This model is a result of merging multiple performant models, aiming to combine their strengths into a single, versatile entity. It is designed to be uncensored, capable of engaging in roleplay scenarios, and performing reasoning tasks. Its relatively small size (1.7 billion parameters) makes it efficient while still offering notable capabilities.
-
Roleplay: Generating dynamic and engaging dialogues for various character roles.
-
General Text Generation: Creating diverse forms of text, from creative writing to informative content.
-
Reasoning Tasks: Assisting with tasks that require logical deduction or problem-solving, within the scope of its parameter size.
-
Research and Exploration: For users interested in exploring the capabilities of merged, uncensored language models.
1models:
2 - model: UnfilteredAI/DAN-Qwen3-1.7B
3 parameters:
4 density: 0.5
5 weight: 0.5
6 - model: mlabonne/Qwen3-1.7B-abliterated+bunnycore/Qwen3-1.7b-Rp-lora
7 parameters:
8 density: 0.5
9 weight: 0.3
10 - model: POLARIS-Project/Polaris-1.7B-Preview
11 parameters:
12 density: 0.3
13 weight: 0.3
14merge_method: ties
15base_model: mlabonne/Qwen3-1.7B-abliterated
16parameters:
17 normalize: true
18dtype: float16
