README.md

metadata

language:
  - en
  - zh
license: apache-2.0
base_model:
  - sthenno/tempesthenno-icy-0130
metrics:
  - accuracy
model-index:
  - name: tempesthenno-kto-0205-ckpt80
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 62.18
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 50.1
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 37.99
            name: exact match
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 19.69
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 19.84
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 48.65
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=sthenno/tempesthenno-kto-0205-ckpt80
          name: Open LLM Leaderboard

sthenno/tempesthenno-kto-0205-ckpt80

update: now checking for evaluations without chat templates

tempesthenno-icy-0130

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the SCE merge method using sthenno/tempesthenno-nuslerp-0124 as a base.

Models Merged

The following models were included in the merge:

• sthenno/tempesthenno-icy-0130-01
• sthenno/tempesthenno-icy-0130-02
• sthenno/tempesthenno-icy-0130-03

Configuration

The following YAML configuration was used to produce this model:

name: tempesthenno-icy-0130
merge_method: sce
parameters:
  select_topk: 0.8
  normalize: true
dtype: float32
out_dtype: bfloat16
base_model: sthenno/tempesthenno-nuslerp-0124
tokenizer:
  source: base
chat_template: chatml
models:
  - model: sthenno/tempesthenno-icy-0130-01
  - model: sthenno/tempesthenno-icy-0130-02
  - model: sthenno/tempesthenno-icy-0130-03

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	39.74
IFEval (0-Shot)	62.18
BBH (3-Shot)	50.10
MATH Lvl 5 (4-Shot)	37.99
GPQA (0-shot)	19.69
MuSR (0-shot)	19.84
MMLU-PRO (5-shot)	48.65