---
language:
- en
- es
license: apache-2.0
datasets:
- Intel/orca_dpo_pairs
model-index:
- name: Barcenas-Tiny-1.1b-DPO
  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: 36.26
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      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: 61.2
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      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: 24.83
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      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: 37.45
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      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: 60.93
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      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: 2.05
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Danielbrdz/Barcenas-Tiny-1.1b-DPO
      name: Open LLM Leaderboard
---
Barcenas Tiny 1.1b DPO

It is a model based on the famous TinyLlama/TinyLlama-1.1B-Chat-v1.0 and trained with DPO using the Intel/orca_dpo_pairs dataset.

With its reinforcement based training we hope to improve the Tiny model in a huge way and have a better model with better responses with a small size and accessible to most people.

Many thanks to Maxime Labonne (mlabonne) for his tutorial on how to train a LLM model using DPO, without his tutorial this model would not have been possible.

Made with ❤️ in Guadalupe, Nuevo Leon, Mexico 🇲🇽
# [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_Danielbrdz__Barcenas-Tiny-1.1b-DPO)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |37.12|
|AI2 Reasoning Challenge (25-Shot)|36.26|
|HellaSwag (10-Shot)              |61.20|
|MMLU (5-Shot)                    |24.83|
|TruthfulQA (0-shot)              |37.45|
|Winogrande (5-shot)              |60.93|
|GSM8k (5-shot)                   | 2.05|