Lei Chen

> How can I use YoWo model to generate videoMAP as per `evaluate_multisports.py`, which requires `video_detections.pkl`? > Open source code for YoWo doesn't generate `video_detections.pkl`. > It would be great...

> I am trying to reproduce the result using YOWO? Can you let me know which anchors you used for YOWO benchmarks? The anchors we used are [0.70458, 1.18803, 1.26654,...

> Which pretrained model you used to initialize the 3D and 2D model in YoWo architecture? We use Darknet-19 [weights](http://pjreddie.com/media/files/yolo.weights) as our 2D pretrained model and resnext-101-kinetics.pth from [Google drive](https://drive.google.com/drive/folders/1zvl89AgFAApbH0At-gMuZSeQB_LpNP-M)...

I modified the learning rate and classes in ucf24.cfg. ------------------ 原始邮件 ------------------ 发件人: "MCG-NJU/MultiSports" ***@***.***>; 发送时间: 2022年2月16日(星期三) 上午10:24 ***@***.***>; ***@***.******@***.***>; 主题: Re: [MCG-NJU/MultiSports] Anchors used for YoWo Benchmark (Issue #16) @MiaSanLei sorry to ask...

+1

> > [@JackeyHRan](https://github.com/JackeyHRan) transformers 4.45.1 遇到了同样的问题，降级到4.40.0似乎可以正常保存权重 > > 感谢，我去尝试一下 请问降到4.40.0能够解决default_config问题吗，我试了一下会报新的错误ImportError: cannot import name 'HybridCache' from 'transformers.cache_utils'，可能跟liger_kernel版本有关，请问您降transformers后有改liger_kernel版本吗，谢谢

InternLM-XComposer2-4KHD-7B has demonstrated remarkable performance, achieving state-of-the-art results in our MindBench benchmark. We hope to integrate these results into your repository.