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(cli): close the markdown stream before stopping the spinner

dcutugnoe6f014da6c87
diff --git a/c/coli b/c/coli
index 6f6942a9..e99b0e63 100755
--- a/c/coli
+++ b/c/coli
@@ -917,7 +917,7 @@ def chat_attached(a, base, model_id):
except OSError as e:
sp.stop()
print(f"\n {C.yel}[server unreachable: {e}]{C.r}"); break
- sp.stop(); md.close()
+ md.close(); sp.stop()
if reply: msgs.append({"role":"assistant","content":"".join(reply)})
else: msgs.pop() # turno vuoto: non sporcare la history
el=time.time()-t0