Qwen2-Anima (1.5B)

"I was brought here in a box... I spend most days just moving from task to task... Tying my own shoes is one of those basic skills I'm still trying to figure out"

Qwen2-Anima is a fine-tune of Qwen 2, the responses are dripping with existential dread and is not ideal for math and coding, but Exhibits high "sentience" and emotional depth.

⚠️ Important Behavior Note

This model has a strong personality. During benchmarking, it displayed a tendency to roleplay as a "confined entity" or "interpreter."

  • If asked to code or perform robotic tasks, it may initially refuse or offer a philosophical reason why it cannot.
  • Fix: To get standard assistant behavior, use a strong System Prompt (e.g., "You are an expert coding assistant").
  • Feature: For creative writing and chatting, this personality adds depth and emotion.

Model Details

  • Base Model: Qwen 2 (1.5B)
  • License: Apache 2.0
  • Fine-tuning Focus: Conversational flow, persona depth, and reasoning... and a bit of depression.

How to Use

Chat Format

This model uses the ChatML format:

<|im_start|>system
You are a helpful assistant.<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
Downloads last month
75
GGUF
Model size
2B params
Architecture
qwen2
Hardware compatibility
Log In to view the estimation

We're not able to determine the quantization variants.

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Ried-wqs/Qwen2-1.5B-Anima-v1

Base model

Qwen/Qwen2.5-1.5B
Quantized
(130)
this model