Sebenzisa i-TensorFlow ne-Keras uqeqesho kunye nokuqikelela kwi-NVIDIA Tesla P40 (24 GB VRAM, uyilo lwePascal). Amaseva e-GPU ayifumaneki ngoku; qhagamshelana nathi ukuze uqinisekiswe xa zikhona.
$ pip install tensorflow[and-cuda]
# Umzekelo wofakelo lomyalelo
I-TensorFlow yinkqubo yokufunda i-machine evulekileyo ye-Google. Kwiseva ye-GPU, i-TensorFlow iqhuba imisebenzi yayo kwi-GPU nge-CUDA.
I-TensorFlow isebenzisa i-GPU nge-CUDA xa ufaka umqhubi ohambelanayo nohlobo lwe-CUDA.
Yakha iimodeli nge-Keras API phezulu kwe-GPU ephambili.
I-Monitor training with TensorBoard on your own server.
Ukubonelela ngeemodeli eziqeqeshiweyo nge-TensorFlow Serving.
TensorFlow inokusebenzisa i NVIDIA GPU ukukhawulezisa umsebenzi wayo. VPS.org GPU abancedisi abafumanekanga ukucela ngoku; qhagamshelana nathi ukuva xa bekhona. TensorFlow nguqeqesho / uxinzelelo lomsebenzi olunokulungiswa ngokulula. Icwangcisa imisebenzi ehamba ixesha elide: yenza i snapshot yakho ye VPS rhoqo kwaye ugcine ikopi yesisindo esiqeqeshiweyo ngaphandle kwe VPS.org.
Abancedisi be-GPU abafumanekanga ukucela ngoku. Namhlanje ungaqhuba i-TensorFlow kwi-CPU eqhelekileyo, apho isebenza khona ngokucothayo kune-GPU: SSH ngaphakathi kwaye uqhube i-pip install tensorflow[and-cuda].
Uqeqesho lwe-VRAM lulawulwa yimeko yokuphucula kunye nokuqhuba. Ugqibo olupheleleyo lwe-7B LLM lufuna malunga ne-24 ukuya kwi-48 GB; i-LoRA / i-QLoRA ilungele i-8 ukuya kwi-16 GB.
Ixabiso le-GPU liza kupapashwa xa iiseva ze-GPU zingafunyanwa. Xhulumanisa nathi ukuba ufuna ukuva xa oku kwenzeka.
Ewe. Isincedisi ngasinye se-VPS.org sifika sinee-root access ezipheleleyo, ngoko ungaqhuba ezinye izixhobo kunye ne-TensorFlow ngaphakathi kwe-RAM yeseva kunye nokugcina.
Yonke i VPS.org umncedisi ifika nge root SSH unikezelo olupheleleyo, ngoko ungafaka abaqhubi, iincwadi ze TensorFlow nokuba ufuna njani.
Hayi. Akukho mncedisi ufika nexesha lokuqhuba le-CUDA okanye umqhubi we-NVIDIA ofakwe ngaphambili, kwaye abancedisi be-GPU abafumanekanga ukucela ngoku. Ufaka uguqulelo lwe-TensorFlow olufunayo ngokwakho.
Ewe. Abancedisi be-VPS.org basebenza njengezimatshini ezibonakalayo eziqhubekayo, azikho iziganeko ezikhawulezayo: iimodyuli ze-TensorFlow, iinkqubo zokumiselwa kwenkqubo kunye nedata zihlala kwi-SSD yokugcina yeseva de uzicime.
Gcina i-data esebenza kwi-SSD yomncedisi ukufikelela ngokukhawuleza ngexesha lokusebenza kwe-TensorFlow, kwaye ukope i-artifacts egqityiweyo (isisindo, u Generations, i-embeddings) ukugcinwa ngaphandle kwe-VPS.org; iifoto ezikhawulezayo kunye nokhuseleko luhlala kwiqela elinye lokugcina.
Unganyusa nayiphi na iseva ye-VPS.org ibe yinkqubo enkulu ukusuka kwiqela lenjongo yolawulo; ubungakanani obutsha buqala kwakhona kwaye ukufaka kwe-TensorFlow kugqitha. Iiseva ze-GPU azifumanekanga ngoku.
Ewe. Iindawo zokugcina zihlawulwe zidityaniswe (ngosuku +30% yexabiso lenkqubo, igcinwe iintsuku ezi-7; ngeveki +20%, igcinwe iintsuku ezi-28) kwaye iifoto ezikhawulezayo zikhululekile, eziluncedo phambi kwe TensorFlow ende iqhuba. Zonke zihlala kwiqela lokugcina elifanayo njengeseva.
Nxibelelana noncedo ngaphakathi kweentsuku ezingama-30 zokuthengiswa kwakho kokuqala ukuze ucele ubuyiselo. Iiseva ze-GPU azithengwanga, ngoko akukho mfuneko yokuthumela i-GPU.
Ubungakanani be-GPU bumiselwe. Sixelele malunga nobunzima bomsebenzi wakho kwaye siza kuphendula ngokufumaneka nokuthengiswa.