README.md

metadata

license: apache-2.0
model-index:
  - name: GALAXY-XB-v.03
    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: 61.77
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          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: 83.59
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          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: 64.55
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          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: 44.19
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          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: 81.06
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          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: 45.03
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TeeZee/GALAXY-XB-v.03
          name: Open LLM Leaderboard

TeeZee/GALAXY-XB-v.03

Experiment, can DUS be taken one or more steps further?

Technical notes:

• 12 layers removed from both models, 4 more than in original paper but its 1/4 of all layers(48) as per original paper.
• base version of upstage/SOLAR-10.7B-v1.0 used for merge
• no finetuning done yet, this is just a merge, first step in DUS paper
• next step, if evaluation proves that its at least as 'smart' as base model, should be finetuning to 'recover' after merge

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	63.37
AI2 Reasoning Challenge (25-Shot)	61.77
HellaSwag (10-Shot)	83.59
MMLU (5-Shot)	64.55
TruthfulQA (0-shot)	44.19
Winogrande (5-shot)	81.06
GSM8k (5-shot)	45.03

Results

• small quality loss can be observed comparing to base model, as described in the DUS paper
• this merge has best evaluation results, so it will be finetuned to 'recover' from the merge
• finetunig will be done on 5-10% of openorca dataset and full DPO datasets used by SOLAR
• v03 > v01 > v02 - based on average evaluation scores, removing 1/4 of total layers seems to be the correct way to scale DUS