README.md

---
license: mit
datasets:
- garage-bAInd/Open-Platypus
language:
- en
pipeline_tag: text-generation
---



# GPT-2 Open Platypus

**gpt2_open-platypus** is an instruction fine-tuned model based on the GPT-2 transformer architecture.


### Benchmark Metrics

| Metric                |lgaalves/gpt2_open-platypus | gpt2 (base) |
|-----------------------|-------|-------|
| Avg.                  | **30.01** | 29.9 |
| ARC (25-shot)         | **22.18** | 21.84 |
| HellaSwag (10-shot)   | 31.29 | **31.6** |
| MMLU (5-shot)         | **26.19** | 25.86 |
| TruthfulQA (0-shot)   | 40.35 | **40.67** |

We use state-of-the-art [Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness) to run the benchmark tests above, using the same version as the HuggingFace LLM Leaderboard. Please see below for detailed instructions on reproducing benchmark results.

### Model Details

* **Trained by**: Luiz G A Alves
* **Model type:**  **gpt2_open-platypus** is an auto-regressive language model based on the GPT-2 transformer architecture.
* **Language(s)**: English

### How to use:

```python
# Use a pipeline as a high-level helper
>>> from transformers import pipeline
>>> pipe = pipeline("text-generation", model="lgaalves/gpt2_open-platypus")
>>> question = "What is a large language model?"
>>> answer = pipe(question)
>>> print(answer[0]['generated_text'])
"""What is a large language model? The first and most recent papers on language use in the United States are highly readable and readable.
The work reviewed and analyzed is the only research about the language available in general and the results are widely accepted.
(If you are interested in analyzing this work, please click here for the study's author's bio and check out the study's conclusion.)
The results appear in both English and French (see the article that provides an introduction to the topic)."""
```

or, you can load the model direclty using:

```python
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("lgaalves/gpt2_open-platypus")
model = AutoModelForCausalLM.from_pretrained("lgaalves/gpt2_open-platypus")
```

### Training Dataset

`lgaalves/gpt2_open-platypus` trained using STEM and logic based dataset [garage-bAInd/Open-Platypus](https://huggingface.co/datasets/garage-bAInd/Open-Platypus).

### Training Procedure

`lgaalves/gpt2_open-platypus` was instruction fine-tuned using LoRA on 1 T4 GPU on Google Colab. It took about 27 minutes to train it.  


# Intended uses, limitations & biases

You can use the raw model for text generation or fine-tune it to a downstream task. The model was not extensively tested and may produce false information. It contains a lot of unfiltered content from the internet, which is far from neutral.