README.md

metadata

license: cc-by-nc-4.0
tags:
  - merge
  - mergekit
  - lazymergekit
base_model:
  - mlabonne/OmniTruthyBeagle-7B-v0
  - mlabonne/NeuBeagle-7B
  - mlabonne/NeuralOmniBeagle-7B
model-index:
  - name: Monarch-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: 73.04
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-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: 89.03
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-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: 64.41
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-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: 77.35
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-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: 84.61
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-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: 69.07
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Monarch-7B
          name: Open LLM Leaderboard

Monarch-7B

Update 13/02/24: Monarch-7B is the best-performing model on the YALL leaderboard.

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

• mlabonne/OmniTruthyBeagle-7B-v0
• mlabonne/NeuBeagle-7B
• mlabonne/NeuralOmniBeagle-7B

🏆 Evaluation

The evaluation was performed using LLM AutoEval on Nous suite. See the entire leaderboard here.

Model	Average	AGIEval	GPT4All	TruthfulQA	Bigbench
Monarch-7B 📄	62.68	45.48	77.07	78.04	50.14
teknium/OpenHermes-2.5-Mistral-7B 📄	52.42	42.75	72.99	52.99	40.94
mlabonne/NeuralHermes-2.5-Mistral-7B 📄	53.51	43.67	73.24	55.37	41.76
mlabonne/NeuralBeagle14-7B 📄	60.25	46.06	76.77	70.32	47.86
eren23/dpo-binarized-NeuralTrix-7B 📄	62.5	44.57	76.34	79.81	49.27
CultriX/NeuralTrix-7B-dpo 📄	62.5	44.61	76.33	79.8	49.24

🧩 Configuration

models:
  - model: mistralai/Mistral-7B-v0.1
    # no parameters necessary for base model
  - model: mlabonne/OmniTruthyBeagle-7B-v0
    parameters:
      density: 0.65
      weight: 0.36
  - model: mlabonne/NeuBeagle-7B
    parameters:
      density: 0.6
      weight: 0.34
  - model: mlabonne/NeuralOmniBeagle-7B 
    parameters:
      density: 0.6
      weight: 0.3
merge_method: dare_ties
base_model: mistralai/Mistral-7B-v0.1
parameters:
  int8_mask: true
dtype: bfloat16
random_seed: 0

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "mlabonne/Monarch-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.	76.25
AI2 Reasoning Challenge (25-Shot)	73.04
HellaSwag (10-Shot)	89.03
MMLU (5-Shot)	64.41
TruthfulQA (0-shot)	77.35
Winogrande (5-shot)	84.61
GSM8k (5-shot)	69.07