Kushanda Jupyter notebooks paNVIDIA Tesla P40 (24 GB VRAM, Pascal architecture) kune interaction data science uye ML basa. GPU servers hazvigone kupihwa; taura nesu kuti uzive kana zvave.
$ pip install jupyterlab torch && jupyter lab --ip=0.0.0.0 --allow-root
# Mufananidzo wekuisa command
Jupyter notebooks pane GPU server inokupa iwe nharaunda inowirirana uko notebook cells inogona kushandisa GPU. Yakakosha kune eksperimeniti uye prototyping.
Usashandisa GPU kubva notebook cells kana CUDA uye yako framework yaiswa.
JupyterLab ne terminal, faira browser uye zvigadziriso.
Unogona kuisa PyTorch, TensorFlow, scikit-kudzidza, pandas uye chero chinhu chimwe zvaunoda.
Kusvika notebooks yako kubva browser; chengetedza navo ne token, password kana SSH tunnel.
Jupyter inogona kushandisa NVIDIA GPU kuti iite basa rayo. VPS.org GPU servers hazvina kuwanikwa kuti utore; taura nesu kuti uzive kana vachiri. Jupyter inowanzo GPU-yakakwira basa. Ita shuwa kuti software yako ine CUDA rutsigiro uye kuti yako driver / runtime versions inosangana nezvinodiwa zvebasa re Jupyter.
GPU servers hadzina kuwanikwa kuti utore. Nhasi unogona kuisa Jupyter pane yakajairika CPU package, uko inoita basa rakakura kupfuura pane GPU: SSH in uye shandisa pip install jupyterlab torch && jupyter lab --ip=0.0.0.0 --allow-root.
VRAM zvinodiwa Jupyter kuenderana nemhando uye batch size; tarisa chirongwa ruzivo rwezvinoreva.
GPU mutengo achaburitswa kana GPU servers anogona kupihwa. Tibate kana iwe uchida kuziva kana ichiitika.
Yega. Every VPS.org server inouya nefull root access, saka unogona kutamba zvimwe zvinhu pamwe Jupyter mukati server's RAM uye kuchengetera.
Every VPS.org server inouya nepfupi root SSH kuwanikwa, saka unogona kuisa madhiraivha, mabhuku uye Jupyter asi iwe unofanirwa.
Hapana server inouya neCUDA runtime kana NVIDIA mutyairi akaiswa, uye GPU servers hazvina kuwanikwa kuti aite kusvika parizvino. Unogona kuisa zvinyorwa Jupyter zvinoda iwe pachako.
VPS.org servers ndezvemapurojekiti akaomarara, kwete ephemeral instances: Jupyter models, configs uye data inogara paSSD storage yeserver kusvika wabvisa.
Tenga data rekushandisa pane SSD ye server yekuwedzera kuwanikwa kwenguva pfupi panguva ye Jupyter runs, uye kopa zvakapera zviwanikwa (zviyero, zvizvarwa, embeddings) kune yekuchengetera kunze kwe VPS.org; snapshots uye backups vanogara pamwechete kuchengetera cluster.
Unogona kusimudzira chero VPS.org server kune yakakura chirongwa kubva kudzora panel; resize rinotangazve uye yako Jupyter kuiswa inotora pamusoro. GPU servers hazvina kuwanikwa kusvika parizvino.
Yeah. Backups ndivo akabhadhara add-on (zuva nezuva + 30% yemari chirongwa, akachengetwa 7 mazuva; zuva nezuva + 20%, akachengetwa 28 mazuva) uye snapshots vakasununguka, izvo zvinokosha pamberi Jupyter mazuva. Both stay on the same storage cluster as the server.
Contact support within 30 days of your first purchase to request a refund. GPU servers are not sold yet, so no GPU order is involved.
GPU kugona kuri kuwanikwa chete. Tiudzai nezvebasa rako uye tichakuudza kuti ripi rinowanikwa uye ripi risiri.