Qianli Ma

Hi @Breakthrough, thanks for your great work! Can you help me?😭

> @MqLeet in v0.6.1 I added `-map 0` to the ffmpeg args as part of fixing #271. It seems to cause issues for this particular video, I am not entirely...

@jon-chuang @Mohan2351999 Hi, have you obtained good generation results? I used the training method of ADD, but the generated images have color issues, such as oversaturation... Just like this ...

My datasets version is 2.19.0

I also wonder whether Janus-Pro-7B has this ability!

> I have changed the pipeline and can do "txt+img -> img" work, but the model doesn't work well, so looking up to the upgrade from janus. Hi @Youngon ,...

alright, thanks for your time! 🥰

Hi @LWang1016 , I notice that you use VBench tools to evaluate Opensorav1-1, which will not get the results we published in the leaderboard. We recommend you to use [VBench-Long](https://github.com/Vchitect/VBench/tree/master/vbench2_beta_long)...

Hi, @LWang1016, we will close this issue. Feel free to contact us if you have other questions.

> import torch.serialization from deepspeed.runtime.zero.config import ZeroStageEnum from deepspeed.runtime.fp16.loss_scaler import LossScaler from deepspeed.runtime.zero.stage_1_and_2 import DeepSpeedZeroOptimizer from deepspeed.runtime.zero.stage3 import DeepSpeedZeroOptimizer_Stage3 > > # 添加所有需要的类到安全列表 > torch.serialization.add_safe_globals([ ZeroStageEnum, LossScaler, DeepSpeedZeroOptimizer, DeepSpeedZeroOptimizer_Stage3 ])...