vortex-3b / README.md
Abhaykoul's picture
Update README.md
9a13257 verified
metadata
language:
  - en
license: other
tags:
  - HelpingAI
  - vortex
datasets:
  - OEvortex/Vortex-50k
license_name: helpingai
license_link: LICENSE.md
pipeline_tag: text-generation
model-index:
  - name: vortex-3b
    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: 31.91
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          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: 56.89
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          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: 27.32
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          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: 37.39
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          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: 60.14
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          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: 0.91
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=OEvortex/vortex-3b
          name: Open LLM Leaderboard

Vortex 3b Model Overview

vortex-3b is a 2.78 billion parameter causal language model created by OEvortex that is derived from EleutherAI's Pythia-2.8b and fine-tuned on Vortex-50k dataset'

from transformers import pipeline

# Initialize the pipeline
pipe = pipeline("text-generation", model="OEvortex/vortex-3b")

# Use the pipeline
text = "Once upon a time"
generated_text = pipe(text, max_length=100, do_sample=True)[0]['generated_text']

print(generated_text)

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric vortex 3b vortex 3b-v2 dolly-v2-3b pythia-2.8b-deduped
Avg. 35.76 37.46 25.26 36.72
AI2 Reasoning Challenge (25-Shot) 31.91 39.68 22.83 36.26
HellaSwag (10-Shot) 56.89 65.04 26.55 60.66
MMLU (5-Shot) 27.32 25.09 24.7 26.78
TruthfulQA (0-shot) 37.39 33.80 0 35.56
Winogrande (5-shot) 60.14 59.12 59.43 60.22
GSM8k (5-shot) 0.91 2.05 1.86 0.83