- include detailed runtime instructions for Windows and macOS
- include instructions for agreeing to use Hugging Face repositories
- always create directories by default when run interactively
- model downloader now supports multiple folder names for known models
- improve logging in sd.py
* fix attention OOM in xformers
* allow passing attention mask in flux attention
* allow an attn_mask in flux
* attn masks can be done using replace patches instead of a separate dict
* fix return types
* fix return order
* enumerate
* patch the right keys
* arg names
* fix a silly bug
* fix xformers masks
* replace match with if, elif, else
* mask with image_ref_size
* remove unused import
* remove unused import 2
* fix pytorch/xformers attention
This corrects a weird inconsistency with skip_reshape.
It also allows masks of various shapes to be passed, which will be
automtically expanded (in a memory-efficient way) to a size that is
compatible with xformers or pytorch sdpa respectively.
* fix mask shapes
Now the only symptom of code messing up and keeping references to a model
object when it should not will be endless prints in the log instead of the
next workflow crashing ComfyUI.
* Added hook_patches to ModelPatcher for weights (model)
* Initial changes to calc_cond_batch to eventually support hook_patches
* Added current_patcher property to BaseModel
* Consolidated add_hook_patches_as_diffs into add_hook_patches func, fixed fp8 support for model-as-lora feature
* Added call to initialize_timesteps on hooks in process_conds func, and added call prepare current keyframe on hooks in calc_cond_batch
* Added default_conds support in calc_cond_batch func
* Added initial set of hook-related nodes, added code to register hooks for loras/model-as-loras, small renaming/refactoring
* Made CLIP work with hook patches
* Added initial hook scheduling nodes, small renaming/refactoring
* Fixed MaxSpeed and default conds implementations
* Added support for adding weight hooks that aren't registered on the ModelPatcher at sampling time
* Made Set Clip Hooks node work with hooks from Create Hook nodes, began work on better Create Hook Model As LoRA node
* Initial work on adding 'model_as_lora' lora type to calculate_weight
* Continued work on simpler Create Hook Model As LoRA node, started to implement ModelPatcher callbacks, attachments, and additional_models
* Fix incorrect ref to create_hook_patches_clone after moving function
* Added injections support to ModelPatcher + necessary bookkeeping, added additional_models support in ModelPatcher, conds, and hooks
* Added wrappers to ModelPatcher to facilitate standardized function wrapping
* Started scaffolding for other hook types, refactored get_hooks_from_cond to organize hooks by type
* Fix skip_until_exit logic bug breaking injection after first run of model
* Updated clone_has_same_weights function to account for new ModelPatcher properties, improved AutoPatcherEjector usage in partially_load
* Added WrapperExecutor for non-classbound functions, added calc_cond_batch wrappers
* Refactored callbacks+wrappers to allow storing lists by id
* Added forward_timestep_embed_patch type, added helper functions on ModelPatcher for emb_patch and forward_timestep_embed_patch, added helper functions for removing callbacks/wrappers/additional_models by key, added custom_should_register prop to hooks
* Added get_attachment func on ModelPatcher
* Implement basic MemoryCounter system for determing with cached weights due to hooks should be offloaded in hooks_backup
* Modified ControlNet/T2IAdapter get_control function to receive transformer_options as additional parameter, made the model_options stored in extra_args in inner_sample be a clone of the original model_options instead of same ref
* Added create_model_options_clone func, modified type annotations to use __future__ so that I can use the better type annotations
* Refactored WrapperExecutor code to remove need for WrapperClassExecutor (now gone), added sampler.sample wrapper (pending review, will likely keep but will see what hacks this could currently let me get rid of in ACN/ADE)
* Added Combine versions of Cond/Cond Pair Set Props nodes, renamed Pair Cond to Cond Pair, fixed default conds never applying hooks (due to hooks key typo)
* Renamed Create Hook Model As LoRA nodes to make the test node the main one (more changes pending)
* Added uuid to conds in CFGGuider and uuids to transformer_options to allow uniquely identifying conds in batches during sampling
* Fixed models not being unloaded properly due to current_patcher reference; the current ComfyUI model cleanup code requires that nothing else has a reference to the ModelPatcher instances
* Fixed default conds not respecting hook keyframes, made keyframes not reset cache when strength is unchanged, fixed Cond Set Default Combine throwing error, fixed model-as-lora throwing error during calculate_weight after a recent ComfyUI update, small refactoring/scaffolding changes for hooks
* Changed CreateHookModelAsLoraTest to be the new CreateHookModelAsLora, rename old ones as 'direct' and will be removed prior to merge
* Added initial support within CLIP Text Encode (Prompt) node for scheduling weight hook CLIP strength via clip_start_percent/clip_end_percent on conds, added schedule_clip toggle to Set CLIP Hooks node, small cleanup/fixes
* Fix range check in get_hooks_for_clip_schedule so that proper keyframes get assigned to corresponding ranges
* Optimized CLIP hook scheduling to treat same strength as same keyframe
* Less fragile memory management.
* Make encode_from_tokens_scheduled call cleaner, rollback change in model_patcher.py for hook_patches_backup dict
* Fix issue.
* Remove useless function.
* Prevent and detect some types of memory leaks.
* Run garbage collector when switching workflow if needed.
* Moved WrappersMP/CallbacksMP/WrapperExecutor to patcher_extension.py
* Refactored code to store wrappers and callbacks in transformer_options, added apply_model and diffusion_model.forward wrappers
* Fix issue.
* Refactored hooks in calc_cond_batch to be part of get_area_and_mult tuple, added extra_hooks to ControlBase to allow custom controlnets w/ hooks, small cleanup and renaming
* Fixed inconsistency of results when schedule_clip is set to False, small renaming/typo fixing, added initial support for ControlNet extra_hooks to work in tandem with normal cond hooks, initial work on calc_cond_batch merging all subdicts in returned transformer_options
* Modified callbacks and wrappers so that unregistered types can be used, allowing custom_nodes to have their own unique callbacks/wrappers if desired
* Updated different hook types to reflect actual progress of implementation, initial scaffolding for working WrapperHook functionality
* Fixed existing weight hook_patches (pre-registered) not working properly for CLIP
* Removed Register/Direct hook nodes since they were present only for testing, removed diff-related weight hook calculation as improved_memory removes unload_model_clones and using sample time registered hooks is less hacky
* Added clip scheduling support to all other native ComfyUI text encoding nodes (sdxl, flux, hunyuan, sd3)
* Made WrapperHook functional, added another wrapper/callback getter, added ON_DETACH callback to ModelPatcher
* Made opt_hooks append by default instead of replace, renamed comfy.hooks set functions to be more accurate
* Added apply_to_conds to Set CLIP Hooks, modified relevant code to allow text encoding to automatically apply hooks to output conds when apply_to_conds is set to True
* Fix cached_hook_patches not respecting target_device/memory_counter results
* Fixed issue with setting weights from hooks instead of copying them, added additional memory_counter check when caching hook patches
* Remove unnecessary torch.no_grad calls for hook patches
* Increased MemoryCounter minimum memory to leave free by *2 until a better way to get inference memory estimate of currently loaded models exists
* For encode_from_tokens_scheduled, allow start_percent and end_percent in add_dict to limit which scheduled conds get encoded for optimization purposes
* Removed a .to call on results of calculate_weight in patch_hook_weight_to_device that was screwing up the intermediate results for fp8 prior to being passed into stochastic_rounding call
* Made encode_from_tokens_scheduled work when no hooks are set on patcher
* Small cleanup of comments
* Turn off hook patch caching when only 1 hook present in sampling, replace some current_hook = None with calls to self.patch_hooks(None) instead to avoid a potential edge case
* On Cond/Cond Pair nodes, removed opt_ prefix from optional inputs
* Allow both FLOATS and FLOAT for floats_strength input
* Revert change, does not work
* Made patch_hook_weight_to_device respect set_func and convert_func
* Make discard_model_sampling True by default
* Add changes manually from 'master' so merge conflict resolution goes more smoothly
* Cleaned up text encode nodes with just a single clip.encode_from_tokens_scheduled call
* Make sure encode_from_tokens_scheduled will respect use_clip_schedule on clip
* Made nodes in nodes_hooks be marked as experimental (beta)
* Add get_nested_additional_models for cases where additional_models could have their own additional_models, and add robustness for circular additional_models references
* Made finalize_default_conds area math consistent with other sampling code
* Changed 'opt_hooks' input of Cond/Cond Pair Set Default Combine nodes to 'hooks'
* Remove a couple old TODO's and a no longer necessary workaround
* Less fragile memory management.
* Fix issue.
* Remove useless function.
* Prevent and detect some types of memory leaks.
* Run garbage collector when switching workflow if needed.
* Fix issue.
Add a way to reshape lora weights.
Allow weight patches to all weight not just .weight and .bias
Add a way for a lora to set a weight to a specific value.
- use their logger when running interactively
- move the extra nodes files to where this fork expects them
- add the mochi checkpoints to known models
- add a mochi workflow test
- Adding the output paths now correctly registers a relative path,
i.e., outputs/loras and models/lorals will now be searched on all
your base paths
- Adding absolute paths with models/ works better
- All the base paths and directories are queried better
To use:
"Load CLIP" node with t5xxl + type mochi
"Load Diffusion Model" node with the mochi dit file.
"Load VAE" with the mochi vae file.
EmptyMochiLatentVideo node for the latent.
euler + linear_quadratic in the KSampler node.
- Validation errors that occur early in the lifecycle of prompt
execution now get propagated to their callers in the
EmbeddedComfyClient. This includes error messages about missing node
classes.
- The execution context now includes the node_id and the prompt_id
- Latent previews are now sent with a node_id. This is not backwards
compatible with old frontends.
- Dependency execution errors are now modeled correctly.
- Distributed progress encodes image previews with node and prompt IDs.
- Typing for models
- The frontend was updated to use node IDs with previews
- Improvements to torch.compile experiments
- Some controlnet_aux nodes were upstreamed
- Experimental support for sage attention on Linux
- Diffusers loader now supports model indices
- Transformers model management now aligns with updates to ComfyUI
- Flux layers correctly use unbind
- Add float8 support for model loading in more places
- Experimental quantization approaches from Quanto and torchao
- Model upscaling interacts with memory management better
This update also disables ROCm testing because it isn't reliable enough
on consumer hardware. ROCm is not really supported by the 7600.
* Override user directory.
* Use overridden user directory.
* Remove prints.
* Remove references to global user_files.
* Remove unused replace_folder function.
* Remove newline.
* Remove global during get_user_directory.
* Add validation.
* Update sampling.py
* Update samplers.py
* my bad
* "fix" the sampler
* Update samplers.py
* i named it wrong
* minor sampling improvements
mainly using a dynamic rho value (hey this sounds a lot like smea!!!)
* revert rho change
rho? r? its just 1/2
This is a format with keys like:
text_encoders.clip_l.transformer.text_model.encoder.layers.9.self_attn.v_proj.lora_up.weight
Instead of waiting for me to add support for specific lora formats you can
convert your text encoder loras to this format instead.
If you want to see an example save a text encoder lora with the SaveLora
node with the commit right after this one.
* Add route for getting output logs
* Include ComfyUI version
* Move to own function
* Changed to memory logger
* Unify logger setup logic
* Fix get version git fallback
---------
Co-authored-by: pythongosssss <125205205+pythongosssss@users.noreply.github.com>
This code automatically forces upcasting attention for MacOS versions 14.5 and 14.6. My computer returns the string "14.6.1" for `platform.mac_ver()[0]`, so this generalizes the comparison to catch more versions.
I am running MacOS Sonoma 14.6.1 (latest version) and was seeing black image generation on previously functional workflows after recent software updates. This PR solved the issue for me.
See comfyanonymous/ComfyUI#3521
When generating images with fp8_e4_m3 Flux and batch size >1, using --fast, ComfyUI throws a "view size is not compatible with input tensor's size and stride" error pointing at the first of these two calls to view.
As reshape is semantically equivalent to view except for working on a broader set of inputs, there should be no downside to changing this. The only difference is that it clones the underlying data in cases where .view would error out. I have confirmed that the output still looks as expected, but cannot confirm that no mutable use is made of the tensors anywhere.
Note that --fast is only marginally faster than the default.
* Add Flux model support for InstantX style controlnet residuals
* Refactor Flux controlnet residual step to a separate method
* Rollback minor change
* New format for applying controlnet residuals: input->double_blocks, output->single_blocks
* Adjust XLabs Flux controlnet to fit new syntax of applying Flux controlnet residuals
* Remove unnecessary import and minor style change
Optimizations that might break things/lower quality will be put behind
this flag first and might be enabled by default in the future.
Currently the only optimization is float8_e4m3fn matrix multiplication on
4000/ADA series Nvidia cards or later. If you have one of these cards you
will see a speed boost when using fp8_e4m3fn flux for example.
--reserve-vram 1.0 for example will make ComfyUI try to keep 1GB vram free.
This can also be useful if workflows are failing because of OOM errors but
in that case please report it if --reserve-vram improves your situation.
* Add Flux model support for InstantX style controlnet residuals
* Refactor Flux controlnet residual step to a separate method
* Rollback minor change
* New format for applying controlnet residuals: input->double_blocks, output->single_blocks
* Adjust XLabs Flux controlnet to fit new syntax of applying Flux controlnet residuals
* Remove unnecessary import and minor style change
* Execution Model Inversion
This PR inverts the execution model -- from recursively calling nodes to
using a topological sort of the nodes. This change allows for
modification of the node graph during execution. This allows for two
major advantages:
1. The implementation of lazy evaluation in nodes. For example, if a
"Mix Images" node has a mix factor of exactly 0.0, the second image
input doesn't even need to be evaluated (and visa-versa if the mix
factor is 1.0).
2. Dynamic expansion of nodes. This allows for the creation of dynamic
"node groups". Specifically, custom nodes can return subgraphs that
replace the original node in the graph. This is an incredibly
powerful concept. Using this functionality, it was easy to
implement:
a. Components (a.k.a. node groups)
b. Flow control (i.e. while loops) via tail recursion
c. All-in-one nodes that replicate the WebUI functionality
d. and more
All of those were able to be implemented entirely via custom nodes,
so those features are *not* a part of this PR. (There are some
front-end changes that should occur before that functionality is
made widely available, particularly around variant sockets.)
The custom nodes associated with this PR can be found at:
https://github.com/BadCafeCode/execution-inversion-demo-comfyui
Note that some of them require that variant socket types ("*") be
enabled.
* Allow `input_info` to be of type `None`
* Handle errors (like OOM) more gracefully
* Add a command-line argument to enable variants
This allows the use of nodes that have sockets of type '*' without
applying a patch to the code.
* Fix an overly aggressive assertion.
This could happen when attempting to evaluate `IS_CHANGED` for a node
during the creation of the cache (in order to create the cache key).
* Fix Pyright warnings
* Add execution model unit tests
* Fix issue with unused literals
Behavior should now match the master branch with regard to undeclared
inputs. Undeclared inputs that are socket connections will be used while
undeclared inputs that are literals will be ignored.
* Make custom VALIDATE_INPUTS skip normal validation
Additionally, if `VALIDATE_INPUTS` takes an argument named `input_types`,
that variable will be a dictionary of the socket type of all incoming
connections. If that argument exists, normal socket type validation will
not occur. This removes the last hurdle for enabling variant types
entirely from custom nodes, so I've removed that command-line option.
I've added appropriate unit tests for these changes.
* Fix example in unit test
This wouldn't have caused any issues in the unit test, but it would have
bugged the UI if someone copy+pasted it into their own node pack.
* Use fstrings instead of '%' formatting syntax
* Use custom exception types.
* Display an error for dependency cycles
Previously, dependency cycles that were created during node expansion
would cause the application to quit (due to an uncaught exception). Now,
we'll throw a proper error to the UI. We also make an attempt to 'blame'
the most relevant node in the UI.
* Add docs on when ExecutionBlocker should be used
* Remove unused functionality
* Rename ExecutionResult.SLEEPING to PENDING
* Remove superfluous function parameter
* Pass None for uneval inputs instead of default
This applies to `VALIDATE_INPUTS`, `check_lazy_status`, and lazy values
in evaluation functions.
* Add a test for mixed node expansion
This test ensures that a node that returns a combination of expanded
subgraphs and literal values functions correctly.
* Raise exception for bad get_node calls.
* Minor refactor of IsChangedCache.get
* Refactor `map_node_over_list` function
* Fix ui output for duplicated nodes
* Add documentation on `check_lazy_status`
* Add file for execution model unit tests
* Clean up Javascript code as per review
* Improve documentation
Converted some comments to docstrings as per review
* Add a new unit test for mixed lazy results
This test validates that when an output list is fed to a lazy node, the
node will properly evaluate previous nodes that are needed by any inputs
to the lazy node.
No code in the execution model has been changed. The test already
passes.
* Allow kwargs in VALIDATE_INPUTS functions
When kwargs are used, validation is skipped for all inputs as if they
had been mentioned explicitly.
* List cached nodes in `execution_cached` message
This was previously just bugged in this PR.
* add support for HunYuanDit ControlNet
* fix hunyuandit controlnet
* fix typo in hunyuandit controlnet
* fix typo in hunyuandit controlnet
* fix code format style
* add control_weight support for HunyuanDit Controlnet
* use control_weights in HunyuanDit Controlnet
* fix typo
The keys are just: model.full.model.key.name.lora_up.weight
It is supported by all comfyui supported models.
Now people can just convert loras to this format instead of having to ask
for me to implement them.
- move workflows to distinct json files
- add the comfy-org workflows for testing
- fix issues where workflows from windows users would not be compatible
with backends running on linux or macos in light of separator
differences. Because this codebase uses get_or_download wherever
checkpoints, models, etc. are used, this is the only place where the
comparison is gracefully handled for downloading. Validation code
will correctly convert backslashes to forward slashes, assuming that
100% of the places they are used and when comparing with a list, they
are intended to be paths and not strict symbols
- huggingface cache is now better used on platforms that support
symlinking and the files you are requesting already exist in the
cache
- absolute imports were changed to relative in the correct places
- StringEnumRequestParameter has a special case in validation
- fix model_management whitespace issue
- fix comfy.ops references
