leejet authored and GitHub committed 2026-10-30 00:38:14 +08:00
1 parent e7b7ec6d6f
commit 37bc88b5e9
17 files changed
+367 -210

No files matched your search

+1 -1
View File
@@ -158,7 +158,7 @@ namespace sd {
for (const auto& entry : copies) {
const size_t source_index = ggml_hash_find(&graph->visited_hash_set, entry.first);
if (source_index != GGML_HASHSET_FULL && ggml_bitset_get(graph->visited_hash_set.used, source_index)) {
const size_t copy_index = ggml_hash_insert(&copy_graph->visited_hash_set, entry.second);
const size_t copy_index = ggml_hash_insert(&copy_graph->visited_hash_set, entry.second);
copy_graph->use_counts[copy_index] = graph->use_counts[source_index];
}
}