Jiacheng Liu
@liujch1998.bsky.social
🎓 PhD student @uwcse @uwnlp. 🛩 Private pilot. Previously: 🧑💻 @oculus, 🎓 @IllinoisCS. 📖 🥾 🚴♂️ 🎵 ♠️
OLMoTrace is powered by my previous work infini-gram, with some innovative algorithmic twists. Really proud to turn an academic research project into a real LLM product, it’s been a truly amazing experience.
Check out infini-gram: infini-gram.io
Check out infini-gram: infini-gram.io
April 9, 2025 at 3:56 PM
OLMoTrace is powered by my previous work infini-gram, with some innovative algorithmic twists. Really proud to turn an academic research project into a real LLM product, it’s been a truly amazing experience.
Check out infini-gram: infini-gram.io
Check out infini-gram: infini-gram.io
Want to predict the task performance of LMs before pretraining them?
We develop task scaling laws and model ladders, which predict the accuracy on individual tasks by OLMo 2 7B & 13B models within 2 points of absolute error. The cost is 1% of the compute used to pretrain them.
We develop task scaling laws and model ladders, which predict the accuracy on individual tasks by OLMo 2 7B & 13B models within 2 points of absolute error. The cost is 1% of the compute used to pretrain them.
December 9, 2024 at 5:07 PM
Want to predict the task performance of LMs before pretraining them?
We develop task scaling laws and model ladders, which predict the accuracy on individual tasks by OLMo 2 7B & 13B models within 2 points of absolute error. The cost is 1% of the compute used to pretrain them.
We develop task scaling laws and model ladders, which predict the accuracy on individual tasks by OLMo 2 7B & 13B models within 2 points of absolute error. The cost is 1% of the compute used to pretrain them.