Bark
Free.ai (self-hosted)
·
tts
·
~500 Ii-token clip
·
4.6
I- 5 Abasebenzisi
Bark yi an Umbhalo-usuka-kwilizwi eyenziwe ngu Suno AI. Enamandla kakhulu kwi Podcasts, dialogue with emotion, creative audio scenes.. I-self-hosted kwi-Free.ai GPUs — iqhuba ngokukhululekileyo kwi-token pool yakho yosuku (500 tokens Iqhosha le-Clip). Ikhutshwe phantsi kwe MIT - ukusetyenziswa korhwebo kuvunyelwe kwi Free.ai.
Iimodeli ezifanayo
Imibuzo ebuzwa rhoqo
Bark ixhasa uluhlu olubanzi lweelwimi. Uluhlu oluchanekileyo luxhomekeke kwinjini; ifomu kule phepha ivuma nawuphi na umbhalo kwaye i-engine izakuveza kwiilwimi zayo ezixhaswayo. Bona / voice / kwi-multi-engine epheleleyo ekhethiweyo ukuba ufuna ulwimi oluthile.
Injini ezininzi zibonisa isiNgesi- saseMelika esingenakuthelekiswa nanto ngokungagqibekanga kunye nendawo- efanelekileyo yesivakalisi seelwimi ezingasazi- isiNgesi. Iinjini eziphezulu zingabonisa izivakalisi ezitshintshayo - Cola umzekelo wokuthelekiswa.
SSML inkxaso itshintsha nge injini. Ukuphumla, iprosody, ne tags yokungaqhelekanga zihlolwe kwinjini ezininzi eziphezulu zezinye ezinomphathi wabo. Umbhalo ocacileyo usoloko usebenza - akukho phawulo lufunekayo.
Ukusasazwa kwe-TTS kufumaneka kwi-premium engines nge /v1/tts/ API endpoint nge stream=true. I-web UI kule phepha ibuyisela i-clip epheleleyo xa ukubonakaliswa kugqityiwe.
Bark iqhuba kwi-GPU yethu. Uhlobo lutsala kwi-pool yakho yasimahla yansuku zonke kuqala. Xa lugqityiwe, i-token ehlawulweyo iqala kwi-$5 → 200,000 i-token. Kunzima ~5 i-token nganye yophawu, umyinge 100 nge-clip nganye.
Ukufikelela kwiimpawu ezingama-5,000 ngesicelo ngasinye kwi-web UI. Kwiinxalenye ezide (iincwadi zesandi, izihloko ezipheleleyo), sebenzisa /voice/audiobook/ equka i-chunks ne-stitches ngokuzenzekelayo, okanye thumela i-API kwi-loop.
Ewe - UTHENGA uluhlu lwamagama kwi /v1/tts/batch/, okanye sebenzisa indawo yokusebenza UI kwi /workspace/ ukudibanisa i TTS kwindlela ende yokuqhuba (umzekelo, guqulela → thetha → stitch).
Ewe - UTHENGA umbhalo kwi /v1/tts/ ngemodeli "Bark" (okanye i-slug kule phepha). Ibuyisela i-WAV okanye i-MP3. Bona /api/ ukubhekisa okupheleleyo + i-SDK iincam.
Eli phepha libhalwe-ku-ukuthetha, hayi ukuclonelwa kwelizwi - ilizwi limiselwe kwinjini. Ukuklonela kwelizwi (ukulayisha umyalezo obhekisa kuyo), bona /voice/clone/, ekufuneka uthathe ilungelo lelizwi okanye ube negunya elibhaliweyo elicacileyo.
Iinjini ezikhoyo ziqhuba kwi Free.ai- iikona ze GPUs; akukho nto ishiya abancedisi bethu. Iinjini eziphezulu zidlulisa umbhalo kubaboneleli bemodeli abaphezulu phantsi kwe DPA yethu. Asiqeqeshi kwiingxelo zakho kwaye asithengi i data.
Ewe - i-Free.ai inikezela ngokusetyenziswa korhwebo kwesandi esiveliswe. I-layisenisi ephantsi ye-engine (i-Apache 2.0, MIT, okanye iimeko zomboneleli) ibonisiwe phezulu kwaye kwiphepha lobhekiso lemodeli; kwinkqubo le nto ithetha ukuba i-voiceovers, i-ad, i-podcasts, kunye neenkqubo ze-apps zonke zikwi-ambiti.
Ewe - imisebenzi engaphumelelanga ibuyiselwe ngokuzenzekelayo kumbhali (iqela lemihla ngemihla okanye iimpawu ezihlawulwe). Ukuba ubuyiselo alubonakalanga ngomhla omnye, thumela i-imeyili contact@free.ai.