Yuyang Zhao

@tsenst Hi, when I finetune the model on MPI-Sintel with your options The loss and epe are all 'nan' Did you meet this problem?

@xianshunw @Blcony Hi, I try to fine-tune or train on MPISintel, but the loss and epe are all ''nan' like this `2019-09-17 00:36:04 Iter 1000 [Train]: loss=nan, epe=nan, lr=0.000100, samples/sec=6.4,...

> > @xianshunw @Blcony Hi, I try to fine-tune or train on MPISintel, but the loss and epe are all ''nan' > > like this > > `2019-09-17 00:36:04 Iter...

I meet the same problem, and have no idea how to solve it

Thanks a lot, I find it. momentum – the value used for the running_mean and running_var computation. **Can be set to None for cumulative moving average (i.e. simple average)** So...

Hi, I tried momentum=0.1 and it performed bad. Momentum=False is a better choice.

May I ask where the `zero123-xl.ckpt` is? In the [hugging face model](https://huggingface.co/cvlab/zero123-weights), I can only find `105000.ckpt` and `165000.ckpt`.

Hi, thanks for your interest in our work. The training code for coco-20i is available now!

Hi, I think it is the problem of openpose checkpoint [here](https://github.com/TencentARC/T2I-Adapter/blob/b324d9f74b3736a35f0178e2a26713bfb100d83d/ldm/modules/extra_condition/openpose/api.py#L13), which is broken. While the openpose checkpoint in [ControlNet repo](https://github.com/lllyasviel/ControlNet/blob/16ea3b5379c1e78a4bc8e3fc9cae8d65c42511b1/annotator/openpose/__init__.py#L12) works for me. @MC-E Hi Chong, thanks for your...

> Due to the policy update of Microsoft OneDrive, the shared folder exceeds the limited size of the new policy, we are working on it, and will solve it ASAP!...