README.md

metadata

license: apache-2.0
model-index:
  - name: dopeyplats-1.1b-2T-v1
    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: 33.11
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          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: 54.31
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          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: 24.55
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          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: 39.26
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          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: 58.8
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          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: 1.67
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=vihangd/dopeyplats-1.1b-2T-v1
          name: Open LLM Leaderboard

DopeyPlats-1.1b V2

An experimental finetune of TinyLLaMA 1.1b 2T with Alpaca-QLoRA with some dpo goodness

Datasets

Trained on alpca style datasets

Prompt Template

Uses alpaca style prompt template # [Open LLM Leaderboard Evaluation Results](https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard) Detailed results can be found [here](https://huggingface.co./datasets/open-llm-leaderboard/details_vihangd__dopeyplats-1.1b-2T-v1)

Metric	Value
Avg.	35.28
AI2 Reasoning Challenge (25-Shot)	33.11
HellaSwag (10-Shot)	54.31
MMLU (5-Shot)	24.55
TruthfulQA (0-shot)	39.26
Winogrande (5-shot)	58.80
GSM8k (5-shot)	1.67