Merge 7333281698 into 2ff3104f70

comfy/hooks.py

@@ -16,91 +16,132 @@ import comfy.model_management
 import comfy.patcher_extension
 from node_helpers import conditioning_set_values
 
+# #######################################################################################################
+# Hooks explanation
+# -------------------
+# The purpose of hooks is to allow conds to influence sampling without the need for ComfyUI core code to
+# make explicit special cases like it does for ControlNet and GLIGEN.
+#
+# This is necessary for nodes/features that are intended for use with masked or scheduled conds, or those
+# that should run special code when a 'marked' cond is used in sampling.
+# #######################################################################################################
+
 class EnumHookMode(enum.Enum):
+    '''
+    Priority of hook memory optimization vs. speed, mostly related to WeightHooks.
+
+    MinVram: No caching will occur for any operations related to hooks.
+    MaxSpeed: Excess VRAM (and RAM, once VRAM is sufficiently depleted) will be used to cache hook weights when switching hook groups.
+    '''
     MinVram = "minvram"
     MaxSpeed = "maxspeed"
 
 class EnumHookType(enum.Enum):
+    '''
+    Hook types, each of which has different expected behavior.
+    '''
     Weight = "weight"
-    Patch = "patch"
     ObjectPatch = "object_patch"
-    AddModels = "add_models"
-    Callbacks = "callbacks"
-    Wrappers = "wrappers"
-    SetInjections = "add_injections"
+    AdditionalModels = "add_models"
+    TransformerOptions = "transformer_options"
+    Injections = "add_injections"
 
 class EnumWeightTarget(enum.Enum):
     Model = "model"
     Clip = "clip"
 
+class EnumHookScope(enum.Enum):
+    '''
+    Determines if hook should be limited in its influence over sampling.
+
+    AllConditioning: hook will affect all conds used in sampling.
+    HookedOnly: hook will only affect the conds it was attached to.
+    '''
+    AllConditioning = "all_conditioning"
+    HookedOnly = "hooked_only"
+
+
 class _HookRef:
     pass
 
-# NOTE: this is an example of how the should_register function should look
-def default_should_register(hook: 'Hook', model: 'ModelPatcher', model_options: dict, target: EnumWeightTarget, registered: list[Hook]):
+
+def default_should_register(hook: Hook, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+    '''Example for how custom_should_register function can look like.'''
     return True
 
 
+def create_target_dict(target: EnumWeightTarget=None, **kwargs) -> dict[str]:
+    '''Creates base dictionary for use with Hooks' target param.'''
+    d = {}
+    if target is not None:
+        d['target'] = target
+    d.update(kwargs)
+    return d
+
+
 class Hook:
     def __init__(self, hook_type: EnumHookType=None, hook_ref: _HookRef=None, hook_id: str=None,
-                 hook_keyframe: 'HookKeyframeGroup'=None):
+                 hook_keyframe: HookKeyframeGroup=None, hook_scope=EnumHookScope.AllConditioning):
         self.hook_type = hook_type
+        '''Enum identifying the general class of this hook.'''
         self.hook_ref = hook_ref if hook_ref else _HookRef()
+        '''Reference shared between hook clones that have the same value. Should NOT be modified.'''
         self.hook_id = hook_id
+        '''Optional string ID to identify hook; useful if need to consolidate duplicates at registration time.'''
         self.hook_keyframe = hook_keyframe if hook_keyframe else HookKeyframeGroup()
+        '''Keyframe storage that can be referenced to get strength for current sampling step.'''
+        self.hook_scope = hook_scope
+        '''Scope of where this hook should apply in terms of the conds used in sampling run.'''
         self.custom_should_register = default_should_register
-        self.auto_apply_to_nonpositive = False
+        '''Can be overriden with a compatible function to decide if this hook should be registered without the need to override .should_register'''
 
     @property
     def strength(self):
         return self.hook_keyframe.strength
 
-    def initialize_timesteps(self, model: 'BaseModel'):
+    def initialize_timesteps(self, model: BaseModel):
         self.reset()
         self.hook_keyframe.initialize_timesteps(model)
 
     def reset(self):
         self.hook_keyframe.reset()
 
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: Hook = subtype()
+    def clone(self):
+        c: Hook = self.__class__()
         c.hook_type = self.hook_type
         c.hook_ref = self.hook_ref
         c.hook_id = self.hook_id
         c.hook_keyframe = self.hook_keyframe
+        c.hook_scope = self.hook_scope
         c.custom_should_register = self.custom_should_register
-        # TODO: make this do something
-        c.auto_apply_to_nonpositive = self.auto_apply_to_nonpositive
         return c
 
-    def should_register(self, model: 'ModelPatcher', model_options: dict, target: EnumWeightTarget, registered: list[Hook]):
-        return self.custom_should_register(self, model, model_options, target, registered)
+    def should_register(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        return self.custom_should_register(self, model, model_options, target_dict, registered)
 
-    def add_hook_patches(self, model: 'ModelPatcher', model_options: dict, target: EnumWeightTarget, registered: list[Hook]):
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
         raise NotImplementedError("add_hook_patches should be defined for Hook subclasses")
 
-    def on_apply(self, model: 'ModelPatcher', transformer_options: dict[str]):
-        pass
-
-    def on_unapply(self, model: 'ModelPatcher', transformer_options: dict[str]):
-        pass
-
-    def __eq__(self, other: 'Hook'):
+    def __eq__(self, other: Hook):
         return self.__class__ == other.__class__ and self.hook_ref == other.hook_ref
 
     def __hash__(self):
         return hash(self.hook_ref)
 
 class WeightHook(Hook):
+    '''
+    Hook responsible for tracking weights to be applied to some model/clip.
+
+    Note, value of hook_scope is ignored and is treated as HookedOnly.
+    '''
     def __init__(self, strength_model=1.0, strength_clip=1.0):
-        super().__init__(hook_type=EnumHookType.Weight)
+        super().__init__(hook_type=EnumHookType.Weight, hook_scope=EnumHookScope.HookedOnly)
         self.weights: dict = None
         self.weights_clip: dict = None
         self.need_weight_init = True
         self._strength_model = strength_model
         self._strength_clip = strength_clip
+        self.hook_scope = EnumHookScope.HookedOnly # this value does not matter for WeightHooks, just for docs
 
     @property
     def strength_model(self):
@@ -110,36 +151,36 @@ class WeightHook(Hook):
     def strength_clip(self):
         return self._strength_clip * self.strength
 
-    def add_hook_patches(self, model: 'ModelPatcher', model_options: dict, target: EnumWeightTarget, registered: list[Hook]):
-        if not self.should_register(model, model_options, target, registered):
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        if not self.should_register(model, model_options, target_dict, registered):
             return False
         weights = None
-        if target == EnumWeightTarget.Model:
-            strength = self._strength_model
-        else:
+
+        target = target_dict.get('target', None)
+        if target == EnumWeightTarget.Clip:
             strength = self._strength_clip
+        else:
+            strength = self._strength_model
 
         if self.need_weight_init:
             key_map = {}
-            if target == EnumWeightTarget.Model:
-                key_map = comfy.lora.model_lora_keys_unet(model.model, key_map)
-            else:
+            if target == EnumWeightTarget.Clip:
                 key_map = comfy.lora.model_lora_keys_clip(model.model, key_map)
+            else:
+                key_map = comfy.lora.model_lora_keys_unet(model.model, key_map)
             weights = comfy.lora.load_lora(self.weights, key_map, log_missing=False)
         else:
-            if target == EnumWeightTarget.Model:
-                weights = self.weights
-            else:
+            if target == EnumWeightTarget.Clip:
                 weights = self.weights_clip
+            else:
+                weights = self.weights
         model.add_hook_patches(hook=self, patches=weights, strength_patch=strength)
-        registered.append(self)
+        registered.add(self)
         return True
         # TODO: add logs about any keys that were not applied
 
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: WeightHook = super().clone(subtype)
+    def clone(self):
+        c: WeightHook = super().clone()
         c.weights = self.weights
         c.weights_clip = self.weights_clip
         c.need_weight_init = self.need_weight_init
@@ -147,127 +188,158 @@ class WeightHook(Hook):
         c._strength_clip = self._strength_clip
         return c
 
-class PatchHook(Hook):
-    def __init__(self):
-        super().__init__(hook_type=EnumHookType.Patch)
-        self.patches: dict = None
-
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: PatchHook = super().clone(subtype)
-        c.patches = self.patches
-        return c
-    # TODO: add functionality
-
 class ObjectPatchHook(Hook):
-    def __init__(self):
+    def __init__(self, object_patches: dict[str]=None,
+                 hook_scope=EnumHookScope.AllConditioning):
         super().__init__(hook_type=EnumHookType.ObjectPatch)
-        self.object_patches: dict = None
+        self.object_patches = object_patches
+        self.hook_scope = hook_scope
 
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: ObjectPatchHook = super().clone(subtype)
+    def clone(self):
+        c: ObjectPatchHook = super().clone()
         c.object_patches = self.object_patches
         return c
-    # TODO: add functionality
 
-class AddModelsHook(Hook):
-    def __init__(self, key: str=None, models: list['ModelPatcher']=None):
-        super().__init__(hook_type=EnumHookType.AddModels)
-        self.key = key
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        raise NotImplementedError("ObjectPatchHook is not supported yet in ComfyUI.")
+
+class AdditionalModelsHook(Hook):
+    '''
+    Hook responsible for telling model management any additional models that should be loaded.
+
+    Note, value of hook_scope is ignored and is treated as AllConditioning.
+    '''
+    def __init__(self, models: list[ModelPatcher]=None, key: str=None):
+        super().__init__(hook_type=EnumHookType.AdditionalModels)
         self.models = models
-        self.append_when_same = True
-
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: AddModelsHook = super().clone(subtype)
-        c.key = self.key
-        c.models = self.models.copy() if self.models else self.models
-        c.append_when_same = self.append_when_same
-        return c
-    # TODO: add functionality
-
-class CallbackHook(Hook):
-    def __init__(self, key: str=None, callback: Callable=None):
-        super().__init__(hook_type=EnumHookType.Callbacks)
         self.key = key
-        self.callback = callback
 
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: CallbackHook = super().clone(subtype)
+    def clone(self):
+        c: AdditionalModelsHook = super().clone()
+        c.models = self.models.copy() if self.models else self.models
         c.key = self.key
-        c.callback = self.callback
-        return c
-    # TODO: add functionality
-
-class WrapperHook(Hook):
-    def __init__(self, wrappers_dict: dict[str, dict[str, dict[str, list[Callable]]]]=None):
-        super().__init__(hook_type=EnumHookType.Wrappers)
-        self.wrappers_dict = wrappers_dict
-
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: WrapperHook = super().clone(subtype)
-        c.wrappers_dict = self.wrappers_dict
         return c
 
-    def add_hook_patches(self, model: 'ModelPatcher', model_options: dict, target: EnumWeightTarget, registered: list[Hook]):
-        if not self.should_register(model, model_options, target, registered):
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        if not self.should_register(model, model_options, target_dict, registered):
             return False
-        add_model_options = {"transformer_options": self.wrappers_dict}
-        comfy.patcher_extension.merge_nested_dicts(model_options, add_model_options, copy_dict1=False)
-        registered.append(self)
+        registered.add(self)
         return True
 
-class SetInjectionsHook(Hook):
-    def __init__(self, key: str=None, injections: list['PatcherInjection']=None):
-        super().__init__(hook_type=EnumHookType.SetInjections)
+class TransformerOptionsHook(Hook):
+    '''
+    Hook responsible for adding wrappers, callbacks, patches, or anything else related to transformer_options.
+    '''
+    def __init__(self, transformers_dict: dict[str, dict[str, dict[str, list[Callable]]]]=None,
+                 hook_scope=EnumHookScope.AllConditioning):
+        super().__init__(hook_type=EnumHookType.TransformerOptions)
+        self.transformers_dict = transformers_dict
+        self.hook_scope = hook_scope
+        self._skip_adding = False
+        '''Internal value used to avoid double load of transformer_options when hook_scope is AllConditioning.'''
+
+    def clone(self):
+        c: TransformerOptionsHook = super().clone()
+        c.transformers_dict = self.transformers_dict
+        c._skip_adding = self._skip_adding
+        return c
+
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        if not self.should_register(model, model_options, target_dict, registered):
+            return False
+        # NOTE: to_load_options will be used to manually load patches/wrappers/callbacks from hooks
+        self._skip_adding = False
+        if self.hook_scope == EnumHookScope.AllConditioning:
+            add_model_options = {"transformer_options": self.transformers_dict,
+                                 "to_load_options": self.transformers_dict}
+            # skip_adding if included in AllConditioning to avoid double loading
+            self._skip_adding = True
+        else:
+            add_model_options = {"to_load_options": self.transformers_dict}
+        registered.add(self)
+        comfy.patcher_extension.merge_nested_dicts(model_options, add_model_options, copy_dict1=False)
+        return True
+
+    def on_apply_hooks(self, model: ModelPatcher, transformer_options: dict[str]):
+        if not self._skip_adding:
+            comfy.patcher_extension.merge_nested_dicts(transformer_options, self.transformers_dict, copy_dict1=False)
+
+WrapperHook = TransformerOptionsHook
+'''Only here for backwards compatibility, WrapperHook is identical to TransformerOptionsHook.'''
+
+class InjectionsHook(Hook):
+    def __init__(self, key: str=None, injections: list[PatcherInjection]=None,
+                 hook_scope=EnumHookScope.AllConditioning):
+        super().__init__(hook_type=EnumHookType.Injections)
         self.key = key
         self.injections = injections
+        self.hook_scope = hook_scope
 
-    def clone(self, subtype: Callable=None):
-        if subtype is None:
-            subtype = type(self)
-        c: SetInjectionsHook = super().clone(subtype)
+    def clone(self):
+        c: InjectionsHook = super().clone()
         c.key = self.key
         c.injections = self.injections.copy() if self.injections else self.injections
         return c
 
-    def add_hook_injections(self, model: 'ModelPatcher'):
-        # TODO: add functionality
-        pass
+    def add_hook_patches(self, model: ModelPatcher, model_options: dict, target_dict: dict[str], registered: HookGroup):
+        raise NotImplementedError("InjectionsHook is not supported yet in ComfyUI.")
 
 class HookGroup:
+    '''
+    Stores groups of hooks, and allows them to be queried by type.
+
+    To prevent breaking their functionality, never modify the underlying self.hooks or self._hook_dict vars directly;
+    always use the provided functions on HookGroup.
+    '''
     def __init__(self):
         self.hooks: list[Hook] = []
+        self._hook_dict: dict[EnumHookType, list[Hook]] = {}
+
+    def __len__(self):
+        return len(self.hooks)
 
     def add(self, hook: Hook):
         if hook not in self.hooks:
             self.hooks.append(hook)
+            self._hook_dict.setdefault(hook.hook_type, []).append(hook)
+
+    def remove(self, hook: Hook):
+        if hook in self.hooks:
+            self.hooks.remove(hook)
+            self._hook_dict[hook.hook_type].remove(hook)
+
+    def get_type(self, hook_type: EnumHookType):
+        return self._hook_dict.get(hook_type, [])
 
     def contains(self, hook: Hook):
         return hook in self.hooks
 
+    def is_subset_of(self, other: HookGroup):
+        self_hooks = set(self.hooks)
+        other_hooks = set(other.hooks)
+        return self_hooks.issubset(other_hooks)
+
+    def new_with_common_hooks(self, other: HookGroup):
+        c = HookGroup()
+        for hook in self.hooks:
+            if other.contains(hook):
+                c.add(hook.clone())
+        return c
+
     def clone(self):
         c = HookGroup()
         for hook in self.hooks:
             c.add(hook.clone())
         return c
 
-    def clone_and_combine(self, other: 'HookGroup'):
+    def clone_and_combine(self, other: HookGroup):
         c = self.clone()
         if other is not None:
             for hook in other.hooks:
                 c.add(hook.clone())
         return c
 
-    def set_keyframes_on_hooks(self, hook_kf: 'HookKeyframeGroup'):
+    def set_keyframes_on_hooks(self, hook_kf: HookKeyframeGroup):
         if hook_kf is None:
             hook_kf = HookKeyframeGroup()
         else:
@@ -275,18 +347,11 @@ class HookGroup:
         for hook in self.hooks:
             hook.hook_keyframe = hook_kf
 
-    def get_dict_repr(self):
-        d: dict[EnumHookType, dict[Hook, None]] = {}
-        for hook in self.hooks:
-            with_type = d.setdefault(hook.hook_type, {})
-            with_type[hook] = None
-        return d
-
     def get_hooks_for_clip_schedule(self):
         scheduled_hooks: dict[WeightHook, list[tuple[tuple[float,float], HookKeyframe]]] = {}
-        for hook in self.hooks:
         # only care about WeightHooks, for now
-            if hook.hook_type == EnumHookType.Weight:
+        for hook in self.get_type(EnumHookType.Weight):
+            hook: WeightHook
             hook_schedule = []
             # if no hook keyframes, assign default value
             if len(hook.hook_keyframe.keyframes) == 0:
@@ -336,7 +401,7 @@ class HookGroup:
             hook.reset()
 
     @staticmethod
-    def combine_all_hooks(hooks_list: list['HookGroup'], require_count=0) -> 'HookGroup':
+    def combine_all_hooks(hooks_list: list[HookGroup], require_count=0) -> HookGroup:
         actual: list[HookGroup] = []
         for group in hooks_list:
             if group is not None:
@@ -433,7 +498,7 @@ class HookKeyframeGroup:
         c._set_first_as_current()
         return c
 
-    def initialize_timesteps(self, model: 'BaseModel'):
+    def initialize_timesteps(self, model: BaseModel):
         for keyframe in self.keyframes:
             keyframe.start_t = model.model_sampling.percent_to_sigma(keyframe.start_percent)
 
@@ -522,6 +587,17 @@ def get_sorted_list_via_attr(objects: list, attr: str) -> list:
         sorted_list.extend(object_list)
     return sorted_list
 
+def create_transformer_options_from_hooks(model: ModelPatcher, hooks: HookGroup,  transformer_options: dict[str]=None):
+    # if no hooks or is not a ModelPatcher for sampling, return empty dict
+    if hooks is None or model.is_clip:
+        return {}
+    if transformer_options is None:
+        transformer_options = {}
+    for hook in hooks.get_type(EnumHookType.TransformerOptions):
+        hook: TransformerOptionsHook
+        hook.on_apply_hooks(model, transformer_options)
+    return transformer_options
+
 def create_hook_lora(lora: dict[str, torch.Tensor], strength_model: float, strength_clip: float):
     hook_group = HookGroup()
     hook = WeightHook(strength_model=strength_model, strength_clip=strength_clip)
@@ -548,7 +624,7 @@ def create_hook_model_as_lora(weights_model, weights_clip, strength_model: float
     hook.need_weight_init = False
     return hook_group
 
-def get_patch_weights_from_model(model: 'ModelPatcher', discard_model_sampling=True):
+def get_patch_weights_from_model(model: ModelPatcher, discard_model_sampling=True):
     if model is None:
         return None
     patches_model: dict[str, torch.Tensor] = model.model.state_dict()
@@ -560,7 +636,7 @@ def get_patch_weights_from_model(model: 'ModelPatcher', discard_model_sampling=T
     return patches_model
 
 # NOTE: this function shows how to register weight hooks directly on the ModelPatchers
-def load_hook_lora_for_models(model: 'ModelPatcher', clip: 'CLIP', lora: dict[str, torch.Tensor],
+def load_hook_lora_for_models(model: ModelPatcher, clip: CLIP, lora: dict[str, torch.Tensor],
                               strength_model: float, strength_clip: float):
     key_map = {}
     if model is not None:
@@ -612,24 +688,26 @@ def _combine_hooks_from_values(c_dict: dict[str, HookGroup], values: dict[str, H
     else:
         c_dict[hooks_key] = cache[hooks_tuple]
 
-def conditioning_set_values_with_hooks(conditioning, values={}, append_hooks=True):
+def conditioning_set_values_with_hooks(conditioning, values={}, append_hooks=True,
+                                       cache: dict[tuple[HookGroup, HookGroup], HookGroup]=None):
     c = []
-    hooks_combine_cache: dict[tuple[HookGroup, HookGroup], HookGroup] = {}
+    if cache is None:
+        cache = {}
     for t in conditioning:
         n = [t[0], t[1].copy()]
         for k in values:
             if append_hooks and k == 'hooks':
-                _combine_hooks_from_values(n[1], values, hooks_combine_cache)
+                _combine_hooks_from_values(n[1], values, cache)
             else:
                 n[1][k] = values[k]
         c.append(n)
 
     return c
 
-def set_hooks_for_conditioning(cond, hooks: HookGroup, append_hooks=True):
+def set_hooks_for_conditioning(cond, hooks: HookGroup, append_hooks=True, cache: dict[tuple[HookGroup, HookGroup], HookGroup]=None):
     if hooks is None:
         return cond
-    return conditioning_set_values_with_hooks(cond, {'hooks': hooks}, append_hooks=append_hooks)
+    return conditioning_set_values_with_hooks(cond, {'hooks': hooks}, append_hooks=append_hooks, cache=cache)
 
 def set_timesteps_for_conditioning(cond, timestep_range: tuple[float,float]):
     if timestep_range is None:
@@ -664,9 +742,10 @@ def combine_with_new_conds(conds: list, new_conds: list):
 def set_conds_props(conds: list, strength: float, set_cond_area: str,
                    mask: torch.Tensor=None, hooks: HookGroup=None, timesteps_range: tuple[float,float]=None, append_hooks=True):
     final_conds = []
+    cache = {}
     for c in conds:
         # first, apply lora_hook to conditioning, if provided
-        c = set_hooks_for_conditioning(c, hooks, append_hooks=append_hooks)
+        c = set_hooks_for_conditioning(c, hooks, append_hooks=append_hooks, cache=cache)
         # next, apply mask to conditioning
         c = set_mask_for_conditioning(cond=c, mask=mask, strength=strength, set_cond_area=set_cond_area)
         # apply timesteps, if present
@@ -678,9 +757,10 @@ def set_conds_props(conds: list, strength: float, set_cond_area: str,
 def set_conds_props_and_combine(conds: list, new_conds: list, strength: float=1.0, set_cond_area: str="default",
                                mask: torch.Tensor=None, hooks: HookGroup=None, timesteps_range: tuple[float,float]=None, append_hooks=True):
     combined_conds = []
+    cache = {}
     for c, masked_c in zip(conds, new_conds):
         # first, apply lora_hook to new conditioning, if provided
-        masked_c = set_hooks_for_conditioning(masked_c, hooks, append_hooks=append_hooks)
+        masked_c = set_hooks_for_conditioning(masked_c, hooks, append_hooks=append_hooks, cache=cache)
         # next, apply mask to new conditioning, if provided
         masked_c = set_mask_for_conditioning(cond=masked_c, mask=mask, set_cond_area=set_cond_area, strength=strength)
         # apply timesteps, if present
@@ -692,9 +772,10 @@ def set_conds_props_and_combine(conds: list, new_conds: list, strength: float=1.
 def set_default_conds_and_combine(conds: list, new_conds: list,
                                    hooks: HookGroup=None, timesteps_range: tuple[float,float]=None, append_hooks=True):
     combined_conds = []
+    cache = {}
     for c, new_c in zip(conds, new_conds):
         # first, apply lora_hook to new conditioning, if provided
-        new_c = set_hooks_for_conditioning(new_c, hooks, append_hooks=append_hooks)
+        new_c = set_hooks_for_conditioning(new_c, hooks, append_hooks=append_hooks, cache=cache)
         # next, add default_cond key to cond so that during sampling, it can be identified
         new_c = conditioning_set_values(new_c, {'default': True})
         # apply timesteps, if present

comfy/latent_formats.py

@@ -382,3 +382,7 @@ class HunyuanVideo(LatentFormat):
     ]
 
     latent_rgb_factors_bias = [ 0.0259, -0.0192, -0.0761]
+
+class Cosmos1CV8x8x8(LatentFormat):
+    latent_channels = 16
+    latent_dimensions = 3

comfy/ldm/cosmos/blocks.py

@@ -0,0 +1,804 @@
+# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import math
+from typing import Optional
+import logging
+
+import numpy as np
+import torch
+from einops import rearrange, repeat
+from einops.layers.torch import Rearrange
+from torch import nn
+
+from comfy.ldm.modules.diffusionmodules.mmdit import RMSNorm
+from comfy.ldm.modules.attention import optimized_attention
+
+
+def apply_rotary_pos_emb(
+    t: torch.Tensor,
+    freqs: torch.Tensor,
+) -> torch.Tensor:
+    t_ = t.reshape(*t.shape[:-1], 2, -1).movedim(-2, -1).unsqueeze(-2).float()
+    t_out = freqs[..., 0] * t_[..., 0] + freqs[..., 1] * t_[..., 1]
+    t_out = t_out.movedim(-1, -2).reshape(*t.shape).type_as(t)
+    return t_out
+
+
+def get_normalization(name: str, channels: int, weight_args={}):
+    if name == "I":
+        return nn.Identity()
+    elif name == "R":
+        return RMSNorm(channels, elementwise_affine=True, eps=1e-6, **weight_args)
+    else:
+        raise ValueError(f"Normalization {name} not found")
+
+
+class BaseAttentionOp(nn.Module):
+    def __init__(self):
+        super().__init__()
+
+
+class Attention(nn.Module):
+    """
+    Generalized attention impl.
+
+    Allowing for both self-attention and cross-attention configurations depending on whether a `context_dim` is provided.
+    If `context_dim` is None, self-attention is assumed.
+
+    Parameters:
+        query_dim (int): Dimension of each query vector.
+        context_dim (int, optional): Dimension of each context vector. If None, self-attention is assumed.
+        heads (int, optional): Number of attention heads. Defaults to 8.
+        dim_head (int, optional): Dimension of each head. Defaults to 64.
+        dropout (float, optional): Dropout rate applied to the output of the attention block. Defaults to 0.0.
+        attn_op (BaseAttentionOp, optional): Custom attention operation to be used instead of the default.
+        qkv_bias (bool, optional): If True, adds a learnable bias to query, key, and value projections. Defaults to False.
+        out_bias (bool, optional): If True, adds a learnable bias to the output projection. Defaults to False.
+        qkv_norm (str, optional): A string representing normalization strategies for query, key, and value projections.
+                                  Defaults to "SSI".
+        qkv_norm_mode (str, optional): A string representing normalization mode for query, key, and value projections.
+                                        Defaults to 'per_head'. Only support 'per_head'.
+
+    Examples:
+        >>> attn = Attention(query_dim=128, context_dim=256, heads=4, dim_head=32, dropout=0.1)
+        >>> query = torch.randn(10, 128)  # Batch size of 10
+        >>> context = torch.randn(10, 256)  # Batch size of 10
+        >>> output = attn(query, context)  # Perform the attention operation
+
+    Note:
+        https://github.com/MatthieuTPHR/diffusers/blob/d80b531ff8060ec1ea982b65a1b8df70f73aa67c/src/diffusers/models/attention.py#L223
+    """
+
+    def __init__(
+        self,
+        query_dim: int,
+        context_dim=None,
+        heads=8,
+        dim_head=64,
+        dropout=0.0,
+        attn_op: Optional[BaseAttentionOp] = None,
+        qkv_bias: bool = False,
+        out_bias: bool = False,
+        qkv_norm: str = "SSI",
+        qkv_norm_mode: str = "per_head",
+        backend: str = "transformer_engine",
+        qkv_format: str = "bshd",
+        weight_args={},
+        operations=None,
+    ) -> None:
+        super().__init__()
+
+        self.is_selfattn = context_dim is None  # self attention
+
+        inner_dim = dim_head * heads
+        context_dim = query_dim if context_dim is None else context_dim
+
+        self.heads = heads
+        self.dim_head = dim_head
+        self.qkv_norm_mode = qkv_norm_mode
+        self.qkv_format = qkv_format
+
+        if self.qkv_norm_mode == "per_head":
+            norm_dim = dim_head
+        else:
+            raise ValueError(f"Normalization mode {self.qkv_norm_mode} not found, only support 'per_head'")
+
+        self.backend = backend
+
+        self.to_q = nn.Sequential(
+            operations.Linear(query_dim, inner_dim, bias=qkv_bias, **weight_args),
+            get_normalization(qkv_norm[0], norm_dim),
+        )
+        self.to_k = nn.Sequential(
+            operations.Linear(context_dim, inner_dim, bias=qkv_bias, **weight_args),
+            get_normalization(qkv_norm[1], norm_dim),
+        )
+        self.to_v = nn.Sequential(
+            operations.Linear(context_dim, inner_dim, bias=qkv_bias, **weight_args),
+            get_normalization(qkv_norm[2], norm_dim),
+        )
+
+        self.to_out = nn.Sequential(
+            operations.Linear(inner_dim, query_dim, bias=out_bias, **weight_args),
+            nn.Dropout(dropout),
+        )
+
+    def cal_qkv(
+        self, x, context=None, mask=None, rope_emb=None, **kwargs
+    ) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        del kwargs
+
+
+        """
+        self.to_q, self.to_k, self.to_v are nn.Sequential with projection + normalization layers.
+        Before 07/24/2024, these modules normalize across all heads.
+        After 07/24/2024, to support tensor parallelism and follow the common practice in the community,
+        we support to normalize per head.
+        To keep the checkpoint copatibility with the previous code,
+        we keep the nn.Sequential but call the projection and the normalization layers separately.
+        We use a flag `self.qkv_norm_mode` to control the normalization behavior.
+        The default value of `self.qkv_norm_mode` is "per_head", which means we normalize per head.
+        """
+        if self.qkv_norm_mode == "per_head":
+            q = self.to_q[0](x)
+            context = x if context is None else context
+            k = self.to_k[0](context)
+            v = self.to_v[0](context)
+            q, k, v = map(
+                lambda t: rearrange(t, "s b (n c) -> b n s c", n=self.heads, c=self.dim_head),
+                (q, k, v),
+            )
+        else:
+            raise ValueError(f"Normalization mode {self.qkv_norm_mode} not found, only support 'per_head'")
+
+        q = self.to_q[1](q)
+        k = self.to_k[1](k)
+        v = self.to_v[1](v)
+        if self.is_selfattn and rope_emb is not None:  # only apply to self-attention!
+            q = apply_rotary_pos_emb(q, rope_emb)
+            k = apply_rotary_pos_emb(k, rope_emb)
+        return q, k, v
+
+    def cal_attn(self, q, k, v, mask=None):
+        out = optimized_attention(q, k, v, self.heads, skip_reshape=True, mask=mask, skip_output_reshape=True)
+        out = rearrange(out, " b n s c -> s b (n c)")
+        return self.to_out(out)
+
+    def forward(
+        self,
+        x,
+        context=None,
+        mask=None,
+        rope_emb=None,
+        **kwargs,
+    ):
+        """
+        Args:
+            x (Tensor): The query tensor of shape [B, Mq, K]
+            context (Optional[Tensor]): The key tensor of shape [B, Mk, K] or use x as context [self attention] if None
+        """
+        q, k, v = self.cal_qkv(x, context, mask, rope_emb=rope_emb, **kwargs)
+        return self.cal_attn(q, k, v, mask)
+
+
+class FeedForward(nn.Module):
+    """
+    Transformer FFN with optional gating
+
+    Parameters:
+        d_model (int): Dimensionality of input features.
+        d_ff (int): Dimensionality of the hidden layer.
+        dropout (float, optional): Dropout rate applied after the activation function. Defaults to 0.1.
+        activation (callable, optional): The activation function applied after the first linear layer.
+                                         Defaults to nn.ReLU().
+        is_gated (bool, optional): If set to True, incorporates gating mechanism to the feed-forward layer.
+                                   Defaults to False.
+        bias (bool, optional): If set to True, adds a bias to the linear layers. Defaults to True.
+
+    Example:
+        >>> ff = FeedForward(d_model=512, d_ff=2048)
+        >>> x = torch.randn(64, 10, 512)  # Example input tensor
+        >>> output = ff(x)
+        >>> print(output.shape)  # Expected shape: (64, 10, 512)
+    """
+
+    def __init__(
+        self,
+        d_model: int,
+        d_ff: int,
+        dropout: float = 0.1,
+        activation=nn.ReLU(),
+        is_gated: bool = False,
+        bias: bool = False,
+        weight_args={},
+        operations=None,
+    ) -> None:
+        super().__init__()
+
+        self.layer1 = operations.Linear(d_model, d_ff, bias=bias, **weight_args)
+        self.layer2 = operations.Linear(d_ff, d_model, bias=bias, **weight_args)
+
+        self.dropout = nn.Dropout(dropout)
+        self.activation = activation
+        self.is_gated = is_gated
+        if is_gated:
+            self.linear_gate = operations.Linear(d_model, d_ff, bias=False, **weight_args)
+
+    def forward(self, x: torch.Tensor):
+        g = self.activation(self.layer1(x))
+        if self.is_gated:
+            x = g * self.linear_gate(x)
+        else:
+            x = g
+        assert self.dropout.p == 0.0, "we skip dropout"
+        return self.layer2(x)
+
+
+class GPT2FeedForward(FeedForward):
+    def __init__(self, d_model: int, d_ff: int, dropout: float = 0.1, bias: bool = False, weight_args={}, operations=None):
+        super().__init__(
+            d_model=d_model,
+            d_ff=d_ff,
+            dropout=dropout,
+            activation=nn.GELU(),
+            is_gated=False,
+            bias=bias,
+            weight_args=weight_args,
+            operations=operations,
+        )
+
+    def forward(self, x: torch.Tensor):
+        assert self.dropout.p == 0.0, "we skip dropout"
+
+        x = self.layer1(x)
+        x = self.activation(x)
+        x = self.layer2(x)
+
+        return x
+
+
+def modulate(x, shift, scale):
+    return x * (1 + scale.unsqueeze(1)) + shift.unsqueeze(1)
+
+
+class Timesteps(nn.Module):
+    def __init__(self, num_channels):
+        super().__init__()
+        self.num_channels = num_channels
+
+    def forward(self, timesteps):
+        half_dim = self.num_channels // 2
+        exponent = -math.log(10000) * torch.arange(half_dim, dtype=torch.float32, device=timesteps.device)
+        exponent = exponent / (half_dim - 0.0)
+
+        emb = torch.exp(exponent)
+        emb = timesteps[:, None].float() * emb[None, :]
+
+        sin_emb = torch.sin(emb)
+        cos_emb = torch.cos(emb)
+        emb = torch.cat([cos_emb, sin_emb], dim=-1)
+
+        return emb
+
+
+class TimestepEmbedding(nn.Module):
+    def __init__(self, in_features: int, out_features: int, use_adaln_lora: bool = False, weight_args={}, operations=None):
+        super().__init__()
+        logging.debug(
+            f"Using AdaLN LoRA Flag:  {use_adaln_lora}. We enable bias if no AdaLN LoRA for backward compatibility."
+        )
+        self.linear_1 = operations.Linear(in_features, out_features, bias=not use_adaln_lora, **weight_args)
+        self.activation = nn.SiLU()
+        self.use_adaln_lora = use_adaln_lora
+        if use_adaln_lora:
+            self.linear_2 = operations.Linear(out_features, 3 * out_features, bias=False, **weight_args)
+        else:
+            self.linear_2 = operations.Linear(out_features, out_features, bias=True, **weight_args)
+
+    def forward(self, sample: torch.Tensor) -> torch.Tensor:
+        emb = self.linear_1(sample)
+        emb = self.activation(emb)
+        emb = self.linear_2(emb)
+
+        if self.use_adaln_lora:
+            adaln_lora_B_3D = emb
+            emb_B_D = sample
+        else:
+            emb_B_D = emb
+            adaln_lora_B_3D = None
+
+        return emb_B_D, adaln_lora_B_3D
+
+
+class FourierFeatures(nn.Module):
+    """
+    Implements a layer that generates Fourier features from input tensors, based on randomly sampled
+    frequencies and phases. This can help in learning high-frequency functions in low-dimensional problems.
+
+    [B] -> [B, D]
+
+    Parameters:
+        num_channels (int): The number of Fourier features to generate.
+        bandwidth (float, optional): The scaling factor for the frequency of the Fourier features. Defaults to 1.
+        normalize (bool, optional): If set to True, the outputs are scaled by sqrt(2), usually to normalize
+                                    the variance of the features. Defaults to False.
+
+    Example:
+        >>> layer = FourierFeatures(num_channels=256, bandwidth=0.5, normalize=True)
+        >>> x = torch.randn(10, 256)  # Example input tensor
+        >>> output = layer(x)
+        >>> print(output.shape)  # Expected shape: (10, 256)
+    """
+
+    def __init__(self, num_channels, bandwidth=1, normalize=False):
+        super().__init__()
+        self.register_buffer("freqs", 2 * np.pi * bandwidth * torch.randn(num_channels), persistent=True)
+        self.register_buffer("phases", 2 * np.pi * torch.rand(num_channels), persistent=True)
+        self.gain = np.sqrt(2) if normalize else 1
+
+    def forward(self, x, gain: float = 1.0):
+        """
+        Apply the Fourier feature transformation to the input tensor.
+
+        Args:
+            x (torch.Tensor): The input tensor.
+            gain (float, optional): An additional gain factor applied during the forward pass. Defaults to 1.
+
+        Returns:
+            torch.Tensor: The transformed tensor, with Fourier features applied.
+        """
+        in_dtype = x.dtype
+        x = x.to(torch.float32).ger(self.freqs.to(torch.float32)).add(self.phases.to(torch.float32))
+        x = x.cos().mul(self.gain * gain).to(in_dtype)
+        return x
+
+
+class PatchEmbed(nn.Module):
+    """
+    PatchEmbed is a module for embedding patches from an input tensor by applying either 3D or 2D convolutional layers,
+    depending on the . This module can process inputs with temporal (video) and spatial (image) dimensions,
+    making it suitable for video and image processing tasks. It supports dividing the input into patches
+    and embedding each patch into a vector of size `out_channels`.
+
+    Parameters:
+    - spatial_patch_size (int): The size of each spatial patch.
+    - temporal_patch_size (int): The size of each temporal patch.
+    - in_channels (int): Number of input channels. Default: 3.
+    - out_channels (int): The dimension of the embedding vector for each patch. Default: 768.
+    - bias (bool): If True, adds a learnable bias to the output of the convolutional layers. Default: True.
+    """
+
+    def __init__(
+        self,
+        spatial_patch_size,
+        temporal_patch_size,
+        in_channels=3,
+        out_channels=768,
+        bias=True,
+        weight_args={},
+        operations=None,
+    ):
+        super().__init__()
+        self.spatial_patch_size = spatial_patch_size
+        self.temporal_patch_size = temporal_patch_size
+
+        self.proj = nn.Sequential(
+            Rearrange(
+                "b c (t r) (h m) (w n) -> b t h w (c r m n)",
+                r=temporal_patch_size,
+                m=spatial_patch_size,
+                n=spatial_patch_size,
+            ),
+            operations.Linear(
+                in_channels * spatial_patch_size * spatial_patch_size * temporal_patch_size, out_channels, bias=bias, **weight_args
+            ),
+        )
+        self.out = nn.Identity()
+
+    def forward(self, x):
+        """
+        Forward pass of the PatchEmbed module.
+
+        Parameters:
+        - x (torch.Tensor): The input tensor of shape (B, C, T, H, W) where
+            B is the batch size,
+            C is the number of channels,
+            T is the temporal dimension,
+            H is the height, and
+            W is the width of the input.
+
+        Returns:
+        - torch.Tensor: The embedded patches as a tensor, with shape b t h w c.
+        """
+        assert x.dim() == 5
+        _, _, T, H, W = x.shape
+        assert H % self.spatial_patch_size == 0 and W % self.spatial_patch_size == 0
+        assert T % self.temporal_patch_size == 0
+        x = self.proj(x)
+        return self.out(x)
+
+
+class FinalLayer(nn.Module):
+    """
+    The final layer of video DiT.
+    """
+
+    def __init__(
+        self,
+        hidden_size,
+        spatial_patch_size,
+        temporal_patch_size,
+        out_channels,
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        weight_args={},
+        operations=None,
+    ):
+        super().__init__()
+        self.norm_final = operations.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6, **weight_args)
+        self.linear = operations.Linear(
+            hidden_size, spatial_patch_size * spatial_patch_size * temporal_patch_size * out_channels, bias=False, **weight_args
+        )
+        self.hidden_size = hidden_size
+        self.n_adaln_chunks = 2
+        self.use_adaln_lora = use_adaln_lora
+        if use_adaln_lora:
+            self.adaLN_modulation = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(hidden_size, adaln_lora_dim, bias=False, **weight_args),
+                operations.Linear(adaln_lora_dim, self.n_adaln_chunks * hidden_size, bias=False, **weight_args),
+            )
+        else:
+            self.adaLN_modulation = nn.Sequential(
+                nn.SiLU(), operations.Linear(hidden_size, self.n_adaln_chunks * hidden_size, bias=False, **weight_args)
+            )
+
+    def forward(
+        self,
+        x_BT_HW_D,
+        emb_B_D,
+        adaln_lora_B_3D: Optional[torch.Tensor] = None,
+    ):
+        if self.use_adaln_lora:
+            assert adaln_lora_B_3D is not None
+            shift_B_D, scale_B_D = (self.adaLN_modulation(emb_B_D) + adaln_lora_B_3D[:, : 2 * self.hidden_size]).chunk(
+                2, dim=1
+            )
+        else:
+            shift_B_D, scale_B_D = self.adaLN_modulation(emb_B_D).chunk(2, dim=1)
+
+        B = emb_B_D.shape[0]
+        T = x_BT_HW_D.shape[0] // B
+        shift_BT_D, scale_BT_D = repeat(shift_B_D, "b d -> (b t) d", t=T), repeat(scale_B_D, "b d -> (b t) d", t=T)
+        x_BT_HW_D = modulate(self.norm_final(x_BT_HW_D), shift_BT_D, scale_BT_D)
+
+        x_BT_HW_D = self.linear(x_BT_HW_D)
+        return x_BT_HW_D
+
+
+class VideoAttn(nn.Module):
+    """
+    Implements video attention with optional cross-attention capabilities.
+
+    This module processes video features while maintaining their spatio-temporal structure. It can perform
+    self-attention within the video features or cross-attention with external context features.
+
+    Parameters:
+        x_dim (int): Dimension of input feature vectors
+        context_dim (Optional[int]): Dimension of context features for cross-attention. None for self-attention
+        num_heads (int): Number of attention heads
+        bias (bool): Whether to include bias in attention projections. Default: False
+        qkv_norm_mode (str): Normalization mode for query/key/value projections. Must be "per_head". Default: "per_head"
+        x_format (str): Format of input tensor. Must be "BTHWD". Default: "BTHWD"
+
+    Input shape:
+        - x: (T, H, W, B, D) video features
+        - context (optional): (M, B, D) context features for cross-attention
+        where:
+            T: temporal dimension
+            H: height
+            W: width
+            B: batch size
+            D: feature dimension
+            M: context sequence length
+    """
+
+    def __init__(
+        self,
+        x_dim: int,
+        context_dim: Optional[int],
+        num_heads: int,
+        bias: bool = False,
+        qkv_norm_mode: str = "per_head",
+        x_format: str = "BTHWD",
+        weight_args={},
+        operations=None,
+    ) -> None:
+        super().__init__()
+        self.x_format = x_format
+
+        self.attn = Attention(
+            x_dim,
+            context_dim,
+            num_heads,
+            x_dim // num_heads,
+            qkv_bias=bias,
+            qkv_norm="RRI",
+            out_bias=bias,
+            qkv_norm_mode=qkv_norm_mode,
+            qkv_format="sbhd",
+            weight_args=weight_args,
+            operations=operations,
+        )
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        context: Optional[torch.Tensor] = None,
+        crossattn_mask: Optional[torch.Tensor] = None,
+        rope_emb_L_1_1_D: Optional[torch.Tensor] = None,
+    ) -> torch.Tensor:
+        """
+        Forward pass for video attention.
+
+        Args:
+            x (Tensor): Input tensor of shape (B, T, H, W, D) or (T, H, W, B, D) representing batches of video data.
+            context (Tensor): Context tensor of shape (B, M, D) or (M, B, D),
+            where M is the sequence length of the context.
+            crossattn_mask (Optional[Tensor]): An optional mask for cross-attention mechanisms.
+            rope_emb_L_1_1_D (Optional[Tensor]):
+            Rotary positional embedding tensor of shape (L, 1, 1, D). L == THW for current video training.
+
+        Returns:
+            Tensor: The output tensor with applied attention, maintaining the input shape.
+        """
+
+        x_T_H_W_B_D = x
+        context_M_B_D = context
+        T, H, W, B, D = x_T_H_W_B_D.shape
+        x_THW_B_D = rearrange(x_T_H_W_B_D, "t h w b d -> (t h w) b d")
+        x_THW_B_D = self.attn(
+            x_THW_B_D,
+            context_M_B_D,
+            crossattn_mask,
+            rope_emb=rope_emb_L_1_1_D,
+        )
+        x_T_H_W_B_D = rearrange(x_THW_B_D, "(t h w) b d -> t h w b d", h=H, w=W)
+        return x_T_H_W_B_D
+
+
+def adaln_norm_state(norm_state, x, scale, shift):
+    normalized = norm_state(x)
+    return normalized * (1 + scale) + shift
+
+
+class DITBuildingBlock(nn.Module):
+    """
+    A building block for the DiT (Diffusion Transformer) architecture that supports different types of
+    attention and MLP operations with adaptive layer normalization.
+
+    Parameters:
+        block_type (str): Type of block - one of:
+            - "cross_attn"/"ca": Cross-attention
+            - "full_attn"/"fa": Full self-attention
+            - "mlp"/"ff": MLP/feedforward block
+        x_dim (int): Dimension of input features
+        context_dim (Optional[int]): Dimension of context features for cross-attention
+        num_heads (int): Number of attention heads
+        mlp_ratio (float): MLP hidden dimension multiplier. Default: 4.0
+        bias (bool): Whether to use bias in layers. Default: False
+        mlp_dropout (float): Dropout rate for MLP. Default: 0.0
+        qkv_norm_mode (str): QKV normalization mode. Default: "per_head"
+        x_format (str): Input tensor format. Default: "BTHWD"
+        use_adaln_lora (bool): Whether to use AdaLN-LoRA. Default: False
+        adaln_lora_dim (int): Dimension for AdaLN-LoRA. Default: 256
+    """
+
+    def __init__(
+        self,
+        block_type: str,
+        x_dim: int,
+        context_dim: Optional[int],
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        bias: bool = False,
+        mlp_dropout: float = 0.0,
+        qkv_norm_mode: str = "per_head",
+        x_format: str = "BTHWD",
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        weight_args={},
+        operations=None
+    ) -> None:
+        block_type = block_type.lower()
+
+        super().__init__()
+        self.x_format = x_format
+        if block_type in ["cross_attn", "ca"]:
+            self.block = VideoAttn(
+                x_dim,
+                context_dim,
+                num_heads,
+                bias=bias,
+                qkv_norm_mode=qkv_norm_mode,
+                x_format=self.x_format,
+                weight_args=weight_args,
+                operations=operations,
+            )
+        elif block_type in ["full_attn", "fa"]:
+            self.block = VideoAttn(
+                x_dim, None, num_heads, bias=bias, qkv_norm_mode=qkv_norm_mode, x_format=self.x_format, weight_args=weight_args, operations=operations
+            )
+        elif block_type in ["mlp", "ff"]:
+            self.block = GPT2FeedForward(x_dim, int(x_dim * mlp_ratio), dropout=mlp_dropout, bias=bias, weight_args=weight_args, operations=operations)
+        else:
+            raise ValueError(f"Unknown block type: {block_type}")
+
+        self.block_type = block_type
+        self.use_adaln_lora = use_adaln_lora
+
+        self.norm_state = nn.LayerNorm(x_dim, elementwise_affine=False, eps=1e-6)
+        self.n_adaln_chunks = 3
+        if use_adaln_lora:
+            self.adaLN_modulation = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(x_dim, adaln_lora_dim, bias=False, **weight_args),
+                operations.Linear(adaln_lora_dim, self.n_adaln_chunks * x_dim, bias=False, **weight_args),
+            )
+        else:
+            self.adaLN_modulation = nn.Sequential(nn.SiLU(), operations.Linear(x_dim, self.n_adaln_chunks * x_dim, bias=False, **weight_args))
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        emb_B_D: torch.Tensor,
+        crossattn_emb: torch.Tensor,
+        crossattn_mask: Optional[torch.Tensor] = None,
+        rope_emb_L_1_1_D: Optional[torch.Tensor] = None,
+        adaln_lora_B_3D: Optional[torch.Tensor] = None,
+    ) -> torch.Tensor:
+        """
+        Forward pass for dynamically configured blocks with adaptive normalization.
+
+        Args:
+            x (Tensor): Input tensor of shape (B, T, H, W, D) or (T, H, W, B, D).
+            emb_B_D (Tensor): Embedding tensor for adaptive layer normalization modulation.
+            crossattn_emb (Tensor): Tensor for cross-attention blocks.
+            crossattn_mask (Optional[Tensor]): Optional mask for cross-attention.
+            rope_emb_L_1_1_D (Optional[Tensor]):
+            Rotary positional embedding tensor of shape (L, 1, 1, D). L == THW for current video training.
+
+        Returns:
+            Tensor: The output tensor after processing through the configured block and adaptive normalization.
+        """
+        if self.use_adaln_lora:
+            shift_B_D, scale_B_D, gate_B_D = (self.adaLN_modulation(emb_B_D) + adaln_lora_B_3D).chunk(
+                self.n_adaln_chunks, dim=1
+            )
+        else:
+            shift_B_D, scale_B_D, gate_B_D = self.adaLN_modulation(emb_B_D).chunk(self.n_adaln_chunks, dim=1)
+
+        shift_1_1_1_B_D, scale_1_1_1_B_D, gate_1_1_1_B_D = (
+            shift_B_D.unsqueeze(0).unsqueeze(0).unsqueeze(0),
+            scale_B_D.unsqueeze(0).unsqueeze(0).unsqueeze(0),
+            gate_B_D.unsqueeze(0).unsqueeze(0).unsqueeze(0),
+        )
+
+        if self.block_type in ["mlp", "ff"]:
+            x = x + gate_1_1_1_B_D * self.block(
+                adaln_norm_state(self.norm_state, x, scale_1_1_1_B_D, shift_1_1_1_B_D),
+            )
+        elif self.block_type in ["full_attn", "fa"]:
+            x = x + gate_1_1_1_B_D * self.block(
+                adaln_norm_state(self.norm_state, x, scale_1_1_1_B_D, shift_1_1_1_B_D),
+                context=None,
+                rope_emb_L_1_1_D=rope_emb_L_1_1_D,
+            )
+        elif self.block_type in ["cross_attn", "ca"]:
+            x = x + gate_1_1_1_B_D * self.block(
+                adaln_norm_state(self.norm_state, x, scale_1_1_1_B_D, shift_1_1_1_B_D),
+                context=crossattn_emb,
+                crossattn_mask=crossattn_mask,
+                rope_emb_L_1_1_D=rope_emb_L_1_1_D,
+            )
+        else:
+            raise ValueError(f"Unknown block type: {self.block_type}")
+
+        return x
+
+
+class GeneralDITTransformerBlock(nn.Module):
+    """
+    A wrapper module that manages a sequence of DITBuildingBlocks to form a complete transformer layer.
+    Each block in the sequence is specified by a block configuration string.
+
+    Parameters:
+        x_dim (int): Dimension of input features
+        context_dim (int): Dimension of context features for cross-attention blocks
+        num_heads (int): Number of attention heads
+        block_config (str): String specifying block sequence (e.g. "ca-fa-mlp" for cross-attention,
+                          full-attention, then MLP)
+        mlp_ratio (float): MLP hidden dimension multiplier. Default: 4.0
+        x_format (str): Input tensor format. Default: "BTHWD"
+        use_adaln_lora (bool): Whether to use AdaLN-LoRA. Default: False
+        adaln_lora_dim (int): Dimension for AdaLN-LoRA. Default: 256
+
+    The block_config string uses "-" to separate block types:
+        - "ca"/"cross_attn": Cross-attention block
+        - "fa"/"full_attn": Full self-attention block
+        - "mlp"/"ff": MLP/feedforward block
+
+    Example:
+        block_config = "ca-fa-mlp" creates a sequence of:
+        1. Cross-attention block
+        2. Full self-attention block
+        3. MLP block
+    """
+
+    def __init__(
+        self,
+        x_dim: int,
+        context_dim: int,
+        num_heads: int,
+        block_config: str,
+        mlp_ratio: float = 4.0,
+        x_format: str = "BTHWD",
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        weight_args={},
+        operations=None
+    ):
+        super().__init__()
+        self.blocks = nn.ModuleList()
+        self.x_format = x_format
+        for block_type in block_config.split("-"):
+            self.blocks.append(
+                DITBuildingBlock(
+                    block_type,
+                    x_dim,
+                    context_dim,
+                    num_heads,
+                    mlp_ratio,
+                    x_format=self.x_format,
+                    use_adaln_lora=use_adaln_lora,
+                    adaln_lora_dim=adaln_lora_dim,
+                    weight_args=weight_args,
+                    operations=operations,
+                )
+            )
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        emb_B_D: torch.Tensor,
+        crossattn_emb: torch.Tensor,
+        crossattn_mask: Optional[torch.Tensor] = None,
+        rope_emb_L_1_1_D: Optional[torch.Tensor] = None,
+        adaln_lora_B_3D: Optional[torch.Tensor] = None,
+        extra_per_block_pos_emb: Optional[torch.Tensor] = None,
+    ) -> torch.Tensor:
+        if extra_per_block_pos_emb is not None:
+            x = x + extra_per_block_pos_emb
+        for block in self.blocks:
+            x = block(
+                x,
+                emb_B_D,
+                crossattn_emb,
+                crossattn_mask,
+                rope_emb_L_1_1_D=rope_emb_L_1_1_D,
+                adaln_lora_B_3D=adaln_lora_B_3D,
+            )
+        return x

comfy/ldm/cosmos/cosmos_tokenizer/patching.py

@@ -0,0 +1,355 @@
+# SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""The patcher and unpatcher implementation for 2D and 3D data.
+
+The idea of Haar wavelet is to compute LL, LH, HL, HH component as two 1D convolutions.
+One on the rows and one on the columns.
+For example, in 1D signal, we have [a, b], then the low-freq compoenent is [a + b] / 2 and high-freq is [a - b] / 2.
+We can use a 1D convolution with kernel [1, 1] and stride 2 to represent the L component.
+For H component, we can use a 1D convolution with kernel [1, -1] and stride 2.
+Although in principle, we typically only do additional Haar wavelet over the LL component. But here we do it for all
+   as we need to support downsampling for more than 2x.
+For example, 4x downsampling can be done by 2x Haar and additional 2x Haar, and the shape would be.
+   [3, 256, 256] -> [12, 128, 128] -> [48, 64, 64]
+"""
+
+import torch
+import torch.nn.functional as F
+from einops import rearrange
+
+_WAVELETS = {
+    "haar": torch.tensor([0.7071067811865476, 0.7071067811865476]),
+    "rearrange": torch.tensor([1.0, 1.0]),
+}
+_PERSISTENT = False
+
+
+class Patcher(torch.nn.Module):
+    """A module to convert image tensors into patches using torch operations.
+
+    The main difference from `class Patching` is that this module implements
+    all operations using torch, rather than python or numpy, for efficiency purpose.
+
+    It's bit-wise identical to the Patching module outputs, with the added
+    benefit of being torch.jit scriptable.
+    """
+
+    def __init__(self, patch_size=1, patch_method="haar"):
+        super().__init__()
+        self.patch_size = patch_size
+        self.patch_method = patch_method
+        self.register_buffer(
+            "wavelets", _WAVELETS[patch_method], persistent=_PERSISTENT
+        )
+        self.range = range(int(torch.log2(torch.tensor(self.patch_size)).item()))
+        self.register_buffer(
+            "_arange",
+            torch.arange(_WAVELETS[patch_method].shape[0]),
+            persistent=_PERSISTENT,
+        )
+        for param in self.parameters():
+            param.requires_grad = False
+
+    def forward(self, x):
+        if self.patch_method == "haar":
+            return self._haar(x)
+        elif self.patch_method == "rearrange":
+            return self._arrange(x)
+        else:
+            raise ValueError("Unknown patch method: " + self.patch_method)
+
+    def _dwt(self, x, mode="reflect", rescale=False):
+        dtype = x.dtype
+        h = self.wavelets.to(device=x.device)
+
+        n = h.shape[0]
+        g = x.shape[1]
+        hl = h.flip(0).reshape(1, 1, -1).repeat(g, 1, 1)
+        hh = (h * ((-1) ** self._arange.to(device=x.device))).reshape(1, 1, -1).repeat(g, 1, 1)
+        hh = hh.to(dtype=dtype)
+        hl = hl.to(dtype=dtype)
+
+        x = F.pad(x, pad=(n - 2, n - 1, n - 2, n - 1), mode=mode).to(dtype)
+        xl = F.conv2d(x, hl.unsqueeze(2), groups=g, stride=(1, 2))
+        xh = F.conv2d(x, hh.unsqueeze(2), groups=g, stride=(1, 2))
+        xll = F.conv2d(xl, hl.unsqueeze(3), groups=g, stride=(2, 1))
+        xlh = F.conv2d(xl, hh.unsqueeze(3), groups=g, stride=(2, 1))
+        xhl = F.conv2d(xh, hl.unsqueeze(3), groups=g, stride=(2, 1))
+        xhh = F.conv2d(xh, hh.unsqueeze(3), groups=g, stride=(2, 1))
+
+        out = torch.cat([xll, xlh, xhl, xhh], dim=1)
+        if rescale:
+            out = out / 2
+        return out
+
+    def _haar(self, x):
+        for _ in self.range:
+            x = self._dwt(x, rescale=True)
+        return x
+
+    def _arrange(self, x):
+        x = rearrange(
+            x,
+            "b c (h p1) (w p2) -> b (c p1 p2) h w",
+            p1=self.patch_size,
+            p2=self.patch_size,
+        ).contiguous()
+        return x
+
+
+class Patcher3D(Patcher):
+    """A 3D discrete wavelet transform for video data, expects 5D tensor, i.e. a batch of videos."""
+
+    def __init__(self, patch_size=1, patch_method="haar"):
+        super().__init__(patch_method=patch_method, patch_size=patch_size)
+        self.register_buffer(
+            "patch_size_buffer",
+            patch_size * torch.ones([1], dtype=torch.int32),
+            persistent=_PERSISTENT,
+        )
+
+    def _dwt(self, x, wavelet, mode="reflect", rescale=False):
+        dtype = x.dtype
+        h = self.wavelets.to(device=x.device)
+
+        n = h.shape[0]
+        g = x.shape[1]
+        hl = h.flip(0).reshape(1, 1, -1).repeat(g, 1, 1)
+        hh = (h * ((-1) ** self._arange.to(device=x.device))).reshape(1, 1, -1).repeat(g, 1, 1)
+        hh = hh.to(dtype=dtype)
+        hl = hl.to(dtype=dtype)
+
+        # Handles temporal axis.
+        x = F.pad(
+            x, pad=(max(0, n - 2), n - 1, n - 2, n - 1, n - 2, n - 1), mode=mode
+        ).to(dtype)
+        xl = F.conv3d(x, hl.unsqueeze(3).unsqueeze(4), groups=g, stride=(2, 1, 1))
+        xh = F.conv3d(x, hh.unsqueeze(3).unsqueeze(4), groups=g, stride=(2, 1, 1))
+
+        # Handles spatial axes.
+        xll = F.conv3d(xl, hl.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1))
+        xlh = F.conv3d(xl, hh.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1))
+        xhl = F.conv3d(xh, hl.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1))
+        xhh = F.conv3d(xh, hh.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1))
+
+        xlll = F.conv3d(xll, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xllh = F.conv3d(xll, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xlhl = F.conv3d(xlh, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xlhh = F.conv3d(xlh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xhll = F.conv3d(xhl, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xhlh = F.conv3d(xhl, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xhhl = F.conv3d(xhh, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+        xhhh = F.conv3d(xhh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2))
+
+        out = torch.cat([xlll, xllh, xlhl, xlhh, xhll, xhlh, xhhl, xhhh], dim=1)
+        if rescale:
+            out = out / (2 * torch.sqrt(torch.tensor(2.0)))
+        return out
+
+    def _haar(self, x):
+        xi, xv = torch.split(x, [1, x.shape[2] - 1], dim=2)
+        x = torch.cat([xi.repeat_interleave(self.patch_size, dim=2), xv], dim=2)
+        for _ in self.range:
+            x = self._dwt(x, "haar", rescale=True)
+        return x
+
+    def _arrange(self, x):
+        xi, xv = torch.split(x, [1, x.shape[2] - 1], dim=2)
+        x = torch.cat([xi.repeat_interleave(self.patch_size, dim=2), xv], dim=2)
+        x = rearrange(
+            x,
+            "b c (t p1) (h p2) (w p3) -> b (c p1 p2 p3) t h w",
+            p1=self.patch_size,
+            p2=self.patch_size,
+            p3=self.patch_size,
+        ).contiguous()
+        return x
+
+
+class UnPatcher(torch.nn.Module):
+    """A module to convert patches into image tensorsusing torch operations.
+
+    The main difference from `class Unpatching` is that this module implements
+    all operations using torch, rather than python or numpy, for efficiency purpose.
+
+    It's bit-wise identical to the Unpatching module outputs, with the added
+    benefit of being torch.jit scriptable.
+    """
+
+    def __init__(self, patch_size=1, patch_method="haar"):
+        super().__init__()
+        self.patch_size = patch_size
+        self.patch_method = patch_method
+        self.register_buffer(
+            "wavelets", _WAVELETS[patch_method], persistent=_PERSISTENT
+        )
+        self.range = range(int(torch.log2(torch.tensor(self.patch_size)).item()))
+        self.register_buffer(
+            "_arange",
+            torch.arange(_WAVELETS[patch_method].shape[0]),
+            persistent=_PERSISTENT,
+        )
+        for param in self.parameters():
+            param.requires_grad = False
+
+    def forward(self, x):
+        if self.patch_method == "haar":
+            return self._ihaar(x)
+        elif self.patch_method == "rearrange":
+            return self._iarrange(x)
+        else:
+            raise ValueError("Unknown patch method: " + self.patch_method)
+
+    def _idwt(self, x, wavelet="haar", mode="reflect", rescale=False):
+        dtype = x.dtype
+        h = self.wavelets.to(device=x.device)
+        n = h.shape[0]
+
+        g = x.shape[1] // 4
+        hl = h.flip([0]).reshape(1, 1, -1).repeat([g, 1, 1])
+        hh = (h * ((-1) ** self._arange.to(device=x.device))).reshape(1, 1, -1).repeat(g, 1, 1)
+        hh = hh.to(dtype=dtype)
+        hl = hl.to(dtype=dtype)
+
+        xll, xlh, xhl, xhh = torch.chunk(x.to(dtype), 4, dim=1)
+
+        # Inverse transform.
+        yl = torch.nn.functional.conv_transpose2d(
+            xll, hl.unsqueeze(3), groups=g, stride=(2, 1), padding=(n - 2, 0)
+        )
+        yl += torch.nn.functional.conv_transpose2d(
+            xlh, hh.unsqueeze(3), groups=g, stride=(2, 1), padding=(n - 2, 0)
+        )
+        yh = torch.nn.functional.conv_transpose2d(
+            xhl, hl.unsqueeze(3), groups=g, stride=(2, 1), padding=(n - 2, 0)
+        )
+        yh += torch.nn.functional.conv_transpose2d(
+            xhh, hh.unsqueeze(3), groups=g, stride=(2, 1), padding=(n - 2, 0)
+        )
+        y = torch.nn.functional.conv_transpose2d(
+            yl, hl.unsqueeze(2), groups=g, stride=(1, 2), padding=(0, n - 2)
+        )
+        y += torch.nn.functional.conv_transpose2d(
+            yh, hh.unsqueeze(2), groups=g, stride=(1, 2), padding=(0, n - 2)
+        )
+
+        if rescale:
+            y = y * 2
+        return y
+
+    def _ihaar(self, x):
+        for _ in self.range:
+            x = self._idwt(x, "haar", rescale=True)
+        return x
+
+    def _iarrange(self, x):
+        x = rearrange(
+            x,
+            "b (c p1 p2) h w -> b c (h p1) (w p2)",
+            p1=self.patch_size,
+            p2=self.patch_size,
+        )
+        return x
+
+
+class UnPatcher3D(UnPatcher):
+    """A 3D inverse discrete wavelet transform for video wavelet decompositions."""
+
+    def __init__(self, patch_size=1, patch_method="haar"):
+        super().__init__(patch_method=patch_method, patch_size=patch_size)
+
+    def _idwt(self, x, wavelet="haar", mode="reflect", rescale=False):
+        dtype = x.dtype
+        h = self.wavelets.to(device=x.device)
+
+        g = x.shape[1] // 8  # split into 8 spatio-temporal filtered tesnors.
+        hl = h.flip([0]).reshape(1, 1, -1).repeat([g, 1, 1])
+        hh = (h * ((-1) ** self._arange.to(device=x.device))).reshape(1, 1, -1).repeat(g, 1, 1)
+        hl = hl.to(dtype=dtype)
+        hh = hh.to(dtype=dtype)
+
+        xlll, xllh, xlhl, xlhh, xhll, xhlh, xhhl, xhhh = torch.chunk(x, 8, dim=1)
+
+        # Height height transposed convolutions.
+        xll = F.conv_transpose3d(
+            xlll, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+        xll += F.conv_transpose3d(
+            xllh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+
+        xlh = F.conv_transpose3d(
+            xlhl, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+        xlh += F.conv_transpose3d(
+            xlhh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+
+        xhl = F.conv_transpose3d(
+            xhll, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+        xhl += F.conv_transpose3d(
+            xhlh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+
+        xhh = F.conv_transpose3d(
+            xhhl, hl.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+        xhh += F.conv_transpose3d(
+            xhhh, hh.unsqueeze(2).unsqueeze(3), groups=g, stride=(1, 1, 2)
+        )
+
+        # Handles width transposed convolutions.
+        xl = F.conv_transpose3d(
+            xll, hl.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1)
+        )
+        xl += F.conv_transpose3d(
+            xlh, hh.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1)
+        )
+        xh = F.conv_transpose3d(
+            xhl, hl.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1)
+        )
+        xh += F.conv_transpose3d(
+            xhh, hh.unsqueeze(2).unsqueeze(4), groups=g, stride=(1, 2, 1)
+        )
+
+        # Handles time axis transposed convolutions.
+        x = F.conv_transpose3d(
+            xl, hl.unsqueeze(3).unsqueeze(4), groups=g, stride=(2, 1, 1)
+        )
+        x += F.conv_transpose3d(
+            xh, hh.unsqueeze(3).unsqueeze(4), groups=g, stride=(2, 1, 1)
+        )
+
+        if rescale:
+            x = x * (2 * torch.sqrt(torch.tensor(2.0)))
+        return x
+
+    def _ihaar(self, x):
+        for _ in self.range:
+            x = self._idwt(x, "haar", rescale=True)
+        x = x[:, :, self.patch_size - 1 :, ...]
+        return x
+
+    def _iarrange(self, x):
+        x = rearrange(
+            x,
+            "b (c p1 p2 p3) t h w -> b c (t p1) (h p2) (w p3)",
+            p1=self.patch_size,
+            p2=self.patch_size,
+            p3=self.patch_size,
+        )
+        x = x[:, :, self.patch_size - 1 :, ...]
+        return x

comfy/ldm/cosmos/cosmos_tokenizer/utils.py

@@ -0,0 +1,120 @@
+# SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Shared utilities for the networks module."""
+
+from typing import Any
+
+import torch
+from einops import pack, rearrange, unpack
+
+
+import comfy.ops
+ops = comfy.ops.disable_weight_init
+
+def time2batch(x: torch.Tensor) -> tuple[torch.Tensor, int]:
+    batch_size = x.shape[0]
+    return rearrange(x, "b c t h w -> (b t) c h w"), batch_size
+
+
+def batch2time(x: torch.Tensor, batch_size: int) -> torch.Tensor:
+    return rearrange(x, "(b t) c h w -> b c t h w", b=batch_size)
+
+
+def space2batch(x: torch.Tensor) -> tuple[torch.Tensor, int]:
+    batch_size, height = x.shape[0], x.shape[-2]
+    return rearrange(x, "b c t h w -> (b h w) c t"), batch_size, height
+
+
+def batch2space(x: torch.Tensor, batch_size: int, height: int) -> torch.Tensor:
+    return rearrange(x, "(b h w) c t -> b c t h w", b=batch_size, h=height)
+
+
+def cast_tuple(t: Any, length: int = 1) -> Any:
+    return t if isinstance(t, tuple) else ((t,) * length)
+
+
+def replication_pad(x):
+    return torch.cat([x[:, :, :1, ...], x], dim=2)
+
+
+def divisible_by(num: int, den: int) -> bool:
+    return (num % den) == 0
+
+
+def is_odd(n: int) -> bool:
+    return not divisible_by(n, 2)
+
+
+def nonlinearity(x):
+    return x * torch.sigmoid(x)
+
+
+def Normalize(in_channels, num_groups=32):
+    return ops.GroupNorm(
+        num_groups=num_groups, num_channels=in_channels, eps=1e-6, affine=True
+    )
+
+
+class CausalNormalize(torch.nn.Module):
+    def __init__(self, in_channels, num_groups=1):
+        super().__init__()
+        self.norm = ops.GroupNorm(
+            num_groups=num_groups,
+            num_channels=in_channels,
+            eps=1e-6,
+            affine=True,
+        )
+        self.num_groups = num_groups
+
+    def forward(self, x):
+        # if num_groups !=1, we apply a spatio-temporal groupnorm for backward compatibility purpose.
+        # All new models should use num_groups=1, otherwise causality is not guaranteed.
+        if self.num_groups == 1:
+            x, batch_size = time2batch(x)
+            return batch2time(self.norm(x), batch_size)
+        return self.norm(x)
+
+
+def exists(v):
+    return v is not None
+
+
+def default(*args):
+    for arg in args:
+        if exists(arg):
+            return arg
+    return None
+
+
+def pack_one(t, pattern):
+    return pack([t], pattern)
+
+
+def unpack_one(t, ps, pattern):
+    return unpack(t, ps, pattern)[0]
+
+
+def round_ste(z: torch.Tensor) -> torch.Tensor:
+    """Round with straight through gradients."""
+    zhat = z.round()
+    return z + (zhat - z).detach()
+
+
+def log(t, eps=1e-5):
+    return t.clamp(min=eps).log()
+
+
+def entropy(prob):
+    return (-prob * log(prob)).sum(dim=-1)

comfy/ldm/cosmos/model.py

@@ -0,0 +1,510 @@
+# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+A general implementation of adaln-modulated VIT-like~(DiT) transformer for video processing.
+"""
+
+from typing import Optional, Tuple
+
+import torch
+from einops import rearrange
+from torch import nn
+from torchvision import transforms
+
+from enum import Enum
+import logging
+
+from comfy.ldm.modules.diffusionmodules.mmdit import RMSNorm
+
+from .blocks import (
+    FinalLayer,
+    GeneralDITTransformerBlock,
+    PatchEmbed,
+    TimestepEmbedding,
+    Timesteps,
+)
+
+from .position_embedding import LearnablePosEmbAxis, VideoRopePosition3DEmb
+
+
+class DataType(Enum):
+    IMAGE = "image"
+    VIDEO = "video"
+
+
+class GeneralDIT(nn.Module):
+    """
+    A general implementation of adaln-modulated VIT-like~(DiT) transformer for video processing.
+
+    Args:
+        max_img_h (int): Maximum height of the input images.
+        max_img_w (int): Maximum width of the input images.
+        max_frames (int): Maximum number of frames in the video sequence.
+        in_channels (int): Number of input channels (e.g., RGB channels for color images).
+        out_channels (int): Number of output channels.
+        patch_spatial (tuple): Spatial resolution of patches for input processing.
+        patch_temporal (int): Temporal resolution of patches for input processing.
+        concat_padding_mask (bool): If True, includes a mask channel in the input to handle padding.
+        block_config (str): Configuration of the transformer block. See Notes for supported block types.
+        model_channels (int): Base number of channels used throughout the model.
+        num_blocks (int): Number of transformer blocks.
+        num_heads (int): Number of heads in the multi-head attention layers.
+        mlp_ratio (float): Expansion ratio for MLP blocks.
+        block_x_format (str): Format of input tensor for transformer blocks ('BTHWD' or 'THWBD').
+        crossattn_emb_channels (int): Number of embedding channels for cross-attention.
+        use_cross_attn_mask (bool): Whether to use mask in cross-attention.
+        pos_emb_cls (str): Type of positional embeddings.
+        pos_emb_learnable (bool): Whether positional embeddings are learnable.
+        pos_emb_interpolation (str): Method for interpolating positional embeddings.
+        affline_emb_norm (bool): Whether to normalize affine embeddings.
+        use_adaln_lora (bool): Whether to use AdaLN-LoRA.
+        adaln_lora_dim (int): Dimension for AdaLN-LoRA.
+        rope_h_extrapolation_ratio (float): Height extrapolation ratio for RoPE.
+        rope_w_extrapolation_ratio (float): Width extrapolation ratio for RoPE.
+        rope_t_extrapolation_ratio (float): Temporal extrapolation ratio for RoPE.
+        extra_per_block_abs_pos_emb (bool): Whether to use extra per-block absolute positional embeddings.
+        extra_per_block_abs_pos_emb_type (str): Type of extra per-block positional embeddings.
+        extra_h_extrapolation_ratio (float): Height extrapolation ratio for extra embeddings.
+        extra_w_extrapolation_ratio (float): Width extrapolation ratio for extra embeddings.
+        extra_t_extrapolation_ratio (float): Temporal extrapolation ratio for extra embeddings.
+
+    Notes:
+        Supported block types in block_config:
+        * cross_attn, ca: Cross attention
+        * full_attn: Full attention on all flattened tokens
+        * mlp, ff: Feed forward block
+    """
+
+    def __init__(
+        self,
+        max_img_h: int,
+        max_img_w: int,
+        max_frames: int,
+        in_channels: int,
+        out_channels: int,
+        patch_spatial: tuple,
+        patch_temporal: int,
+        concat_padding_mask: bool = True,
+        # attention settings
+        block_config: str = "FA-CA-MLP",
+        model_channels: int = 768,
+        num_blocks: int = 10,
+        num_heads: int = 16,
+        mlp_ratio: float = 4.0,
+        block_x_format: str = "BTHWD",
+        # cross attention settings
+        crossattn_emb_channels: int = 1024,
+        use_cross_attn_mask: bool = False,
+        # positional embedding settings
+        pos_emb_cls: str = "sincos",
+        pos_emb_learnable: bool = False,
+        pos_emb_interpolation: str = "crop",
+        affline_emb_norm: bool = False,  # whether or not to normalize the affine embedding
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        rope_h_extrapolation_ratio: float = 1.0,
+        rope_w_extrapolation_ratio: float = 1.0,
+        rope_t_extrapolation_ratio: float = 1.0,
+        extra_per_block_abs_pos_emb: bool = False,
+        extra_per_block_abs_pos_emb_type: str = "sincos",
+        extra_h_extrapolation_ratio: float = 1.0,
+        extra_w_extrapolation_ratio: float = 1.0,
+        extra_t_extrapolation_ratio: float = 1.0,
+        image_model=None,
+        device=None,
+        dtype=None,
+        operations=None,
+    ) -> None:
+        super().__init__()
+        self.max_img_h = max_img_h
+        self.max_img_w = max_img_w
+        self.max_frames = max_frames
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.patch_spatial = patch_spatial
+        self.patch_temporal = patch_temporal
+        self.num_heads = num_heads
+        self.num_blocks = num_blocks
+        self.model_channels = model_channels
+        self.use_cross_attn_mask = use_cross_attn_mask
+        self.concat_padding_mask = concat_padding_mask
+        # positional embedding settings
+        self.pos_emb_cls = pos_emb_cls
+        self.pos_emb_learnable = pos_emb_learnable
+        self.pos_emb_interpolation = pos_emb_interpolation
+        self.affline_emb_norm = affline_emb_norm
+        self.rope_h_extrapolation_ratio = rope_h_extrapolation_ratio
+        self.rope_w_extrapolation_ratio = rope_w_extrapolation_ratio
+        self.rope_t_extrapolation_ratio = rope_t_extrapolation_ratio
+        self.extra_per_block_abs_pos_emb = extra_per_block_abs_pos_emb
+        self.extra_per_block_abs_pos_emb_type = extra_per_block_abs_pos_emb_type.lower()
+        self.extra_h_extrapolation_ratio = extra_h_extrapolation_ratio
+        self.extra_w_extrapolation_ratio = extra_w_extrapolation_ratio
+        self.extra_t_extrapolation_ratio = extra_t_extrapolation_ratio
+        self.dtype = dtype
+        weight_args = {"device": device, "dtype": dtype}
+
+        in_channels = in_channels + 1 if concat_padding_mask else in_channels
+        self.x_embedder = PatchEmbed(
+            spatial_patch_size=patch_spatial,
+            temporal_patch_size=patch_temporal,
+            in_channels=in_channels,
+            out_channels=model_channels,
+            bias=False,
+            weight_args=weight_args,
+            operations=operations,
+        )
+
+        self.build_pos_embed(device=device)
+        self.block_x_format = block_x_format
+        self.use_adaln_lora = use_adaln_lora
+        self.adaln_lora_dim = adaln_lora_dim
+        self.t_embedder = nn.ModuleList(
+            [Timesteps(model_channels),
+             TimestepEmbedding(model_channels, model_channels, use_adaln_lora=use_adaln_lora, weight_args=weight_args, operations=operations),]
+        )
+
+        self.blocks = nn.ModuleDict()
+
+        for idx in range(num_blocks):
+            self.blocks[f"block{idx}"] = GeneralDITTransformerBlock(
+                x_dim=model_channels,
+                context_dim=crossattn_emb_channels,
+                num_heads=num_heads,
+                block_config=block_config,
+                mlp_ratio=mlp_ratio,
+                x_format=self.block_x_format,
+                use_adaln_lora=use_adaln_lora,
+                adaln_lora_dim=adaln_lora_dim,
+                weight_args=weight_args,
+                operations=operations,
+            )
+
+        if self.affline_emb_norm:
+            logging.debug("Building affine embedding normalization layer")
+            self.affline_norm = RMSNorm(model_channels, elementwise_affine=True, eps=1e-6)
+        else:
+            self.affline_norm = nn.Identity()
+
+        self.final_layer = FinalLayer(
+            hidden_size=self.model_channels,
+            spatial_patch_size=self.patch_spatial,
+            temporal_patch_size=self.patch_temporal,
+            out_channels=self.out_channels,
+            use_adaln_lora=self.use_adaln_lora,
+            adaln_lora_dim=self.adaln_lora_dim,
+            weight_args=weight_args,
+            operations=operations,
+        )
+
+    def build_pos_embed(self, device=None):
+        if self.pos_emb_cls == "rope3d":
+            cls_type = VideoRopePosition3DEmb
+        else:
+            raise ValueError(f"Unknown pos_emb_cls {self.pos_emb_cls}")
+
+        logging.debug(f"Building positional embedding with {self.pos_emb_cls} class, impl {cls_type}")
+        kwargs = dict(
+            model_channels=self.model_channels,
+            len_h=self.max_img_h // self.patch_spatial,
+            len_w=self.max_img_w // self.patch_spatial,
+            len_t=self.max_frames // self.patch_temporal,
+            is_learnable=self.pos_emb_learnable,
+            interpolation=self.pos_emb_interpolation,
+            head_dim=self.model_channels // self.num_heads,
+            h_extrapolation_ratio=self.rope_h_extrapolation_ratio,
+            w_extrapolation_ratio=self.rope_w_extrapolation_ratio,
+            t_extrapolation_ratio=self.rope_t_extrapolation_ratio,
+            device=device,
+        )
+        self.pos_embedder = cls_type(
+            **kwargs,
+        )
+
+        if self.extra_per_block_abs_pos_emb:
+            assert self.extra_per_block_abs_pos_emb_type in [
+                "learnable",
+            ], f"Unknown extra_per_block_abs_pos_emb_type {self.extra_per_block_abs_pos_emb_type}"
+            kwargs["h_extrapolation_ratio"] = self.extra_h_extrapolation_ratio
+            kwargs["w_extrapolation_ratio"] = self.extra_w_extrapolation_ratio
+            kwargs["t_extrapolation_ratio"] = self.extra_t_extrapolation_ratio
+            kwargs["device"] = device
+            self.extra_pos_embedder = LearnablePosEmbAxis(
+                **kwargs,
+            )
+
+    def prepare_embedded_sequence(
+        self,
+        x_B_C_T_H_W: torch.Tensor,
+        fps: Optional[torch.Tensor] = None,
+        padding_mask: Optional[torch.Tensor] = None,
+        latent_condition: Optional[torch.Tensor] = None,
+        latent_condition_sigma: Optional[torch.Tensor] = None,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor]]:
+        """
+        Prepares an embedded sequence tensor by applying positional embeddings and handling padding masks.
+
+        Args:
+            x_B_C_T_H_W (torch.Tensor): video
+            fps (Optional[torch.Tensor]): Frames per second tensor to be used for positional embedding when required.
+                                    If None, a default value (`self.base_fps`) will be used.
+            padding_mask (Optional[torch.Tensor]): current it is not used
+
+        Returns:
+            Tuple[torch.Tensor, Optional[torch.Tensor]]:
+                - A tensor of shape (B, T, H, W, D) with the embedded sequence.
+                - An optional positional embedding tensor, returned only if the positional embedding class
+                (`self.pos_emb_cls`) includes 'rope'. Otherwise, None.
+
+        Notes:
+            - If `self.concat_padding_mask` is True, a padding mask channel is concatenated to the input tensor.
+            - The method of applying positional embeddings depends on the value of `self.pos_emb_cls`.
+            - If 'rope' is in `self.pos_emb_cls` (case insensitive), the positional embeddings are generated using
+                the `self.pos_embedder` with the shape [T, H, W].
+            - If "fps_aware" is in `self.pos_emb_cls`, the positional embeddings are generated using the
+            `self.pos_embedder` with the fps tensor.
+            - Otherwise, the positional embeddings are generated without considering fps.
+        """
+        if self.concat_padding_mask:
+            if padding_mask is not None:
+                padding_mask = transforms.functional.resize(
+                    padding_mask, list(x_B_C_T_H_W.shape[-2:]), interpolation=transforms.InterpolationMode.NEAREST
+                )
+            else:
+                padding_mask = torch.zeros((x_B_C_T_H_W.shape[0], 1, x_B_C_T_H_W.shape[-2], x_B_C_T_H_W.shape[-1]), dtype=x_B_C_T_H_W.dtype, device=x_B_C_T_H_W.device)
+
+            x_B_C_T_H_W = torch.cat(
+                [x_B_C_T_H_W, padding_mask.unsqueeze(1).repeat(1, 1, x_B_C_T_H_W.shape[2], 1, 1)], dim=1
+            )
+        x_B_T_H_W_D = self.x_embedder(x_B_C_T_H_W)
+
+        if self.extra_per_block_abs_pos_emb:
+            extra_pos_emb = self.extra_pos_embedder(x_B_T_H_W_D, fps=fps, device=x_B_C_T_H_W.device)
+        else:
+            extra_pos_emb = None
+
+        if "rope" in self.pos_emb_cls.lower():
+            return x_B_T_H_W_D, self.pos_embedder(x_B_T_H_W_D, fps=fps, device=x_B_C_T_H_W.device), extra_pos_emb
+
+        if "fps_aware" in self.pos_emb_cls:
+            x_B_T_H_W_D = x_B_T_H_W_D + self.pos_embedder(x_B_T_H_W_D, fps=fps, device=x_B_C_T_H_W.device)  # [B, T, H, W, D]
+        else:
+            x_B_T_H_W_D = x_B_T_H_W_D + self.pos_embedder(x_B_T_H_W_D, device=x_B_C_T_H_W.device)  # [B, T, H, W, D]
+
+        return x_B_T_H_W_D, None, extra_pos_emb
+
+    def decoder_head(
+        self,
+        x_B_T_H_W_D: torch.Tensor,
+        emb_B_D: torch.Tensor,
+        crossattn_emb: torch.Tensor,
+        origin_shape: Tuple[int, int, int, int, int],  # [B, C, T, H, W]
+        crossattn_mask: Optional[torch.Tensor] = None,
+        adaln_lora_B_3D: Optional[torch.Tensor] = None,
+    ) -> torch.Tensor:
+        del crossattn_emb, crossattn_mask
+        B, C, T_before_patchify, H_before_patchify, W_before_patchify = origin_shape
+        x_BT_HW_D = rearrange(x_B_T_H_W_D, "B T H W D -> (B T) (H W) D")
+        x_BT_HW_D = self.final_layer(x_BT_HW_D, emb_B_D, adaln_lora_B_3D=adaln_lora_B_3D)
+        # This is to ensure x_BT_HW_D has the correct shape because
+        # when we merge T, H, W into one dimension, x_BT_HW_D has shape (B * T * H * W, 1*1, D).
+        x_BT_HW_D = x_BT_HW_D.view(
+            B * T_before_patchify // self.patch_temporal,
+            H_before_patchify // self.patch_spatial * W_before_patchify // self.patch_spatial,
+            -1,
+        )
+        x_B_D_T_H_W = rearrange(
+            x_BT_HW_D,
+            "(B T) (H W) (p1 p2 t C) -> B C (T t) (H p1) (W p2)",
+            p1=self.patch_spatial,
+            p2=self.patch_spatial,
+            H=H_before_patchify // self.patch_spatial,
+            W=W_before_patchify // self.patch_spatial,
+            t=self.patch_temporal,
+            B=B,
+        )
+        return x_B_D_T_H_W
+
+    def forward_before_blocks(
+        self,
+        x: torch.Tensor,
+        timesteps: torch.Tensor,
+        crossattn_emb: torch.Tensor,
+        crossattn_mask: Optional[torch.Tensor] = None,
+        fps: Optional[torch.Tensor] = None,
+        image_size: Optional[torch.Tensor] = None,
+        padding_mask: Optional[torch.Tensor] = None,
+        scalar_feature: Optional[torch.Tensor] = None,
+        data_type: Optional[DataType] = DataType.VIDEO,
+        latent_condition: Optional[torch.Tensor] = None,
+        latent_condition_sigma: Optional[torch.Tensor] = None,
+        **kwargs,
+    ) -> torch.Tensor:
+        """
+        Args:
+            x: (B, C, T, H, W) tensor of spatial-temp inputs
+            timesteps: (B, ) tensor of timesteps
+            crossattn_emb: (B, N, D) tensor of cross-attention embeddings
+            crossattn_mask: (B, N) tensor of cross-attention masks
+        """
+        del kwargs
+        assert isinstance(
+            data_type, DataType
+        ), f"Expected DataType, got {type(data_type)}. We need discuss this flag later."
+        original_shape = x.shape
+        x_B_T_H_W_D, rope_emb_L_1_1_D, extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D = self.prepare_embedded_sequence(
+            x,
+            fps=fps,
+            padding_mask=padding_mask,
+            latent_condition=latent_condition,
+            latent_condition_sigma=latent_condition_sigma,
+        )
+        # logging affline scale information
+        affline_scale_log_info = {}
+
+        timesteps_B_D, adaln_lora_B_3D = self.t_embedder[1](self.t_embedder[0](timesteps.flatten()).to(x.dtype))
+        affline_emb_B_D = timesteps_B_D
+        affline_scale_log_info["timesteps_B_D"] = timesteps_B_D.detach()
+
+        if scalar_feature is not None:
+            raise NotImplementedError("Scalar feature is not implemented yet.")
+
+        affline_scale_log_info["affline_emb_B_D"] = affline_emb_B_D.detach()
+        affline_emb_B_D = self.affline_norm(affline_emb_B_D)
+
+        if self.use_cross_attn_mask:
+            if crossattn_mask is not None and not torch.is_floating_point(crossattn_mask):
+                crossattn_mask = (crossattn_mask - 1).to(x.dtype) * torch.finfo(x.dtype).max
+            crossattn_mask = crossattn_mask[:, None, None, :]  # .to(dtype=torch.bool)  # [B, 1, 1, length]
+        else:
+            crossattn_mask = None
+
+        if self.blocks["block0"].x_format == "THWBD":
+            x = rearrange(x_B_T_H_W_D, "B T H W D -> T H W B D")
+            if extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D is not None:
+                extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D = rearrange(
+                    extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D, "B T H W D -> T H W B D"
+                )
+            crossattn_emb = rearrange(crossattn_emb, "B M D -> M B D")
+
+            if crossattn_mask:
+                crossattn_mask = rearrange(crossattn_mask, "B M -> M B")
+
+        elif self.blocks["block0"].x_format == "BTHWD":
+            x = x_B_T_H_W_D
+        else:
+            raise ValueError(f"Unknown x_format {self.blocks[0].x_format}")
+        output = {
+            "x": x,
+            "affline_emb_B_D": affline_emb_B_D,
+            "crossattn_emb": crossattn_emb,
+            "crossattn_mask": crossattn_mask,
+            "rope_emb_L_1_1_D": rope_emb_L_1_1_D,
+            "adaln_lora_B_3D": adaln_lora_B_3D,
+            "original_shape": original_shape,
+            "extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D": extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D,
+        }
+        return output
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        timesteps: torch.Tensor,
+        context: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        # crossattn_emb: torch.Tensor,
+        # crossattn_mask: Optional[torch.Tensor] = None,
+        fps: Optional[torch.Tensor] = None,
+        image_size: Optional[torch.Tensor] = None,
+        padding_mask: Optional[torch.Tensor] = None,
+        scalar_feature: Optional[torch.Tensor] = None,
+        data_type: Optional[DataType] = DataType.VIDEO,
+        latent_condition: Optional[torch.Tensor] = None,
+        latent_condition_sigma: Optional[torch.Tensor] = None,
+        condition_video_augment_sigma: Optional[torch.Tensor] = None,
+        **kwargs,
+    ):
+        """
+        Args:
+            x: (B, C, T, H, W) tensor of spatial-temp inputs
+            timesteps: (B, ) tensor of timesteps
+            crossattn_emb: (B, N, D) tensor of cross-attention embeddings
+            crossattn_mask: (B, N) tensor of cross-attention masks
+            condition_video_augment_sigma: (B,) used in lvg(long video generation), we add noise with this sigma to
+                augment condition input, the lvg model will condition on the condition_video_augment_sigma value;
+                we need forward_before_blocks pass to the forward_before_blocks function.
+        """
+
+        crossattn_emb = context
+        crossattn_mask = attention_mask
+
+        inputs = self.forward_before_blocks(
+            x=x,
+            timesteps=timesteps,
+            crossattn_emb=crossattn_emb,
+            crossattn_mask=crossattn_mask,
+            fps=fps,
+            image_size=image_size,
+            padding_mask=padding_mask,
+            scalar_feature=scalar_feature,
+            data_type=data_type,
+            latent_condition=latent_condition,
+            latent_condition_sigma=latent_condition_sigma,
+            condition_video_augment_sigma=condition_video_augment_sigma,
+            **kwargs,
+        )
+        x, affline_emb_B_D, crossattn_emb, crossattn_mask, rope_emb_L_1_1_D, adaln_lora_B_3D, original_shape = (
+            inputs["x"],
+            inputs["affline_emb_B_D"],
+            inputs["crossattn_emb"],
+            inputs["crossattn_mask"],
+            inputs["rope_emb_L_1_1_D"],
+            inputs["adaln_lora_B_3D"],
+            inputs["original_shape"],
+        )
+        extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D = inputs["extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D"].to(x.dtype)
+        if extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D is not None:
+            assert (
+                x.shape == extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D.shape
+            ), f"{x.shape} != {extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D.shape} {original_shape}"
+
+        for _, block in self.blocks.items():
+            assert (
+                self.blocks["block0"].x_format == block.x_format
+            ), f"First block has x_format {self.blocks[0].x_format}, got {block.x_format}"
+
+            x = block(
+                x,
+                affline_emb_B_D,
+                crossattn_emb,
+                crossattn_mask,
+                rope_emb_L_1_1_D=rope_emb_L_1_1_D,
+                adaln_lora_B_3D=adaln_lora_B_3D,
+                extra_per_block_pos_emb=extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D,
+            )
+
+        x_B_T_H_W_D = rearrange(x, "T H W B D -> B T H W D")
+
+        x_B_D_T_H_W = self.decoder_head(
+            x_B_T_H_W_D=x_B_T_H_W_D,
+            emb_B_D=affline_emb_B_D,
+            crossattn_emb=None,
+            origin_shape=original_shape,
+            crossattn_mask=None,
+            adaln_lora_B_3D=adaln_lora_B_3D,
+        )
+
+        return x_B_D_T_H_W

comfy/ldm/cosmos/position_embedding.py

@@ -0,0 +1,207 @@
+# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import List, Optional
+
+import torch
+from einops import rearrange, repeat
+from torch import nn
+import math
+
+
+def normalize(x: torch.Tensor, dim: Optional[List[int]] = None, eps: float = 0) -> torch.Tensor:
+    """
+    Normalizes the input tensor along specified dimensions such that the average square norm of elements is adjusted.
+
+    Args:
+        x (torch.Tensor): The input tensor to normalize.
+        dim (list, optional): The dimensions over which to normalize. If None, normalizes over all dimensions except the first.
+        eps (float, optional): A small constant to ensure numerical stability during division.
+
+    Returns:
+        torch.Tensor: The normalized tensor.
+    """
+    if dim is None:
+        dim = list(range(1, x.ndim))
+    norm = torch.linalg.vector_norm(x, dim=dim, keepdim=True, dtype=torch.float32)
+    norm = torch.add(eps, norm, alpha=math.sqrt(norm.numel() / x.numel()))
+    return x / norm.to(x.dtype)
+
+
+class VideoPositionEmb(nn.Module):
+    def forward(self, x_B_T_H_W_C: torch.Tensor, fps=Optional[torch.Tensor], device=None) -> torch.Tensor:
+        """
+        It delegates the embedding generation to generate_embeddings function.
+        """
+        B_T_H_W_C = x_B_T_H_W_C.shape
+        embeddings = self.generate_embeddings(B_T_H_W_C, fps=fps, device=device)
+
+        return embeddings
+
+    def generate_embeddings(self, B_T_H_W_C: torch.Size, fps=Optional[torch.Tensor], device=None):
+        raise NotImplementedError
+
+
+class VideoRopePosition3DEmb(VideoPositionEmb):
+    def __init__(
+        self,
+        *,  # enforce keyword arguments
+        head_dim: int,
+        len_h: int,
+        len_w: int,
+        len_t: int,
+        base_fps: int = 24,
+        h_extrapolation_ratio: float = 1.0,
+        w_extrapolation_ratio: float = 1.0,
+        t_extrapolation_ratio: float = 1.0,
+        device=None,
+        **kwargs,  # used for compatibility with other positional embeddings; unused in this class
+    ):
+        del kwargs
+        super().__init__()
+        self.register_buffer("seq", torch.arange(max(len_h, len_w, len_t), dtype=torch.float, device=device))
+        self.base_fps = base_fps
+        self.max_h = len_h
+        self.max_w = len_w
+
+        dim = head_dim
+        dim_h = dim // 6 * 2
+        dim_w = dim_h
+        dim_t = dim - 2 * dim_h
+        assert dim == dim_h + dim_w + dim_t, f"bad dim: {dim} != {dim_h} + {dim_w} + {dim_t}"
+        self.register_buffer(
+            "dim_spatial_range",
+            torch.arange(0, dim_h, 2, device=device)[: (dim_h // 2)].float() / dim_h,
+            persistent=False,
+        )
+        self.register_buffer(
+            "dim_temporal_range",
+            torch.arange(0, dim_t, 2, device=device)[: (dim_t // 2)].float() / dim_t,
+            persistent=False,
+        )
+
+        self.h_ntk_factor = h_extrapolation_ratio ** (dim_h / (dim_h - 2))
+        self.w_ntk_factor = w_extrapolation_ratio ** (dim_w / (dim_w - 2))
+        self.t_ntk_factor = t_extrapolation_ratio ** (dim_t / (dim_t - 2))
+
+    def generate_embeddings(
+        self,
+        B_T_H_W_C: torch.Size,
+        fps: Optional[torch.Tensor] = None,
+        h_ntk_factor: Optional[float] = None,
+        w_ntk_factor: Optional[float] = None,
+        t_ntk_factor: Optional[float] = None,
+        device=None,
+    ):
+        """
+        Generate embeddings for the given input size.
+
+        Args:
+            B_T_H_W_C (torch.Size): Input tensor size (Batch, Time, Height, Width, Channels).
+            fps (Optional[torch.Tensor], optional): Frames per second. Defaults to None.
+            h_ntk_factor (Optional[float], optional): Height NTK factor. If None, uses self.h_ntk_factor.
+            w_ntk_factor (Optional[float], optional): Width NTK factor. If None, uses self.w_ntk_factor.
+            t_ntk_factor (Optional[float], optional): Time NTK factor. If None, uses self.t_ntk_factor.
+
+        Returns:
+            Not specified in the original code snippet.
+        """
+        h_ntk_factor = h_ntk_factor if h_ntk_factor is not None else self.h_ntk_factor
+        w_ntk_factor = w_ntk_factor if w_ntk_factor is not None else self.w_ntk_factor
+        t_ntk_factor = t_ntk_factor if t_ntk_factor is not None else self.t_ntk_factor
+
+        h_theta = 10000.0 * h_ntk_factor
+        w_theta = 10000.0 * w_ntk_factor
+        t_theta = 10000.0 * t_ntk_factor
+
+        h_spatial_freqs = 1.0 / (h_theta**self.dim_spatial_range.to(device=device))
+        w_spatial_freqs = 1.0 / (w_theta**self.dim_spatial_range.to(device=device))
+        temporal_freqs = 1.0 / (t_theta**self.dim_temporal_range.to(device=device))
+
+        B, T, H, W, _ = B_T_H_W_C
+        uniform_fps = (fps is None) or isinstance(fps, (int, float)) or (fps.min() == fps.max())
+        assert (
+            uniform_fps or B == 1 or T == 1
+        ), "For video batch, batch size should be 1 for non-uniform fps. For image batch, T should be 1"
+        assert (
+            H <= self.max_h and W <= self.max_w
+        ), f"Input dimensions (H={H}, W={W}) exceed the maximum dimensions (max_h={self.max_h}, max_w={self.max_w})"
+        half_emb_h = torch.outer(self.seq[:H].to(device=device), h_spatial_freqs)
+        half_emb_w = torch.outer(self.seq[:W].to(device=device), w_spatial_freqs)
+
+        # apply sequence scaling in temporal dimension
+        if fps is None:  # image case
+            half_emb_t = torch.outer(self.seq[:T].to(device=device), temporal_freqs)
+        else:
+            half_emb_t = torch.outer(self.seq[:T].to(device=device) / fps * self.base_fps, temporal_freqs)
+
+        half_emb_h = torch.stack([torch.cos(half_emb_h), -torch.sin(half_emb_h), torch.sin(half_emb_h), torch.cos(half_emb_h)], dim=-1)
+        half_emb_w = torch.stack([torch.cos(half_emb_w), -torch.sin(half_emb_w), torch.sin(half_emb_w), torch.cos(half_emb_w)], dim=-1)
+        half_emb_t = torch.stack([torch.cos(half_emb_t), -torch.sin(half_emb_t), torch.sin(half_emb_t), torch.cos(half_emb_t)], dim=-1)
+
+        em_T_H_W_D = torch.cat(
+            [
+                repeat(half_emb_t, "t d x -> t h w d x", h=H, w=W),
+                repeat(half_emb_h, "h d x -> t h w d x", t=T, w=W),
+                repeat(half_emb_w, "w d x -> t h w d x", t=T, h=H),
+            ]
+            , dim=-2,
+        )
+
+        return rearrange(em_T_H_W_D, "t h w d (i j) -> (t h w) d i j", i=2, j=2).float()
+
+
+class LearnablePosEmbAxis(VideoPositionEmb):
+    def __init__(
+        self,
+        *,  # enforce keyword arguments
+        interpolation: str,
+        model_channels: int,
+        len_h: int,
+        len_w: int,
+        len_t: int,
+        device=None,
+        **kwargs,
+    ):
+        """
+        Args:
+            interpolation (str): we curretly only support "crop", ideally when we need extrapolation capacity, we should adjust frequency or other more advanced methods. they are not implemented yet.
+        """
+        del kwargs  # unused
+        super().__init__()
+        self.interpolation = interpolation
+        assert self.interpolation in ["crop"], f"Unknown interpolation method {self.interpolation}"
+
+        self.pos_emb_h = nn.Parameter(torch.empty(len_h, model_channels, device=device))
+        self.pos_emb_w = nn.Parameter(torch.empty(len_w, model_channels, device=device))
+        self.pos_emb_t = nn.Parameter(torch.empty(len_t, model_channels, device=device))
+
+
+    def generate_embeddings(self, B_T_H_W_C: torch.Size, fps=Optional[torch.Tensor], device=None) -> torch.Tensor:
+        B, T, H, W, _ = B_T_H_W_C
+        if self.interpolation == "crop":
+            emb_h_H = self.pos_emb_h[:H].to(device=device)
+            emb_w_W = self.pos_emb_w[:W].to(device=device)
+            emb_t_T = self.pos_emb_t[:T].to(device=device)
+            emb = (
+                repeat(emb_t_T, "t d-> b t h w d", b=B, h=H, w=W)
+                + repeat(emb_h_H, "h d-> b t h w d", b=B, t=T, w=W)
+                + repeat(emb_w_W, "w d-> b t h w d", b=B, t=T, h=H)
+            )
+            assert list(emb.shape)[:4] == [B, T, H, W], f"bad shape: {list(emb.shape)[:4]} != {B, T, H, W}"
+        else:
+            raise ValueError(f"Unknown interpolation method {self.interpolation}")
+
+        return normalize(emb, dim=-1, eps=1e-6)

comfy/ldm/cosmos/vae.py

@@ -0,0 +1,124 @@
+# SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""The causal continuous video tokenizer with VAE or AE formulation for 3D data.."""
+
+import logging
+import torch
+from torch import nn
+from enum import Enum
+
+from .cosmos_tokenizer.layers3d import (
+    EncoderFactorized,
+    DecoderFactorized,
+    CausalConv3d,
+)
+
+
+class IdentityDistribution(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    def forward(self, parameters):
+        return parameters, (torch.tensor([0.0]), torch.tensor([0.0]))
+
+
+class GaussianDistribution(torch.nn.Module):
+    def __init__(self, min_logvar: float = -30.0, max_logvar: float = 20.0):
+        super().__init__()
+        self.min_logvar = min_logvar
+        self.max_logvar = max_logvar
+
+    def sample(self, mean, logvar):
+        std = torch.exp(0.5 * logvar)
+        return mean + std * torch.randn_like(mean)
+
+    def forward(self, parameters):
+        mean, logvar = torch.chunk(parameters, 2, dim=1)
+        logvar = torch.clamp(logvar, self.min_logvar, self.max_logvar)
+        return self.sample(mean, logvar), (mean, logvar)
+
+
+class ContinuousFormulation(Enum):
+    VAE = GaussianDistribution
+    AE = IdentityDistribution
+
+
+class CausalContinuousVideoTokenizer(nn.Module):
+    def __init__(
+        self, z_channels: int, z_factor: int, latent_channels: int, **kwargs
+    ) -> None:
+        super().__init__()
+        self.name = kwargs.get("name", "CausalContinuousVideoTokenizer")
+        self.latent_channels = latent_channels
+        self.sigma_data = 0.5
+
+        # encoder_name = kwargs.get("encoder", Encoder3DType.BASE.name)
+        self.encoder = EncoderFactorized(
+            z_channels=z_factor * z_channels, **kwargs
+        )
+        if kwargs.get("temporal_compression", 4) == 4:
+            kwargs["channels_mult"] = [2, 4]
+        # decoder_name = kwargs.get("decoder", Decoder3DType.BASE.name)
+        self.decoder = DecoderFactorized(
+            z_channels=z_channels, **kwargs
+        )
+
+        self.quant_conv = CausalConv3d(
+            z_factor * z_channels,
+            z_factor * latent_channels,
+            kernel_size=1,
+            padding=0,
+        )
+        self.post_quant_conv = CausalConv3d(
+            latent_channels, z_channels, kernel_size=1, padding=0
+        )
+
+        # formulation_name = kwargs.get("formulation", ContinuousFormulation.AE.name)
+        self.distribution = IdentityDistribution()  # ContinuousFormulation[formulation_name].value()
+
+        num_parameters = sum(param.numel() for param in self.parameters())
+        logging.info(f"model={self.name}, num_parameters={num_parameters:,}")
+        logging.info(
+            f"z_channels={z_channels}, latent_channels={self.latent_channels}."
+        )
+
+        latent_temporal_chunk = 16
+        self.latent_mean = nn.Parameter(torch.zeros([self.latent_channels * latent_temporal_chunk], dtype=torch.float32))
+        self.latent_std = nn.Parameter(torch.ones([self.latent_channels * latent_temporal_chunk], dtype=torch.float32))
+
+
+    def encode(self, x):
+        h = self.encoder(x)
+        moments = self.quant_conv(h)
+        z, posteriors = self.distribution(moments)
+        latent_ch = z.shape[1]
+        latent_t = z.shape[2]
+        dtype = z.dtype
+        mean = self.latent_mean.view(latent_ch, -1)[:, : latent_t].reshape([1, latent_ch, -1, 1, 1]).to(dtype=dtype, device=z.device)
+        std = self.latent_std.view(latent_ch, -1)[:, : latent_t].reshape([1, latent_ch, -1, 1, 1]).to(dtype=dtype, device=z.device)
+        return ((z - mean) / std) * self.sigma_data
+
+    def decode(self, z):
+        in_dtype = z.dtype
+        latent_ch = z.shape[1]
+        latent_t = z.shape[2]
+        mean = self.latent_mean.view(latent_ch, -1)[:, : latent_t].reshape([1, latent_ch, -1, 1, 1]).to(dtype=in_dtype, device=z.device)
+        std = self.latent_std.view(latent_ch, -1)[:, : latent_t].reshape([1, latent_ch, -1, 1, 1]).to(dtype=in_dtype, device=z.device)
+
+        z = z / self.sigma_data
+        z = z * std + mean
+        z = self.post_quant_conv(z)
+        return self.decoder(z)
+

comfy/ldm/modules/attention.py

@@ -89,7 +89,7 @@ class FeedForward(nn.Module):
 def Normalize(in_channels, dtype=None, device=None):
     return torch.nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True, dtype=dtype, device=device)
 
-def attention_basic(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_basic(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     attn_precision = get_attn_precision(attn_precision)
 
     if skip_reshape:
@@ -142,6 +142,13 @@ def attention_basic(q, k, v, heads, mask=None, attn_precision=None, skip_reshape
     sim = sim.softmax(dim=-1)
 
     out = einsum('b i j, b j d -> b i d', sim.to(v.dtype), v)
+
+    if skip_output_reshape:
+        out = (
+            out.unsqueeze(0)
+            .reshape(b, heads, -1, dim_head)
+        )
+    else:
         out = (
             out.unsqueeze(0)
             .reshape(b, heads, -1, dim_head)
@@ -151,7 +158,7 @@ def attention_basic(q, k, v, heads, mask=None, attn_precision=None, skip_reshape
     return out
 
 
-def attention_sub_quad(query, key, value, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_sub_quad(query, key, value, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     attn_precision = get_attn_precision(attn_precision)
 
     if skip_reshape:
@@ -215,11 +222,13 @@ def attention_sub_quad(query, key, value, heads, mask=None, attn_precision=None,
     )
 
     hidden_states = hidden_states.to(dtype)
-
+    if skip_output_reshape:
+        hidden_states = hidden_states.unflatten(0, (-1, heads))
+    else:
         hidden_states = hidden_states.unflatten(0, (-1, heads)).transpose(1,2).flatten(start_dim=2)
     return hidden_states
 
-def attention_split(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_split(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     attn_precision = get_attn_precision(attn_precision)
 
     if skip_reshape:
@@ -326,6 +335,12 @@ def attention_split(q, k, v, heads, mask=None, attn_precision=None, skip_reshape
 
     del q, k, v
 
+    if skip_output_reshape:
+        r1 = (
+            r1.unsqueeze(0)
+            .reshape(b, heads, -1, dim_head)
+        )
+    else:
         r1 = (
             r1.unsqueeze(0)
             .reshape(b, heads, -1, dim_head)
@@ -342,7 +357,7 @@ try:
 except:
     pass
 
-def attention_xformers(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_xformers(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     b = q.shape[0]
     dim_head = q.shape[-1]
     # check to make sure xformers isn't broken
@@ -395,6 +410,9 @@ def attention_xformers(q, k, v, heads, mask=None, attn_precision=None, skip_resh
 
     out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=mask)
 
+    if skip_output_reshape:
+        out = out.permute(0, 2, 1, 3)
+    else:
         out = (
             out.reshape(b, -1, heads * dim_head)
         )
@@ -408,7 +426,7 @@ else:
     SDP_BATCH_LIMIT = 2**31
 
 
-def attention_pytorch(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_pytorch(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     if skip_reshape:
         b, _, _, dim_head = q.shape
     else:
@@ -429,6 +447,7 @@ def attention_pytorch(q, k, v, heads, mask=None, attn_precision=None, skip_resha
 
     if SDP_BATCH_LIMIT >= b:
         out = torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=mask, dropout_p=0.0, is_causal=False)
+        if not skip_output_reshape:
             out = (
                 out.transpose(1, 2).reshape(b, -1, heads * dim_head)
             )
@@ -450,7 +469,7 @@ def attention_pytorch(q, k, v, heads, mask=None, attn_precision=None, skip_resha
     return out
 
 
-def attention_sage(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False):
+def attention_sage(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False):
     if skip_reshape:
         b, _, _, dim_head = q.shape
         tensor_layout="HND"
@@ -473,9 +492,13 @@ def attention_sage(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=
 
     out = sageattn(q, k, v, attn_mask=mask, is_causal=False, tensor_layout=tensor_layout)
     if tensor_layout == "HND":
+        if not skip_output_reshape:
             out = (
                 out.transpose(1, 2).reshape(b, -1, heads * dim_head)
             )
+    else:
+        if skip_output_reshape:
+            out = out.transpose(1, 2)
         else:
             out = out.reshape(b, -1, heads * dim_head)
     return out

comfy/model_base.py

@@ -33,6 +33,7 @@ import comfy.ldm.audio.embedders
 import comfy.ldm.flux.model
 import comfy.ldm.lightricks.model
 import comfy.ldm.hunyuan_video.model
+import comfy.ldm.cosmos.model
 
 import comfy.model_management
 import comfy.patcher_extension
@@ -856,3 +857,19 @@ class HunyuanVideo(BaseModel):
             out['c_crossattn'] = comfy.conds.CONDRegular(cross_attn)
         out['guidance'] = comfy.conds.CONDRegular(torch.FloatTensor([kwargs.get("guidance", 6.0)]))
         return out
+
+class CosmosVideo(BaseModel):
+    def __init__(self, model_config, model_type=ModelType.EDM, device=None):
+        super().__init__(model_config, model_type, device=device, unet_model=comfy.ldm.cosmos.model.GeneralDIT)
+
+    def extra_conds(self, **kwargs):
+        out = super().extra_conds(**kwargs)
+        attention_mask = kwargs.get("attention_mask", None)
+        if attention_mask is not None:
+            out['attention_mask'] = comfy.conds.CONDRegular(attention_mask)
+        cross_attn = kwargs.get("cross_attn", None)
+        if cross_attn is not None:
+            out['c_crossattn'] = comfy.conds.CONDRegular(cross_attn)
+
+        out['fps'] = comfy.conds.CONDConstant(kwargs.get("frame_rate", None))
+        return out

comfy/model_detection.py

@@ -239,6 +239,50 @@ def detect_unet_config(state_dict, key_prefix):
             dit_config["micro_condition"] = False
         return dit_config
 
+    if '{}blocks.block0.blocks.0.block.attn.to_q.0.weight'.format(key_prefix) in state_dict_keys:
+        dit_config = {}
+        dit_config["image_model"] = "cosmos"
+        dit_config["max_img_h"] = 240
+        dit_config["max_img_w"] = 240
+        dit_config["max_frames"] = 128
+        dit_config["in_channels"] = 16
+        dit_config["out_channels"] = 16
+        dit_config["patch_spatial"] = 2
+        dit_config["patch_temporal"] = 1
+        dit_config["model_channels"] = state_dict['{}blocks.block0.blocks.0.block.attn.to_q.0.weight'.format(key_prefix)].shape[0]
+        dit_config["block_config"] = "FA-CA-MLP"
+        dit_config["concat_padding_mask"] = True
+        dit_config["pos_emb_cls"] = "rope3d"
+        dit_config["pos_emb_learnable"] = False
+        dit_config["pos_emb_interpolation"] = "crop"
+        dit_config["block_x_format"] = "THWBD"
+        dit_config["affline_emb_norm"] = True
+        dit_config["use_adaln_lora"] = True
+        dit_config["adaln_lora_dim"] = 256
+
+        if dit_config["model_channels"] == 4096:
+            # 7B
+            dit_config["num_blocks"] = 28
+            dit_config["num_heads"] = 32
+            dit_config["extra_per_block_abs_pos_emb"] = True
+            dit_config["rope_h_extrapolation_ratio"] = 1.0
+            dit_config["rope_w_extrapolation_ratio"] = 1.0
+            dit_config["rope_t_extrapolation_ratio"] = 2.0
+            dit_config["extra_per_block_abs_pos_emb_type"] = "learnable"
+        else:  # 5120
+            # 14B
+            dit_config["num_blocks"] = 36
+            dit_config["num_heads"] = 40
+            dit_config["extra_per_block_abs_pos_emb"] = True
+            dit_config["rope_h_extrapolation_ratio"] = 2.0
+            dit_config["rope_w_extrapolation_ratio"] = 2.0
+            dit_config["rope_t_extrapolation_ratio"] = 2.0
+            dit_config["extra_h_extrapolation_ratio"] = 2.0
+            dit_config["extra_w_extrapolation_ratio"] = 2.0
+            dit_config["extra_t_extrapolation_ratio"] = 2.0
+            dit_config["extra_per_block_abs_pos_emb_type"] = "learnable"
+        return dit_config
+
     if '{}input_blocks.0.0.weight'.format(key_prefix) not in state_dict_keys:
         return None
 
@@ -393,6 +437,7 @@ def model_config_from_unet(state_dict, unet_key_prefix, use_base_if_no_match=Fal
 def unet_prefix_from_state_dict(state_dict):
     candidates = ["model.diffusion_model.", #ldm/sgm models
                   "model.model.", #audio models
+                  "net.", #cosmos
                   ]
     counts = {k: 0 for k in candidates}
     for k in state_dict:

comfy/model_patcher.py

@@ -210,7 +210,7 @@ class ModelPatcher:
         self.injections: dict[str, list[PatcherInjection]] = {}
 
         self.hook_patches: dict[comfy.hooks._HookRef] = {}
-        self.hook_patches_backup: dict[comfy.hooks._HookRef] = {}
+        self.hook_patches_backup: dict[comfy.hooks._HookRef] = None
         self.hook_backup: dict[str, tuple[torch.Tensor, torch.device]] = {}
         self.cached_hook_patches: dict[comfy.hooks.HookGroup, dict[str, torch.Tensor]] = {}
         self.current_hooks: Optional[comfy.hooks.HookGroup] = None
@@ -282,7 +282,7 @@ class ModelPatcher:
             n.injections[k] = i.copy()
         # hooks
         n.hook_patches = create_hook_patches_clone(self.hook_patches)
-        n.hook_patches_backup = create_hook_patches_clone(self.hook_patches_backup)
+        n.hook_patches_backup = create_hook_patches_clone(self.hook_patches_backup) if self.hook_patches_backup else self.hook_patches_backup
         for group in self.cached_hook_patches:
             n.cached_hook_patches[group] = {}
             for k in self.cached_hook_patches[group]:
@@ -855,6 +855,9 @@ class ModelPatcher:
         if key in self.injections:
             self.injections.pop(key)
 
+    def get_injections(self, key: str):
+        return self.injections.get(key, None)
+
     def set_additional_models(self, key: str, models: list['ModelPatcher']):
         self.additional_models[key] = models
 
@@ -925,9 +928,9 @@ class ModelPatcher:
             callback(self, timestep)
 
     def restore_hook_patches(self):
-        if len(self.hook_patches_backup) > 0:
+        if self.hook_patches_backup is not None:
             self.hook_patches = self.hook_patches_backup
-            self.hook_patches_backup = {}
+            self.hook_patches_backup = None
 
     def set_hook_mode(self, hook_mode: comfy.hooks.EnumHookMode):
         self.hook_mode = hook_mode
@@ -953,25 +956,26 @@ class ModelPatcher:
         if reset_current_hooks:
             self.patch_hooks(None)
 
-    def register_all_hook_patches(self, hooks_dict: dict[comfy.hooks.EnumHookType, dict[comfy.hooks.Hook, None]], target: comfy.hooks.EnumWeightTarget, model_options: dict=None):
+    def register_all_hook_patches(self, hooks: comfy.hooks.HookGroup, target_dict: dict[str], model_options: dict=None,
+                                  registered: comfy.hooks.HookGroup = None):
         self.restore_hook_patches()
-        registered_hooks: list[comfy.hooks.Hook] = []
-        # handle WrapperHooks, if model_options provided
-        if model_options is not None:
-            for hook in hooks_dict.get(comfy.hooks.EnumHookType.Wrappers, {}):
-                hook.add_hook_patches(self, model_options, target, registered_hooks)
+        if registered is None:
+            registered = comfy.hooks.HookGroup()
         # handle WeightHooks
         weight_hooks_to_register: list[comfy.hooks.WeightHook] = []
-        for hook in hooks_dict.get(comfy.hooks.EnumHookType.Weight, {}):
+        for hook in hooks.get_type(comfy.hooks.EnumHookType.Weight):
             if hook.hook_ref not in self.hook_patches:
                 weight_hooks_to_register.append(hook)
+            else:
+                registered.add(hook)
         if len(weight_hooks_to_register) > 0:
             # clone hook_patches to become backup so that any non-dynamic hooks will return to their original state
             self.hook_patches_backup = create_hook_patches_clone(self.hook_patches)
             for hook in weight_hooks_to_register:
-                hook.add_hook_patches(self, model_options, target, registered_hooks)
+                hook.add_hook_patches(self, model_options, target_dict, registered)
         for callback in self.get_all_callbacks(CallbacksMP.ON_REGISTER_ALL_HOOK_PATCHES):
-            callback(self, hooks_dict, target)
+            callback(self, hooks, target_dict, model_options, registered)
+        return registered
 
     def add_hook_patches(self, hook: comfy.hooks.WeightHook, patches, strength_patch=1.0, strength_model=1.0):
         with self.use_ejected():
@@ -1022,11 +1026,11 @@ class ModelPatcher:
     def apply_hooks(self, hooks: comfy.hooks.HookGroup, transformer_options: dict=None, force_apply=False):
         # TODO: return transformer_options dict with any additions from hooks
         if self.current_hooks == hooks and (not force_apply or (not self.is_clip and hooks is None)):
-            return {}
+            return comfy.hooks.create_transformer_options_from_hooks(self, hooks, transformer_options)
         self.patch_hooks(hooks=hooks)
         for callback in self.get_all_callbacks(CallbacksMP.ON_APPLY_HOOKS):
             callback(self, hooks)
-        return {}
+        return comfy.hooks.create_transformer_options_from_hooks(self, hooks, transformer_options)
 
     def patch_hooks(self, hooks: comfy.hooks.HookGroup):
         with self.use_ejected():

comfy/sampler_helpers.py

@@ -24,15 +24,13 @@ def get_models_from_cond(cond, model_type):
                 models += [c[model_type]]
     return models
 
-def get_hooks_from_cond(cond, hooks_dict: dict[comfy.hooks.EnumHookType, dict[comfy.hooks.Hook, None]]):
+def get_hooks_from_cond(cond, full_hooks: comfy.hooks.HookGroup):
     # get hooks from conds, and collect cnets so they can be checked for extra_hooks
     cnets: list[ControlBase] = []
     for c in cond:
         if 'hooks' in c:
             for hook in c['hooks'].hooks:
-                hook: comfy.hooks.Hook
-                with_type = hooks_dict.setdefault(hook.hook_type, {})
-                with_type[hook] = None
+                full_hooks.add(hook)
         if 'control' in c:
             cnets.append(c['control'])
 
@@ -50,10 +48,9 @@ def get_hooks_from_cond(cond, hooks_dict: dict[comfy.hooks.EnumHookType, dict[co
     extra_hooks = comfy.hooks.HookGroup.combine_all_hooks(hooks_list)
     if extra_hooks is not None:
         for hook in extra_hooks.hooks:
-            with_type = hooks_dict.setdefault(hook.hook_type, {})
-            with_type[hook] = None
+            full_hooks.add(hook)
 
-    return hooks_dict
+    return full_hooks
 
 def convert_cond(cond):
     out = []
@@ -73,13 +70,11 @@ def get_additional_models(conds, dtype):
     cnets: list[ControlBase] = []
     gligen = []
     add_models = []
-    hooks: dict[comfy.hooks.EnumHookType, dict[comfy.hooks.Hook, None]] = {}
 
     for k in conds:
         cnets += get_models_from_cond(conds[k], "control")
         gligen += get_models_from_cond(conds[k], "gligen")
         add_models += get_models_from_cond(conds[k], "additional_models")
-        get_hooks_from_cond(conds[k], hooks)
 
     control_nets = set(cnets)
 
@@ -90,11 +85,20 @@ def get_additional_models(conds, dtype):
         inference_memory += m.inference_memory_requirements(dtype)
 
     gligen = [x[1] for x in gligen]
-    hook_models = [x.model for x in hooks.get(comfy.hooks.EnumHookType.AddModels, {}).keys()]
-    models = control_models + gligen + add_models + hook_models
+    models = control_models + gligen + add_models
 
     return models, inference_memory
 
+def get_additional_models_from_model_options(model_options: dict[str]=None):
+    """loads additional models from registered AddModels hooks"""
+    models = []
+    if model_options is not None and "registered_hooks" in model_options:
+        registered: comfy.hooks.HookGroup = model_options["registered_hooks"]
+        for hook in registered.get_type(comfy.hooks.EnumHookType.AdditionalModels):
+            hook: comfy.hooks.AdditionalModelsHook
+            models.extend(hook.models)
+    return models
+
 def cleanup_additional_models(models):
     """cleanup additional models that were loaded"""
     for m in models:
@@ -102,9 +106,10 @@ def cleanup_additional_models(models):
             m.cleanup()
 
 
-def prepare_sampling(model: 'ModelPatcher', noise_shape, conds):
-    real_model: 'BaseModel' = None
+def prepare_sampling(model: ModelPatcher, noise_shape, conds, model_options=None):
+    real_model: BaseModel = None
     models, inference_memory = get_additional_models(conds, model.model_dtype())
+    models += get_additional_models_from_model_options(model_options)
     models += model.get_nested_additional_models()  # TODO: does this require inference_memory update?
     memory_required = model.memory_required([noise_shape[0] * 2] + list(noise_shape[1:])) + inference_memory
     minimum_memory_required = model.memory_required([noise_shape[0]] + list(noise_shape[1:])) + inference_memory
@@ -123,12 +128,35 @@ def cleanup_models(conds, models):
     cleanup_additional_models(set(control_cleanup))
 
 def prepare_model_patcher(model: 'ModelPatcher', conds, model_options: dict):
+    '''
+    Registers hooks from conds.
+    '''
     # check for hooks in conds - if not registered, see if can be applied
-    hooks = {}
+    hooks = comfy.hooks.HookGroup()
     for k in conds:
         get_hooks_from_cond(conds[k], hooks)
     # add wrappers and callbacks from ModelPatcher to transformer_options
     model_options["transformer_options"]["wrappers"] = comfy.patcher_extension.copy_nested_dicts(model.wrappers)
     model_options["transformer_options"]["callbacks"] = comfy.patcher_extension.copy_nested_dicts(model.callbacks)
-    # register hooks on model/model_options
-    model.register_all_hook_patches(hooks, comfy.hooks.EnumWeightTarget.Model, model_options)
+    # begin registering hooks
+    registered = comfy.hooks.HookGroup()
+    target_dict = comfy.hooks.create_target_dict(comfy.hooks.EnumWeightTarget.Model)
+    # handle all TransformerOptionsHooks
+    for hook in hooks.get_type(comfy.hooks.EnumHookType.TransformerOptions):
+        hook: comfy.hooks.TransformerOptionsHook
+        hook.add_hook_patches(model, model_options, target_dict, registered)
+    # handle all AddModelsHooks
+    for hook in hooks.get_type(comfy.hooks.EnumHookType.AdditionalModels):
+        hook: comfy.hooks.AdditionalModelsHook
+        hook.add_hook_patches(model, model_options, target_dict, registered)
+    # handle all WeightHooks by registering on ModelPatcher
+    model.register_all_hook_patches(hooks, target_dict, model_options, registered)
+    # add registered_hooks onto model_options for further reference
+    if len(registered) > 0:
+        model_options["registered_hooks"] = registered
+    # merge original wrappers and callbacks with hooked wrappers and callbacks
+    to_load_options: dict[str] = model_options.setdefault("to_load_options", {})
+    for wc_name in ["wrappers", "callbacks"]:
+        comfy.patcher_extension.merge_nested_dicts(to_load_options.setdefault(wc_name, {}), model_options["transformer_options"][wc_name],
+                                                    copy_dict1=False)
+    return to_load_options

comfy/samplers.py

@@ -810,6 +810,33 @@ def preprocess_conds_hooks(conds: dict[str, list[dict[str]]]):
             for cond in conds_to_modify:
                 cond['hooks'] = hooks
 
+def filter_registered_hooks_on_conds(conds: dict[str, list[dict[str]]], model_options: dict[str]):
+    '''Modify 'hooks' on conds so that only hooks that were registered remain. Properly accounts for
+    HookGroups that have the same reference.'''
+    registered: comfy.hooks.HookGroup = model_options.get('registered_hooks', None)
+    # if None were registered, make sure all hooks are cleaned from conds
+    if registered is None:
+        for k in conds:
+            for kk in conds[k]:
+                kk.pop('hooks', None)
+        return
+    # find conds that contain hooks to be replaced - group by common HookGroup refs
+    hook_replacement: dict[comfy.hooks.HookGroup, list[dict]] = {}
+    for k in conds:
+        for kk in conds[k]:
+            hooks: comfy.hooks.HookGroup = kk.get('hooks', None)
+            if hooks is not None:
+                if not hooks.is_subset_of(registered):
+                    to_replace = hook_replacement.setdefault(hooks, [])
+                    to_replace.append(kk)
+    # for each hook to replace, create a new proper HookGroup and assign to all common conds
+    for hooks, conds_to_modify in hook_replacement.items():
+        new_hooks = hooks.new_with_common_hooks(registered)
+        if len(new_hooks) == 0:
+            new_hooks = None
+        for kk in conds_to_modify:
+            kk['hooks'] = new_hooks
+
 
 def get_total_hook_groups_in_conds(conds: dict[str, list[dict[str]]]):
     hooks_set = set()
@@ -819,9 +846,58 @@ def get_total_hook_groups_in_conds(conds: dict[str, list[dict[str]]]):
     return len(hooks_set)
 
 
+def cast_to_load_options(model_options: dict[str], device=None, dtype=None):
+    '''
+    If any patches from hooks, wrappers, or callbacks have .to to be called, call it.
+    '''
+    if model_options is None:
+        return
+    to_load_options = model_options.get("to_load_options", None)
+    if to_load_options is None:
+        return
+
+    casts = []
+    if device is not None:
+        casts.append(device)
+    if dtype is not None:
+        casts.append(dtype)
+    # if nothing to apply, do nothing
+    if len(casts) == 0:
+        return
+
+    # try to call .to on patches
+    if "patches" in to_load_options:
+        patches = to_load_options["patches"]
+        for name in patches:
+            patch_list = patches[name]
+            for i in range(len(patch_list)):
+                if hasattr(patch_list[i], "to"):
+                    for cast in casts:
+                        patch_list[i] = patch_list[i].to(cast)
+    if "patches_replace" in to_load_options:
+        patches = to_load_options["patches_replace"]
+        for name in patches:
+            patch_list = patches[name]
+            for k in patch_list:
+                if hasattr(patch_list[k], "to"):
+                    for cast in casts:
+                        patch_list[k] = patch_list[k].to(cast)
+    # try to call .to on any wrappers/callbacks
+    wrappers_and_callbacks = ["wrappers", "callbacks"]
+    for wc_name in wrappers_and_callbacks:
+        if wc_name in to_load_options:
+            wc: dict[str, list] = to_load_options[wc_name]
+            for wc_dict in wc.values():
+                for wc_list in wc_dict.values():
+                    for i in range(len(wc_list)):
+                        if hasattr(wc_list[i], "to"):
+                            for cast in casts:
+                                wc_list[i] = wc_list[i].to(cast)
+
+
 class CFGGuider:
-    def __init__(self, model_patcher):
-        self.model_patcher: 'ModelPatcher' = model_patcher
+    def __init__(self, model_patcher: ModelPatcher):
+        self.model_patcher = model_patcher
         self.model_options = model_patcher.model_options
         self.original_conds = {}
         self.cfg = 1.0
@@ -861,7 +937,7 @@ class CFGGuider:
         return self.inner_model.process_latent_out(samples.to(torch.float32))
 
     def outer_sample(self, noise, latent_image, sampler, sigmas, denoise_mask=None, callback=None, disable_pbar=False, seed=None):
-        self.inner_model, self.conds, self.loaded_models = comfy.sampler_helpers.prepare_sampling(self.model_patcher, noise.shape, self.conds)
+        self.inner_model, self.conds, self.loaded_models = comfy.sampler_helpers.prepare_sampling(self.model_patcher, noise.shape, self.conds, self.model_options)
         device = self.model_patcher.load_device
 
         if denoise_mask is not None:
@@ -870,6 +946,7 @@ class CFGGuider:
         noise = noise.to(device)
         latent_image = latent_image.to(device)
         sigmas = sigmas.to(device)
+        cast_to_load_options(self.model_options, device=device, dtype=self.model_patcher.model_dtype())
 
         try:
             self.model_patcher.pre_run()
@@ -899,6 +976,7 @@ class CFGGuider:
             if get_total_hook_groups_in_conds(self.conds) <= 1:
                 self.model_patcher.hook_mode = comfy.hooks.EnumHookMode.MinVram
             comfy.sampler_helpers.prepare_model_patcher(self.model_patcher, self.conds, self.model_options)
+            filter_registered_hooks_on_conds(self.conds, self.model_options)
             executor = comfy.patcher_extension.WrapperExecutor.new_class_executor(
                 self.outer_sample,
                 self,
@@ -906,6 +984,7 @@ class CFGGuider:
             )
             output = executor.execute(noise, latent_image, sampler, sigmas, denoise_mask, callback, disable_pbar, seed)
         finally:
+            cast_to_load_options(self.model_options, device=self.model_patcher.offload_device)
             self.model_options = orig_model_options
             self.model_patcher.hook_mode = orig_hook_mode
             self.model_patcher.restore_hook_patches()

comfy/sd.py

@@ -11,6 +11,7 @@ from .ldm.cascade.stage_c_coder import StageC_coder
 from .ldm.audio.autoencoder import AudioOobleckVAE
 import comfy.ldm.genmo.vae.model
 import comfy.ldm.lightricks.vae.causal_video_autoencoder
+import comfy.ldm.cosmos.vae
 import yaml
 import math
 
@@ -34,6 +35,7 @@ import comfy.text_encoders.long_clipl
 import comfy.text_encoders.genmo
 import comfy.text_encoders.lt
 import comfy.text_encoders.hunyuan_video
+import comfy.text_encoders.cosmos
 
 import comfy.model_patcher
 import comfy.lora
@@ -376,6 +378,19 @@ class VAE:
                 self.memory_used_decode = lambda shape, dtype: (1500 * shape[2] * shape[3] * shape[4] * (4 * 8 * 8)) * model_management.dtype_size(dtype)
                 self.memory_used_encode = lambda shape, dtype: (900 * max(shape[2], 2) * shape[3] * shape[4]) * model_management.dtype_size(dtype)
                 self.working_dtypes = [torch.bfloat16, torch.float16, torch.float32]
+            elif "decoder.unpatcher3d.wavelets" in sd:
+                self.upscale_ratio = (lambda a: max(0, a * 8 - 7), 8, 8)
+                self.upscale_index_formula = (8, 8, 8)
+                self.downscale_ratio = (lambda a: max(0, math.floor((a + 7) / 8)), 8, 8)
+                self.downscale_index_formula = (8, 8, 8)
+                self.latent_dim = 3
+                self.latent_channels = 16
+                ddconfig = {'z_channels': 16, 'latent_channels': self.latent_channels, 'z_factor': 1, 'resolution': 1024, 'in_channels': 3, 'out_channels': 3, 'channels': 128, 'channels_mult': [2, 4, 4], 'num_res_blocks': 2, 'attn_resolutions': [32], 'dropout': 0.0, 'patch_size': 4, 'num_groups': 1, 'temporal_compression': 8, 'spacial_compression': 8}
+                self.first_stage_model = comfy.ldm.cosmos.vae.CausalContinuousVideoTokenizer(**ddconfig)
+                #TODO: these values are a bit off because this is not a standard VAE
+                self.memory_used_decode = lambda shape, dtype: (220 * shape[2] * shape[3] * shape[4] * (8 * 8 * 8)) * model_management.dtype_size(dtype)
+                self.memory_used_encode = lambda shape, dtype: (500 * max(shape[2], 2) * shape[3] * shape[4]) * model_management.dtype_size(dtype)
+                self.working_dtypes = [torch.bfloat16, torch.float32]
             else:
                 logging.warning("WARNING: No VAE weights detected, VAE not initalized.")
                 self.first_stage_model = None
@@ -641,6 +656,7 @@ class CLIPType(Enum):
     LTXV = 8
     HUNYUAN_VIDEO = 9
     PIXART = 10
+    COSMOS = 11
 
 
 def load_clip(ckpt_paths, embedding_directory=None, clip_type=CLIPType.STABLE_DIFFUSION, model_options={}):
@@ -658,6 +674,7 @@ class TEModel(Enum):
     T5_XL = 5
     T5_BASE = 6
     LLAMA3_8 = 7
+    T5_XXL_OLD = 8
 
 def detect_te_model(sd):
     if "text_model.encoder.layers.30.mlp.fc1.weight" in sd:
@@ -672,6 +689,8 @@ def detect_te_model(sd):
             return TEModel.T5_XXL
         elif weight.shape[-1] == 2048:
             return TEModel.T5_XL
+    if 'encoder.block.23.layer.1.DenseReluDense.wi.weight' in sd:
+        return TEModel.T5_XXL_OLD
     if "encoder.block.0.layer.0.SelfAttention.k.weight" in sd:
         return TEModel.T5_BASE
     if "model.layers.0.post_attention_layernorm.weight" in sd:
@@ -681,9 +700,10 @@ def detect_te_model(sd):
 
 def t5xxl_detect(clip_data):
     weight_name = "encoder.block.23.layer.1.DenseReluDense.wi_1.weight"
+    weight_name_old = "encoder.block.23.layer.1.DenseReluDense.wi.weight"
 
     for sd in clip_data:
-        if weight_name in sd:
+        if weight_name in sd or weight_name_old in sd:
             return comfy.text_encoders.sd3_clip.t5_xxl_detect(sd)
 
     return {}
@@ -740,6 +760,9 @@ def load_text_encoder_state_dicts(state_dicts=[], embedding_directory=None, clip
             else: #CLIPType.MOCHI
                 clip_target.clip = comfy.text_encoders.genmo.mochi_te(**t5xxl_detect(clip_data))
                 clip_target.tokenizer = comfy.text_encoders.genmo.MochiT5Tokenizer
+        elif te_model == TEModel.T5_XXL_OLD:
+            clip_target.clip = comfy.text_encoders.cosmos.te(**t5xxl_detect(clip_data))
+            clip_target.tokenizer = comfy.text_encoders.cosmos.CosmosT5Tokenizer
         elif te_model == TEModel.T5_XL:
             clip_target.clip = comfy.text_encoders.aura_t5.AuraT5Model
             clip_target.tokenizer = comfy.text_encoders.aura_t5.AuraT5Tokenizer

comfy/supported_models.py

@@ -14,6 +14,7 @@ import comfy.text_encoders.flux
 import comfy.text_encoders.genmo
 import comfy.text_encoders.lt
 import comfy.text_encoders.hunyuan_video
+import comfy.text_encoders.cosmos
 
 from . import supported_models_base
 from . import latent_formats
@@ -823,6 +824,37 @@ class HunyuanVideo(supported_models_base.BASE):
         hunyuan_detect = comfy.text_encoders.hunyuan_video.llama_detect(state_dict, "{}llama.transformer.".format(pref))
         return supported_models_base.ClipTarget(comfy.text_encoders.hunyuan_video.HunyuanVideoTokenizer, comfy.text_encoders.hunyuan_video.hunyuan_video_clip(**hunyuan_detect))
 
-models = [Stable_Zero123, SD15_instructpix2pix, SD15, SD20, SD21UnclipL, SD21UnclipH, SDXL_instructpix2pix, SDXLRefiner, SDXL, SSD1B, KOALA_700M, KOALA_1B, Segmind_Vega, SD_X4Upscaler, Stable_Cascade_C, Stable_Cascade_B, SV3D_u, SV3D_p, SD3, StableAudio, AuraFlow, PixArtAlpha, PixArtSigma, HunyuanDiT, HunyuanDiT1, FluxInpaint, Flux, FluxSchnell, GenmoMochi, LTXV, HunyuanVideo]
+class Cosmos(supported_models_base.BASE):
+    unet_config = {
+        "image_model": "cosmos",
+    }
+
+    sampling_settings = {
+        "sigma_data": 0.5,
+        "sigma_max": 80.0,
+        "sigma_min": 0.002,
+    }
+
+    unet_extra_config = {}
+    latent_format = latent_formats.Cosmos1CV8x8x8
+
+    memory_usage_factor = 2.4 #TODO
+
+    supported_inference_dtypes = [torch.bfloat16, torch.float16, torch.float32] #TODO
+
+    vae_key_prefix = ["vae."]
+    text_encoder_key_prefix = ["text_encoders."]
+
+    def get_model(self, state_dict, prefix="", device=None):
+        out = model_base.CosmosVideo(self, device=device)
+        return out
+
+    def clip_target(self, state_dict={}):
+        pref = self.text_encoder_key_prefix[0]
+        t5_detect = comfy.text_encoders.sd3_clip.t5_xxl_detect(state_dict, "{}t5xxl.transformer.".format(pref))
+        return supported_models_base.ClipTarget(comfy.text_encoders.cosmos.CosmosT5Tokenizer, comfy.text_encoders.cosmos.te(**t5_detect))
+
+
+models = [Stable_Zero123, SD15_instructpix2pix, SD15, SD20, SD21UnclipL, SD21UnclipH, SDXL_instructpix2pix, SDXLRefiner, SDXL, SSD1B, KOALA_700M, KOALA_1B, Segmind_Vega, SD_X4Upscaler, Stable_Cascade_C, Stable_Cascade_B, SV3D_u, SV3D_p, SD3, StableAudio, AuraFlow, PixArtAlpha, PixArtSigma, HunyuanDiT, HunyuanDiT1, FluxInpaint, Flux, FluxSchnell, GenmoMochi, LTXV, HunyuanVideo, Cosmos]
 
 models += [SVD_img2vid]

comfy/text_encoders/cosmos.py

@@ -0,0 +1,42 @@
+from comfy import sd1_clip
+import comfy.text_encoders.t5
+import os
+from transformers import T5TokenizerFast
+
+
+class T5XXLModel(sd1_clip.SDClipModel):
+    def __init__(self, device="cpu", layer="last", layer_idx=None, dtype=None, attention_mask=True, model_options={}):
+        textmodel_json_config = os.path.join(os.path.dirname(os.path.realpath(__file__)), "t5_old_config_xxl.json")
+        t5xxl_scaled_fp8 = model_options.get("t5xxl_scaled_fp8", None)
+        if t5xxl_scaled_fp8 is not None:
+            model_options = model_options.copy()
+            model_options["scaled_fp8"] = t5xxl_scaled_fp8
+
+        super().__init__(device=device, layer=layer, layer_idx=layer_idx, textmodel_json_config=textmodel_json_config, dtype=dtype, special_tokens={"end": 1, "pad": 0}, model_class=comfy.text_encoders.t5.T5, enable_attention_masks=attention_mask, return_attention_masks=attention_mask, zero_out_masked=attention_mask, model_options=model_options)
+
+class CosmosT5XXL(sd1_clip.SD1ClipModel):
+    def __init__(self, device="cpu", dtype=None, model_options={}):
+        super().__init__(device=device, dtype=dtype, name="t5xxl", clip_model=T5XXLModel, model_options=model_options)
+
+
+class T5XXLTokenizer(sd1_clip.SDTokenizer):
+    def __init__(self, embedding_directory=None, tokenizer_data={}):
+        tokenizer_path = os.path.join(os.path.dirname(os.path.realpath(__file__)), "t5_tokenizer")
+        super().__init__(tokenizer_path, embedding_directory=embedding_directory, pad_with_end=False, embedding_size=1024, embedding_key='t5xxl', tokenizer_class=T5TokenizerFast, has_start_token=False, pad_to_max_length=False, max_length=99999999, min_length=512)
+
+
+class CosmosT5Tokenizer(sd1_clip.SD1Tokenizer):
+    def __init__(self, embedding_directory=None, tokenizer_data={}):
+        super().__init__(embedding_directory=embedding_directory, tokenizer_data=tokenizer_data, clip_name="t5xxl", tokenizer=T5XXLTokenizer)
+
+
+def te(dtype_t5=None, t5xxl_scaled_fp8=None):
+    class CosmosTEModel_(CosmosT5XXL):
+        def __init__(self, device="cpu", dtype=None, model_options={}):
+            if t5xxl_scaled_fp8 is not None and "t5xxl_scaled_fp8" not in model_options:
+                model_options = model_options.copy()
+                model_options["t5xxl_scaled_fp8"] = t5xxl_scaled_fp8
+            if dtype is None:
+                dtype = dtype_t5
+            super().__init__(device=device, dtype=dtype, model_options=model_options)
+    return CosmosTEModel_

comfy/text_encoders/t5_old_config_xxl.json

@@ -0,0 +1,22 @@
+{
+  "d_ff": 65536,
+  "d_kv": 128,
+  "d_model": 1024,
+  "decoder_start_token_id": 0,
+  "dropout_rate": 0.1,
+  "eos_token_id": 1,
+  "dense_act_fn": "relu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": false,
+  "layer_norm_epsilon": 1e-06,
+  "model_type": "t5",
+  "num_decoder_layers": 24,
+  "num_heads": 128,
+  "num_layers": 24,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_num_buckets": 32,
+  "tie_word_embeddings": false,
+  "vocab_size": 32128
+}

comfy_extras/nodes_cosmos.py

@@ -0,0 +1,23 @@
+import nodes
+import torch
+import comfy.model_management
+
+class EmptyCosmosLatentVideo:
+    @classmethod
+    def INPUT_TYPES(s):
+        return {"required": { "width": ("INT", {"default": 1280, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 16}),
+                              "height": ("INT", {"default": 704, "min": 16, "max": nodes.MAX_RESOLUTION, "step": 16}),
+                              "length": ("INT", {"default": 121, "min": 1, "max": nodes.MAX_RESOLUTION, "step": 8}),
+                              "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096})}}
+    RETURN_TYPES = ("LATENT",)
+    FUNCTION = "generate"
+
+    CATEGORY = "latent/video"
+
+    def generate(self, width, height, length, batch_size=1):
+        latent = torch.zeros([batch_size, 16, ((length - 1) // 8) + 1, height // 8, width // 8], device=comfy.model_management.intermediate_device())
+        return ({"samples":latent}, )
+
+NODE_CLASS_MAPPINGS = {
+    "EmptyCosmosLatentVideo": EmptyCosmosLatentVideo,
+}

comfy_extras/nodes_hooks.py

@@ -246,7 +246,7 @@ class SetClipHooks:
     CATEGORY = "advanced/hooks/clip"
     FUNCTION = "apply_hooks"
 
-    def apply_hooks(self, clip: 'CLIP', schedule_clip: bool, apply_to_conds: bool, hooks: comfy.hooks.HookGroup=None):
+    def apply_hooks(self, clip: CLIP, schedule_clip: bool, apply_to_conds: bool, hooks: comfy.hooks.HookGroup=None):
         if hooks is not None:
             clip = clip.clone()
             if apply_to_conds:
@@ -255,7 +255,7 @@ class SetClipHooks:
             clip.use_clip_schedule = schedule_clip
             if not clip.use_clip_schedule:
                 clip.patcher.forced_hooks.set_keyframes_on_hooks(None)
-            clip.patcher.register_all_hook_patches(hooks.get_dict_repr(), comfy.hooks.EnumWeightTarget.Clip)
+            clip.patcher.register_all_hook_patches(hooks, comfy.hooks.create_target_dict(comfy.hooks.EnumWeightTarget.Clip))
         return (clip,)
 
 class ConditioningTimestepsRange:

nodes.py

@@ -912,7 +912,7 @@ class CLIPLoader:
     @classmethod
     def INPUT_TYPES(s):
         return {"required": { "clip_name": (folder_paths.get_filename_list("text_encoders"), ),
-                              "type": (["stable_diffusion", "stable_cascade", "sd3", "stable_audio", "mochi", "ltxv", "pixart"], ),
+                              "type": (["stable_diffusion", "stable_cascade", "sd3", "stable_audio", "mochi", "ltxv", "pixart", "cosmos"], ),
                               },
                 "optional": {
                               "device": (["default", "cpu"], {"advanced": True}),
@@ -922,7 +922,7 @@ class CLIPLoader:
 
     CATEGORY = "advanced/loaders"
 
-    DESCRIPTION = "[Recipes]\n\nstable_diffusion: clip-l\nstable_cascade: clip-g\nsd3: t5 / clip-g / clip-l\nstable_audio: t5\nmochi: t5"
+    DESCRIPTION = "[Recipes]\n\nstable_diffusion: clip-l\nstable_cascade: clip-g\nsd3: t5 / clip-g / clip-l\nstable_audio: t5\nmochi: t5\ncosmos: old t5 xxl"
 
     def load_clip(self, clip_name, type="stable_diffusion", device="default"):
         if type == "stable_cascade":
@@ -2225,6 +2225,7 @@ def init_builtin_extra_nodes():
         "nodes_lt.py",
         "nodes_hooks.py",
         "nodes_load_3d.py",
+        "nodes_cosmos.py",
     ]
 
     import_failed = []