Quantization made by Richard Erkhov.

orthogonal-2x7B-base - GGUF

Model creator: https://huggingface.co/LoSboccacc/
Original model: https://huggingface.co/LoSboccacc/orthogonal-2x7B-base/

Name	Quant method	Size
orthogonal-2x7B-base.Q2_K.gguf	Q2_K	4.43GB
orthogonal-2x7B-base.IQ3_XS.gguf	IQ3_XS	4.95GB
orthogonal-2x7B-base.IQ3_S.gguf	IQ3_S	5.22GB
orthogonal-2x7B-base.Q3_K_S.gguf	Q3_K_S	5.2GB
orthogonal-2x7B-base.IQ3_M.gguf	IQ3_M	5.35GB
orthogonal-2x7B-base.Q3_K.gguf	Q3_K	5.78GB
orthogonal-2x7B-base.Q3_K_M.gguf	Q3_K_M	5.78GB
orthogonal-2x7B-base.Q3_K_L.gguf	Q3_K_L	6.27GB
orthogonal-2x7B-base.IQ4_XS.gguf	IQ4_XS	6.5GB
orthogonal-2x7B-base.Q4_0.gguf	Q4_0	6.78GB
orthogonal-2x7B-base.IQ4_NL.gguf	IQ4_NL	6.85GB
orthogonal-2x7B-base.Q4_K_S.gguf	Q4_K_S	6.84GB
orthogonal-2x7B-base.Q4_K.gguf	Q4_K	7.25GB
orthogonal-2x7B-base.Q4_K_M.gguf	Q4_K_M	7.25GB
orthogonal-2x7B-base.Q4_1.gguf	Q4_1	7.52GB
orthogonal-2x7B-base.Q5_0.gguf	Q5_0	8.26GB
orthogonal-2x7B-base.Q5_K_S.gguf	Q5_K_S	8.26GB
orthogonal-2x7B-base.Q5_K.gguf	Q5_K	8.51GB
orthogonal-2x7B-base.Q5_K_M.gguf	Q5_K_M	8.51GB
orthogonal-2x7B-base.Q5_1.gguf	Q5_1	9.01GB
orthogonal-2x7B-base.Q6_K.gguf	Q6_K	9.84GB
orthogonal-2x7B-base.Q8_0.gguf	Q8_0	12.75GB

Original model description:

model-index: - name: orthogonal-2x7B-base 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: 66.89 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base 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: 85.54 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base 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: 62.49 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base 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: 66.0 source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base 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: 77.03 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base 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: 50.8 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=LoSboccacc/orthogonal-2x7B-base name: Open LLM Leaderboard

    base_model: mistralai/Mistral-7B-Instruct-v0.2
    gate_mode: hidden # one of "hidden", "cheap_embed", or "random"
    dtype: bfloat16 # output dtype (float32, float16, or bfloat16)
    experts:
      - source_model: SanjiWatsuki/Silicon-Maid-7B
        positive_prompts:
            - "roleplay"
      - source_model: mistralai/Mistral-7B-Instruct-v0.2
        positive_prompts:
            - "chat"

chatml format

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	68.13
AI2 Reasoning Challenge (25-Shot)	66.89
HellaSwag (10-Shot)	85.54
MMLU (5-Shot)	62.49
TruthfulQA (0-shot)	66.00
Winogrande (5-shot)	77.03
GSM8k (5-shot)	50.80