GEHIRNS

community

https://gehirns.pages.dev/

AI & ML interests

None defined yet.

Recent Activity

Vasiniks updated a Space about 2 hours ago

Vasiniks updated a collection about 2 hours ago

Vasiniks updated a model about 2 hours ago

GEHIRNS/magi-1.2-4B

View all activity

updated a Space about 2 hours ago

README

updated a collection about 2 hours ago

MAGI 1.2

Bits fitted to each model by measurement. magi-1.2-2B: loop-DPO + 2B map, density 1.305. magi-1.2-4B: 4B map, density 0.968. • 2 items • Updated about 2 hours ago

updated a model about 2 hours ago

GEHIRNS/magi-1.2-4B

Text Generation • 4B • Updated about 2 hours ago

published a model about 2 hours ago

GEHIRNS/magi-1.2-4B

Text Generation • 4B • Updated about 2 hours ago

updated a collection 1 day ago

MAGI 1.2

Bits fitted to each model by measurement. magi-1.2-2B: loop-DPO + 2B map, density 1.305. magi-1.2-4B: 4B map, density 0.968. • 2 items • Updated about 2 hours ago

updated a model 1 day ago

GEHIRNS/magi-1.2-2B

Text Generation • 2B • Updated 1 day ago • 27 • 1

published a model 1 day ago

GEHIRNS/magi-1.2-2B

Text Generation • 2B • Updated 1 day ago • 27 • 1

updated a model 1 day ago

GEHIRNS/magi-1.0-4B

Text Generation • 4B • Updated 1 day ago • 1.46k • 5

updated a collection 1 day ago

MAGI 1.1

The magi-1.0 recipe plus a LoRA that stops loops and finishes thinking. magi-1.1-4B: 81.7%, density 0.939. magi-1.1-2B: density 1.149. • 2 items • Updated 1 day ago

updated a model 1 day ago

GEHIRNS/magi-1.1-4B

Text Generation • 4B • Updated 1 day ago • 360 • 1

published a model 1 day ago

GEHIRNS/magi-1.1-4B

Text Generation • 4B • Updated 1 day ago • 360 • 1

updated 2 models 2 days ago

GEHIRNS/magi-1.0-2B

Text Generation • 2B • Updated 1 day ago • 536

GEHIRNS/magi-1.1-2B

Text Generation • 2B • Updated 1 day ago • 738 • 1

updated a collection 2 days ago

MAGI 1.0

Training-free mixed-precision Qwen3.5 for 12 GB GPUs: magi-1.0-4B (96% of BF16's score in 30% of the memory) and magi-1.0-2B. • 2 items • Updated 1 day ago

published a model 2 days ago

GEHIRNS/magi-1.0-2B

Text Generation • 2B • Updated 1 day ago • 536

updated 2 collections 2 days ago

MAGI 1.1

The magi-1.0 recipe plus a LoRA that stops loops and finishes thinking. magi-1.1-4B: 81.7%, density 0.939. magi-1.1-2B: density 1.149. • 2 items • Updated 1 day ago

MAGI 1.0

Training-free mixed-precision Qwen3.5 for 12 GB GPUs: magi-1.0-4B (96% of BF16's score in 30% of the memory) and magi-1.0-2B. • 2 items • Updated 1 day ago

published a model 2 days ago

GEHIRNS/magi-1.1-2B

Text Generation • 2B • Updated 1 day ago • 738 • 1

updated a collection 3 days ago

SEELE

Training-free pruned and quantized Jeffan-4B-2010 decision model. At 8K context: 98% of its score in 56% of the GPU memory, 1.4x faster. • 1 item • Updated 3 days ago

updated a model 3 days ago

GEHIRNS/seele-1.0-3.3B

Text Classification • 3B • Updated 2 days ago • 593 • 1