Downloads · 30 days
8
20% of all-time downloads
hamishivi/llama-3.1-tulu-3-arena-hard-939k-sft
llama-3.1-tulu-3-arena-hard-939k-sft is a machine learning model from hamishivi. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as llama3.1.
This is a model trained on 939k samples selected by RDS+ using Arena Hard samples from the Tulu 3 unfiltered dataset. For more details, please see the paper Practical Large-Scale Data Selection for Instruction Tuning…
Downloads · 30 days
8
20% of all-time downloads
All-time downloads
41
Public
Repo size
17.2 GB
Likes
0
Public
Click a slice to open those files.
.bin16.1 GB · 93%
From the Hugging Face model README
This is a model trained on 939k samples selected by RDS+ using Arena Hard samples from the Tulu 3 unfiltered dataset. For more details, please see the paper Practical Large-Scale Data Selection for Instruction Tuning and associated codebase.
<center> <img src="https://huggingface.co/hamishivi/tulu-2-multitask-rrmax-326k-sft/resolve/main/image.png" alt="Practical Large-Scale Data Selection for Instruction Tuning logo" width="200px"/> </center>For more results and analysis, please see our paper.
| Method | MMLU | GSM8k | BBH | TydiQA | Codex | Squad | AlpacaEval | Average |
|---|---|---|---|---|---|---|---|---|
| Random (unbal.) | 61.6 | 81.2 | 66.8 | 71.1 | 76.4 | 89.7 | 75.6 | 74.6 |
| Random (bal.) | 62.1 | 76.0 | 68.6 | 68.8 | 87.2 | 87.4 | 72.4 | 74.7 |
| Tulu 3 SFT | 62.2 | 74.3 | 68.2 | 67.4 | 83.8 | 85.5 | 71.9 | 73.3 |
| RDS+ | 62.5 | 77.6 | 66.6 | 72.1 | 83.8 | 90.2 | 80.2 | 76.1 |
| RDS+ - Arena Hard (this model) | 57.0 | 78.7 | 59.7 | 49.4 | 75.7 | 66.3 | 84.5 | 67.3 |
The model is trained to use the following format (note the newlines):
<|user|>
Your message here!
<|assistant|>
For best results, format all inputs in this manner. Make sure to include a newline after <|assistant|>, this can affect generation quality quite a bit.
We have included a chat template in the tokenizer implementing this template.
These models have not been aligned to generate safe completions, so the model can produce problematic outputs (especially when prompted to do so).
If you find this model or data is useful in your work, please cite it with:
@misc{ivison2025data,
title={{Practical Large-Scale Data Selection for Instruction Tuning}},
author={{Hamish Ivison and Muru Zhang and Faeze Brahman and Pang Wei Koh and Pradeep Dasigi}}
year={2025},
url={https://arxiv.org/abs/2503.01807},
eprint={2503.01807},
archivePrefix={arXiv},
primaryClass={cs.CL}
}