MagiD-24B

MagiD-24B

MagiD-24B is a merge of the following models using LazyMergekit:

🧩 Configuration

models:
  - model: unsloth/Mistral-Small-3.2-24B-Instruct-2506
  - model: win10/MagiDevs-24B-2506-Vision
merge_method: sce
base_model: unsloth/Mistral-Small-3.2-24B-Instruct-2506
tokenizer_source: base
parameters:
  select_topk: 1.0
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "win10/MagiD-24B"
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"])

Ways to support me:

1.ko-fi:

https://ko-fi.com/ogodwin10

1. Non-Taiwan users:

https://www.paypal.com/ncp/payment/6KQGRXF5KZJCY

1. Japan:

https://booth.pm/zh-tw/items/6997257

1. Send me a private message (commission me)