Phr00t

/

LTX2-Rapid-Merges

like 258

Read the note, as instructed.

KJNodes required (make sure it is updated).

Read the note, as instructed.

KJNodes required (make sure it is updated).

thanks Phr00t, nodes issue resolved as your instruction, still one question here need your help, i can't found where to download google-gemma-3-12b-it-fp8-scaled, i only found google-gemma-3-12b-it-fp8 on Huggingface, but it's huge, could you show me where to get google-gemma-3-12b-it-fp8-scaled? thanks

Read the note, as instructed.

KJNodes required (make sure it is updated).

thanks Phr00t, nodes issue resolved as your instruction, still one question here need your help, i can't found where to download google-gemma-3-12b-it-fp8-scaled, i only found google-gemma-3-12b-it-fp8 on Huggingface, but it's huge, could you show me where to get google-gemma-3-12b-it-fp8-scaled? thanks

Read the note, as instructed.

KJNodes required (make sure it is updated).

thanks Phr00t, nodes issue resolved as your instruction, still one question here need your help, i can't found where to download google-gemma-3-12b-it-fp8-scaled, i only found google-gemma-3-12b-it-fp8 on Huggingface, but it's huge, could you show me where to get google-gemma-3-12b-it-fp8-scaled? thanks

found it here , https://huggingface.co/Comfy-Org/ltx-2/tree/main/split_files/text_encoders

Read the note, as instructed.

KJNodes required (make sure it is updated).

thanks Phr00t, nodes issue resolved as your instruction, still one question here need your help, i can't found where to download google-gemma-3-12b-it-fp8-scaled, i only found google-gemma-3-12b-it-fp8 on Huggingface, but it's huge, could you show me where to get google-gemma-3-12b-it-fp8-scaled? thanks

found it here , https://huggingface.co/Comfy-Org/ltx-2/tree/main/split_files/text_encoders

I will add this link to the model page!

hello phr00t! After numerous attempts this custom node(LTXVNormalizingSampler) is not getting installed I tried everything comfyui update bts, manager updates, install via git, download as zip, rest of the custom nodes got installed easily through custom node manager

hello phr00t! After numerous attempts this custom node(LTXVNormalizingSampler) is not getting installed I tried everything comfyui update bts, manager updates, install via git, download as zip, rest of the custom nodes got installed easily through custom node manager

You need to install and update ComfyUI-LTXVideo nodes to the latest. That is where that one is.

i tried to run the workflow many times, but everytime it shown an error as following, i have all the nodes installed properly and used the exactly the same model, VAE, CLIPs as shown in the Phr00t's original workflow, and i run it with offical latest version desktop comfyUI, i don't know why this issue happened.
"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj",

i tried to run the workflow many times, but everytime it shown an error as following, i have all the nodes installed properly and used the exactly the same model, VAE, CLIPs as shown in the Phr00t's original workflow, and i run it with offical latest version desktop comfyUI, i don't know why this issue happened.
"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj",

You left a screenshot in the LTX-2-Thread which indicates that you don't run the latest official version: your ComfyUI version is v.0.8.2, the current one is 0.9.2. Your frontendversion is v.135.9, not 1.36.14. But you are using the desktop version, which is the newest one: v0.7.2 but the desktop version doesn't get daily updates. If you don't know how to install into your custom-nodes for the Comfy desktop version, you have to wait a few days or maybe just another day. If that happens don't forget to check the node in "properties panel", the DualClip loader should have the newest version number, but the one in your workflow doesn't update itself. Replace it with one out of the "nodes"-library (left panel).

i tried to run the workflow many times, but everytime it shown an error as following, i have all the nodes installed properly and used the exactly the same model, VAE, CLIPs as shown in the Phr00t's original workflow, and i run it with offical latest version desktop comfyUI, i don't know why this issue happened.
"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj",

You left a screenshot in the LTX-2-Thread which indicates that you don't run the latest official version: your ComfyUI version is v.0.8.2, the current one is 0.9.2. Your frontendversion is v.135.9, not 1.36.14. But you are using the desktop version, which is the newest one: v0.7.2 but the desktop version doesn't get daily updates. If you don't know how to install into your custom-nodes for the Comfy desktop version, you have to wait a few days or maybe just another day. If that happens don't forget to check the node in "properties panel", the DualClip loader should have the newest version number, but the one in your workflow doesn't update itself. Replace it with one out of the "nodes"-library (left panel).

thank you so much for your clear instruction, i know how to install the custom-nodes for desktop version, i can git clone to the custom-nodes folder or just install them with mananger, but i don't know which nodes to update or install to update the version of DualClip node, could you please tell me this? thanks!