metadata

tags:
  - merge
  - mergekit
  - lazymergekit
  - UsernameJustAnother/Nemo-12B-Marlin-v5
  - anthracite-org/magnum-12b-v2
base_model:
  - UsernameJustAnother/Nemo-12B-Marlin-v5
  - anthracite-org/magnum-12b-v2
model-index:
  - name: MagnusIntellectus-12B-v1
    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: 44.21
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          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: 33.26
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          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: 5.14
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          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: 4.59
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          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: 15.18
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          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: 26.9
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=GalrionSoftworks/MagnusIntellectus-12B-v1
          name: Open LLM Leaderboard

MagnusIntellectus-12B-v1

How pleasant, the rocks appear to have made a decent conglomerate. A-.

MagnusIntellectus is a merge of the following models using LazyMergekit:

🧩 Configuration

models:
  - model: UsernameJustAnother/Nemo-12B-Marlin-v5
    parameters:
      density: 0.4
      weight: 0.70
  - model: anthracite-org/magnum-12b-v2
    parameters:
      density: 0.6
      weight: 0.30

merge_method: ties
base_model: UsernameJustAnother/Nemo-12B-Marlin-v5
parameters:
  normalize: true
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "GalrionSoftworks/MagnusIntellectus-12B-v1"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	21.55
IFEval (0-Shot)	44.21
BBH (3-Shot)	33.26
MATH Lvl 5 (4-Shot)	5.14
GPQA (0-shot)	4.59
MuSR (0-shot)	15.18
MMLU-PRO (5-shot)	26.90