MLIC icon indicating copy to clipboard operation
MLIC copied to clipboard

Multi-Reference Entropy Models for Learned Image Compression

Results 10 MLIC issues
Sort by recently updated
recently updated
newest added

请问训练mlic++的时候大概需要多长时间使得模型收敛呢? 感觉在训练的时候这么多模块会一定程度上延长训练时间。

Currently, the repo only has MLIC++, so where can I see the MLIC and MLIC+ model code?

fine tuning时,报如下错误,是对训练集的图片尺寸有要求吗? 24-04-10 10:36:36.398 - INFO: Learning rate: 0.0001 Traceback (most recent call last): File "/home/zhanghh/work/JiangWeibeta/MLIC/MLIC++/playground/train.py", line 162, in main() File "/home/zhanghh/work/JiangWeibeta/MLIC/MLIC++/playground/train.py", line 124, in main current_step = train_one_epoch( File...

请问可以上传MLIC和MLIC+的代码吗?

utils文件夹是否缺少__init__文件呢?另外,能否提供一下requirements.txt呢?感谢!

我自己实现数据并行会报错如下: Traceback (most recent call last): File "/home/user01/hzz/MLIC++/train.py", line 158, in main() File "/home/user01/hzz/MLIC++/train.py", line 119, in main current_step = train_one_epoch( File "/home/user01/hzz/MLIC++/utils/training.py", line 18, in train_one_epoch out_net = model(d)...

作者您好,在进行训练自己的模型的过程中报错 此外,我打印了模型中最后部分的一些参数形状如下所示:(这里我采用的是老版本的预训练模型,并将对应模块换成Old版本)

May I ask how the bpp value is calculated when you use VTM17.1 as an anchor? Also, can you share the calculation method for your BD-rate value?