Missing Models: text_encoders, diffusion_models

[juanca]

I am trying out a template for text to video. I was able to download the missing models but can't seem to figure out where to place them.

Missing Models

When loading the graph, the following models were not found:

- text_encoders / umt5_xxl_fp8_e4m3fn_scaled.safetensors
- diffusion_models / wan2.1_t2v_1.3B_fp16.safetensors

Chatgpt / google doesn't seem to give correct answers on where to place these downloaded files (why can't comfyui webui / comfyui-manager download these to the correct place?).

I tried:

• /data/models/text_encoders (had to create this directory)
• /data/models/diffusion_models (had to create this directory)
• /data/models/unet (had to create this directory)
• /data/models/clip
• /data/models/Stable-diffusion
• equivalent in: /stable-diffusion/models/

Nothing seems to work.

Whenever I place files in these directories and try to run the workflow. I get the following error message:

Prompt outputs failed validation: 
UNETLoader:
    - Value not in list: unet_name: 'wan2.1_t2v_1.3B_fp16.safetensors' not in []
CLIPLoader:
    - Value not in list: clip_name: 'umt5_xxl_fp8_e4m3fn_scaled.safetensors' not in []

Where am I suppose to place these files? How would I know where to place them? How do I reload comfyui after moving files? Do I always need to restart the docker container?

I am choosing to open this issue here because it seems like it has to do with how /data/ is mapped into comfyui.

[juanca]

I found the answer for clips.

root@da23463ab9d1:/data/models# cat /stable-diffusion/extra_model_paths.yaml
a111:
  base_path: /data

  checkpoints: models/Stable-diffusion
  configs: models/Stable-diffusion
  vae: models/VAE
  loras: models/Lora
  upscale_models: |
    models/RealESRGAN
    models/ESRGAN
    models/SwinIR
    models/GFPGAN
  hypernetworks: models/hypernetworks
  controlnet: models/ControlNet
  gligen: models/GLIGEN
  clip: models/CLIPEncoder
  embeddings: embeddings

Looks like we renamed the clip folder to CLIPEncoder.

I am making an attempt at documenting these paths.

I am still unsure about the unet files. What kind are those?

[JimmyAlan]

I am having the same issue but with different models, i have been trying all day to find out where to place the missing files to no avail.

[juanca]

Which kind of models?

[JimmyAlan]

the wan2.2 5b ones, i already found the folders, but now i am having different trouble with a node, but i won't mention it here as it is out of the purview of this github page.

[JimmyAlan]

If you haven't found where to place all the files in Pinokio. I think that's what you're using or was. I am not sure, but I tried every dang type of comfy, from the stand alone windows version, to pinokio, to comfyui-zluda. All of them except for the portable one, which I am dreading, after fussing with the zluda version with downloading everything and not having work was frustrating, but anyway here's all the folder locations: C:\pinokio\drive\drives\peers\d1757282761620 the umt file goes into the clip folder, the others are self-explanatory.