PAIN_LeDestroy-7B-task_arithmetic

PAIN_LeDestroy-7B-task_arithmetic is a merge of the following models using LazyMergekit:

• ajibawa-2023/Code-Mistral-7B
• fblgit/UNA-dolphin-2.6-mistral-7b-dpo-laser
• DreadPoor/AllOverButTheCrying-7B-slerp

🧩 Configuration

merge_method: task_arithmetic
base_model: DreadPoor/CeramicMaiden-7B-Slerp
models:
  - model: DreadPoor/CeramicMaiden-7B-Slerp
  - model: ajibawa-2023/Code-Mistral-7B
    parameters:
      weight: 0.28
  - model: fblgit/UNA-dolphin-2.6-mistral-7b-dpo-laser
    parameters:
      weight: 0.28
  - model: DreadPoor/AllOverButTheCrying-7B-slerp
    parameters:
      weight: 0.14
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "DreadPoor/PAIN_LeDestroy-7B-task_arithmetic"
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"])

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	71.47
AI2 Reasoning Challenge (25-Shot)	68.26
HellaSwag (10-Shot)	86.61
MMLU (5-Shot)	65.08
TruthfulQA (0-shot)	63.90
Winogrande (5-shot)	81.85
GSM8k (5-shot)	63.15