Rune-14b

Rune-14b is a merge of the following models using LazyMergekit:

• Quazim0t0/time-14b-stock
• Quazim0t0/Mithril-14B-sce

🧩 Configuration

base_model: Quazim0t0/Mithril-14B-sce
dtype: bfloat16
merge_method: slerp
parameters:
  t:
  - filter: self_attn
    value: [0.0, 0.5, 0.3, 0.7, 1.0]
  - filter: mlp
    value: [1.0, 0.5, 0.7, 0.3, 0.0]
  - value: 0.5
slices:
- sources:
  - layer_range: [0, 40]
    model: Quazim0t0/time-14b-stock
  - layer_range: [0, 40]
    model: Quazim0t0/Mithril-14B-sce

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Quazim0t0/Rune-14b"
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.	41.82
IFEval (0-Shot)	70.16
BBH (3-Shot)	56.05
MATH Lvl 5 (4-Shot)	45.85
GPQA (0-shot)	13.53
MuSR (0-shot)	16.29
MMLU-PRO (5-shot)	49.02