metadata

language:
  - en
license: llama2
tags:
  - information retrieval
  - reranker
inference: false
model-index:
  - name: rank_vicuna_7b_v1_fp16
    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: 44.62
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          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: 65.67
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          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: 44.14
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          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: 45.13
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          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: 66.61
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          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: 0
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=castorini/rank_vicuna_7b_v1_fp16
          name: Open LLM Leaderboard

RankVicuna (FP16) Model Card

Model Details

RankVicuna is a chat assistant trained by fine-tuning Llama 2 on user-shared conversations collected from ShareGPT.

Developed by: Castorini
Model type: An auto-regressive language model based on the transformer architecture
License: Llama 2 Community License Agreement
Finetuned from base model: Llama 2

This specific model is a 7B variant and is trained with data augmentation. It is also worth noting that it is converted to FP16.

Model Sources

Repository: https://github.com/castorini/rank_llm
Paper: https://arxiv.org/abs/2309.15088

Uses

The primary use of RankVicuna is research at the intersection of large language models and retrieval. The primary intended users of the model are researchers and hobbyists in natural language processing and information retrieval.

Training Details

RankVicuna is finetuned from lmsys/vicuna-7b-v1.5 with supervised instruction fine-tuning.

Evaluation

RankVicuna is currently evaluated on DL19/DL20. See more details in our paper.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	44.36
AI2 Reasoning Challenge (25-Shot)	44.62
HellaSwag (10-Shot)	65.67
MMLU (5-Shot)	44.14
TruthfulQA (0-shot)	45.13
Winogrande (5-shot)	66.61
GSM8k (5-shot)	0.00