metadata

base_model:
  - Qwen/Qwen2-7B
  - Qwen/Qwen2-7B
license: apache-2.0
tags:
  - moe
  - frankenmoe
  - merge
  - mergekit
  - lazymergekit
  - Qwen/Qwen2-7B

qwen2-merged

qwen2-merged is a Mixture of Experts (MoE) made with the following models using LazyMergekit:

🧩 Configuration

base_model: Qwen/Qwen2-7B
gate_mode: random
architecture: qwen
dtype: bfloat16
experts:
  - source_model: Qwen/Qwen2-7B
    positive_prompts: []

  - source_model: Qwen/Qwen2-7B
    positive_prompts: []

shared_experts:
  - source_model: Qwen/Qwen2-7B
    positive_prompts: []
    residual_scale: 0.1

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Aoyinke/qwen2-merged"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
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"])