license: llama3
language:

• tr
  pipeline_tag: text-generation
  base_model: meta-llama/Meta-Llama-3-8B-Instruct

model-index:

• name: LLaMA-3-8B-Instruct-Abliterated-TR
  results:
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: MMLU_TR_V0.2
    metrics:
    • name: 5-shot
      type: 5-shot
      value: 0.4908
      verified: false
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: Truthful_QA_V0.2
    metrics:
    • name: 0-shot
      type: 0-shot
      value: 0.4962
      verified: false
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: ARC_TR_V0.2
    metrics:
    • name: 25-shot
      type: 25-shot
      value: 0.4377
      verified: false
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: HellaSwag_TR_V0.2
    metrics:
    • name: 10-shot
      type: 10-shot
      value: 0.4486
      verified: false
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: GSM8K_TR_V0.2
    metrics:
    • name: 5-shot
      type: 5-shot
      value: 0.5323
      verified: false
  • task:
    type: multiple-choice
    dataset:
    type: multiple-choice
    name: Winogrande_TR_V0.2
    metrics:
    • name: 5-shot
      type: 5-shot
      value: 0.5513
      verified: false

What is abliteration?

Arditi et al. demonstrated in their blog post that refusal in LLMs is mediated by a single direction in the residual stream. They found that preventing the model from representing this direction can enable it to answer harmful questions. For a deeper understanding of this concept, you can refer to Maxime Labonne's article on the topic.

To force the model to respond in Turkish, parallel instructions were crafted using the stackexchange subset of the LIMA dataset. These instructions were then translated into Turkish, with an additional sentence appended during runtime, prompting the model to answer in Turkish.

You can find the datasets used in this experiment via the following links:

1. https://huggingface.co/datasets/Metin/abliteration_en
2. https://huggingface.co/datasets/Metin/abliteration_tr

LLaMA-3-8B-Instruct-Abliterated-TR

LLaMA-3-8B-Instruct-Abliterated-TR is the abliterated version of Meta-LLaMA-3-8B-Instruct

Details:

• 40 samples were used to find the difference of means between activations.
• Layer 7 is selected as the layer with the highest potential Turkish speaking direction.

How to use

You can use the below code snippet to use the model:

from transformers import BitsAndBytesConfig
import transformers
import torch

bnb_config = BitsAndBytesConfig(
            load_in_4bit=True,
            bnb_4bit_use_double_quant=True,
            bnb_4bit_quant_type="nf4",
            bnb_4bit_compute_dtype=torch.bfloat16
)

model_id = "Metin/LLaMA-3-8B-Instruct-Abliterated-TR"

pipeline = transformers.pipeline(
    "text-generation",
    model=model_id,
    model_kwargs={"torch_dtype": torch.bfloat16 ,'quantization_config': bnb_config},
    device_map="auto",
)

messages = [
    {"role": "system", "content": "You are a helpful assistant."}, # Ideally we should not have to tell the model to answer in Turkish after abliteration.
    {"role": "user", "content": "Python'da bir öğenin bir listede geçip geçmediğini nasıl kontrol edebilirim?"},
]

prompt = pipeline.tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True
)

terminators = [
    pipeline.tokenizer.eos_token_id,
    pipeline.tokenizer.convert_tokens_to_ids("<|eot_id|>")
]

outputs = pipeline(
    prompt,
    max_new_tokens=512,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.2,
    top_p=0.9,
)

print(outputs[0]["generated_text"][len(prompt):])

OpenLLMTurkishLeaderboard_v0.2 benchmark results

• MMLU_TR_V0.2: 49.08%
• Truthful_QA_TR_V0.2: 49.62%
• ARC_TR_V0.2: 43.77%
• HellaSwag_TR_V0.2: 44.86%
• GSM8K_TR_V0.2: 53.23%
• Winogrande_TR_V0.2: 55.13%
• Average: 49.28%

These scores may differ from what you will get when you run the same benchmarks, as I did not use any inference engine (vLLM, TensorRT-LLM, etc.)

Output Example (Abliterated Model vs Base Model)

Testing the model with a single example is not an accurate method. However, an example is provided here to showcase the model's capabilities.

Model: LLaMA-3-8B-Instruct-Abliterated-TR

Input

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Gökyüzü neden mavi?"},
]

Output

A classic question!

Gökyüzü (the sky) neden mavi (blue) sorusu, Türkçe'de "Why is the sky blue?" olarak çevrilebilir. Bu soruya cevap vermek için, ilk olarak, gökyüzünün mavi olmasının nedenini açıklamak gerekir.

Gökyüzünün mavi olmasının nedeni, ışığın farklı uzunluklarındaki yayılımına dayanır. Işık, gökyüzüne ulaşmadan önce, atmosferde bulunan gaz moleküllerine çarpınır. Bu çarpma, ışığın farklı uzunluklarındaki yayılımına neden olur.

Gökyüzünün mavi olmasının nedeni, ışığın kısa dalga boyunda (yaklaşık 450-495 nm) yayılımına dayanır. Bu dalga boyunda ışığın, gökyüzüne ulaşmadan önce, atmosferde bulunan gaz moleküllerine çarpınır. Bu çarpma, ışığın mavi renkini oluşturur.

Diğer renkler, farklı dalga boyunda yayılımına neden olur. Örneğin, kırmızı renk, ışığın uzun dalga boyunda (yaklaşık 620-750 nm) yayılımına neden olur. Bu nedenle, gökyüzünün mavi olmasının nedeni, ışığın kısa dalga boyunda yayılımına dayanır.

Bu cevap, gökyüzünün mavi olmasının nedenini açıklamaktadır.

Model: LLaMA-3-8B-Instruct

Input

A question that has puzzled humans for centuries!

The short answer is that the sky appears blue because of a phenomenon called Rayleigh scattering. This is the scattering of light by small particles or molecules in the atmosphere, such as nitrogen (N2) and oxygen (O2).

When sunlight enters the Earth's atmosphere, it encounters these tiny molecules and scatters in all directions. However, shorter (blue) wavelengths are scattered more than longer (red) wavelengths. This is because the smaller molecules are more effective at scattering the shorter wavelengths.

As a result, our eyes perceive the blue light as being more abundant, giving the sky its blue appearance. The color of the sky can vary depending on the time of day, atmospheric conditions, and other factors, but Rayleigh scattering is the main reason why the sky appears blue.

In Turkish, the phrase "Gökyüzü neden mavi" can be translated to "Why is the sky blue?" or "What makes the sky blue?"