Vateye

Training Script for CLIP

1

Hi, is there any script or config that use CLIP as the initialization?

RuntimeError: cuda runtime error (59) : device-side assert triggered at /pytorch/torch/lib/THC/generated/../THCReduceAll.cuh:339 terminate called after throwing an instance of 'std::runtime_error' what(): cuda runtime error (59) : device-side assert triggered at /pytorch/torch/lib/THC/generic/THCStorage.c:184

15

dataset loading [0/3570] dataset loading [1000/3570] dataset loading [2000/3570] dataset loading [3000/3570] dataset loading [0/1530] dataset loading [1000/1530] run train at epoch 1 Epoch: [1][1/112] Time 4.807 (4.807) Data 2.836...

Weight Decay

3

Hi, as stated in the issue, the ALPRO does use weight decay. But I did not find the process that passing the parameter "weight_decay" during the optimizer initialization. `optimizer =...

ios7

12

Will you redesign Airfloat?About new flat style?

How can I get the now playing song'name and art cover?In which file?

About data augmentation for pretraining

2

Hi, I have noticed that RandomResizedCrop is adopted for pretraining with the scaling parameter (0.2, 1). I am a bit curious about the choice of this scaling parameter since if...

Code Release of BEiT3

8

Hi, I would like to know when the code of BEiT3 will be released since I have noticed the repo has not been updated for about a month. Hope to...

I find an awesome work that uses pure LLaMA to understand visual information and support multi-modal chatting. The repo is called mPLUG-Owl, and it states better than miniGPT-4 and LLaVA...

The tanh activation function vqkd

1

**Describe** Model I am using (BEiTv2): I have noticed that both `encode_task_layer` and `decode_task_layer` adopt the `Tanh()` function as the activation. Why not using other activation function like `GELU()` or...

Multi Label

14

Hi, is it possible to use top-k entropy loss for multi-label classification problem? Which each of the gt_label can be the top-K?