Turn off FlashInfer for mlc runtime

mlc-ai · Jun 7, 2024 · 7934130 · 7934130
1 parent 5f163aa
commit 7934130
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/scripts/build_mlc_llm_wheel_manylinux.sh b/scripts/build_mlc_llm_wheel_manylinux.sh
@@ -110,7 +110,7 @@ elif [[ ${GPU} == cuda* ]]; then
 	echo set\(USE_CUBLAS ON\) >>config.cmake
 	echo set\(USE_THRUST ON\) >>config.cmake
 	echo set\(USE_NCCL ON\) >>config.cmake
-	echo set\(USE_FLASHINFER ON\) >>config.cmake
+	echo set\(USE_FLASHINFER OFF\) >>config.cmake
 	echo set\(FLASHINFER_ENABLE_FP8 OFF\) >>config.cmake
 	echo set\(FLASHINFER_ENABLE_BF16 OFF\) >>config.cmake
 	echo set\(FLASHINFER_GEN_GROUP_SIZES 1 4 6 8\) >>config.cmake