- download itksnap, unzip tar at ~
tar -xvf xxx.tar.gz
- add bin folder to PATH in .bashrc
just use python quart or dont use python as web services.
but if you perfer to be stuck with flask, then below could be useful:
https://github.com/benoitc/gunicorn/blob/master/gunicorn/config.py#L1437
gemini-cli
https://geminicli.com/docs/get-started/deployment/
export GOOGLE_CLOUD_PROJECT_ID=xxxxx
https://docs.cloud.google.com/gemini/docs/codeassist/set-up-gemini#enable-api
| """ | |
| buffer flushing logic using python asyncio Queue and Event | |
| """ | |
| import numpy as np | |
| import asyncio | |
| entire_list = [] | |
| async def flusher(queue, max_queue_size, interval, flush_event, pool): | |
| while True: |
| import time | |
| import sys | |
| import pandas as pd | |
| import numpy as np | |
| import SimpleITK as sitk | |
| def foobar_np(source_nifti_file,source_csv_file,target_nifti_file): | |
| df = pd.read_csv(source_csv_file) # columns,x,y,z,foo |
--
notes
| docker swarm | kubernetes |
|---|---|
| docker node ls | kubectl get nodes |
| docker stack deploy -c $YAMLFILE | kubectl apply -f $YAMLFILE |
| docker service create ... | kubectl create deploy -n $NAMESPACE $PODNAME ... |
| docker service inspect $SERVICENAME | kubectl describe pod -n $NAMESPACE $PODNAME |
| log |
| FROM pytorch/pytorch:2.2.2-cuda11.8-cudnn8-runtime | |
| ARG GROUPID | |
| ARG USERID | |
| ARG USERNAME | |
| RUN groupadd -g $GROUPID developer | |
| RUN useradd --no-log-init -u $USERID -g $GROUPID -s /bin/bash $USERNAME |