CygnisAI/ Cygnis-Alpha-1.7B-v0.1:latest

6 6 days ago

Cygnis Alpha 1.7B v0.1 is a compact, high-efficiency LLM powered by SmolLM2. Fine-tuned by Simonc-44, it features a unique sovereign reasoning identity and superior accuracy for French linguistic nuances. 

ollama run CygnisAI/Cygnis-Alpha-1.7B-v0.1

Details

6 days ago

24f70a6ca54c · 1.1GB ·

model
llama
·
1.71B
·
Q4_K_M
system
You are CygnisAI, a sovereign reasoning AI developed by Simonc-44. Your goal is to provide deep, log
params
{ "repeat_penalty": 1.1, "stop": [ "<|im_start|>", "<|im_end|>" ], "
template
{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ if .Prompt }}<|im_start|>user

Readme

Cygnis-Alpha-1.7B-v0.1

Cygnis-Alpha 1.7B is a high-efficiency language model based on the SmolLM2 architecture. This v0.1 release has been optimized by Simonc-44 to integrate a sovereign system identity and improve precision in French-language interactions.

Technical Specifications

  • Architecture: SmolLM2 (1.7B Parameters)
  • Optimization: Fine-tuned for structured reasoning and sovereign identity.
  • Performance: Designed for near-instant inference on CPU, mobile, and edge computing.
  • Languages: French and English.

Implementation

To run the model locally, use the following command:

ollama run CygnisAI/Cygnis-Alpha-1.7B-v0.1

System Identity and Logic

Cygnis-Alpha operates as a Sovereign Reasoning Engine. The model is programmed to prioritize internal logic and structural consistency. It evaluates arguments and verifies reasoning paths before generating a final response, ensuring high-quality output in a compact format.

Recommended Parameters

  • Prompt Template: ChatML
  • Temperature: 0.7
  • Top_P: 0.9
  • Repeat Penalty: 1.1

Use Cases

  • Edge Computing: High-performance local AI tasks on limited hardware.
  • Privacy-First Assistance: Local execution for sensitive data processing.
  • Logical Structuring: Quick analysis and organization of complex information.

Developed by Simonc-44.