ttthhh

Results 8 comments of ttthhh

all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)

same problem too.

all the input arrays must have same number of dimensions, but the array at index 0 has 1 dimension(s) and the array at index 1 has 0 dimension(s)

same problem too.

Not getting correct outputs

@ricshaw Hi, I have the similar results with you. I think maybe there are bugs in the code. Have you fixed it?

Not getting correct outputs

@aunaik But flipCode=1 means horizontal flippoing in cv2.

Results are not satisfactory as shown

@DRAhmadFaraz I have similar results. Have you figure out what's the problem? I think there are some bugs in the code.

deepseek-vl-7b模型使用deepspeed的ZeRo3报错

The same question

prepare_rotary_positional_embeddings training and inference are different

> pipeline里的代码由于一些历史原因写起来有点啰嗦，建议以training的代码作为参考。虽然实现方式有区别，但是最后的结果都是一样的。 @OleehyO 同样的疑问，看上去training的get_3d_rotary_pos_embed中是根据当前的宽高生成的位置编码，而pipeline中的1.0版本是根据base的宽高做了crop和插值，1.5版本貌似是基于base_size的宽高做了外推。如果当前的宽高不等于base的宽高，这三种方式得到的位置编码应该都是不一样的。能解释一下吗 @cdfan0627 有测试过这里的影响吗

wan i2v-14b-720p蒸馏模型问题

> Was the model removed? [Wan2.1-I2V-14B-720P-StepDistill-CfgDistill-Lightx2v](https://huggingface.co/lightx2v/Wan2.1-I2V-14B-720P-StepDistill-CfgDistill-Lightx2v) > > I see the link as blank same question