ShareAI
AI Prosumer
JV
Basa

Model

Dokumentasi API Gabung minangka Provider

8 model

Rega per 1M token

cognitivecomputations/dolphin3

Dolphin 3.0 Llama 3.1 8B 🐬 minangka generasi sabanjure saka seri Dolphin model instruksi sing disetel kanggo dadi model lokal tujuan umum paling apik, ngaktifake coding, matematika, agentik, function calling, lan kasus panggunaan umum.

5 tag model131K konteksTukar token

cognitivecomputations/megadolphin

MegaDolphin-2.2-120b minangka transformasi Dolphin-2.2-70b sing digawe kanthi interleaving model karo awake dhewe.

19 tag modelKonteks 4KTukar token

cognitivecomputations/dolphin-phi

Model Dolphin tanpa sensor 2.7B dening Eric Hartford, adhedhasar model basa Phi dening Microsoft Research.

15 tag modelKonteks 2KTukar token

cognitivecomputations/dolphincoder

Varian uncensored 7B lan 15B saka kulawarga model Dolphin sing unggul ing coding, adhedhasar StarCoder2.

35 tag model16K konteksTukar token

cognitivecomputations/tinydolphin

Model eksperimen parameter 1.1B sing dilatih ing dataset Dolphin 2.8 anyar dening Eric Hartford lan adhedhasar TinyLlama.

18 tag modelKonteks 4KTukar token

cognitivecomputations/dolphin-llama3

Dolphin 2.9 minangka model anyar kanthi ukuran 8B lan 70B dening Eric Hartford adhedhasar Llama 3 sing nduweni macem-macem katrampilan instruksi, obrolan, lan coding.

53 tag modelKonteks 8KTukar token

cognitivecomputations/dolphin-mixtral

Model fine-tuned uncensored, 8x7b lan 8x22b adhedhasar model Mixture of Experts Mixtral sing unggul ing tugas coding. Digawe dening Eric Hartford.

87 tag modelKonteks 33KTukar token

Saringan

Penyedia
Kreator
Akses
Tanpa Server