From 40517b593b23870cf351a05a1d53e930cea6a58d Mon Sep 17 00:00:00 2001
From: Yanbin Jiang <jybsuper@gmail.com>
Date: Wed, 15 Jul 2026 22:19:49 -0700
Subject: [PATCH] [docs] Inkling cookbook: LoRA cells require
 --disable-prefill-cuda-graph (#31418)

---
 docs_new/src/snippets/configs/thinkingmachines/inkling.jsx | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/docs_new/src/snippets/configs/thinkingmachines/inkling.jsx b/docs_new/src/snippets/configs/thinkingmachines/inkling.jsx
index 6ee62b5c7..ec67776d7 100644
--- a/docs_new/src/snippets/configs/thinkingmachines/inkling.jsx
+++ b/docs_new/src/snippets/configs/thinkingmachines/inkling.jsx
@@ -871,6 +871,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -904,6 +905,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -937,6 +939,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -970,6 +973,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -1004,6 +1008,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -1040,6 +1045,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",
@@ -1071,6 +1077,7 @@ export const config = {
         "--reasoning-parser inkling",
         "--tool-call-parser inkling",
         "--enable-lora",
+        "--disable-prefill-cuda-graph",
         "--lora-backend triton",
         "--lora-use-virtual-experts",
         "--max-loras-per-batch {{MAX_LORAS}}",