PommesPeter

I would like to do some translation work first because I am also new to the DI-engine. Meanwhile, I also try to learn something about RL or MARL for my...

Hi Bruce, Could you give more error messages when you are using our code? @BruceLeeeee

> TypeError: DiT_Llama.**init**() got an unexpected keyword argument 'max_seq_len' Okay, we will check the training code again to fix this problem.

> Hi, @PommesPeter I tried to finetune the model, but no matter how I set it up the image size or batch size or precision (I remove the 'max_seq_len' argument),...

@BruceLeeeee we use `precision: bf16` and `grad_precision: fp32`, you can try these hyper-parameters, and we will check the `model.py` code whether is correct.

> > Traceback (most recent call last): > > File "/mnt/new_sfs_turbo/lsh2/train_projects/Lumina-T2X/lumina_t2i/train.py", line 753, in > > main(args) > > File "/mnt/new_sfs_turbo/lsh2/train_projects/Lumina-T2X/lumina_t2i/train.py", line 514, in main > > loss_dict = transport.training_losses(model,...

Hi @JincanDeng , We have supported Lumina-Next in the diffusers, check the instructions at [here](https://github.com/Alpha-VLLM/Lumina-T2X/blob/main/README.md#fast-demo)

``` Miss Mexico portrait of the most beautiful mexican woman, Exquisite detail, 30-megapixel, 4k, 85-mm-lens, sharp-focus, f:8, ISO 100, shutter-speed 1:125, diffuse-back-lighting, award-winning photograph, small-catchlight, High-sharpness, facial-symmetry, 8k --q 2...

你好，感谢您的贡献，这里是需要使用 apex 的 FusedRMSNorm，不然训练的时候会爆显存，推理的时候可不用。

> > 你好，感谢您的贡献，这里是需要使用 apex 的 FusedRMSNorm，不然训练的时候会爆显存，推理的时候可不用。 > > 您好！ > > * 现在的model.py是只引用了apex。因为readme里面写的apex是可选的，如果没装apex的话，这样直接推理会报错 >  > * components.py里是这样import的方式才是正确的。所以改成在model里直接import components就行了。 >  > > 或者您可能需要修改readme，或者加是否引用apex的参数 好的，感谢您修复代码的问题，这是我们 release 时候的问题，稍后再补充一些修改就可以 merge 了。