Flan T5 Base Arxiv Math Question Answering

🚀 FLAN - T5 Math Question - Answering Model

A FLAN - T5 model trained on arxiv math question - answering data for research purposes.

🚀 Quick Start

This is a FLAN - T5 model trained on [ArtifactAI/arxiv - math - instruct - 50k](https://huggingface.co/datasets/ArtifactAI/arxiv - math - instruct - 50k). It's for research only and should not be used in production settings as the output is highly unreliable.

✨ Features

• Model Type: Language model
• Language: English
• License: Apache 2.0
• Related Models: [All FLAN - T5 Checkpoints](https://huggingface.co/models?search=flan - t5)

📦 Installation

No specific installation steps are provided in the original README. If you want to use the model, you need to install the transformers library and other necessary dependencies according to the usage examples.

💻 Usage Examples

Basic Usage

Running the model on a CPU

from transformers import T5Tokenizer, T5ForConditionalGeneration

tokenizer = T5Tokenizer.from_pretrained("ArtifactAI/flan-t5-base-arxiv-math-question-answering")
model = T5ForConditionalGeneration.from_pretrained("ArtifactAI/flan-t5-base-arxiv-math-question-answering")

input_text = "What is the spectral isolation of bi-invariant metrics?"
input_ids = tokenizer(input_text, return_tensors="pt").input_ids

outputs = model.generate(input_ids)
print(tokenizer.decode(outputs[0]))

Running the model on a GPU

# pip install accelerate
from transformers import T5Tokenizer, T5ForConditionalGeneration

tokenizer = T5Tokenizer.from_pretrained("ArtifactAI/flan-t5-base-arxiv-math-question-answering")
model = T5ForConditionalGeneration.from_pretrained("ArtifactAI/flan-t5-base-arxiv-math-question-answering", device_map="auto")

input_text = "What is the spectral isolation of bi-invariant metrics?"
input_ids = tokenizer(input_text, return_tensors="pt").input_ids.to("cuda")

outputs = model.generate(input_ids)
print(tokenizer.decode(outputs[0]))

Advanced Usage

Running the model in an HF pipeline (FP16)

# load model and tokenizer from huggingface hub with pipeline
qa = pipeline("summarization", model="ArtifactAI/flan-t5-base-arxiv-math-question-answering")

query = "What is the spectral isolation of bi-invariant metrics?"
print(f"query: {query}")
res = qa("answer: " + query)

print(f"{res[0]['summary_text']}")

📚 Documentation

Training Details

Training Data

The model was trained on [ArtifactAI/arxiv - math - instruct - 50k](https://huggingface.co/datasets/ArtifactAI/arxiv - math - instruct - 50k), a dataset of question/answer pairs. Questions are generated using the t5 - base model, while the answers are generated using the GPT - 3.5 - turbo model.

📄 License

This model is licensed under the Apache 2.0 license.

📄 Citation

@misc{flan-t5-base-arxiv-math-question-answering,
    title={flan-t5-base-arxiv-math-question-answering},
    author={Matthew Kenney},
    year={2023}
}