Al_th
@alth.fr
Applied Math Ph.D, R&D engineer (Image processing, numerical modeling, Machine Learning) in the healthcare sector. #MLsky
Also cooking (Pâté en croûte maker) and slowly learning guitar.
Alth.fr
@althcuisine on Instagram
@AlthCuisine on YouTube
FR/EN
Also cooking (Pâté en croûte maker) and slowly learning guitar.
Alth.fr
@althcuisine on Instagram
@AlthCuisine on YouTube
FR/EN
Some time ago, I DM'd @dorialexander.bsky.social about a similar (yet somewhat diff) idea :
While there is a point in fixing the generated tokens, we do squash enormous amount of information by actually looking if the cat is dead or alive.
AFAIK, the issue with diff is fixed context size tho
While there is a point in fixing the generated tokens, we do squash enormous amount of information by actually looking if the cat is dead or alive.
AFAIK, the issue with diff is fixed context size tho
February 19, 2025 at 8:45 AM
Some time ago, I DM'd @dorialexander.bsky.social about a similar (yet somewhat diff) idea :
While there is a point in fixing the generated tokens, we do squash enormous amount of information by actually looking if the cat is dead or alive.
AFAIK, the issue with diff is fixed context size tho
While there is a point in fixing the generated tokens, we do squash enormous amount of information by actually looking if the cat is dead or alive.
AFAIK, the issue with diff is fixed context size tho
To be fair, the GRPO optimized model doesnt shout, the RL cheated by having more people speak (as names are capitalized in the dataset I'm using)
(Left is base transformer, right is post GRPO)
(Left is base transformer, right is post GRPO)
February 5, 2025 at 5:22 PM
To be fair, the GRPO optimized model doesnt shout, the RL cheated by having more people speak (as names are capitalized in the dataset I'm using)
(Left is base transformer, right is post GRPO)
(Left is base transformer, right is post GRPO)
Je suis mort de rire : un médecin retraité >>>>n’exerçant plus aucune activité médicale rémunérée <<<< doit toujours filer ses 100 balles a l’ordre des médecins 😂
January 26, 2025 at 8:19 AM
Je suis mort de rire : un médecin retraité >>>>n’exerçant plus aucune activité médicale rémunérée <<<< doit toujours filer ses 100 balles a l’ordre des médecins 😂
2/2 : Sorry, in French (www.alth.fr/article-117)
January 21, 2025 at 1:44 PM
2/2 : Sorry, in French (www.alth.fr/article-117)
Le chiffre de 500mL faisant référence à 10-50 requête (cf picture).
Et du coup l'eau en question est l'eau utilisée pour refroidir les serveurs. On peut difficilement parler d'une ressource perdue (e.g. si l'eau est utilisée pour réchauffer des logements après...)
Et du coup l'eau en question est l'eau utilisée pour refroidir les serveurs. On peut difficilement parler d'une ressource perdue (e.g. si l'eau est utilisée pour réchauffer des logements après...)
January 16, 2025 at 9:59 AM
Le chiffre de 500mL faisant référence à 10-50 requête (cf picture).
Et du coup l'eau en question est l'eau utilisée pour refroidir les serveurs. On peut difficilement parler d'une ressource perdue (e.g. si l'eau est utilisée pour réchauffer des logements après...)
Et du coup l'eau en question est l'eau utilisée pour refroidir les serveurs. On peut difficilement parler d'une ressource perdue (e.g. si l'eau est utilisée pour réchauffer des logements après...)
If you were not aware : most sane compiler should (and do) know how to convert this to the O(1) formula.
They also know how to do it for squares, cubes, and more.
They also know how to do it for squares, cubes, and more.
January 1, 2025 at 1:17 PM
If you were not aware : most sane compiler should (and do) know how to convert this to the O(1) formula.
They also know how to do it for squares, cubes, and more.
They also know how to do it for squares, cubes, and more.
One manifestation of a small hidden "handicap".
I'm color blind. Youtube decided to change the color of its progress bar from red to a gradient (red to purple at the end).
Issue is that I don't see where I'm at precisely now. I don't think anyone can tbh, but it's especially difficult for me...
I'm color blind. Youtube decided to change the color of its progress bar from red to a gradient (red to purple at the end).
Issue is that I don't see where I'm at precisely now. I don't think anyone can tbh, but it's especially difficult for me...
December 24, 2024 at 8:36 AM
One manifestation of a small hidden "handicap".
I'm color blind. Youtube decided to change the color of its progress bar from red to a gradient (red to purple at the end).
Issue is that I don't see where I'm at precisely now. I don't think anyone can tbh, but it's especially difficult for me...
I'm color blind. Youtube decided to change the color of its progress bar from red to a gradient (red to purple at the end).
Issue is that I don't see where I'm at precisely now. I don't think anyone can tbh, but it's especially difficult for me...
I put Trellis, from Microsoft (github.com/microsoft/TR...), up to the test.
From a far-from-perfect photograph (zoomed in, object not fully in FoV), this is the 3D reconstruction of a pâté en croûte. Impressive.
From a far-from-perfect photograph (zoomed in, object not fully in FoV), this is the 3D reconstruction of a pâté en croûte. Impressive.
December 10, 2024 at 11:10 AM
I put Trellis, from Microsoft (github.com/microsoft/TR...), up to the test.
From a far-from-perfect photograph (zoomed in, object not fully in FoV), this is the 3D reconstruction of a pâté en croûte. Impressive.
From a far-from-perfect photograph (zoomed in, object not fully in FoV), this is the 3D reconstruction of a pâté en croûte. Impressive.
Thibault Sauron defending his PhD work.
Great presentation, and opening on great philosophical questions.
- What is visual similarity (and in context of multimodal medical images)
- why does self supervised mono vs multi encoders perform preferentially given end tasks.
#MLsky
Great presentation, and opening on great philosophical questions.
- What is visual similarity (and in context of multimodal medical images)
- why does self supervised mono vs multi encoders perform preferentially given end tasks.
#MLsky
December 5, 2024 at 7:38 AM
Thibault Sauron defending his PhD work.
Great presentation, and opening on great philosophical questions.
- What is visual similarity (and in context of multimodal medical images)
- why does self supervised mono vs multi encoders perform preferentially given end tasks.
#MLsky
Great presentation, and opening on great philosophical questions.
- What is visual similarity (and in context of multimodal medical images)
- why does self supervised mono vs multi encoders perform preferentially given end tasks.
#MLsky
November 17, 2024 at 1:33 PM