Is there an existing issue for this?
What happened?
When using "Checkpoint Name" as an axis in the XYZ Plot script, the wrong checkpoint can be loaded.
Steps to reproduce the problem
This can be difficult to reproduce. In this example, I ran an XYZ plot with 3 similarly named checkpoints:
The ryanblend_v70 checkpoint was loaded instead of the ryanblend_v60 checkpoint. A duplicate image is then generated for the last checkpoint.
If the model name is set in the filename, ryanblend_v70 appears twice:
The final grid is incorrect:
I can load the ryanblend_v60 model and generate an image when not using XYZ Plot. Here's the example with the same prompt and seed value compared to ryanblend_v70:
I've had the same problem with other checkpoints.
What should have happened?
The correct checkpoint should be loaded.
Commit where the problem happens
version: v1.2.0 • python: 3.10.6 • torch: 2.0.0+cu118 • xformers: N/A • gradio: 3.29.0 • checkpoint: 5b53c2e02e
What platforms do you use to access the UI ?
Windows
What browsers do you use to access the UI ?
Mozilla Firefox
Command Line Arguments
List of extensions
No
Console logs
venv "H:\GitHub\stable-diffusion-webui\venv\Scripts\Python.exe"
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)]
Version: v1.2.0
Commit hash: b08500cec8a791ef20082628b49b17df833f5dda
Installing requirements
Launching Web UI with arguments:
No module 'xformers'. Proceeding without it.
*** "Disable all extensions" option was set, will not load any extensions ***
Loading weights [6e8859ee58] from H:\GitHub\stable-diffusion-webui\models\Stable-diffusion\StableDiffusion\StableDiffusion_v15-fp16.safetensors
Creating model from config: H:\GitHub\stable-diffusion-webui\configs\v1-inference.yaml
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
Loading VAE weights specified in settings: H:\GitHub\stable-diffusion-webui\models\VAE\vae-ft-mse-840000-ema-pruned.safetensors
Running on local URL: http://127.0.0.1:7860
To create a public link, set `share=True` in `launch()`.
Startup time: 7.4s (import torch: 1.3s, import gradio: 0.8s, import ldm: 0.4s, other imports: 0.7s, list SD models: 1.1s, create ui: 2.9s, gradio launch: 0.1s).
Applying cross attention optimization (Doggettx).
Textual inversion embeddings loaded(1): negative_hand-neg
Model loaded in 4.1s (load weights from disk: 0.3s, create model: 0.8s, apply weights to model: 0.8s, apply half(): 0.6s, load VAE: 0.3s, move model to device: 0.6s, load textual inversion embeddings: 0.8s).
X/Y/Z plot will create 3 images on 1 3x1 grid. (Total steps to process: 60)
Loading weights [ce31f6a078] from H:\GitHub\stable-diffusion-webui\models\Stable-diffusion\ryanblend_v50.safetensorst/s]
Loading VAE weights specified in settings: cached vae-ft-mse-840000-ema-pruned.safetensors
Applying cross attention optimization (Doggettx).
Weights loaded in 0.7s (apply weights to model: 0.3s, move model to device: 0.3s).
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [00:02<00:00, 7.78it/s]
Loading weights [5b53c2e02e] from H:\GitHub\stable-diffusion-webui\models\Stable-diffusion\ryanblend_v70.safetensors/s]
Loading VAE weights specified in settings: cached vae-ft-mse-840000-ema-pruned.safetensors
Applying cross attention optimization (Doggettx).
Weights loaded in 1.1s (apply weights to model: 0.3s, load VAE: 0.4s, move model to device: 0.3s).
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [00:01<00:00, 11.27it/s]
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [00:01<00:00, 11.26it/s]
Loading weights [6e8859ee58] from H:\GitHub\stable-diffusion-webui\models\Stable-diffusion\StableDiffusion\StableDiffusion_v15-fp16.safetensors
Loading VAE weights specified in settings: cached vae-ft-mse-840000-ema-pruned.safetensors
Applying cross attention optimization (Doggettx).
Weights loaded in 1.0s (load weights from disk: 0.1s, apply weights to model: 0.3s, load VAE: 0.3s, move model to device: 0.3s).
Total progress: 100%|██████████████████████████████████████████████████████████████████| 60/60 [00:11<00:00, 5.02it/s]
Total progress: 100%|██████████████████████████████████████████████████████████████████| 60/60 [00:11<00:00, 11.20it/s]Additional information
No response
Is there an existing issue for this?
What happened?
When using "Checkpoint Name" as an axis in the XYZ Plot script, the wrong checkpoint can be loaded.
Steps to reproduce the problem
This can be difficult to reproduce. In this example, I ran an XYZ plot with 3 similarly named checkpoints:
The ryanblend_v70 checkpoint was loaded instead of the ryanblend_v60 checkpoint. A duplicate image is then generated for the last checkpoint.
If the model name is set in the filename, ryanblend_v70 appears twice:
The final grid is incorrect:
I can load the ryanblend_v60 model and generate an image when not using XYZ Plot. Here's the example with the same prompt and seed value compared to ryanblend_v70:
I've had the same problem with other checkpoints.
What should have happened?
The correct checkpoint should be loaded.
Commit where the problem happens
version: v1.2.0 • python: 3.10.6 • torch: 2.0.0+cu118 • xformers: N/A • gradio: 3.29.0 • checkpoint: 5b53c2e02e
What platforms do you use to access the UI ?
Windows
What browsers do you use to access the UI ?
Mozilla Firefox
Command Line Arguments
List of extensions
No
Console logs
Additional information
No response