Safetensors
English
llava
video-retrieval
text-to-video-search
multimodal-embedding
TARA / shared /run /run.sh
bpiyush's picture
Update TARA to latest Tarsier2 checkpoint and runnable demo.
7daf628
raw
history blame contribute delete
335 Bytes
CUDA_VISIBLE_DEVICES=0 bash shared/run/extract_feat_dinov2_ego4d.sh 0 110000 &
CUDA_VISIBLE_DEVICES=1 bash shared/run/extract_feat_dinov2_ego4d.sh 110000 220000 &
CUDA_VISIBLE_DEVICES=2 bash shared/run/extract_feat_dinov2_ego4d.sh 220000 330000 &
CUDA_VISIBLE_DEVICES=3 bash shared/run/extract_feat_dinov2_ego4d.sh 330000 440000 & wait