Yoel Zimmermann
@yzimmermann.bsky.social
Exploring how the smallest shapes the biggest @ ETH Zurich.
Somewhere between Physics, Chemistry, and ML.
Somewhere between Physics, Chemistry, and ML.
P.S. If you work in the flavor/food/fragrance industry and are laughing about our 15k dataset, reach out to hello@kvantlabs.ai. We would love to build an even better model for you!
December 11, 2024 at 2:35 PM
P.S. If you work in the flavor/food/fragrance industry and are laughing about our 15k dataset, reach out to hello@kvantlabs.ai. We would love to build an even better model for you!
Preprint: chemrxiv.org/engage/chemr...
Dataset: huggingface.co/datasets/Far...
It has been fun working on this with @leif7ieben.bsky.social, @heseng.bsky.social, Philipp Pestlin, Franz Görlich!
Dataset: huggingface.co/datasets/Far...
It has been fun working on this with @leif7ieben.bsky.social, @heseng.bsky.social, Philipp Pestlin, Franz Görlich!
A Chemical Language Model for Molecular Taste Prediction
Determining molecular taste remains a significant challenge in food science. Here, we present FART (Flavor Analysis and Recognition Transformer), a chemical language model capable of predicting molecu...
chemrxiv.org
December 11, 2024 at 2:34 PM
Preprint: chemrxiv.org/engage/chemr...
Dataset: huggingface.co/datasets/Far...
It has been fun working on this with @leif7ieben.bsky.social, @heseng.bsky.social, Philipp Pestlin, Franz Görlich!
Dataset: huggingface.co/datasets/Far...
It has been fun working on this with @leif7ieben.bsky.social, @heseng.bsky.social, Philipp Pestlin, Franz Görlich!
Trained on the largest publicly available dataset (15,025 unique compounds), we achieve >90% accuracy - outperforming existing binary classifiers that focus on only one taste class (e.g. sweet/non-sweet).
Try it out here: fartlabs-fart.hf.space
Try it out here: fartlabs-fart.hf.space
December 11, 2024 at 2:25 PM
Trained on the largest publicly available dataset (15,025 unique compounds), we achieve >90% accuracy - outperforming existing binary classifiers that focus on only one taste class (e.g. sweet/non-sweet).
Try it out here: fartlabs-fart.hf.space
Try it out here: fartlabs-fart.hf.space
Is 15 minutes supposed to be a lot?
November 28, 2024 at 8:38 PM
Is 15 minutes supposed to be a lot?
CERN also came to my mind 😅
November 26, 2024 at 9:43 PM
CERN also came to my mind 😅
In principle you can do a static evaluation on a holdout set but I have no clue if 4k datapoints is enough.
November 24, 2024 at 11:26 AM
In principle you can do a static evaluation on a holdout set but I have no clue if 4k datapoints is enough.
Cool work! How about doing RLHF on your data?
November 24, 2024 at 10:46 AM
Cool work! How about doing RLHF on your data?