#Bagel 7B MoT

6 messages · Page 1 of 1 (latest)

severe heart
#

wow it has image generation, better benchmark scores than FLUX-1-dev

#

"It is finetuned from Qwen2.5-7B-Instruct and siglip-so400m-14-384-flash-attn2 model, and uses the FLUX.1-schnell VAE model"

#

also is a reasoning model

#

they don't provide any text benchmarks

obtuse pebble
#

Wait, this quite interesting..