An error message of “Error: permute(sparse_coo): number of dimensions in the tensor input does not match the length of the desired ordering of dimensions i.e. input.dim() = 2 is not equal to len(dims) = 3” appears in the message field in the traintrain and Lora cannot be created in Forge.

[noha2310]

The following is the text after the TRAINTRAIN run.

---

Start Training!
Loading weights [5353d90e0c] from /content/drive/MyDrive/sd/stable-diffusion-webui-forge/models/Stable-diffusion/anyloraCheckpoint_bakedvaeBlessedFp16.safetensors
model_type EPS
UNet ADM Dimension 0
Using xformers attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using xformers attention in VAE
extra {'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_l.logit_scale'}
left over keys: dict_keys(['alphas_cumprod', 'alphas_cumprod_prev', 'betas', 'log_one_minus_alphas_cumprod', 'posterior_log_variance_clipped', 'posterior_mean_coef1', 'posterior_mean_coef2', 'posterior_variance', 'sqrt_alphas_cumprod', 'sqrt_one_minus_alphas_cumprod', 'sqrt_recip_alphas_cumprod', 'sqrt_recipm1_alphas_cumprod'])
loaded straight to GPU
To load target model BaseModel
Begin to load 1 model
[Memory Management] Current Free GPU Memory (MB) = 13313.25634765625
[Memory Management] Model Memory (MB) = 0.00762939453125
[Memory Management] Minimal Inference Memory (MB) = 1024.0
[Memory Management] Estimated Remaining GPU Memory (MB) = 12289.248718261719
Moving model(s) has taken 0.02 seconds
To load target model SD1ClipModel
Begin to load 1 model
[Memory Management] Current Free GPU Memory (MB) = 13313.24755859375
[Memory Management] Model Memory (MB) = 454.2076225280762
[Memory Management] Minimal Inference Memory (MB) = 1024.0
[Memory Management] Estimated Remaining GPU Memory (MB) = 11835.039936065674
Moving model(s) has taken 0.11 seconds
Model loaded in 12.4s (unload existing model: 2.9s, load weights from disk: 0.4s, forge load real models: 8.8s, calculate empty prompt: 0.1s).
Preparing the Model...
Enabling Xformers
Preparing image latents and text-conditional...
max bucket sizes : [(512, 512)]
sub bucket sizes : []
Traceback (most recent call last):
File "/content/drive/MyDrive/sd/stable-diffusion-webui-forge/extensions/sd-webui-traintrain/trainer/train.py", line 195, in train_main
result = train_lora(t)
File "/content/drive/MyDrive/sd/stable-diffusion-webui-forge/extensions/sd-webui-traintrain/trainer/train.py", line 226, in train_lora
dataloaders = dataset.make_dataloaders(t)
File "/content/drive/MyDrive/sd/stable-diffusion-webui-forge/extensions/sd-webui-traintrain/trainer/dataset.py", line 15, in make_dataloaders
load_resize_image_and_text(t) #画像を読み込み、画像サイズごとに振り分け、リサイズ、テキストの読み込み
File "/content/drive/MyDrive/sd/stable-diffusion-webui-forge/extensions/sd-webui-traintrain/trainer/dataset.py", line 241, in load_resize_image_and_text
resized, alpha_mask = resize_and_crop(ar_error, image, *max, t.image_disable_upscale)
File "/content/drive/MyDrive/sd/stable-diffusion-webui-forge/extensions/sd-webui-traintrain/trainer/dataset.py", line 231, in resize_and_crop
tensor = torch.from_numpy(np.array(image)).permute(2, 0, 1).float() / 255.0
RuntimeError: permute(sparse_coo): number of dimensions in the tensor input does not match the length of the desired ordering of dimensions i.e. input.dim() = 2 is not equal to len(dims) = 3

Loading weights [9807d1172b] from /content/drive/MyDrive/sd/stable-diffusion-webui-forge/models/Stable-diffusion/anyloracleanlinearmix_v10.safetensors
model_type EPS
UNet ADM Dimension 0
Using xformers attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using xformers attention in VAE
extra {'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_l.logit_scale'}
left over keys: dict_keys(['model_ema.decay', 'model_ema.num_updates'])
loaded straight to GPU
To load target model BaseModel
Begin to load 1 model
[Memory Management] Current Free GPU Memory (MB) = 13294.35791015625
[Memory Management] Model Memory (MB) = 0.00762939453125
[Memory Management] Minimal Inference Memory (MB) = 1024.0
[Memory Management] Estimated Remaining GPU Memory (MB) = 12270.350280761719
Moving model(s) has taken 0.02 seconds
To load target model SD1ClipModel
Begin to load 1 model
[Memory Management] Current Free GPU Memory (MB) = 13294.34912109375
[Memory Management] Model Memory (MB) = 454.2076225280762
[Memory Management] Minimal Inference Memory (MB) = 1024.0
[Memory Management] Estimated Remaining GPU Memory (MB) = 11816.141498565674
Moving model(s) has taken 0.10 seconds
Model loaded in 21.2s (load weights from disk: 1.2s, forge load real models: 19.6s, calculate empty prompt: 0.2s).

---

That's all.
Reinstalling traintrain did not solve the problem and searching did not find a solution.

How can I solve this problem?

[jurgens1206]

Hello.

I was also having trouble because the same error occurred, but when I reviewed the image data I was trying to learn, I realized that grayscale was mixed in, and I converted it to RGB, and it worked properly.

I will post the address of the blog referenced below.

Well then.

https://yukr.hatenablog.com/entry/2020/05/03/170502