Jilong Wang comments

Repositories
Issues
Comments

Results 4 comments of


                                            Jilong Wang

why are sp and sn both decreasing during training process?

> Thanks. > I find this as well. > MNIST example is my simple test case. Maybe you can try some other experiments. Thanks for your reply!! i also tried...

OOM and slow tokenization after upgrade LLaMA-Factory

same problem

在opengait上使用单卡训练步态识别的模型时，得到的准确率全是2%，这是怎么回事呢？

单卡可以不用改动代码，直接改动train和test的命令即可，CUDA_VISBLE_DEVICES改成0，nproc_per_node改成1

Deepgaitv2在Gait3D准确率问题，使用deepgaitv2-p3d训练gait3d时，rank-1准确率只有60%,而论文里为74%，配置文件都是按照原始配置的，请问是什么原因呢请大神帮忙解答一下，谢谢！

你这不是训练到了3w iter的结果吗 config里需要训练到6w iter