search

db0 / nataili

GNU Affero General Public License v3.0
13 stars 35 forks source link

img2img prompt weight problem? The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1 #26

Open jug-dev opened 1 year ago

jug-dev commented 1 year ago 
TRACE      | 2023-03-25 03:22:34.099801 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:22:42.063516 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '2460182306', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F135D82AD0>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:22:42.065511 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:22:53.433121 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '2987815359', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F45AAEB040>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:22:53.435116 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:22:59.398197 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '3467785280', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1EC91739360>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:22:59.400191 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:23:01.424005 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '1049846903', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F3C3B69D20>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:23:01.426997 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:23:02.566947 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '849254711', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F3412DB640>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:23:02.567954 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:23:03.615141 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '1875858139', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F2FB8F3250>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:23:03.617140 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:23:05.843430 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '1951882132', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F1A9C827D0>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:23:05.844396 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:23:12.533016 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '534180506', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F1A9C823B0>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:23:12.535011 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:26:10.551235 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '165397835', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1EC92F989A0>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:26:10.553230 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1

ERROR      | 2023-03-25 03:26:13.394278 | worker.jobs.stable_diffusion:start_job:373 - Something went wrong when processing request. Please check your trace.log file for the full stack trace. Payload: {'prompt': 'PROMPT REDACTED', 'height': 640, 'width': 640, 'seed': '2390131036', 'tiling': False, 'n_iter': 1, 'batch_size': 1, 'save_individual_images': False, 'save_grid': False, 'ddim_steps': 50, 'sampler_name': 'k_euler_karras', 'cfg_scale': 5.0, 'denoising_strength': 0.2, 'init_img': <PIL.Image.Image image mode=RGB size=512x512 at 0x1F135B27490>, 'request_type': 'img2img', 'model': 'Dreamlike Photoreal'}
TRACE      | 2023-03-25 03:26:13.396273 | worker.jobs.stable_diffusion:start_job:379 - Traceback (most recent call last):
  File "F:\AI\AI-Horde-Worker\worker\jobs\stable_diffusion.py", line 355, in start_job
    generator.generate(**gen_payload)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 628, in generate
    [
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\compvis.py", line 629, in <listcomp>
    get_learned_conditioning_with_prompt_weights(prompt, model, clip_skip)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 153, in get_learned_conditioning_with_prompt_weights
    current_prompt_c = update_conditioning(
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\util\cast.py", line 29, in wrap
    return amp.autocast(device_type="cuda", dtype=dtype)(no_grad()(func))(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\amp\autocast_mode.py", line 14, in decorate_autocast
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "F:\AI\AI-Horde-Worker\conda\envs\windows\lib\site-packages\nataili\stable_diffusion\prompt_weights.py", line 70, in update_conditioning
    current_prompt_c += (weight - 1.0) * subprompt_contribution_to_c
RuntimeError: The size of tensor a (154) must match the size of tensor b (231) at non-singleton dimension 1
ResidentChief commented 1 year ago

looks like one side of the prompt is bigger than the other, but by that point the tensors should have been fixed if they were mismatched... hmmm...