README.md

metadata

language:
  - en
license: mit
base_model:
  - mistralai/Mistral-7B-v0.1
datasets:
  - argilla/ultrafeedback-binarized-preferences-cleaned
pipeline_tag: text-generation
model-index:
  - name: Mistral-ORPO-β
    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
            name: normalized accuracy
            value: 61.18
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - 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
            name: normalized accuracy
            value: 84.03
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - 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: 47.69
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - 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
            name: accuracy
            value: 39.8
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - 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
            name: accuracy
            value: 63.26
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - 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
            name: accuracy
            value: 79.24
        source:
          name: Open LLM Leaderboard
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaist-ai%2Fmistral-orpo-beta
      - task:
          type: text-generation
        dataset:
          name: AlpacaEval 1
          type: AlpacaEval
        metrics:
          - type: AlpacaEval 1.0
            value: 91.16%
            name: Win Rate
        source:
          url: https://tatsu-lab.github.io/alpaca_eval/
          name: Leaderboard
      - task:
          type: text-generation
        dataset:
          name: AlpacaEval 2
          type: AlpacaEval
        metrics:
          - type: AlpacaEval 2.0
            value: 12.57%
            name: Win Rate
        source:
          url: https://tatsu-lab.github.io/alpaca_eval/
          name: Leaderboard
      - task:
          type: text-generation
        dataset:
          name: MT-Bench
          type: MT-Bench
        metrics:
          - type: MT-Bench
            value: 7.322
            name: Score
        source:
          url: https://github.com/lm-sys/FastChat/blob/main/fastchat/llm_judge/
          name: self-reported

Mistral-ORPO-β (7B)

Mistral-ORPO is a fine-tuned version of mistralai/Mistral-7B-v0.1 using the odds ratio preference optimization (ORPO). With ORPO, the model directly learns the preference without the supervised fine-tuning warmup phase. Mistral-ORPO-β is fine-tuned exclusively on the 61k instances of the cleaned version of UltraFeedback, argilla/ultrafeedback-binarized-preferences-cleaned, by Argilla.

• Github Repository: https://github.com/xfactlab/orpo

👍 Model Performance

1) AlpacaEval & MT-Bench

Model Name	Size	Align	MT-Bench	AlpacaEval 1.0	AlpacaEval 2.0
Mistral-ORPO-⍺	7B	ORPO	7.23	87.92	11.33
Mistral-ORPO-β	7B	ORPO	7.32	91.41	12.20
Zephyr β	7B	DPO	7.34	90.60	10.99
TULU-2-DPO	13B	DPO	7.00	89.5	10.12
Llama-2-Chat	7B	RLHF	6.27	71.37	4.96
Llama-2-Chat	13B	RLHF	6.65	81.09	7.70

2) IFEval

Model Type	Prompt-Strict	Prompt-Loose	Inst-Strict	Inst-Loose
Mistral-ORPO-⍺	0.5009	0.5083	0.5995	0.6163
Mistral-ORPO-β	0.5287	0.5564	0.6355	0.6619

🗺️ MT-Bench by Category

🖥️ Inference

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("kaist-ai/mistral-orpo-beta")
tokenizer = AutoTokenizer.from_pretrained("kaist-ai/mistral-orpo-beta")

# Apply chat template
query = [{'role': 'user', 'content': 'Hi! How are you doing?'}]
prompt = tokenizer.apply_chat_template(query, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors='pt')

# Generation with specific configurations
output = model.generate(
  **inputs,
  max_new_tokens=128,
  do_sample=True,
  temperature=0.7
)
response = tokenizer.batch_decode(output)

#<|user|>
#Hi! How are you doing?</s>
#<|assistant|>
#I'm doing well, thank you! How are you?</s>

📎 Citation

@misc{hong2024orpo,
      title={ORPO: Monolithic Preference Optimization without Reference Model}, 
      author={Jiwoo Hong and Noah Lee and James Thorne},
      year={2024},
      eprint={2403.07691},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}