Introducing 🦣 MAmmoTH2, a game-changer in improving the reasoning abilities of large language models (LLMs) through innovative instruction tuning. By efficiently harvesting 10 million instruction-response pairs from the pre-training web corpus, we've developed MAmmoTH2 models that significantly boost performance on reasoning benchmarks. For instance, MAmmoTH2-7B (Mistral) sees its performance soar from 11% to 36.7% on MATH and from 36% to 68.4% on GSM8K, all without training on any domain-specific data. Further training on public instruction tuning datasets yields MAmmoTH2-Plus, setting new standards in reasoning and chatbot benchmarks. Our work presents a cost-effective approach to acquiring large-scale, high-quality instruction data, offering a fresh perspective on enhancing LLM reasoning abilities.
The models are fine-tuned with the WEBINSTRUCT dataset using the original Llama-3, Mistral and Mistal models as base models. The training procedure varies for different models based on their sizes. Check out our paper for more details.
Evaluation
The models are evaluated using open-ended and multiple-choice math problems from several datasets. Here are the results:
You can use the models through Huggingface's Transformers library. Use the pipeline function to create a text-generation pipeline with the model of your choice, then feed in a math problem to get the solution.
Check our Github repo for more advanced use:
https://github.com/TIGER-AI-Lab/MAmmoTH2
Limitations
We've tried our best to build math generalist models. However, we acknowledge that the models' performance may vary based on the complexity and specifics of the math problem. Still not all mathematical fields can be covered comprehensively.
Citation
If you use the models, data, or code from this project, please cite the original paper:
@article{yue2024mammoth2,
title={MAmmoTH2: Scaling Instructions from the Web},
author={Yue, Xiang and Zheng, Tuney and Zhang, Ge and Chen, Wenhu},
journal={arXiv preprint arXiv:2405.03548},
year={2024}
}
Runs of TIGER-Lab MAmmoTH2-7B-Plus on huggingface.co
12.3K
Total runs
0
24-hour runs
0
3-day runs
191
7-day runs
594
30-day runs
More Information About MAmmoTH2-7B-Plus huggingface.co Model
MAmmoTH2-7B-Plus huggingface.co is an AI model on huggingface.co that provides MAmmoTH2-7B-Plus's model effect (), which can be used instantly with this TIGER-Lab MAmmoTH2-7B-Plus model. huggingface.co supports a free trial of the MAmmoTH2-7B-Plus model, and also provides paid use of the MAmmoTH2-7B-Plus. Support call MAmmoTH2-7B-Plus model through api, including Node.js, Python, http.
MAmmoTH2-7B-Plus huggingface.co is an online trial and call api platform, which integrates MAmmoTH2-7B-Plus's modeling effects, including api services, and provides a free online trial of MAmmoTH2-7B-Plus, you can try MAmmoTH2-7B-Plus online for free by clicking the link below.
TIGER-Lab MAmmoTH2-7B-Plus online free url in huggingface.co:
MAmmoTH2-7B-Plus is an open source model from GitHub that offers a free installation service, and any user can find MAmmoTH2-7B-Plus on GitHub to install. At the same time, huggingface.co provides the effect of MAmmoTH2-7B-Plus install, users can directly use MAmmoTH2-7B-Plus installed effect in huggingface.co for debugging and trial. It also supports api for free installation.