wyx
> 这里：https://github.com/THUDM/VisualGLM-6B/blob/main/model/infer_util.py#L21 > > 同时还要在web_demo.py的main函数之前`from finetune_visualglm import FineTuneVisualGLMModel` 改了下还是不行 ``` For torch.distributed users or loading model parallel models, set environment variables RANK, WORLD_SIZE and LOCAL_RANK. /data/home/yaokj5/anaconda3/envs/vglm/lib/python3.10/site-packages/torch/nn/init.py:405: UserWarning: Initializing zero-element tensors...
get Error: unknown flag: --policy-config-file
@southquist use ``` services: scheduler: extra_args: config: /etc/kubernetes/scheduler-policy-config.yaml ``` but kubescheduler can't start get error ``` {"log":"I0202 02:01:24.426117 1 requestheader_controller.go:244] Loaded a new request header values for RequestHeaderAuthRequestController\n","stream":"stderr","time":"2023-02-02T02:01:24.426433785Z"} {"log":"E0202 02:01:24.427547...
Oh change `/etc/kubernetes/scheduler.conf` to `/etc/kubernetes/ssl/kubecfg-kube-scheduler.yaml` in **/etc/kubernetes/scheduler-policy-config.yaml** is Ok
Finally I follow this https://github.com/kserve/kserve/issues/924 and use [volcano](https://github.com/volcano-sh/volcano/blob/master/docs/user-guide/how_to_use_gpu_sharing.md) and [Time-Slicing GPUs in Kubernetes](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/gpu-sharing.html) to solve gpu-mem share. This is [my note](https://www.wyx.cloudns.asia/blog/2022/04/15/GPU-%E9%80%82%E9%85%8D)
> Yea... Can someone suggest what I've missed..... How to fixed pending state?
just delete nodeSelector in gpushare-schd-extender.yaml
https://github.com/AliyunContainerService/gpushare-scheduler-extender/issues/80 参考这个
@Jeremy-boo I use [harbor-online-installer-v2.3.5.tgz](https://github.com/goharbor/harbor/releases/download/v2.3.5/harbor-online-installer-v2.3.5.tgz) with the arm image to start. start harbor ``` ./install.sh --with-chartmuseum ``` In docker-compose.yml ``` registry: image: goharbor/registry-photon:dev-arm container_name: registry restart: always cap_drop: - ALL cap_add:...