Junlei Zhang

Hello! Could you please tell me how to do the cutout operation?

1

Dear: Thank you very much for your code. I noticed that there is a cutout parameter. But you do not implement it. Could you please tell me how to or...

Hello，in the file dataset.py， there are two "elif dataset == reduced_imagenet"。 Is this a mistake?

Hello, I noticed that you used the color jitter and lighting during training the imagenet. But normally we will not apply these two data augment for a fair comparison. Could...

Hi，When I train the network, the used GPU memery keep going up?

4

Hi, Thank for your nice code. It is really a beautiful project. When I am trying to train the dataset refcoco, the used GPU memery keep going up. Initially, the...

下载了60个文件但是只有239GB

10

请问一下，我下载了60个文件，但是只有239GB，这个正常么？非常感谢

hello, thank you for your code. Could you please tell me that the max_sequence_length of the GLM-130B model? I tired to set the max_sequence_length in the config/model_glm_130b.sh to 4096 but...

Hi, I've completed the course and uploaded the training report. Could you please help me review it? I'm a bit anxious to get this dataset. Thank you very much.

Does some properties (like the temperature of an obj) will be changed if the seed is fixed?

2

Hello, thank you very much for your work. I have a question. If for a specific task, I set the random number at the beginning of the environment initialization using...

Why you discard to token-to-token attention in your model

3

Dears, thank you for your work. I found that in your model, you split the attention into two modes: ( num_head/2). And all these two modes use the compressed sequence....

[New Model]: Cogagent

2

### The model to consider. https://huggingface.co/THUDM/CogAgent ### The closest model vllm already supports. _No response_ ### What's your difficulty of supporting the model you want? Vision models