A strong, economical, and efficient Mixture-of-Experts language model.
ollama run deepseek-v2
Ollama kurulu değil mi? ollama.com/download — Windows, macOS ve Linux için ücretsiz. İlk çalıştırmada model indirilir, sonrası tamamen çevrimdışıdır.
Boyut büyüdükçe kalite artar, donanım ihtiyacı yükselir. Başlangıç için küçük varyantı deneyin.
| Etiket | Boyut | Bağlam | Girdi | Komut |
|---|---|---|---|---|
| latest | 8.9GB | 160K | Text |
ollama run deepseek-v2:latest
|
| lite | 8.9GB | 160K | Text |
ollama run deepseek-v2:lite
|
| 16b | 8.9GB | 160K | Text |
ollama run deepseek-v2:16b
|
| 236b | 133GB | 4K | Text |
ollama run deepseek-v2:236b
|
| 16b-lite-chat-q2_K | 6.4GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q2_K
|
| 16b-lite-chat-q3_K_S | 7.5GB | 160K | Text |
ollama run deepseek-v2:16b-lite-chat-q3_K_S
|
| 16b-lite-chat-q3_K_M | 8.1GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q3_K_M
|
| 16b-lite-chat-q3_K_L | 8.5GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q3_K_L
|
| 16b-lite-chat-q4_0 | 8.9GB | 160K | Text |
ollama run deepseek-v2:16b-lite-chat-q4_0
|
| 16b-lite-chat-q4_1 | 9.9GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q4_1
|
| 16b-lite-chat-q4_K_S | 9.5GB | 160K | Text |
ollama run deepseek-v2:16b-lite-chat-q4_K_S
|
| 16b-lite-chat-q4_K_M | 10GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q4_K_M
|
| 16b-lite-chat-q5_0 | 11GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q5_0
|
| 16b-lite-chat-q5_1 | 12GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q5_1
|
| 16b-lite-chat-q5_K_S | 11GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q5_K_S
|
| 16b-lite-chat-q5_K_M | 12GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q5_K_M
|
| 16b-lite-chat-q6_K | 14GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q6_K
|
| 16b-lite-chat-q8_0 | 17GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-q8_0
|
| 16b-lite-chat-fp16 | 31GB | 4K | Text |
ollama run deepseek-v2:16b-lite-chat-fp16
|
| 236b-chat-q2_K | 86GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q2_K
|
| 236b-chat-q3_K_S | 102GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q3_K_S
|
| 236b-chat-q3_K_M | 113GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q3_K_M
|
| 236b-chat-q3_K_L | 122GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q3_K_L
|
| 236b-chat-q4_0 | 133GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q4_0
|
| 236b-chat-q4_1 | 148GB | 4K | Text |
ollama run deepseek-v2:236b-chat-q4_1
|
9 varyant daha var — tümü için ollama.com.
Note: this model requires Ollama 0.1.40.
DeepSeek-V2 is a a strong Mixture-of-Experts (MoE) language model characterized by economical training and efficient inference.
Note: this model is bilingual in English and Chinese.
The model comes in two sizes:
ollama run deepseek-v2:16bollama run deepseek-v2:236b