metadata

datasets:
  - alexandreteles/AlpacaToxicQA_ShareGPT
  - Nitral-AI/Active_RP-ShareGPT
  - PJMixers/hieunguyenminh_roleplay-deduped-ShareGPT
  - Nitral-AI/RP_Alignment-ShareGPT
  - Chaser-cz/sonnet35-charcard-roleplay-sharegpt
  - AiCloser/sharegpt_cot_dataset
  - PJMixers/Gryphe_Opus-WritingPrompts-Story2Prompt-ShareGPT
  - priveeai/pippa_sharegpt
  - Locutusque/sharegpt_gpt4_uncensored_cleaned
  - OpenCoder-LLM/opc-sft-stage1
  - OpenCoder-LLM/opc-sft-stage2
  - microsoft/orca-agentinstruct-1M-v1
  - microsoft/orca-math-word-problems-200k
  - NousResearch/hermes-function-calling-v1
  - AI-MO/NuminaMath-CoT
  - AI-MO/NuminaMath-TIR
  - allenai/tulu-3-sft-mixture
  - cognitivecomputations/dolphin-coder
  - HuggingFaceTB/smoltalk
  - cognitivecomputations/samantha-data
  - m-a-p/CodeFeedback-Filtered-Instruction
  - m-a-p/Code-Feedback
base_model:
  - NickyNicky/Llama-1B-GRPO_Final
  - xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora
  - bunnycore/FuseChat-3.2-1B-Creative-RP
  - huihui-ai/Llama-3.2-1B-Instruct-abliterated
  - prithivMLmods/Bellatrix-Tiny-1B-v3
  - cognitivecomputations/Dolphin3.0-Llama3.2-1B
library_name: transformers
tags:
  - mergekit
  - merge
language:
  - es
  - en
license: apache-2.0
pipeline_tag: text-generation
model-index:
  - name: La_Mejor_Mezcla-3.2-1B
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: wis-k/instruction-following-eval
          split: train
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 55.1
            name: averaged accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: SaylorTwift/bbh
          split: test
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 9.41
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: lighteval/MATH-Hard
          split: test
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 8.99
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          split: train
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 1.01
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          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: 0.62
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          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: 9.21
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=Novaciano%2FLa_Mejor_Mezcla-3.2-1B
          name: Open LLM Leaderboard

Mezcla

Esta es una mezcla de modelos de lenguaje pre-entrenados creado a partir de mergekit.

Detalles de la mezcla

Fue creado a partir de los que considero los mejores modelos que he usado de base para mis anteriores creaciones. Cada uno destaca en lo suyo:

Roleplay
GRPO
Uncensored
Abliterated
Gran cantidad de datasets inyectados

Método de Mezcla

Este modelo ha sido mezclado usando el método de mezcla Model Stock usando bunnycore/FuseChat-3.2-1B-Creative-RP como base.

Modelos Mezclados

Los siguientes modelos han sido incluidos en la mezcla:

Configuración

La siguiente configuración YAML ha sido usada para producir el modelo:

models:
  - model: bunnycore/FuseChat-3.2-1B-Creative-RP
  - model: NickyNicky/Llama-1B-GRPO_Final
  - model: prithivMLmods/Bellatrix-Tiny-1B-v3
  - model: xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora
  - model: cognitivecomputations/Dolphin3.0-Llama3.2-1B
  - model: huihui-ai/Llama-3.2-1B-Instruct-abliterated
merge_method: model_stock
base_model: bunnycore/FuseChat-3.2-1B-Creative-RP
dtype: bfloat16
parameters:
  t: [0, 0.5, 1, 0.5, 0]

Open LLM Leaderboard Evaluation Results

Detailed results can be found here! Summarized results can be found here!

Metric	Value (%)
Average	14.06
IFEval (0-Shot)	55.10
BBH (3-Shot)	9.41
MATH Lvl 5 (4-Shot)	8.99
GPQA (0-shot)	1.01
MuSR (0-shot)	0.62
MMLU-PRO (5-shot)	9.21