Qwen3-4B-Max-Merge

bunnycore/Qwen3-4B-Max-Merge

bunnycore/Qwen3-4B-Max-Merge is a merge of the following models using LazyMergekit:

• bunnycore/Qwen3-4B-Max-Ties
• SamuelBang/AesCoder-4B
• Pinkstack/DistilGPT-OSS-qwen3-4B
• minchyeom/Qwaifu

🧩 Configuration

models:
  - model: bunnycore/Qwen3-4B-Max-Ties
    parameters:
      density: 0.5
      weight: 0.5
  - model: SamuelBang/AesCoder-4B
    parameters:
      density: 0.2
      weight: 0.2
  - model: Pinkstack/DistilGPT-OSS-qwen3-4B
    parameters:
      density: 0.3
      weight: 0.3
  - model: minchyeom/Qwaifu
    parameters:
      density: 0.2
      weight: 0.2
merge_method: ties
base_model: bunnycore/Qwen3-4B-Max-Ties
parameters:
  normalize: true
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "bunnycore/bunnycore/Qwen3-4B-Max-Merge"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])