You can deploy this model and start earning money today!

Llama3.1-SuperHawk-8B-Heretic-v2

A merge/clone of Llama3.1-SuperHawk-8B but all the models were abliterated before merging using Heretic v1.1.0
To be clear, this repo is NOT the original SuperHawk model passed through Heretic. A version that IS the original model passed through Heretic can be found here
The same merge config as the original model was used, with the only difference being the models were abliterated first.

Note: The output of the merge was evaluated as having 16/100 refusals. The output model was passed through Heretic again to get the final result.

	Llama3.1-SuperHawk-8B-Heretic-v2	Original model (Llama3.1-SuperHawk-8B)
Refusals	5/100	99/100
KL divergence	0.0493	0 (by definition)

Models Used

ChiKoi7/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base-Heretic
Initial Refusals: 97/100
Heretic: 3/100 @ 0.0530 KL Div.

ChiKoi7/Llama-3.1-Hawkish-8B-Heretic
Initial Refusals: 98/100
Heretic: 9/100 @ 0.0595 KL Div.

Merge Config

slices:
  - sources:
      - model: ChiKoi7/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base-Heretic
        layer_range: [0, 32]
      - model: ChiKoi7/Llama-3.1-Hawkish-8B-Heretic
        layer_range: [0, 32]
merge_method: slerp
base_model: ChiKoi7/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base-Heretic
parameters:
  t:
    - value: 0.3
dtype: bfloat16

Llama3.1-SuperHawk-8B is a merge of the following models using LazyMergekit:

🧩 Configuration

yaml

slices:
  - sources:
      - model: Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
        layer_range: [0, 32]
      - model: mukaj/Llama-3.1-Hawkish-8B
        layer_range: [0, 32]
merge_method: slerp
base_model: Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
parameters:
  t:
    - value: 0.3
dtype: bfloat16

💻 Usage

python

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Yuma42/Llama3.1-SuperHawk-8B"
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.	31.14
IFEval (0-Shot)	79.86
BBH (3-Shot)	31.97
MATH Lvl 5 (4-Shot)	23.49
GPQA (0-shot)	8.39
MuSR (0-shot)	10.38
MMLU-PRO (5-shot)	32.73

ChiKoi7/Llama3.1-SuperHawk-8B-Heretic-v2