haloq38flash — qwen3.8-flash-next on strix halo: converter fix, 91g provenance-verified quant, engine a/b, depth tables through 256k

This commit is contained in:
Julian Beltran
2026-08-31 23:50:55 +10:00
commit a7c7aac004
9 changed files with 672 additions and 0 deletions
+6
View File
@@ -0,0 +1,6 @@
models/
weights
.codegraph
.omo/
wiki-*.parquet
corpus/