mannix/ deepseek-v2-lite-instruct

593 2 years ago

A strong, economical, and efficient Mixture-of-Experts language model.

8 models

deepseek-v2-lite-instruct:latest
8e53acb9158a • 9.9GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q3_k_s
066e613aaf22 • 7.5GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q3_k_m
f9c0fb3c583d • 8.1GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q4_1
8e53acb9158a • 9.9GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q5_1
5426b690d98f • 12GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q5_k_m
72ed1c997652 • 12GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q6_k
8ea247a9e747 • 14GB • 4K context window •
Text input • 2 years ago
deepseek-v2-lite-instruct:q8_0
5678205183eb • 17GB • 4K context window •
Text input • 2 years ago