README.md

metadata

license: apache-2.0
library_name: transformers
tags:
  - mergekit
  - merge
base_model:
  - huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated-v2
  - bunnycore/Qwen-2.5-7B-1M-RRP-v1-lora
model-index:
  - name: DSR1-Distill-Qwen-7B-RP
    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: 36.09
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          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: 19.85
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          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: 48.04
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          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: 9.28
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          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: 8.8
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          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: 22.53
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=Triangle104/DSR1-Distill-Qwen-7B-RP
          name: Open LLM Leaderboard

Merge

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

Quants:

https://huggingface.co/Triangle104/DSR1-Distill-Qwen-7B-RP-Q4_K_M-GGUF

https://huggingface.co/mradermacher/DSR1-Distill-Qwen-7B-RP-GGUF

https://huggingface.co/mradermacher/DSR1-Distill-Qwen-7B-RP-i1-GGUF

Merge Method

This model was merged using the Passthrough merge method using huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated-v2 + bunnycore/Qwen-2.5-7B-1M-RRP-v1-lora as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

base_model: huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated-v2+bunnycore/Qwen-2.5-7B-1M-RRP-v1-lora
dtype: bfloat16
merge_method: passthrough
models:
  - model: huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated-v2+bunnycore/Qwen-2.5-7B-1M-RRP-v1-lora
tokenizer_source: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	24.10
IFEval (0-Shot)	36.09
BBH (3-Shot)	19.85
MATH Lvl 5 (4-Shot)	48.04
GPQA (0-shot)	9.28
MuSR (0-shot)	8.80
MMLU-PRO (5-shot)	22.53