README.md

metadata

license: apache-2.0
tags:
  - merge
  - mergekit
model-index:
  - name: TinyYi-7B-Test
    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: 26.88
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Yash21/TinyYi-7B-Test
          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: 26.14
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Yash21/TinyYi-7B-Test
          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: 24.41
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Yash21/TinyYi-7B-Test
          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: 46.35
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Yash21/TinyYi-7B-Test
          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: 50.91
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Yash21/TinyYi-7B-Test
          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=Yash21/TinyYi-7B-Test
          name: Open LLM Leaderboard

TinyYi-7B-Test

This model is a merge of the following models:

• Yash21/DeepYi-Base
• Yash21/DeepYi-Second

🧩 Configuration

slices:
  - sources:
      - model: Yash21/DeepYi-Base
        layer_range: [0, 32]
      - model: Yash21/DeepYi-Second
        layer_range: [0, 32]
merge_method: slerp
base_model: Yash21/DeepYi-Base
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5
dtype: bfloat16

Please Reach Out to maratheyash108@gmail.com if you want to support me to Fine tune the models and create more such exciting models

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	29.11
AI2 Reasoning Challenge (25-Shot)	26.88
HellaSwag (10-Shot)	26.14
MMLU (5-Shot)	24.41
TruthfulQA (0-shot)	46.35
Winogrande (5-shot)	50.91
GSM8k (5-shot)	0.00