unicom

unicom copied to clipboard

未来会公开数据么

vanpersie32

About kmeans clustering

2

Hi, First thanks for such a great work and making it open. I notice in your paper you mentioned, - you can cluster 400 million samples into 1 million clustering...

zheng-xing

是否会公开完整训练代码？

3

您好，想请问下后续是否有打算公开预训练模型的训练代码呢？非常感谢！

KeyaoZhao

我发现`unicom/examples/retrieval.md`这个例子中没有`model.eval()`. 我用了别的数据做的测试，发现加上`model.eval()`之后会得到不一样的 Cosine 值. 而且不加的话会导致输入两张一样的图片却获得两个不一样的 features.

luoshmg

关于数据集制作

3

您好！我想请问一下在新的小数据集上finetune时可能会遗忘学过的LAION 400M，导致finetune后的模型泛化性能下降。所以我计划在finetune时也加入部分LAION 400M数据，但是我使用LAION 400M聚类到1M时的类id可能和您训练时的不同，这是否会产生冲突呢？请问数据集的这些信息或者原始的制作方法您可以公布下吗？非常感谢~

KeyaoZhao

chenran06

Pretrained for ViT-S/14 or ViT-S/16

5

Thank you for your excellent works! I want to implement this pretrained on my research with limited resource machine. Could you release pretrained weights for ViT-S/14 or ViT-S/16? My email:...

PD-Mera

Training mode

1

I understand that there is an evaluation mode for evaluating the image retrieval using: ``torchrun` retrieval.py --eval --dataset cub --model_name ViT-B/32`` Is there a mode for training the models or...

vioanl

or the device required

Deeperfinder

Bug in CombinedMarginLoss implementation

2

Hi @anxiangsir, Thanks for sharing your work. I have a question about the `forward` pass in [`CombinedMarginLoss`](https://github.com/deepglint/unicom/blob/64e2d64a63e95f91b93f41410f4639b2bd9b216e/partial_fc.py#L49) when running [`sop_vit_b_16.sh`](https://github.com/deepglint/unicom/blob/64e2d64a63e95f91b93f41410f4639b2bd9b216e/scripts/sop_vit_b_16.sh) as an example. In this case, `self.m1 = 1.0`, `self.m2...

MaroonAmor