Uguqulelo lwesandi kunye nevidiyo nge OpenAI Whisper kwi-NVIDIA Tesla P40 (24 GB VRAM, uyilo lwePascal), ngesandi esigcinwe kwisincedisi sakho. Iiseva ze-GPU azikwazi ukuthengwa; qhagamshelana nathi ukuze uqinisekiswe xa zikhona.
$ pip install openai-whisper && whisper audio.mp3 --model large-v3 --device cuda
# Umzekelo wofakelo lomyalelo
OpenAI Whisper yimodeli yokwahlula-hlula ukuthetha-ukuthetha. Ukusebenza kwayo kwi-GPU kukhawuleza kakhulu kune-CPU kwaye igcina umsindo wakho kwiseva yakho.
I-Whisper isebenza ngokukhawuleza kakhulu kwi-GPU kune-CPU.
I-Whisper iguqula ukuthetha kwiilwimi ezili-99.
Isandi sihlala kwiseva yakho.
Umsebenzi kwiilayibrari ezinkulu zesandi kwiqela.
Whisper inokusebenzisa i-NVIDIA GPU ukukhawulezisa umsebenzi wayo. VPS.org GPU abancedisi abafumanekanga ukucela ngoku; qhagamshelana nathi ukuva xa bekhona. Whisper yiaudiyo / umthwalo womsebenzi wokuthetha. CPU-side I/O ibalulekile njenge-GPU compute, ngoko qiniseka ukuba idiski yakho ifunda isantya sokugcina iqhubekeka nokusebenza kweqela.
Abancedisi be-GPU abafumanekanga ukucela ngoku. Namhlanje ungaqhuba i-Whisper kwi-CPU eqhelekileyo, apho isebenza khona ngokucothayo kune-GPU: SSH ngaphakathi kwaye uqhube i-pip install openai-whisper && whisper audio.mp3 --model large-v3 --device cuda.
Iimodeli zesandi zihlala ziVRAM-light. I-Whisper Large ilungele malunga ne-10 GB ye-VRAM, kwaye i-TTS ifuna iimodeli ezi-2 ukuya kwi-8 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-Whisper ngaphakathi kwe-RAM yeseva kunye nokugcina.
Yonke i VPS.org umncedisi ifika nge root SSH unikezelo olupheleleyo, ngoko ungafaka abaqhubi, iincwadi ze Whisper 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-Whisper olufunayo ngokwakho.
Ewe. Abancedisi be-VPS.org basebenza njengezimatshini ezibonakalayo eziqhubekayo, azikho iziganeko ezikhawulezayo: iimodyuli ze-Whisper, iinkqubo zokumiselwa kwenkqubo kunye nedata zihlala kwi-SSD yokugcina yeseva de uzicime.
Gcina i-data esebenza kwi-SSD yomncedisi ukufikelela ngokukhawuleza ngexesha lokusebenza kwe-Whisper, kwaye ukope i-artifacts egqityiweyo (ubunzima, 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-Whisper 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 Whisper 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.