From 3a7cb765aea92153bbe32f652afe83b977108d24 Mon Sep 17 00:00:00 2001 From: Lianhao Lu Date: Thu, 21 Nov 2024 17:45:53 +0800 Subject: [PATCH] Fix audioqna readme for cached MODELDIR (#603) Signed-off-by: Lianhao Lu --- helm-charts/audioqna/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/helm-charts/audioqna/README.md b/helm-charts/audioqna/README.md index 80c62b1b..6ab65a9f 100644 --- a/helm-charts/audioqna/README.md +++ b/helm-charts/audioqna/README.md @@ -30,7 +30,7 @@ helm install audioqna audioqna --set global.HUGGINGFACEHUB_API_TOKEN=${HFTOKEN} 1. If you want to cache the downloaded model for later reuse, please set the bash environment variable `MODELDIR` to an existing directory on the node, then append `--set global.modelUseHostPath=${MODELDIR}` to the `helm install` commands. -2. Make sure your `${MODELDIR}/.locks` is writable to all the users if you want to use the cached downloaded models, i.e. `$ sudo chmod +w ${MODELDIR}/.locks`. +2. Make sure your `${MODELDIR}` and `${MODELDIR}/.locks` is writable to all the users if you want to use the cached downloaded models, i.e. `$ sudo chmod 0777 ${MODELDIR} && sudo chmod 0777 ${MODELDIR}/.locks `. ## Verify