README.md

metadata

license: apache-2.0
tags:
  - merge
  - mergekit
  - lazymergekit
  - argilla/distilabeled-Marcoro14-7B-slerp
  - mlabonne/NeuralMarcoro14-7B
datasets:
  - mlabonne/chatml_dpo_pairs
  - argilla/distilabel-intel-orca-dpo-pairs
model-index:
  - name: NeuDist-Ro-7B
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: AI2 Reasoning Challenge (25-Shot)
          type: ai2_arc
          config: ARC-Challenge
          split: test
          args:
            num_few_shot: 25
        metrics:
          - type: acc_norm
            value: 71.25
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: HellaSwag (10-Shot)
          type: hellaswag
          split: validation
          args:
            num_few_shot: 10
        metrics:
          - type: acc_norm
            value: 87.48
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU (5-Shot)
          type: cais/mmlu
          config: all
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 65.13
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: TruthfulQA (0-shot)
          type: truthful_qa
          config: multiple_choice
          split: validation
          args:
            num_few_shot: 0
        metrics:
          - type: mc2
            value: 64.93
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: Winogrande (5-shot)
          type: winogrande
          config: winogrande_xl
          split: validation
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 82.08
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GSM8k (5-shot)
          type: gsm8k
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 70.96
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=flemmingmiguel/NeuDist-Ro-7B
          name: Open LLM Leaderboard

NeuDist-Ro-7B

NeuDist-Ro-7B is a merge of the following models using LazyMergekit:

• argilla/distilabeled-Marcoro14-7B-slerp
• mlabonne/NeuralMarcoro14-7B

As an experiment to find the best base merge to further fine-tuning, expect a lot of experiments named using parts of the component models until a clear winner emerges in the benchmarks

In this case merging 2 DPOs of the same model

🧩 Configuration

slices:
  - sources:
      - model:  argilla/distilabeled-Marcoro14-7B-slerp
        layer_range: [0, 32]
      - model: mlabonne/NeuralMarcoro14-7B
        layer_range: [0, 32]
merge_method: slerp
base_model: mlabonne/NeuralMarcoro14-7B
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "flemmingmiguel/NeuDist-Ro-7B"
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.	73.64
AI2 Reasoning Challenge (25-Shot)	71.25
HellaSwag (10-Shot)	87.48
MMLU (5-Shot)	65.13
TruthfulQA (0-shot)	64.93
Winogrande (5-shot)	82.08
GSM8k (5-shot)	70.96