Modelpublic

EpistemeAI/gpt-oss-20b-stem-distilled-reasoning

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes27downloads

Model Card

This fine tune comprehensive STEM reasoning model, it covers concept understanding, multi-step deduction, and formula/theorem application across Mathematics, Physics, Chemistry, Computer Science, Engineering, and Life Sciences.

Limitation

—Responsible Use: This model must not be used for exam cheating or for generating misleading scientific claims. In high-risk applications such as Medicine, Chemistry, or Engineering, outputs should always undergo expert review and human verification before being acted upon.

Uploaded finetuned model

—Developed by: EpistemeAI
—License: apache-2.0
—Finetuned from model : unsloth/gpt-oss-20b-unsloth-bnb-4bit

This gpt_oss model was trained 2x faster with Unsloth and Huggingface's TRL library.