From 804d1c71bdd0bf0cce74ec6bf4e599f9613529e7 Mon Sep 17 00:00:00 2001
From: LostRuins Concedo <39025047+LostRuins@users.noreply.github.com>
Date: Wed, 12 Aug 2026 17:56:14 +0800
Subject: [PATCH] Replace free_compute_buffer with runner_done

free_compute_buffer does not seem to release all resources after the gen.

Replaced free_compute_buffer() calls with runner_done() to fix subsequent generations failing due to insufficient memory that was not freed.
---
 src/model/vae/vae.hpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/model/vae/vae.hpp b/src/model/vae/vae.hpp
index 7013f732a..62a6dd066 100644
--- a/src/model/vae/vae.hpp
+++ b/src/model/vae/vae.hpp
@@ -156,7 +156,7 @@ struct VAE : public GGMLRunner {
             output = _compute(n_threads, input, false);
         }
 
-        free_compute_buffer();
+        runner_done();
 
         if (output.empty()) {
             LOG_ERROR("vae encode compute failed");
@@ -207,7 +207,7 @@ struct VAE : public GGMLRunner {
             output = _compute(n_threads, input, true);
         }
 
-        free_compute_buffer();
+        runner_done();
 
         if (output.empty()) {
             LOG_ERROR("vae decode compute failed");