tgxs002 / HPSv2

Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis

tgxs002/HPSv2 Issues

Having issues with using v2.1 inference
Updated a month ago5
hps_version="v2.1" VS hps_version="v2.0"
Updated 2 months ago
Unable to reproduce the preference prediction accuracy results
Updated 2 months ago1
Access for New Test data in README
Updated 2 months ago1
Question about new test set
Updated 2 months ago
The same code and input will have different results in different environments.
Updated 2 months ago2
Reproduction of benchmark
Updated 3 months ago1
hps12 is accepted by cvpr?
Updated 3 months ago
Can I use it to compare different text-image pairs？
Closed 4 months ago2
Question about benchmark DALL-E 2 subset
Closed 4 months ago1
Is there an estimated date of realeasing HPDv2?
Updated 4 months ago9
When will the train data be uploaded?
Closed 4 months ago2
will the training dataset released now?
Closed 4 months ago2
What is the score range?
Updated 4 months ago
Any solution for using HPSv2 evaluation model offline?
Updated 4 months ago4
How to train with our custom data?
Updated 4 months ago5
please help check the use of hpsv2 func
Updated 4 months ago1
the given training data link cannot be downloaded
Closed 4 months ago1
Can it be a metric for image to image Local inpainting
Updated 6 months ago5
Model outputs don't seem to have enough granularity from low precision
Updated 7 months ago
Choos cuda device
Updated 7 months ago1
Where can I download the file ’ImageReward_test.json‘
Updated 8 months ago1
questions about ImageReward and HPS Comparison
Updated 9 months ago1
About SDXL-Base-v1.0 performance
Closed 9 months ago3
Inference speed is too slow, how to optimize it and accelerate inference?
Updated 9 months ago1
ModuleNotFoundError: No module named '_curses'
Closed 10 months ago2
question about downloading the hpsv2 test dataset
Updated 10 months ago3
Do you test in HPSv1 datasets by using HPSv2 checkpoint?
Closed a year ago12
For the test of custom data
Updated a year ago6
format of test.json
Closed a year ago1
About the evaluation of generating real people
Updated a year ago5
Why jointly training a preference predictor with rating data does not help much with preference prediction?
Closed a year ago4
Could you give a hugging face hps v2 predictor checkpoint address？
Updated a year ago6
simple script
Closed a year ago1