2023-05-18 05:52:05 +00:00
|
|
|
model:
|
|
|
|
|
arch: bind_gpt4
|
|
|
|
|
|
|
|
|
|
# Imagebind
|
|
|
|
|
freeze_imagebind: True
|
|
|
|
|
|
|
|
|
|
# Q-Former
|
|
|
|
|
freeze_qformer: True
|
2023-05-22 10:38:05 +00:00
|
|
|
q_former_model: "/mnt/bn/bykang/chixma/data/pretrained_models/blip2_pretrained_flant5xxl.pth"
|
2023-05-18 05:52:05 +00:00
|
|
|
num_query_token: 32
|
|
|
|
|
|
|
|
|
|
# Vicuna
|
2023-05-22 10:38:05 +00:00
|
|
|
llama_model: "/mnt/bn/bykang/chixma/data/pretrained_models/vicuna-7b-v0/"
|
2023-05-18 05:52:05 +00:00
|
|
|
|
|
|
|
|
# generation configs
|
|
|
|
|
prompt: ""
|
|
|
|
|
|
|
|
|
|
preprocess:
|
|
|
|
|
vis_processor:
|
|
|
|
|
train:
|
|
|
|
|
name: "imagebind_vision_train"
|
|
|
|
|
image_size: 224
|
|
|
|
|
eval:
|
|
|
|
|
name: "imagebind_vision_eval"
|
|
|
|
|
image_size: 224
|
|
|
|
|
text_processor:
|
|
|
|
|
train:
|
|
|
|
|
name: "imagebind_caption"
|
|
|
|
|
eval:
|
|
|
|
|
name: "imagebind_caption"
|
