README.md

---
language:
- en
license: apache-2.0
tags:
- merge
model-index:
- name: palmer-002.5
  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: 37.54
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      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: 61.84
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      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: 25.21
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      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: 40.22
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      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: 66.38
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      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: 1.97
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=appvoid/palmer-002.5
      name: Open LLM Leaderboard
---

Creative writing has never been so accesible, palmer goes beyond what it was thought about small language models. This model is a "MErging of Experts" (MEoE) using `palmer-002-2401` as base, biased as an assistant without using any prompts—as a result of these efforts—palmer is better than most 1b language models on most benchmarks, despite being sometimes 40% smaller than its counterparts.

```
	              MMLU     ARC-C    OBQA   HellaSwag  PIQA  Winogrande Average
tinyllama-chat | 0.2470 | 0.3285 | 0.3740 | 0.6037 | 0.7448 | 0.6022 | 0.4833 |
zyte-1b	       | 0.2397 | 0.3353 | 0.3700 | 0.6086 | 0.7541 | 0.5998 | 0.4845 |
palmer-002.5   | 0.2534 | 0.3370 | 0.3740 | 0.6128 | 0.7486 | 0.6535 | 0.4965 |
qwen-1-8       | 0.4536 | 0.3490 | 0.3320 | 0.5876 | 0.7307 | 0.5896 | 0.5070 |
```

This work constitutes, given its compactness, an advancement towards SMLs, easily empowering edge devices such as mobile phones, raspberry pis and automated software/robots. Aditionally, palmer-002.5 deviates its main philosophy from palmer-family to become a more powerful model with more data instead of less.

```
prompt: Reality is but
output: a dream,
And the dreams we make are our reality.

The world is a canvas, painted by our minds,
And we can make it a masterpiece.

So let us create, let us dream,
And let our imagination run wild.

For in our imagination lies our power,
To create a world that is truly our own.
```

You can support me [through kofi](https://ko-fi.com/appvoid)

Note that since this model uses a transformer architecture as any popular language model, its output sometimes contains hallucinations (make mistakes or false statements), and as such, it must be used with caution on sensitive scenarios.
# [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_appvoid__palmer-002.5)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |38.86|
|AI2 Reasoning Challenge (25-Shot)|37.54|
|HellaSwag (10-Shot)              |61.84|
|MMLU (5-Shot)                    |25.21|
|TruthfulQA (0-shot)              |40.22|
|Winogrande (5-shot)              |66.38|
|GSM8k (5-shot)                   | 1.97|