Release multi-bit quantization suite: GGUF (F16, BF16, Q8_0, Q6_K, Q5_K_M, Q4_K_M, Q4_0, Q3_K_M, Q2_K) + PyTorch INT8/INT4/INT2 Safetensors + Ollama Modelfile
8fc2b78 verified | # Official Ollama Modelfile for danAI-55M-Reasoning | |
| FROM ./danAI-55M-Reasoning-Q8_0.gguf | |
| # Model Parameters | |
| PARAMETER temperature 0.2 | |
| PARAMETER top_p 0.9 | |
| PARAMETER top_k 40 | |
| PARAMETER repeat_penalty 1.2 | |
| PARAMETER stop "<|endoftext|>" | |
| PARAMETER stop "<eos>" | |
| PARAMETER stop "User:" | |
| # Official System Prompt | |
| SYSTEM """You are danAI, a helpful, precise, and wise AI assistant with native reasoning and agentic tool-calling capabilities. When solving reasoning or math problems, think step-by-step.""" | |
| # Template | |
| TEMPLATE """{{ if .System }}System: {{ .System }} | |
| {{ end }}{{ range .Messages }}{{ if eq .Role "user" }}User: {{ .Content }} | |
| Assistant: {{ else if eq .Role "assistant" }}{{ .Content }}<|endoftext|>{{ end }}{{ end }}""" | |