Daredevil-7B / README.md
mlabonne's picture
Adding Evaluation Results (#3)
66ada06 verified
metadata
license: cc-by-nc-4.0
tags:
  - merge
  - mergekit
  - lazymergekit
  - samir-fama/SamirGPT-v1
  - abacusai/Slerp-CM-mist-dpo
  - EmbeddedLLM/Mistral-7B-Merge-14-v0.2
base_model:
  - mistralai/Mistral-7B-v0.1
  - samir-fama/SamirGPT-v1
  - abacusai/Slerp-CM-mist-dpo
  - EmbeddedLLM/Mistral-7B-Merge-14-v0.2
model-index:
  - name: Daredevil-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: 69.37
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-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: 87.17
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-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: 65.3
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-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: 64.09
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-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: 81.29
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-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: 72.93
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=mlabonne/Daredevil-7B
          name: Open LLM Leaderboard

Daredevil-7B

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

πŸ† Evaluation

Open LLM Leaderboard

TBD.

Nous

Model AGIEval GPT4All TruthfulQA Bigbench Average
Daredevil-7B 44.85 76.07 64.89 47.07 58.22
OpenHermes-2.5-Mistral-7B 42.75 72.99 52.99 40.94 52.42
NeuralHermes-2.5-Mistral-7B 43.67 73.24 55.37 41.76 53.51
Nous-Hermes-2-SOLAR-10.7B 47.79 74.69 55.92 44.84 55.81
Marcoro14-7B-slerp 44.66 76.24 64.15 45.64 57.67
CatMarcoro14-7B-slerp 45.21 75.91 63.81 47.31 58.06

See the complete evaluation here.

🧩 Configuration

models:
  - model: mistralai/Mistral-7B-v0.1
    # No parameters necessary for base model
  - model: samir-fama/SamirGPT-v1
    parameters:
      density: 0.53
      weight: 0.4
  - model: abacusai/Slerp-CM-mist-dpo
    parameters:
      density: 0.53
      weight: 0.3
  - model: EmbeddedLLM/Mistral-7B-Merge-14-v0.2
    parameters:
      density: 0.53
      weight: 0.3
merge_method: dare_ties
base_model: mistralai/Mistral-7B-v0.1
parameters:
  int8_mask: true
dtype: bfloat16

πŸ’» Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "shadowml/Daredevil-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. 73.36
AI2 Reasoning Challenge (25-Shot) 69.37
HellaSwag (10-Shot) 87.17
MMLU (5-Shot) 65.30
TruthfulQA (0-shot) 64.09
Winogrande (5-shot) 81.29
GSM8k (5-shot) 72.93