issues
search
kingsdigitallab
/
kdl-vqa
Python tool for batch visual question answering (BVQA).
MIT License
0
stars
0
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Pass max image size, and max_new_token
#15
geoffroy-noel-ddh
opened
1 week ago
0
API to run task over series of (response, image) pairs
#14
geoffroy-noel-ddh
opened
1 week ago
0
Show answers and images in a document to help with manual review.
#13
geoffroy-noel-ddh
closed
1 week ago
0
Filter param (--filter) must be lowercase to find match
#12
mchesterkadwell
closed
1 week ago
2
Save the hostname & resources in the log and answer file
#11
geoffroy-noel-ddh
closed
1 week ago
1
Save the compute time in the answer files.
#10
geoffroy-noel-ddh
opened
1 week ago
0
Keep multiple answers per questions in the answer files
#9
geoffroy-noel-ddh
closed
1 week ago
0
not using GPU on HPC environment
#8
geoffroy-noel-ddh
closed
1 week ago
2
Pass a huggingface model qualifier to bvqa instead of a model type.
#7
geoffroy-noel-ddh
closed
1 week ago
0
qwen-vl describer can misbehave with multiple questions per prompt
#6
geoffroy-noel-ddh
closed
1 week ago
1
Missing dependencies for qwen-vl describer
#5
geoffroy-noel-ddh
closed
1 week ago
2
Default describer should not be instantiated if another model is selected.
#4
geoffroy-noel-ddh
closed
1 week ago
0
Consider using a VLM server
#3
geoffroy-noel-ddh
opened
4 weeks ago
1
Lock time out can be too short
#2
geoffroy-noel-ddh
opened
4 weeks ago
0
Automate evaluation of the answers
#1
geoffroy-noel-ddh
closed
1 week ago
1