tutorials/image/z-image/z-image-turbo

[giscus[bot]]

tutorials/image/z-image/z-image-turbo

Z-Image is a 6B parameter efficient image generation foundation model with single-stream diffusion transformer, offering sub-second inference latency.

https://docs.comfy.org/tutorials/image/z-image/z-image-turbo

[Demuzx]

Is it joke?

CLIPLoader
Error(s) in loading state_dict for Llama2:
size mismatch for model.embed_tokens.weight: copying a param with shape torch.Size([151936, 2560]) from checkpoint, the shape in current model is torch.Size([128256, 4096]).
size mismatch for model.layers.0.self_attn.q_proj.weight: copying a param with shape torch.Size([4096, 2560]) from checkpoint, the shape in current model is torch.Size([4096, 4096]).
size mismatch for model.layers.0.self_attn.k_proj.weight: copying a param with shape torch.Size([1024, 2560]) from checkpoint, the shape in current model is torch.Size([1024, 4096]).

[RebootWebsites]

why are you posting llama 2 error for this image model?

[MiBaDK]

That's the errors I get as well

[RebootWebsites]

took a while to reproduce this...
Guys once you load the models into the correct folders. you select the models correctly in the sampler (unet_name, vae, clip ) . You click run once - will fail.

Now, click on the manager (top right), you will see the red restart button. ON TOP OF THAT, click on "Update All Custom Nodes" , once update done - click restart.

Now will work perfect.

[Philnex]

Had the same issue, updating all did it for me.

[Yonghao-lucky]

Is it joke?

CLIPLoader Error(s) in loading state_dict for Llama2: size mismatch for model.embed_tokens.weight: copying a param with shape torch.Size([151936, 2560]) from checkpoint, the shape in current model is torch.Size([128256, 4096]). size mismatch for model.layers.0.self_attn.q_proj.weight: copying a param with shape torch.Size([4096, 2560]) from checkpoint, the shape in current model is torch.Size([4096, 4096]). size mismatch for model.layers.0.self_attn.k_proj.weight: copying a param with shape torch.Size([1024, 2560]) from checkpoint, the shape in current model is torch.Size([1024, 4096]).

How to solve the problem of deploying on Linux servers?