Instructions to use SnurfyAI/is-even with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use SnurfyAI/is-even with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="SnurfyAI/is-even")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("SnurfyAI/is-even") model = AutoModelForSequenceClassification.from_pretrained("SnurfyAI/is-even") - Notebooks
- Google Colab
- Kaggle
Is Even
Is Even is a tiny decoder-only transformer trained on chat-formatted synthetic data from SnurfyAI that predicts whether a decimal input is even.
This repository contains the published model weights and metadata for the released checkpoint.
What The Model Does
- Input: a decimal numeral rendered inside a fixed ChatML prompt
- Output:
truefor even numbers andfalsefor odd numbers - Tokenization: strict digit-level tokenization with a tiny fixed vocabulary
The task is mathematically determined by the final decimal digit, but the model was trained as a small transformer classifier on synthetic examples.
Prompt Format
<|im_start|>user
123456<|im_end|>
<|im_start|>assistant
The expected completion label is true or false.
Training Code
You can train the model yourself using the official SnurfyAI training codebase:
The repository contains the code needed to generate the dataset, train the model, and benchmark saved checkpoints.
Research Paper
The accompanying paper analyzes the model’s generalization behavior, attention structure, and causal dependence on the final digit.
- Downloads last month
- 98