README.md

metadata

language:
  - en
license: apache-2.0
model-index:
  - name: luxia-21.4b-alignment-v1.0
    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: 77.47
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          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: 91.88
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          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: 68.1
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          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: 79.17
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          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: 87.45
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          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: 62.4
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=saltlux/luxia-21.4b-alignment-v1.0
          name: Open LLM Leaderboard

Introduction

We introduce luxia-21.4b-alignment-v1.0, an instruction-tuned and alignment model based on luxia-21.4b. Please refer to the evaluation results table for details.

Instruction Fine-tuning Strategy

We utilize state-of-the-art instruction fine-tuning methods including supervised fine-tuning (SFT) and direct preference optimization (DPO)

Data Contamination Test Results

Results will be updated soon.

Evaluation Results

Results will be updated soon.

Usage Instructions

How to use

# pip install transformers==4.35.2
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("saltlux/luxia-21.4b-alignment-v0.1")
model = AutoModelForCausalLM.from_pretrained(
    "saltlux/luxia-21.4b-alignment-v0.1",
    device_map="auto",
    torch_dtype=torch.float16,
)

License

• saltlux/luxia-21.4b-alignment-v1.0: apache-2.0

Contact Us

Any questions and suggestions are welcomed at the discussion tab.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	77.74
AI2 Reasoning Challenge (25-Shot)	77.47
HellaSwag (10-Shot)	91.88
MMLU (5-Shot)	68.10
TruthfulQA (0-shot)	79.17
Winogrande (5-shot)	87.45
GSM8k (5-shot)	62.40