Michael Kupchick comments

Results 10 comments of


                                            Michael Kupchick

Running command git checkout -q 4054d5aa9ed7c93a255de60d8625b41087b5c022 hangs

Are you using ComfyUI Manager to install the nodes or just git clone?

Weird video out put with ltx 13b model

please download the latest version of fp8 model, latest model can now run natively in ComfyUI without using q8_kernels and Q8 Patch node.

Weird video out put with ltx 13b model

Make sure to update to latest comfy and ComfyUI-LTXVideo as well

Output of the ltxv-2b-0.9.6-dev-04-25.safetensors I2V is glitchi as hell and for T2V some alien fingers are "around"...Pls advice!

Note that distilled version has its own flow and regular i2v flow can't be used with it, did you use the distilled version flow? For the full model please try...

Tensor to CPU/GPU Node

The flow was updated, please take the latest version

SamplerCustomAdvanced LTXVTransformer3D.forward() missing 1 required positional argument: 'attention_mask'

Probably you are not using the model resulting from the configurator node.

13b distilled fp8 crashes with no error message

Is there any error message when it crashes?

LTX Audio VAE is not found.

try to use distilled model and one of the fp8 comfy flags

LTX Audio VAE is not found.

make sure you use flow suitable for distilled model

LTX-2 - Still getting mat1 and mat2 shapes cannot be multiplied (1024x62208 and 188160x3840) even with live preview off.

you should use the gemma model from [here](https://huggingface.co/google/gemma-3-12b-it-qat-q4_0-unquantized) with our flows, we are working on allowing prompt enhancer to work with comfy's gemma loader so it won't be needed.