Long conditioning prompts cause crashes from tensor lengths not matching?

[cchance27]

Updated the name because i realized whats causing this, i'm trying to recreate a workflow and the negative prompt is very long, and it seems that this totally destroys things, it seems like the ksampler is limited to 77 tokens

Workflow: easywf.json
I ran updates for comfy and CMLS

But i'm getting the following error

Loading 1 new model
  0%|                                                                                                                                                                                          | 0/20 [00:00<?, ?it/s]
ERROR:root:!!! Exception during processing !!!
ERROR:root:Traceback (most recent call last):
  File "/Users/user/AI/ComfyUI/execution.py", line 153, in recursive_execute
    output_data, output_ui = get_output_data(obj, input_data_all)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/execution.py", line 83, in get_output_data
    return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/execution.py", line 76, in map_node_over_list
    results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/custom_nodes/ComfyUI-CoreMLSuite/coreml_suite/nodes.py", line 65, in sample
    return super().sample(
           ^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/nodes.py", line 1286, in sample
    return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/nodes.py", line 1256, in common_ksampler
    samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/custom_nodes/ComfyUI-Impact-Pack/modules/impact/sample_error_enhancer.py", line 9, in informative_sample
    return original_sample(*args, **kwargs)  # This code helps interpret error messages that occur within exceptions but does not have any impact on other operations.
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/sample.py", line 100, 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 "/Users/user/AI/ComfyUI/comfy/samplers.py", line 711, in sample
    return sample(self.model, noise, positive, negative, cfg, self.device, sampler, sigmas, self.model_options, latent_image=latent_image, denoise_mask=denoise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 617, in sample
    samples = sampler.sample(model_wrap, sigmas, extra_args, callback, noise, latent_image, denoise_mask, disable_pbar)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 556, in sample
    samples = self.sampler_function(model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar, **self.extra_options)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/k_diffusion/sampling.py", line 655, in sample_dpmpp_3m_sde
    denoised = model(x, sigmas[i] * s_in, **extra_args)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1511, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1520, in _call_impl
    return forward_call(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 277, in forward
    out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, model_options=model_options, seed=seed)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1511, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1520, in _call_impl
    return forward_call(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 267, in forward
    return self.apply_model(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 264, in apply_model
    out = sampling_function(self.inner_model, x, timestep, uncond, cond, cond_scale, model_options=model_options, seed=seed)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 252, in sampling_function
    cond, uncond = calc_cond_uncond_batch(model, cond, uncond, x, timestep, model_options)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/samplers.py", line 230, in calc_cond_uncond_batch
    output = model.apply_model(input_x, timestep_, **c).chunk(batch_chunks)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/comfy/model_base.py", line 73, in apply_model
    model_output = self.diffusion_model(xc, t, context=context, control=control, transformer_options=transformer_options, **extra_conds).float()
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/custom_nodes/ComfyUI-CoreMLSuite/coreml_suite/models.py", line 23, in __call__
    chunked_out = [
                  ^
  File "/Users/user/AI/ComfyUI/custom_nodes/ComfyUI-CoreMLSuite/coreml_suite/models.py", line 25, in <listcomp>
    self.coreml_model(**input_kwargs.coreml_kwargs(self.expected_inputs)),
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/python_coreml_stable_diffusion/coreml_model.py", line 97, in __call__
    self._verify_inputs(**kwargs)
  File "/Users/user/AI/ComfyUI/venv/lib/python3.11/site-packages/python_coreml_stable_diffusion/coreml_model.py", line 90, in _verify_inputs
    raise TypeError(
TypeError: Expected shape (1, 768, 1, 77), got (1, 768, 1, 385) for input: encoder_hidden_states

Prompt executed in 267.22 secondsuser

[aszc-dev]

Yes, there is a token limit for prompts and similarly to sample sizes, it cannot be changed easily. If the prompt is too long, it simply won't work. I think in Apple's implementation of SD pipeline, the input is trunkated if it's too long, and a warning message is displayed in the logs. I was planning to do something similar, but I keep forgetting about it. And in this case, a warning can be easily missed by the user, so I think I'll stick to raising an error, but with more helpful message.

[cchance27]

doesn't this technically exist in standard SD as well, but apps like a1111 use token chunking to batch 75 tokens at a time? not possible to do a similar form of batching on the coreml side?

[cchance27]

a workaround that seems to work though I don't think its how A111 does it, is I can manually split out the longer prompt into multiple embed nodes, and then use conditioning combines on them to end up at a final one.