2023-04-16 22:04:16 +00:00
|
|
|
model:
|
2023-10-12 21:14:34 +00:00
|
|
|
arch: minigpt4
|
2023-04-16 22:04:16 +00:00
|
|
|
|
|
|
|
|
# vit encoder
|
|
|
|
|
image_size: 224
|
|
|
|
|
drop_path_rate: 0
|
|
|
|
|
use_grad_checkpoint: False
|
|
|
|
|
vit_precision: "fp16"
|
|
|
|
|
freeze_vit: True
|
|
|
|
|
freeze_qformer: True
|
|
|
|
|
|
|
|
|
|
# Q-Former
|
|
|
|
|
num_query_token: 32
|
|
|
|
|
|
|
|
|
|
# generation configs
|
|
|
|
|
prompt: ""
|
|
|
|
|
|
2023-10-13 14:11:11 +00:00
|
|
|
llama_model: "please set this value to the path of vicuna model"
|
2023-08-28 18:26:00 +00:00
|
|
|
|
2023-04-16 22:04:16 +00:00
|
|
|
preprocess:
|
|
|
|
|
vis_processor:
|
|
|
|
|
train:
|
|
|
|
|
name: "blip2_image_train"
|
|
|
|
|
image_size: 224
|
|
|
|
|
eval:
|
|
|
|
|
name: "blip2_image_eval"
|
|
|
|
|
image_size: 224
|
|
|
|
|
text_processor:
|
|
|
|
|
train:
|
|
|
|
|
name: "blip_caption"
|
|
|
|
|
eval:
|
|
|
|
|
name: "blip_caption"
|
