Chichu-2.0-500M-Instruct

Chichu 2.0 500M Instruct 🐱

A 500 million parameter language model fine-tuned from Qwen2.5-0.5B-Instruct on the r0b0tlab/qwen3.8-max-glm5.2-kimi-k3-distillation dataset — a multi-teacher distillation corpus covering math, code, reasoning, and instructions.

Named after Chichu the cat. 🐱

Model Details

• Base model: Qwen/Qwen2.5-0.5B-Instruct
• Parameters: 494M (2.16M LoRA adapters trained)
• Training: LoRA fine-tuning (rank=16, alpha=32)
• Context length: 32,768 tokens

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained("Sebastianpro88/Chichu-2.0-500M-Instruct", torch_dtype=torch.float16, device_map="cpu")
tokenizer = AutoTokenizer.from_pretrained("Sebastianpro88/Chichu-2.0-500M-Instruct")

messages = [
    {"role": "system", "content": "You are Chichu 2.0, a language model named after Chichu the cat."},
    {"role": "user", "content": "What is your name?"}
]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt")
out = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
# "My name is Chichu 2.0."