From 8f67f29317d986aa469c6d696394080a9c35e5fb Mon Sep 17 00:00:00 2001
From: Akarshan <akarshan@menlo.ai>
Date: Thu, 11 Sep 2025 13:08:01 +0530
Subject: [PATCH] feat: add support for mmproj offload setting

Expose the new `mmproj_offload` option in the model settings UI and include it in the `ModelPlan` type. The component now collects the offload flag (`result.offloadMmproj`) and queues it with other setting updates to ensure a single atomic change, preventing race conditions when toggling this feature. This enables users to control MMProj offloading directly from the app.
---
 web-app/src/containers/ModelSetting.tsx | 10 ++++++++++
 web-app/src/services/models/types.ts    |  1 +
 2 files changed, 11 insertions(+)

diff --git a/web-app/src/containers/ModelSetting.tsx b/web-app/src/containers/ModelSetting.tsx
index a02b9df80..39a587cbc 100644
--- a/web-app/src/containers/ModelSetting.tsx
+++ b/web-app/src/containers/ModelSetting.tsx
@@ -93,6 +93,16 @@ export function ModelSetting({
           })
         }
 
+        if (
+          model.settings?.mmproj_offload &&
+          result.offloadMmproj !== undefined
+        ) {
+          settingsToUpdate.push({
+            key: 'mmproj_offload',
+            value: result.offloadMmproj,
+          })
+        }
+
         // Apply all settings in a single update to avoid race conditions
         if (settingsToUpdate.length > 0) {
           handleMultipleSettingsChange(settingsToUpdate)
diff --git a/web-app/src/services/models/types.ts b/web-app/src/services/models/types.ts
index b7959227a..b7724fef2 100644
--- a/web-app/src/services/models/types.ts
+++ b/web-app/src/services/models/types.ts
@@ -85,6 +85,7 @@ export interface ModelPlan {
   gpuLayers: number
   maxContextLength: number
   noOffloadKVCache: boolean
+  offloadMmproj: boolean
   mode: 'GPU' | 'Hybrid' | 'CPU' | 'Unsupported'
 }