X_Bee

Percevier_IO mask

2

Hi,guy.I learing you code to attention mask code in perceiver io. ``` if exists(mask): mask = rearrange(mask, 'b ... -> b (...)') max_neg_value = -torch.finfo(sim.dtype).max mask = repeat(mask, 'b j...

Sklearn Pipeline has None estimator,Convert MissingConverter

2

Hi,I used the convert to translate a mutli-ouput sklearn pipeline,Struction as:  `scaler` and `selector` maybe is None Exception result: ``` hummingbird.ml.exceptions.MissingConverter: Unable to find converter for model type ....

Hi,Mrs. Yang.Where is the baseline Demo that you showed in your speech on the evening of the 25th?

How to calculates the blendshape values based on the distance between key points?

1

EXCUSE ME,what is the key points? 468 landmark?

How to update the q90?

1

In this lines,The `self.q90` tensor could not do backwards，How can it to update the parameters? ``` class Generator(nn.Module): def __init__(self, params): super(Generator, self).__init__() self.q50 = nn.Linear(params.d_model, 1) self.q90 = nn.Linear(params.d_model,...

[func prepare_mask_masked_video]( https://github.com/PKU-YuanGroup/Open-Sora-Plan/blob/294993ca78bf65dec1c3b6fb25541432c545eda9/opensora/sample/pipeline_inpaint.py#L605) Assume that after the first frame is given, when the constructed Mask is constructed at the latent level, the latent frame dim result is incorrect; demo_inpaint_prepare_mask_masked_video.py ```...