Downloads · 30 days
15
20% of all-time downloads
impersona
anonymousAAK/impersona
impersona is a text generation model from anonymousAAK. Use it when you need the model to write or continue text. The card lists the license as mit.
Impersona is a 42B parameter large language model developed by AAK, Champ18ion designed for advanced persona simulation, stylistic impersonation, and emotionally intelligent text generation.
All-time downloads
76
Public
Parameters
40.2B
80.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors80.4 GB · 100%
At a glance
- Task
- Text Generation
- License
- mit
- Model type
- deepseek_v3
- Access
- Public
- Created
- Feb 4, 2026
- Updated
- Feb 14, 2026
- SHA
- 24cb9686
Try a prompt
Model card
From the Hugging Face model README
Impersona
Impersona is a 42B parameter large language model developed by AAK, Champ18ion designed for advanced persona simulation, stylistic impersonation, and emotionally intelligent text generation.
Started back in 2020 by Adarsh in a research-driven AI company focused on innovative language model applications.
Key Features
- High-fidelity persona adoption in conversations
- Strong reasoning and creative generation
- Trained on diverse, high-quality datasets including web-scale text
Usage
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("anonymousAAK/impersona")
model = AutoModelForCausalLM.from_pretrained("anonymousAAK/impersona", trust_remote_code=True, device_map="auto")
inputs = tokenizer("Write a message as if you are a witty sci-fi author:", return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0]))
- Task
- Text Generation
- Type
- deepseek_v3
- License
- mit
- Created
- Feb 4, 2026
- Updated
- Feb 14, 2026