milkey/
deepseek-r1-UD

194 9 months ago

(Unsloth Dynamic Quants) DeepSeek's first-generation of reasoning models with comparable performance to OpenAI-o1, 671B MoE full model, not dense distilled models.

Models

View all →

2 models

deepseek-r1-UD:IQ1_S

140GB · 4K context window · Text · 9 months ago

deepseek-r1-UD:IQ1_M

169GB · 4K context window · Text · 9 months ago

Readme

Note: this model requires Ollama 0.5.5 or later.

https://unsloth.ai/blog/deepseekr1-dynamic

https://huggingface.co/unsloth/DeepSeek-R1-GGUF