Skip to content

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

C

fix(glm): label banner bit widths as compute, not storage

JustVugg582693da26ef
diff --git a/c/colibri.c b/c/colibri.c
index a46f5d29..8c391ffb 100644
--- a/c/colibri.c
+++ b/c/colibri.c
@@ -10765,7 +10765,11 @@ int main(int argc, char **argv){
g_tq_codec ? "rotated int4 + Lloyd codebook (4-bit)" : "PolarQuant recursive-polar");
}
}
- printf("== GLM C engine (glm_moe_dsa), cache=%d experts/layer | experts@%d-bit dense@%d-bit | idot: " IDOT_KERNEL " ==\n", cap, ebits, dbits);
+ /* ebits/dbits are the compute (dequant) width the idot kernels run at, not
+ * the stored weight format: a fmt=4 grouped-int4 container still computes at
+ * 8-bit here. Label it as compute so the banner is not misread as a storage
+ * claim (#1183). */
+ printf("== GLM C engine (glm_moe_dsa), cache=%d experts/layer | compute experts@%d-bit dense@%d-bit | idot: " IDOT_KERNEL " ==\n", cap, ebits, dbits);
g_mem_avail_boot = mem_available_gb();
#if !defined(_WIN32)
if(getenv("CLUSTER_WORKERS") && *getenv("CLUSTER_WORKERS")){