Kosinkadink
commented
1 year ago Hey, no worries! Did you install AnimateDiff using comfy manager, or a git clone? And can you try to turn off comfy manager/remove and see if your issue persists?
Closed verosment closed 1 month ago
Kosinkadink
commented
1 year ago Hey, no worries! Did you install AnimateDiff using comfy manager, or a git clone? And can you try to turn off comfy manager/remove and see if your issue persists?
verosment
commented
1 year ago Hi, sadly i have already tried every single one of these. I initially installed with comfy manager, then removed comfy manager and reinstalled using git clone, then went to a brand new fresh installation of ComfyUI and the issue still persisted across all these attempts.
Kosinkadink
commented
1 year ago I will try to replicate your issue. What exactly do you mean by "after 2 generations"? Can you go into detail when it works vs it doesn't?
verosment
commented
1 year ago I can successfully generate 2 animated images before I get the error on the third, the first result usually taking around 230 seconds but the second taking only 50 seconds roughly. Here are my launch parameters if it helps;
--windows-standalone-build --force-fp32 --preview-method auto --normalvram
I also used --disable-xformers but i got the exact same error. I have tried many different launch options but they didn't seem to do anything for the error so i just left them.
Also after the error I can successfully generate normal images without fail
Kosinkadink
commented
1 year ago Gotcha, can you try it without any additional arguments, so just --windows-standalone-build ?
verosment
commented
1 year ago Same error sadly
Kosinkadink
commented
1 year ago Hmm, someone in the other open issue a was able to solve their problem by using a different SD model - I think that SD1.5 inpaint models fail in unique ways, as SD1.5 models (noninpaint) are expected. But assuming that is not the issue, I am going to make a branch that is reverted back to my changes prior to Friday - once I make it, would you be down to git checkout that branch so we can see if this is a new or old issue?
verosment
commented
1 year ago Of course, I'd be more than happy to
Kosinkadink
commented
1 year ago I've made the branch - to switch to it, cd into ComfyUI-AnimateDiff-Evolved, and then do git pull, and then git checkout before-refactor-1. For future reference, you can then switch back to main later using git checkout main.
Kosinkadink
commented
1 year ago Crap, I forgot to push the branch - it's up now, but was not when I posted that comment.
Kosinkadink
commented
1 year ago If you want a simple workflow that will work with that old commit, use this link: https://github.com/Kosinkadink/ComfyUI-AnimateDiff-Evolved/tree/before-refactor-1 (it's the readme from that branch)
verosment
commented
1 year ago I think this fixed the issue, I have so far generated 4 short low-step count gifs without error, but I have noticed it is much slower than it used to be which is quite noticeable on a system like mine (speeds have gone from roughly 3s/it on the main branch to 25/it on this older one)
Kosinkadink
commented
1 year ago Gotcha, so there is a super-rare, hard to reproduce bug in the new refactor. That at least help track down where the problem could lie. This is less of a fix as there are some other underlying issues in the old code that was the whole reason for my refactor, but things should kinda work the same in terms of the animatediff output - you will need to use Load Checkpoint w/ Noise Select though because back then I couldn't figure out how to properly go back and forth between the noise schedules.
I am going to spend part of today looking through to see what could be wrong, but it's so strange that everything just works for pretty much everyone else.
verosment
commented
1 year ago Thank you. Also, you mentioned ComfyUI portable in #51, so I manually installed it but i ran into the exact same issues, which didn't surprise me much considering i always try my best to keep ComfyUI up to date by putting git pull in the start .bat file. But i did notice something this time around, the branch that works for me never goes into lowvram mode, but in the case of the main branch, it crashes after generating one result with lowvram mode.
It will generate one in normally, then the second will be in lowvram which will work, but when it tries to generate the third in lowvram mode that is when the error occurs. I think lowvram mode could be related to the problem somehow.
Here is the first lowvram generation followed by the second lowvram generation, which is the one that errors it out:
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (8) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15V2.ckpt version v2.
loading new
loading in lowvram mode 1761.8534021377563
100%|████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:05<00:00, 5.31s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15V2.ckpt version v2.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
Prompt executed in 37.97 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (8) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15V2.ckpt version v2.
loading new
loading in lowvram mode 1804.9303255081177
0%| | 0/1 [00:00<?, ?it/s]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15V2.ckpt version v2.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
!!! Exception during processing !!!
Traceback (most recent call last):
File "D:\apps\AI\COMFYUI3\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "D:\apps\AI\COMFYUI3\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "D:\apps\AI\COMFYUI3\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "D:\apps\AI\COMFYUI3\ComfyUI\nodes.py", line 1236, in sample
return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise)
File "D:\apps\AI\COMFYUI3\ComfyUI\nodes.py", line 1206, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 161, in animatediff_sample
return wrap_function_to_inject_xformers_bug_info(orig_comfy_sample)(model, *args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\model_utils.py", line 144, in wrapped_function
return function_to_wrap(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\sample.py", line 93, in sample
samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\samplers.py", line 742, in sample
samples = getattr(k_diffusion_sampling, "sample_{}".format(self.sampler))(self.model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\k_diffusion\sampling.py", line 137, in sample_euler
denoised = model(x, sigma_hat * s_in, **extra_args)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\samplers.py", line 323, in forward
out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, cond_concat=cond_concat, model_options=model_options, seed=seed)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\k_diffusion\external.py", line 125, in forward
eps = self.get_eps(input * c_in, self.sigma_to_t(sigma), **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\k_diffusion\external.py", line 151, in get_eps
return self.inner_model.apply_model(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\samplers.py", line 311, in apply_model
out = sampling_function(self.inner_model.apply_model, x, timestep, uncond, cond, cond_scale, cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 527, in sliding_sampling_function
cond, uncond = calc_cond_uncond_batch(model_function, cond, uncond, x, timestep, max_total_area, cond_concat, model_options)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 427, in calc_cond_uncond_batch
output = model_function(input_x, timestep_, **c).chunk(batch_chunks)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\model_base.py", line 63, in apply_model
return self.diffusion_model(xc, t, context=context, y=c_adm, control=control, transformer_options=transformer_options).float()
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\comfy\ldm\modules\diffusionmodules\openaimodel.py", line 634, in forward
h = forward_timestep_embed(self.middle_block, h, emb, context, transformer_options)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 71, in forward_timestep_embed
x = layer(x, context)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 398, in forward
return self.temporal_transformer(input_tensor, encoder_hidden_states, attention_mask)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI3\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 461, in forward
hidden_states = self.norm(hidden_states)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\hooks.py", line 160, in new_forward
args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\hooks.py", line 104, in pre_forward
args, kwargs = hook.pre_forward(module, *args, **kwargs)
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\hooks.py", line 286, in pre_forward
set_module_tensor_to_device(
File "C:\Users\JohnJ\AppData\Local\Programs\Python\Python310\lib\site-packages\accelerate\utils\modeling.py", line 298, in set_module_tensor_to_device
new_value = value.to(device)
NotImplementedError: Cannot copy out of meta tensor; no data!
Prompt executed in 20.81 secondsInteresting, I will see if I can force mine into low vram mode explicitly and replicate the issue, good catch and thanks for the heads up! If I can replicate it, then I can actually do trial and error with code to fix it. Otherwise, we are playing a long silly game of telephone lol
Kosinkadink
commented
1 year ago @Cedri4 I'm adding you to this thread so that we can chat without me having to retype messages.
As veresmont noticed, model loading in lowvram mode appears to be a possible issue with new code. For context, Cedri4 is rocking only 3GB of VRAM, while veresmont is rocking 6GB. Looking at his cedri4's logs, he also gets into low vram mode when using the old code, but does not crash on subsequent generations. So, only the new code is unhappy. What is most weird is that ced crashes on his second run, but veresmont does on his third.
I would like you guys to try out a couple things for me on the main branch. Prerequisites:
What to do:
1) boot up comfy ui 2) load the basic txt2img workflow in the main branch 3) select an SD model to use. We'll call this SD model A. 4) select a motion model to use (anything but 15_v2). We'll call this ADiff model A. 5) run the workflow once. 6) increment the seed 7) run it again. 8) EXPECTED BEHAVIOR: ced should now be crashed. Veresmont, run it once more so you crash too, and make note if the second image has any visible image degradation (or just post em here so I can compare). Copy all command line output, and send it my way as TEST 1 Results. 9) shut down comfy
1) do steps 1-5 from TEST 1. 2) keep the seed the same, but now change to use your second SD model; we'll call this SD model B. DO NOT CHANGE YOUR MOTION MODEL. 3) run it again. 4) EXPECTED BEHAVIOR: this is where things might change. Ced, note if you have crashed now or not. Veresmont, run again and note if you have crashed. In the case you have not crashed yet, run it again with the same SD model B, DO NOT CHANGE THE MOTION MODEL. And note the results. Send the command line output as TEST 2 results. 5) shut down comfy
1) do everything in TEST 2 but instead of switching SD models from A to B, stick with SD model A the whole time, and instead, switch from ADiff model A to ADiff model B after the first run. As before, note differences and try to run until you get the error. Results for this are Test 3 results. 2) remember to shut down comfy
1) If TEST 2 or TEST 3 (or both) yielded results different from TEST 1, repeat the tests that yielded different results, but this time, before running the workflow the nth time, where n is the run that would cause you to crash in TEST 2 or TEST 3, switch the SD model back to A (TEST 2) or switch the ADiff model back to A (TEST 3), and run it. If you don't crash at this point, switch back to models B (based on which test you're in), and run it again. If you still haven't crashed, run it again without changing models until you crash. Note the results as TEST2ALT and/or TEST3ALT. 2) shut down comfy
After we get results for these tests, I will review results and we can do more tests if needed. This will help me track down the issue immensely, and I can't replicate this on my end so this is the only way for me to narrow things down.
Kosinkadink
commented
1 year ago Oh yeah, also, save your generated gifs! Would be good to know if the output changes on runs that don't crash
Cedri4
commented
1 year ago The successful run (first)
The unsuccessful run: (second)
Cedri4
commented
1 year ago Successful run (first)
Successful run (second) It has work in incredient but it was instant I doubt it has change anything and the size of the file is the same as the first
Unsuccessful run (third)
Kosinkadink
commented
1 year ago Thanks for running them! But I think I may have been misunderstood in terms of the terminal output - I wanna see the full comfy UI terminal output since the time comfy boots up, so that I can see all the [AnimateDiffEvo] messages and any of the print statements from comfy - those will let me see what the plugin (and comfy) is trying to do
Cedri4
commented
1 year ago Crap the terminal are already closed
Kosinkadink
commented
1 year ago If it wouldn't be too much trouble, that terminal output from the 3 tests is what I'm most interested in - some of those statements only get printed when certain things happen.
Cedri4
commented
1 year ago For test 3 I have to use model a or b?
Kosinkadink
commented
1 year ago Technically, SD model A, but since you should booting comfy at the beginning of the test, either works. I'd do SD model A to be consistent.
Kosinkadink
commented
1 year ago Hopefully with the terminal output from TEST 1, TEST 2, and TEST 3, I can figure out where things go wrong, and then I can make a branch with more detailed print statements for you guys to run and send me the output of
Cedri4
commented
1 year ago Run 1 is successful Run 2 is successful basically, it was instant and the same as the first run with increment seed. Run 3 is unsuccessful with model A
So this is the whole terminal with the first run 2/3 run with the fixed seed which was instant then 2unsuccessful run
Kosinkadink
commented
1 year ago Did you change the motion module in test 3 between the first run and the second run as the test instructions said?
Kosinkadink
commented
1 year ago Cause looking at the output, it was loading mm-Stabilized_mid.pth every single time, I don't see any change
Kosinkadink
commented
1 year ago And I'd also need the terminal output from tests 1 and 2
Kosinkadink
commented
1 year ago Be sure to follow the intructions to a T, so that I can draw accurate conclusions from what we get printed in the terminal
Kosinkadink
commented
1 year ago And if you need any clarification for any of them, let me know!
verosment
commented
1 year ago Hi, sorry for late response, just woke up. I will get to testing these immediately.
@Cedri4 I'm adding you to this thread so that we can chat without me having to retype messages.
As veresmont noticed, model loading in lowvram mode appears to be a possible issue with new code. For context, Cedri4 is rocking only 3GB of VRAM, while veresmont is rocking 6GB. Looking at his cedri4's logs, he also gets into low vram mode when using the old code, but does not crash on subsequent generations. So, only the new code is unhappy. What is most weird is that ced crashes on his second run, but veresmont does on his third.
I would like you guys to try out a couple things for me on the main branch. Prerequisites:
- have at least 2 SD1.5-compatible checkpoints to use
- have at least 2 motion modules to use
What to do:
TEST 1 (baseline)
- boot up comfy ui
- load the basic txt2img workflow in the main branch
- select an SD model to use. We'll call this SD model A.
- select a motion model to use (anything but 15_v2). We'll call this ADiff model A.
- run the workflow once.
- increment the seed
- run it again.
- EXPECTED BEHAVIOR: ced should now be crashed. Veresmont, run it once more so you crash too, and make note if the second image has any visible image degradation (or just post em here so I can compare). Copy all command line output, and send it my way as TEST 1 Results.
- shut down comfy
TEST 2 (SD model change)
- do steps 1-5 from TEST 1.
- keep the seed the same, but now change to use your second SD model; we'll call this SD model B. DO NOT CHANGE YOUR MOTION MODEL.
- run it again.
- EXPECTED BEHAVIOR: this is where things might change. Ced, note if you have crashed now or not. Veresmont, run again and note if you have crashed. In the case you have not crashed yet, run it again with the same SD model B, DO NOT CHANGE THE MOTION MODEL. And note the results. Send the command line output as TEST 2 results.
- shut down comfy
TEST 3 (ADiff model change)
- do everything in TEST 2 but instead of switching SD models from A to B, stick with SD model A the whole time, and instead, switch from ADiff model A to ADiff model B after the first run. As before, note differences and try to run until you get the error. Results for this are Test 3 results.
- remember to shut down comfy
EXTRA TESTS
- If TEST 2 or TEST 3 (or both) yielded results different from TEST 1, repeat the tests that yielded different results, but this time, before running the workflow the nth time, where n is the run that would cause you to crash in TEST 2 or TEST 3, switch the SD model back to A (TEST 2) or switch the ADiff model back to A (TEST 3), and run it. If you don't crash at this point, switch back to models B (based on which test you're in), and run it again. If you still haven't crashed, run it again without changing models until you crash. Note the results as TEST2ALT and/or TEST3ALT.
- shut down comfy
After we get results for these tests, I will review results and we can do more tests if needed. This will help me track down the issue immensely, and I can't replicate this on my end so this is the only way for me to narrow things down.
Sorry for the late response, i just woke up, I will get to testing these immediately.
Kosinkadink
commented
1 year ago No worries, I appreciate the help!
verosment
commented
1 year ago Is it fine to use animateddiffMotion_V15 (Not V2)?
Kosinkadink
commented
1 year ago yep. technically, the only reason why i say not V2 is because it has an extra little portion to it vs the other models. It still works just fine in AD, but for debugging purposes, I am trying to keep everything constant.
verosment
commented
1 year ago Is it alright if I change the resolution and steps? I only ask because it is running at 182s/it with 512x512 and 20 steps and says it will take 57 minutes for one result and I wish to run it at something like 256x256 at 10 steps, which will still get an intelligible result.
verosment
commented
1 year ago I used the text2image example from the main branch but the only thing I changed was the Load VAE module as I don't have any external VAEs downloaded, so I switched it to use the VAE from the model. (I don't really know what VAEs exactly are to be honest). I also lowered the resolution down to 256x256 and changed steps to 10 so that I could actually run this at a suitable time as I'm going out later today and 1 hour for a single result is a bit out of my time budget right now. If it is required then I will run 512x512 20-step generations later tomorrow.
--Motion Model; animatediffMotion_v15 --Diffusion Model; epicrealism_naturalSin
CMD:
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --force-fp32 --preview-method auto --normalvram --use-pytorch-cross-attention --disable-xformers
Total VRAM 6144 MB, total RAM 16307 MB
Forcing FP32, if this improves things please report it.
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce GTX 1660 : cudaMallocAsync
VAE dtype: torch.float32
Using pytorch cross attention
Using pytorch cross attention
Import times for custom nodes:
0.1 seconds: D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved
Starting server
To see the GUI go to: http://127.0.0.1:8188
got prompt
model_type EPS
adm 0
making attention of type 'vanilla-pytorch' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-pytorch' with 512 in_channels
missing {'cond_stage_model.logit_scale', 'cond_stage_model.text_projection'}
left over keys: dict_keys(['cond_stage_model.transformer.text_model.embeddings.position_ids', 'model_ema.decay', 'model_ema.num_updates'])
[AnimateDiffEvo] - INFO - Loading motion module animatediffMotion_v15.ckpt
loading new
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\_utils.py:776: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly. To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
return self.fget.__get__(instance, owner)()
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [10:49<00:00, 64.96s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:47: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('betas', torch.tensor(betas, dtype=torch.float32))
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:48: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('alphas_cumprod', torch.tensor(alphas_cumprod, dtype=torch.float32))
Prompt executed in 782.55 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1763.1319274902344
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [01:53<00:00, 11.30s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
Prompt executed in 166.46 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1907.7473125457764
0%| | 0/10 [00:01<?, ?it/s]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
!!! Exception during processing !!!
Traceback (most recent call last):
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1236, in sample
return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1206, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 161, in animatediff_sample
return wrap_function_to_inject_xformers_bug_info(orig_comfy_sample)(model, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\model_utils.py", line 144, in wrapped_function
return function_to_wrap(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\sample.py", line 93, in sample
samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 741, in sample
samples = getattr(k_diffusion_sampling, "sample_{}".format(self.sampler))(self.model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\sampling.py", line 137, in sample_euler
denoised = model(x, sigma_hat * s_in, **extra_args)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 322, in forward
out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, cond_concat=cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 125, in forward
eps = self.get_eps(input * c_in, self.sigma_to_t(sigma), **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 151, in get_eps
return self.inner_model.apply_model(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 310, in apply_model
out = sampling_function(self.inner_model.apply_model, x, timestep, uncond, cond, cond_scale, cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 527, in sliding_sampling_function
cond, uncond = calc_cond_uncond_batch(model_function, cond, uncond, x, timestep, max_total_area, cond_concat, model_options)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 427, in calc_cond_uncond_batch
output = model_function(input_x, timestep_, **c).chunk(batch_chunks)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py", line 63, in apply_model
return self.diffusion_model(xc, t, context=context, y=c_adm, control=control, transformer_options=transformer_options).float()
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\ldm\modules\diffusionmodules\openaimodel.py", line 653, in forward
h = forward_timestep_embed(module, h, emb, context, transformer_options, output_shape)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 71, in forward_timestep_embed
x = layer(x, context)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 398, in forward
return self.temporal_transformer(input_tensor, encoder_hidden_states, attention_mask)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 461, in forward
hidden_states = self.norm(hidden_states)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 160, in new_forward
args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 104, in pre_forward
args, kwargs = hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 286, in pre_forward
set_module_tensor_to_device(
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\utils\modeling.py", line 298, in set_module_tensor_to_device
new_value = value.to(device)
NotImplementedError: Cannot copy out of meta tensor; no data!
Prompt executed in 31.14 secondsResult 1:
Result 2:
Result 3: Throws the error.
Kosinkadink
commented
1 year ago Is it alright if I change the resolution and steps? I only ask because it is running at 182s/it with 512x512 and 20 steps and says it will take 57 minutes for one result and I wish to run it at something like 256x256 at 10 steps, which will still get an intelligible result.
I think you can make the size super tiny if you'd like, evne something like 64x64 should work if you are really stretched on time
Kosinkadink
commented
1 year ago and you can have the steps be very low, too
verosment
commented
1 year ago I'll put it to 128 or 192 for the time being, gonna run test 2 now.
verosment
commented
1 year ago I changed the resolution to 128 instead of the 256 resolution I used in Test 1, but everything else should be the same. (Besides the model, of course.)
--Motion Model; animatediffMotion_v15 --Diffusion Model; dreamshaper_8
CMD:
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --force-fp32 --preview-method auto --normalvram --use-pytorch-cross-attention --disable-xformers
Total VRAM 6144 MB, total RAM 16307 MB
Forcing FP32, if this improves things please report it.
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce GTX 1660 : cudaMallocAsync
VAE dtype: torch.float32
Using pytorch cross attention
Using pytorch cross attention
Import times for custom nodes:
0.1 seconds: D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved
Starting server
To see the GUI go to: http://127.0.0.1:8188
got prompt
model_type EPS
adm 0
making attention of type 'vanilla-pytorch' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-pytorch' with 512 in_channels
missing {'cond_stage_model.logit_scale', 'cond_stage_model.text_projection'}
left over keys: dict_keys(['cond_stage_model.transformer.text_model.embeddings.position_ids'])
[AnimateDiffEvo] - INFO - Loading motion module animatediffMotion_v15.ckpt
loading new
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\_utils.py:776: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly. To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
return self.fget.__get__(instance, owner)()
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [01:52<00:00, 11.24s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:47: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('betas', torch.tensor(betas, dtype=torch.float32))
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:48: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('alphas_cumprod', torch.tensor(alphas_cumprod, dtype=torch.float32))
Prompt executed in 154.85 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1763.1319274902344
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [00:20<00:00, 2.02s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
Prompt executed in 23.97 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1907.7473125457764
0%| | 0/10 [00:01<?, ?it/s]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
!!! Exception during processing !!!
Traceback (most recent call last):
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1236, in sample
return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1206, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 161, in animatediff_sample
return wrap_function_to_inject_xformers_bug_info(orig_comfy_sample)(model, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\model_utils.py", line 144, in wrapped_function
return function_to_wrap(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\sample.py", line 93, in sample
samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 741, in sample
samples = getattr(k_diffusion_sampling, "sample_{}".format(self.sampler))(self.model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\sampling.py", line 137, in sample_euler
denoised = model(x, sigma_hat * s_in, **extra_args)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 322, in forward
out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, cond_concat=cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 125, in forward
eps = self.get_eps(input * c_in, self.sigma_to_t(sigma), **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 151, in get_eps
return self.inner_model.apply_model(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 310, in apply_model
out = sampling_function(self.inner_model.apply_model, x, timestep, uncond, cond, cond_scale, cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 527, in sliding_sampling_function
cond, uncond = calc_cond_uncond_batch(model_function, cond, uncond, x, timestep, max_total_area, cond_concat, model_options)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 427, in calc_cond_uncond_batch
output = model_function(input_x, timestep_, **c).chunk(batch_chunks)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py", line 63, in apply_model
return self.diffusion_model(xc, t, context=context, y=c_adm, control=control, transformer_options=transformer_options).float()
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\ldm\modules\diffusionmodules\openaimodel.py", line 653, in forward
h = forward_timestep_embed(module, h, emb, context, transformer_options, output_shape)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 71, in forward_timestep_embed
x = layer(x, context)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 398, in forward
return self.temporal_transformer(input_tensor, encoder_hidden_states, attention_mask)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 461, in forward
hidden_states = self.norm(hidden_states)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 160, in new_forward
args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 104, in pre_forward
args, kwargs = hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 286, in pre_forward
set_module_tensor_to_device(
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\utils\modeling.py", line 298, in set_module_tensor_to_device
new_value = value.to(device)
NotImplementedError: Cannot copy out of meta tensor; no data!
Prompt executed in 2.58 secondsResult 1:
Result 2:
Result 3: Error.
Kosinkadink
commented
1 year ago And to be clear, did you change the SD model between your result 2 and 3 in this test? Not between test 1 and 2 i mean, but between the results in test 2.
verosment
commented
1 year ago No, the model stayed as dreamshaper_8 the whole time in Test 2.
verosment
commented
1 year ago
- do everything in TEST 2 but instead of switching SD models from A to B, stick with SD model A the whole time, and instead, switch from ADiff model A to ADiff model B after the first run. As before, note differences and try to run until you get the error. Results for this are Test 3 results.
Does this mean generating the first one with ADiff model A and then the second with ADiff model B? What about after the second generation, should I switch back to A or keep it on B? Or should I run it all on ADiff model B?
Kosinkadink
commented
1 year ago You'll need to repeat Test 2 before you go for Test 3, but change the SD model between Result 2 and Result 3. Basically, I am testing here to see if the SD model it has in memory gets corrupted due to my code not cleaning it properly for one reason or another.
And Test 3 will have the same SD model between Result 2 and 3, but a different motion model.
verosment
commented
1 year ago So change the SD model on the third generation?
Kosinkadink
commented
1 year ago Yep, exactly, that's Test 2. And on Test 3, you will instead change the motion model.
verosment
commented
1 year ago --Motion Model; animatediffMotion_v15 --Result 1 & 2 Diffusion Model; epicrealism_naturalSin --Result 3 Diffusion Model; dreamshaper_8
CMD:
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --force-fp32 --preview-method auto --normalvram --use-pytorch-cross-attention --disable-xformers
Total VRAM 6144 MB, total RAM 16307 MB
Forcing FP32, if this improves things please report it.
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce GTX 1660 : cudaMallocAsync
VAE dtype: torch.float32
Using pytorch cross attention
Using pytorch cross attention
Import times for custom nodes:
0.1 seconds: D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved
Starting server
To see the GUI go to: http://127.0.0.1:8188
got prompt
model_type EPS
adm 0
making attention of type 'vanilla-pytorch' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-pytorch' with 512 in_channels
missing {'cond_stage_model.text_projection', 'cond_stage_model.logit_scale'}
left over keys: dict_keys(['cond_stage_model.transformer.text_model.embeddings.position_ids', 'model_ema.decay', 'model_ema.num_updates'])
[AnimateDiffEvo] - INFO - Loading motion module animatediffMotion_v15.ckpt
loading new
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\_utils.py:776: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly. To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
return self.fget.__get__(instance, owner)()
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [01:54<00:00, 11.49s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:47: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('betas', torch.tensor(betas, dtype=torch.float32))
D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py:48: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
self.register_buffer('alphas_cumprod', torch.tensor(alphas_cumprod, dtype=torch.float32))
Prompt executed in 152.51 seconds
got prompt
2
3
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1763.1319274902344
100%|██████████████████████████████████████████████████████████████████████████████████| 10/10 [00:20<00:00, 2.01s/it]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
Prompt executed in 23.58 seconds
got prompt
2
2
model_type EPS
adm 0
making attention of type 'vanilla-pytorch' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-pytorch' with 512 in_channels
missing {'cond_stage_model.text_projection', 'cond_stage_model.logit_scale'}
left over keys: dict_keys(['cond_stage_model.transformer.text_model.embeddings.position_ids'])
loading new
[AnimateDiffEvo] - INFO - Regular AnimateDiff activated - latents passed in (16) less or equal to context_length None.
[AnimateDiffEvo] - INFO - Injecting motion module animatediffMotion_v15.ckpt version v1.
loading new
loading in lowvram mode 1907.7473125457764
0%| | 0/10 [00:01<?, ?it/s]
[AnimateDiffEvo] - INFO - Ejecting motion module animatediffMotion_v15.ckpt version v1.
[AnimateDiffEvo] - INFO - Cleaning motion module from unet.
!!! Exception during processing !!!
Traceback (most recent call last):
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1236, in sample
return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1206, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 161, in animatediff_sample
return wrap_function_to_inject_xformers_bug_info(orig_comfy_sample)(model, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\model_utils.py", line 144, in wrapped_function
return function_to_wrap(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\sample.py", line 93, in sample
samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 741, in sample
samples = getattr(k_diffusion_sampling, "sample_{}".format(self.sampler))(self.model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\sampling.py", line 137, in sample_euler
denoised = model(x, sigma_hat * s_in, **extra_args)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 322, in forward
out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, cond_concat=cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 125, in forward
eps = self.get_eps(input * c_in, self.sigma_to_t(sigma), **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\k_diffusion\external.py", line 151, in get_eps
return self.inner_model.apply_model(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 310, in apply_model
out = sampling_function(self.inner_model.apply_model, x, timestep, uncond, cond, cond_scale, cond_concat, model_options=model_options, seed=seed)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 527, in sliding_sampling_function
cond, uncond = calc_cond_uncond_batch(model_function, cond, uncond, x, timestep, max_total_area, cond_concat, model_options)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 427, in calc_cond_uncond_batch
output = model_function(input_x, timestep_, **c).chunk(batch_chunks)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\model_base.py", line 63, in apply_model
return self.diffusion_model(xc, t, context=context, y=c_adm, control=control, transformer_options=transformer_options).float()
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\comfy\ldm\modules\diffusionmodules\openaimodel.py", line 653, in forward
h = forward_timestep_embed(module, h, emb, context, transformer_options, output_shape)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\sampling.py", line 71, in forward_timestep_embed
x = layer(x, context)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 398, in forward
return self.temporal_transformer(input_tensor, encoder_hidden_states, attention_mask)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 165, in new_forward
output = old_forward(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-AnimateDiff-Evolved\animatediff\motion_module.py", line 461, in forward
hidden_states = self.norm(hidden_states)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 160, in new_forward
args, kwargs = module._hf_hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 104, in pre_forward
args, kwargs = hook.pre_forward(module, *args, **kwargs)
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\hooks.py", line 286, in pre_forward
set_module_tensor_to_device(
File "D:\apps\AI\COMFYUI2\ComfyUI_windows_portable\python_embeded\lib\site-packages\accelerate\utils\modeling.py", line 298, in set_module_tensor_to_device
new_value = value.to(device)
NotImplementedError: Cannot copy out of meta tensor; no data!
Prompt executed in 79.64 secondsResult 1:
Result 2:
Result 3: Error.
Kosinkadink
commented
1 year ago Btw, are result 1 and 2 in this test using the same seed?
Kosinkadink
commented
1 year ago wait, that was a dumb question, i guess since the models didnt change here that had to be a different seed
I've been trying to generate videos but after 2 successful generations I get the following error in the console:
This is the first time I've tried AnimateDiff so I'm not sure if it's something I did wrong on my end or if it's greater than that. I'm running this on Windows 11 with a GTX 1660 6GB, 16GB RAM and a Ryzen 5600X with the latest ComfyUI. I've also never opened a GitHub issue before so if there's anything you need please do tell me.
Here's the workflow I'm using (i have 256x256 latent image size so that i could quickly generate while i was trying to figure out what was going on):