My series of fully open, state-of-the-art small mixture-of-experts models.
aquiffoo
aquiffoo
AI & ML interests
thanks for everything.
Recent Activity
upvoted a paper about 20 hours ago
HeavySkill: Heavy Thinking as the Inner Skill in Agentic Harness new activity 1 day ago
mistralai/Mistral-Medium-3.5-128B:Mistral 12B - 34B ?? Please 🥰 liked a model 2 days ago
HuggingFaceTB/nanowhale-100mOrganizations
Solo research around the `mesh` architecture: a novel solution to the problems of MoE
my GPT-2-like model, pretrained from scratch
Models i really really like
-
zai-org/GLM-5
Text Generation • 754B • Updated • 315k • • 2.08k -
zai-org/GLM-4.7-Flash
Text Generation • 31B • Updated • 623k • • 1.72k -
moonshotai/Kimi-K2.5
Image-Text-to-Text • 1.1T • Updated • 1.96M • • 2.78k -
fal/FLUX.2-dev-Turbo
Text-to-Image • Updated • 5.55k • • 368
The second generation of my pretrained models.
My series of fully open, state-of-the-art small mixture-of-experts models.
Models i really really like
-
zai-org/GLM-5
Text Generation • 754B • Updated • 315k • • 2.08k -
zai-org/GLM-4.7-Flash
Text Generation • 31B • Updated • 623k • • 1.72k -
moonshotai/Kimi-K2.5
Image-Text-to-Text • 1.1T • Updated • 1.96M • • 2.78k -
fal/FLUX.2-dev-Turbo
Text-to-Image • Updated • 5.55k • • 368
Solo research around the `mesh` architecture: a novel solution to the problems of MoE
The second generation of my pretrained models.
my GPT-2-like model, pretrained from scratch