Merge branch 'master' of github.com:comfyanonymous/ComfyUI

README.md

@@ -54,12 +54,13 @@ See what ComfyUI can do with the [example workflows](https://comfyanonymous.gith
     - [Flux](https://comfyanonymous.github.io/ComfyUI_examples/flux/)
     - [Lumina Image 2.0](https://comfyanonymous.github.io/ComfyUI_examples/lumina2/)
     - [HiDream](https://comfyanonymous.github.io/ComfyUI_examples/hidream/)
+   - [Cosmos Predict2](https://comfyanonymous.github.io/ComfyUI_examples/cosmos_predict2/)
 - Video Models
     - [Stable Video Diffusion](https://comfyanonymous.github.io/ComfyUI_examples/video/)
     - [Mochi](https://comfyanonymous.github.io/ComfyUI_examples/mochi/)
     - [LTX-Video](https://comfyanonymous.github.io/ComfyUI_examples/ltxv/)
     - [Hunyuan Video](https://comfyanonymous.github.io/ComfyUI_examples/hunyuan_video/)
-    - [Nvidia Cosmos](https://comfyanonymous.github.io/ComfyUI_examples/cosmos/)
+    - [Nvidia Cosmos](https://comfyanonymous.github.io/ComfyUI_examples/cosmos/) and [Cosmos Predict2](https://comfyanonymous.github.io/ComfyUI_examples/cosmos_predict2/)
     - [Wan 2.1](https://comfyanonymous.github.io/ComfyUI_examples/wan/)
 - Audio Models
    - [Stable Audio](https://comfyanonymous.github.io/ComfyUI_examples/audio/)

alembic.ini

@@ -0,0 +1,84 @@
+# A generic, single database configuration.
+
+[alembic]
+# path to migration scripts
+# Use forward slashes (/) also on windows to provide an os agnostic path
+script_location = alembic_db
+
+# template used to generate migration file names; The default value is %%(rev)s_%%(slug)s
+# Uncomment the line below if you want the files to be prepended with date and time
+# see https://alembic.sqlalchemy.org/en/latest/tutorial.html#editing-the-ini-file
+# for all available tokens
+# file_template = %%(year)d_%%(month).2d_%%(day).2d_%%(hour).2d%%(minute).2d-%%(rev)s_%%(slug)s
+
+# sys.path path, will be prepended to sys.path if present.
+# defaults to the current working directory.
+prepend_sys_path = .
+
+# timezone to use when rendering the date within the migration file
+# as well as the filename.
+# If specified, requires the python>=3.9 or backports.zoneinfo library and tzdata library.
+# Any required deps can installed by adding `alembic[tz]` to the pip requirements
+# string value is passed to ZoneInfo()
+# leave blank for localtime
+# timezone =
+
+# max length of characters to apply to the "slug" field
+# truncate_slug_length = 40
+
+# set to 'true' to run the environment during
+# the 'revision' command, regardless of autogenerate
+# revision_environment = false
+
+# set to 'true' to allow .pyc and .pyo files without
+# a source .py file to be detected as revisions in the
+# versions/ directory
+# sourceless = false
+
+# version location specification; This defaults
+# to alembic_db/versions.  When using multiple version
+# directories, initial revisions must be specified with --version-path.
+# The path separator used here should be the separator specified by "version_path_separator" below.
+# version_locations = %(here)s/bar:%(here)s/bat:alembic_db/versions
+
+# version path separator; As mentioned above, this is the character used to split
+# version_locations. The default within new alembic.ini files is "os", which uses os.pathsep.
+# If this key is omitted entirely, it falls back to the legacy behavior of splitting on spaces and/or commas.
+# Valid values for version_path_separator are:
+#
+# version_path_separator = :
+# version_path_separator = ;
+# version_path_separator = space
+# version_path_separator = newline
+#
+# Use os.pathsep. Default configuration used for new projects.
+version_path_separator = os
+
+# set to 'true' to search source files recursively
+# in each "version_locations" directory
+# new in Alembic version 1.10
+# recursive_version_locations = false
+
+# the output encoding used when revision files
+# are written from script.py.mako
+# output_encoding = utf-8
+
+sqlalchemy.url = sqlite:///user/comfyui.db
+
+
+[post_write_hooks]
+# post_write_hooks defines scripts or Python functions that are run
+# on newly generated revision scripts.  See the documentation for further
+# detail and examples
+
+# format using "black" - use the console_scripts runner, against the "black" entrypoint
+# hooks = black
+# black.type = console_scripts
+# black.entrypoint = black
+# black.options = -l 79 REVISION_SCRIPT_FILENAME
+
+# lint with attempts to fix using "ruff" - use the exec runner, execute a binary
+# hooks = ruff
+# ruff.type = exec
+# ruff.executable = %(here)s/.venv/bin/ruff
+# ruff.options = check --fix REVISION_SCRIPT_FILENAME

alembic_db/README.md

@@ -0,0 +1,4 @@
+## Generate new revision
+
+1. Update models in `/app/database/models.py`
+2. Run `alembic revision --autogenerate -m "{your message}"`

alembic_db/env.py

@@ -0,0 +1,64 @@
+from sqlalchemy import engine_from_config
+from sqlalchemy import pool
+
+from alembic import context
+
+# this is the Alembic Config object, which provides
+# access to the values within the .ini file in use.
+config = context.config
+
+
+from app.database.models import Base
+target_metadata = Base.metadata
+
+# other values from the config, defined by the needs of env.py,
+# can be acquired:
+# my_important_option = config.get_main_option("my_important_option")
+# ... etc.
+
+
+def run_migrations_offline() -> None:
+    """Run migrations in 'offline' mode.
+    This configures the context with just a URL
+    and not an Engine, though an Engine is acceptable
+    here as well.  By skipping the Engine creation
+    we don't even need a DBAPI to be available.
+    Calls to context.execute() here emit the given string to the
+    script output.
+    """
+    url = config.get_main_option("sqlalchemy.url")
+    context.configure(
+        url=url,
+        target_metadata=target_metadata,
+        literal_binds=True,
+        dialect_opts={"paramstyle": "named"},
+    )
+
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+def run_migrations_online() -> None:
+    """Run migrations in 'online' mode.
+    In this scenario we need to create an Engine
+    and associate a connection with the context.
+    """
+    connectable = engine_from_config(
+        config.get_section(config.config_ini_section, {}),
+        prefix="sqlalchemy.",
+        poolclass=pool.NullPool,
+    )
+
+    with connectable.connect() as connection:
+        context.configure(
+            connection=connection, target_metadata=target_metadata
+        )
+
+        with context.begin_transaction():
+            context.run_migrations()
+
+
+if context.is_offline_mode():
+    run_migrations_offline()
+else:
+    run_migrations_online()

alembic_db/script.py.mako

@@ -0,0 +1,28 @@
+"""${message}
+
+Revision ID: ${up_revision}
+Revises: ${down_revision | comma,n}
+Create Date: ${create_date}
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+${imports if imports else ""}
+
+# revision identifiers, used by Alembic.
+revision: str = ${repr(up_revision)}
+down_revision: Union[str, None] = ${repr(down_revision)}
+branch_labels: Union[str, Sequence[str], None] = ${repr(branch_labels)}
+depends_on: Union[str, Sequence[str], None] = ${repr(depends_on)}
+
+
+def upgrade() -> None:
+    """Upgrade schema."""
+    ${upgrades if upgrades else "pass"}
+
+
+def downgrade() -> None:
+    """Downgrade schema."""
+    ${downgrades if downgrades else "pass"}

app/database/db.py

@@ -0,0 +1,112 @@
+import logging
+import os
+import shutil
+from app.logger import log_startup_warning
+from utils.install_util import get_missing_requirements_message
+from comfy.cli_args import args
+
+_DB_AVAILABLE = False
+Session = None
+
+
+try:
+    from alembic import command
+    from alembic.config import Config
+    from alembic.runtime.migration import MigrationContext
+    from alembic.script import ScriptDirectory
+    from sqlalchemy import create_engine
+    from sqlalchemy.orm import sessionmaker
+
+    _DB_AVAILABLE = True
+except ImportError as e:
+    log_startup_warning(
+        f"""
+------------------------------------------------------------------------
+Error importing dependencies: {e}
+{get_missing_requirements_message()}
+This error is happening because ComfyUI now uses a local sqlite database.
+------------------------------------------------------------------------
+""".strip()
+    )
+
+
+def dependencies_available():
+    """
+    Temporary function to check if the dependencies are available
+    """
+    return _DB_AVAILABLE
+
+
+def can_create_session():
+    """
+    Temporary function to check if the database is available to create a session
+    During initial release there may be environmental issues (or missing dependencies) that prevent the database from being created
+    """
+    return dependencies_available() and Session is not None
+
+
+def get_alembic_config():
+    root_path = os.path.join(os.path.dirname(__file__), "../..")
+    config_path = os.path.abspath(os.path.join(root_path, "alembic.ini"))
+    scripts_path = os.path.abspath(os.path.join(root_path, "alembic_db"))
+
+    config = Config(config_path)
+    config.set_main_option("script_location", scripts_path)
+    config.set_main_option("sqlalchemy.url", args.database_url)
+
+    return config
+
+
+def get_db_path():
+    url = args.database_url
+    if url.startswith("sqlite:///"):
+        return url.split("///")[1]
+    else:
+        raise ValueError(f"Unsupported database URL '{url}'.")
+
+
+def init_db():
+    db_url = args.database_url
+    logging.debug(f"Database URL: {db_url}")
+    db_path = get_db_path()
+    db_exists = os.path.exists(db_path)
+
+    config = get_alembic_config()
+
+    # Check if we need to upgrade
+    engine = create_engine(db_url)
+    conn = engine.connect()
+
+    context = MigrationContext.configure(conn)
+    current_rev = context.get_current_revision()
+
+    script = ScriptDirectory.from_config(config)
+    target_rev = script.get_current_head()
+
+    if target_rev is None:
+        logging.warning("No target revision found.")
+    elif current_rev != target_rev:
+        # Backup the database pre upgrade
+        backup_path = db_path + ".bkp"
+        if db_exists:
+            shutil.copy(db_path, backup_path)
+        else:
+            backup_path = None
+
+        try:
+            command.upgrade(config, target_rev)
+            logging.info(f"Database upgraded from {current_rev} to {target_rev}")
+        except Exception as e:
+            if backup_path:
+                # Restore the database from backup if upgrade fails
+                shutil.copy(backup_path, db_path)
+                os.remove(backup_path)
+            logging.exception("Error upgrading database: ")
+            raise e
+
+    global Session
+    Session = sessionmaker(bind=engine)
+
+
+def create_session():
+    return Session()

app/database/models.py

@@ -0,0 +1,14 @@
+from sqlalchemy.orm import declarative_base
+
+Base = declarative_base()
+
+
+def to_dict(obj):
+    fields = obj.__table__.columns.keys()
+    return {
+        field: (val.to_dict() if hasattr(val, "to_dict") else val)
+        for field in fields
+        if (val := getattr(obj, field))
+    }
+
+# TODO: Define models here

comfy/__init__.py

@@ -1,3 +1,3 @@
 # This file is automatically generated by the build process when version is
 # updated in pyproject.toml.
-__version__ = "0.3.40"
+__version__ = "0.3.41"

comfy/app/frontend_management.py

@@ -79,9 +79,22 @@ class FrontEndProvider:
         response.raise_for_status()  # Raises an HTTPError if the response was an error
         return response.json()
 
+    @cached_property
+    def latest_prerelease(self) -> Release:
+        """Get the latest pre-release version - even if it's older than the latest release"""
+        release = [release for release in self.all_releases if release["prerelease"]]
+
+        if not release:
+            raise ValueError("No pre-releases found")
+
+        # GitHub returns releases in reverse chronological order, so first is latest
+        return release[0]
+
     def get_release(self, version: str) -> Release:
         if version == "latest":
             return self.latest_release
+        elif version == "prerelease":
+            return self.latest_prerelease
         else:
             for release in self.all_releases:
                 if release["tag_name"] in [version, f"v{version}"]:
@@ -176,7 +189,7 @@ comfyui-workflow-templates is not installed.
         Raises:
             argparse.ArgumentTypeError: If the version string is invalid.
         """
-        VERSION_PATTERN = r"^([a-zA-Z0-9][a-zA-Z0-9-]{0,38})/([a-zA-Z0-9_.-]+)@(v?\d+\.\d+\.\d+|latest)$"
+        VERSION_PATTERN = r"^([a-zA-Z0-9][a-zA-Z0-9-]{0,38})/([a-zA-Z0-9_.-]+)@(v?\d+\.\d+\.\d+[-._a-zA-Z0-9]*|latest|prerelease)$"
         match_result = re.match(VERSION_PATTERN, value)
         if match_result is None:
             raise argparse.ArgumentTypeError(f"Invalid version string: {value}")

comfy/cli_args.py

@@ -260,6 +260,11 @@ def _create_parser() -> EnhancedConfigArgParser:
         help="Set the base URL for the ComfyUI API.  (default: https://api.comfy.org)",
     )
 
+    database_default_path = os.path.abspath(
+        os.path.join(os.path.dirname(__file__), "..", "user", "comfyui.db")
+    )
+    parser.add_argument("--database-url", type=str, default=f"sqlite:///{database_default_path}", help="Specify the database URL, e.g. for an in-memory database you can use 'sqlite:///:memory:'.")
+
     parser.add_argument("--workflows", type=str, nargs='+', default=[], help="Execute the API workflow(s) specified in the provided files. For each workflow, its outputs will be printed to a line to standard out. Application logging will be redirected to standard error. Use `-` to signify standard in.")
 
     # now give plugins a chance to add configuration

comfy/cli_args_types.py

@@ -145,6 +145,7 @@ class Configuration(dict):
         front_end_version (str): Specifies the version of the frontend to be used.
         front_end_root (Optional[str]): The local filesystem path to the directory where the frontend is located. Overrides --front-end-version.
         comfy_api_base (str): Set the base URL for the ComfyUI API. (default: https://api.comfy.org)
+        database_url (str): Specify the database URL, e.g. for an in-memory database you can use 'sqlite:///:memory:'.
     """
 
     def __init__(self, **kwargs):
@@ -258,6 +259,8 @@ class Configuration(dict):
         self.front_end_version: str = "comfyanonymous/ComfyUI@latest"
         self.front_end_root: Optional[str] = None
         self.comfy_api_base: str = "https://api.comfy.org"
+        self.database_url: str = "sqlite:///:memory:"
+
         for key, value in kwargs.items():
             self[key] = value
         # this must always be last

comfy/cmd/folder_paths.py

@@ -1,10 +1,11 @@
 from __future__ import annotations
 
+import time
+
 import collections.abc
 import logging
 import mimetypes
 import os
-import time
 from contextlib import nullcontext
 from functools import reduce
 from pathlib import Path, PurePosixPath
@@ -447,6 +448,30 @@ def filter_files_content_types(files: list[str], content_types: list[Literal["im
     return result
 
 
+def get_input_subfolders() -> list[str]:
+    """Returns a list of all subfolder paths in the input directory, recursively.
+
+    Returns:
+        List of folder paths relative to the input directory, excluding the root directory
+    """
+    input_dir = get_input_directory()
+    folders = []
+
+    try:
+        if not os.path.exists(input_dir):
+            return []
+
+        for root, dirs, _ in os.walk(input_dir):
+            rel_path = os.path.relpath(root, input_dir)
+            if rel_path != ".":  # Only include non-root directories
+                # Normalize path separators to forward slashes
+                folders.append(rel_path.replace(os.sep, '/'))
+
+        return sorted(folders)
+    except FileNotFoundError:
+        return []
+
+
 @_module_properties.getter
 def _cache_helper():
     return nullcontext()
@@ -491,5 +516,6 @@ __all__ = [
     "get_save_image_path",
     "create_directories",
     "invalidate_cache",
-    "filter_files_content_types"
+    "filter_files_content_types",
+    "get_input_subfolders",
 ]

comfy/cmd/folder_paths.pyi

@@ -105,3 +105,6 @@ def invalidate_cache(folder_name: str) -> None: ...
 
 
 def filter_files_content_types(files: List[str], content_types: List[Literal["image", "video", "audio", "model"]]) -> List[str]: ...
+
+
+def get_input_subfolders() -> list[str]: ...

comfy/comfy_types/node_typing.py

@@ -37,6 +37,8 @@ class IO(StrEnum):
     CONTROL_NET = "CONTROL_NET"
     VAE = "VAE"
     MODEL = "MODEL"
+    LORA_MODEL = "LORA_MODEL"
+    LOSS_MAP = "LOSS_MAP"
     CLIP_VISION = "CLIP_VISION"
     CLIP_VISION_OUTPUT = "CLIP_VISION_OUTPUT"
     STYLE_MODEL = "STYLE_MODEL"

comfy/k_diffusion/sampling.py

@@ -1,4 +1,5 @@
 import math
+from functools import partial
 
 import torch
 import torchsde
@@ -143,6 +144,33 @@ class BrownianTreeNoiseSampler:
         return self.tree(t0, t1) / (t1 - t0).abs().sqrt()
 
 
+def sigma_to_half_log_snr(sigma, _model_sampling):
+    """Convert sigma to half-logSNR log(alpha_t / sigma_t)."""
+    if isinstance(_model_sampling, model_sampling.CONST):
+        # log((1 - t) / t) = log((1 - sigma) / sigma)
+        return sigma.logit().neg()
+    return sigma.log().neg()
+
+
+def half_log_snr_to_sigma(half_log_snr, _model_sampling):
+    """Convert half-logSNR log(alpha_t / sigma_t) to sigma."""
+    if isinstance(_model_sampling, model_sampling.CONST):
+        # 1 / (1 + exp(half_log_snr))
+        return half_log_snr.neg().sigmoid()
+    return half_log_snr.neg().exp()
+
+
+def offset_first_sigma_for_snr(sigmas, _model_sampling, percent_offset=1e-4):
+    """Adjust the first sigma to avoid invalid logSNR."""
+    if len(sigmas) <= 1:
+        return sigmas
+    if isinstance(_model_sampling, model_sampling.CONST):
+        if sigmas[0] >= 1:
+            sigmas = sigmas.clone()
+            sigmas[0] = _model_sampling.percent_to_sigma(percent_offset)
+    return sigmas
+
+
 @torch.no_grad()
 def sample_euler(model, x, sigmas, extra_args=None, callback=None, disable=None, s_churn=0., s_tmin=0., s_tmax=float('inf'), s_noise=1.):
     """Implements Algorithm 2 (Euler steps) from Karras et al. (2022)."""
@@ -774,8 +802,12 @@ def sample_dpmpp_2m_sde(model, x, sigmas, extra_args=None, callback=None, disabl
     noise_sampler = BrownianTreeNoiseSampler(x, sigma_min, sigma_max, seed=seed, cpu=True) if noise_sampler is None else noise_sampler
     s_in = x.new_ones([x.shape[0]])
 
+    model_sampling = model.inner_model.model_patcher.get_model_object('model_sampling')
+    lambda_fn = partial(sigma_to_half_log_snr, model_sampling=model_sampling)
+    sigmas = offset_first_sigma_for_snr(sigmas, model_sampling)
+
     old_denoised = None
-    h_last = None
+    h, h_last = None, None
 
     for i in trange(len(sigmas) - 1, disable=disable):
         denoised = model(x, sigmas[i] * s_in, **extra_args)
@@ -787,21 +819,23 @@ def sample_dpmpp_2m_sde(model, x, sigmas, extra_args=None, callback=None, disabl
             h = None
         else:
             # DPM-Solver++(2M) SDE
-            t, s = -sigmas[i].log(), -sigmas[i + 1].log()
-            h = s - t
-            eta_h = eta * h
+            lambda_s, lambda_t = lambda_fn(sigmas[i]), lambda_fn(sigmas[i + 1])
+            h = lambda_t - lambda_s
+            h_eta = h * (eta + 1)
 
-            x = sigmas[i + 1] / sigmas[i] * (-eta_h).exp() * x + (-h - eta_h).expm1().neg() * denoised
+            alpha_t = sigmas[i + 1] * lambda_t.exp()
+
+            x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x + alpha_t * (-h_eta).expm1().neg() * denoised
 
             if old_denoised is not None:
                 r = h_last / h
                 if solver_type == 'heun':
-                    x = x + ((-h - eta_h).expm1().neg() / (-h - eta_h) + 1) * (1 / r) * (denoised - old_denoised)
+                    x = x + alpha_t * ((-h_eta).expm1().neg() / (-h_eta) + 1) * (1 / r) * (denoised - old_denoised)
                 elif solver_type == 'midpoint':
-                    x = x + 0.5 * (-h - eta_h).expm1().neg() * (1 / r) * (denoised - old_denoised)
+                    x = x + 0.5 * alpha_t * (-h_eta).expm1().neg() * (1 / r) * (denoised - old_denoised)
 
-            if eta:
-                x = x + noise_sampler(sigmas[i], sigmas[i + 1]) * sigmas[i + 1] * (-2 * eta_h).expm1().neg().sqrt() * s_noise
+            if eta > 0 and s_noise > 0:
+                x = x + noise_sampler(sigmas[i], sigmas[i + 1]) * sigmas[i + 1] * (-2 * h * eta).expm1().neg().sqrt() * s_noise
 
         old_denoised = denoised
         h_last = h if h is not None else h_last
@@ -821,6 +855,10 @@ def sample_dpmpp_3m_sde(model, x, sigmas, extra_args=None, callback=None, disabl
     noise_sampler = BrownianTreeNoiseSampler(x, sigma_min, sigma_max, seed=seed, cpu=True) if noise_sampler is None else noise_sampler
     s_in = x.new_ones([x.shape[0]])
 
+    model_sampling = model.inner_model.model_patcher.get_model_object('model_sampling')
+    lambda_fn = partial(sigma_to_half_log_snr, model_sampling=model_sampling)
+    sigmas = offset_first_sigma_for_snr(sigmas, model_sampling)
+
     denoised_1, denoised_2 = None, None
     h, h_1, h_2 = None, None, None
 
@@ -832,13 +870,16 @@ def sample_dpmpp_3m_sde(model, x, sigmas, extra_args=None, callback=None, disabl
             # Denoising step
             x = denoised
         else:
-            t, s = -sigmas[i].log(), -sigmas[i + 1].log()
-            h = s - t
+            lambda_s, lambda_t = lambda_fn(sigmas[i]), lambda_fn(sigmas[i + 1])
+            h = lambda_t - lambda_s
             h_eta = h * (eta + 1)
 
-            x = torch.exp(-h_eta) * x + (-h_eta).expm1().neg() * denoised
+            alpha_t = sigmas[i + 1] * lambda_t.exp()
+
+            x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x + alpha_t * (-h_eta).expm1().neg() * denoised
 
             if h_2 is not None:
+                # DPM-Solver++(3M) SDE
                 r0 = h_1 / h
                 r1 = h_2 / h
                 d1_0 = (denoised - denoised_1) / r0
@@ -847,14 +888,15 @@ def sample_dpmpp_3m_sde(model, x, sigmas, extra_args=None, callback=None, disabl
                 d2 = (d1_0 - d1_1) / (r0 + r1)
                 phi_2 = h_eta.neg().expm1() / h_eta + 1
                 phi_3 = phi_2 / h_eta - 0.5
-                x = x + phi_2 * d1 - phi_3 * d2
+                x = x + (alpha_t * phi_2) * d1 - (alpha_t * phi_3) * d2
             elif h_1 is not None:
+                # DPM-Solver++(2M) SDE
                 r = h_1 / h
                 d = (denoised - denoised_1) / r
                 phi_2 = h_eta.neg().expm1() / h_eta + 1
-                x = x + phi_2 * d
+                x = x + (alpha_t * phi_2) * d
 
-            if eta:
+            if eta > 0 and s_noise > 0:
                 x = x + noise_sampler(sigmas[i], sigmas[i + 1]) * sigmas[i + 1] * (-2 * h * eta).expm1().neg().sqrt() * s_noise
 
         denoised_1, denoised_2 = denoised, denoised_1
@@ -1493,12 +1535,12 @@ def sample_er_sde(model, x, sigmas, extra_args=None, callback=None, disable=None
         old_denoised = denoised
     return x
 
+
 @torch.no_grad()
 def sample_seeds_2(model, x, sigmas, extra_args=None, callback=None, disable=None, eta=1., s_noise=1., noise_sampler=None, r=0.5):
-    '''
-    SEEDS-2 - Stochastic Explicit Exponential Derivative-free Solvers (VE Data Prediction) stage 2
-    Arxiv: https://arxiv.org/abs/2305.14267
-    '''
+    """SEEDS-2 - Stochastic Explicit Exponential Derivative-free Solvers (VP Data Prediction) stage 2.
+    arXiv: https://arxiv.org/abs/2305.14267
+    """
     extra_args = {} if extra_args is None else extra_args
     seed = extra_args.get("seed", None)
     noise_sampler = default_noise_sampler(x, seed=seed) if noise_sampler is None else noise_sampler
@@ -1506,6 +1548,11 @@ def sample_seeds_2(model, x, sigmas, extra_args=None, callback=None, disable=Non
 
     inject_noise = eta > 0 and s_noise > 0
 
+    model_sampling = model.inner_model.model_patcher.get_model_object('model_sampling')
+    sigma_fn = partial(half_log_snr_to_sigma, model_sampling=model_sampling)
+    lambda_fn = partial(sigma_to_half_log_snr, model_sampling=model_sampling)
+    sigmas = offset_first_sigma_for_snr(sigmas, model_sampling)
+
     for i in trange(len(sigmas) - 1, disable=disable):
         denoised = model(x, sigmas[i] * s_in, **extra_args)
         if callback is not None:
@@ -1513,80 +1560,96 @@ def sample_seeds_2(model, x, sigmas, extra_args=None, callback=None, disable=Non
         if sigmas[i + 1] == 0:
             x = denoised
         else:
-            t, t_next = -sigmas[i].log(), -sigmas[i + 1].log()
-            h = t_next - t
+            lambda_s, lambda_t = lambda_fn(sigmas[i]), lambda_fn(sigmas[i + 1])
+            h = lambda_t - lambda_s
             h_eta = h * (eta + 1)
-            s = t + r * h
+            lambda_s_1 = lambda_s + r * h
             fac = 1 / (2 * r)
-            sigma_s = s.neg().exp()
+            sigma_s_1 = sigma_fn(lambda_s_1)
+
+            # alpha_t = sigma_t * exp(log(alpha_t / sigma_t)) = sigma_t * exp(lambda_t)
+            alpha_s_1 = sigma_s_1 * lambda_s_1.exp()
+            alpha_t = sigmas[i + 1] * lambda_t.exp()
 
             coeff_1, coeff_2 = (-r * h_eta).expm1(), (-h_eta).expm1()
             if inject_noise:
+                # 0 < r < 1
                 noise_coeff_1 = (-2 * r * h * eta).expm1().neg().sqrt()
-                noise_coeff_2 = ((-2 * r * h * eta).expm1() - (-2 * h * eta).expm1()).sqrt()
-                noise_1, noise_2 = noise_sampler(sigmas[i], sigma_s), noise_sampler(sigma_s, sigmas[i + 1])
+                noise_coeff_2 = (-r * h * eta).exp() * (-2 * (1 - r) * h * eta).expm1().neg().sqrt()
+                noise_1, noise_2 = noise_sampler(sigmas[i], sigma_s_1), noise_sampler(sigma_s_1, sigmas[i + 1])
 
             # Step 1
-            x_2 = (coeff_1 + 1) * x - coeff_1 * denoised
-            if inject_noise:
-                x_2 = x_2 + sigma_s * (noise_coeff_1 * noise_1) * s_noise
-            denoised_2 = model(x_2, sigma_s * s_in, **extra_args)
-
-            # Step 2
-            denoised_d = (1 - fac) * denoised + fac * denoised_2
-            x = (coeff_2 + 1) * x - coeff_2 * denoised_d
-            if inject_noise:
-                x = x + sigmas[i + 1] * (noise_coeff_2 * noise_1 + noise_coeff_1 * noise_2) * s_noise
-    return x
-
-@torch.no_grad()
-def sample_seeds_3(model, x, sigmas, extra_args=None, callback=None, disable=None, eta=1., s_noise=1., noise_sampler=None, r_1=1./3, r_2=2./3):
-    '''
-    SEEDS-3 - Stochastic Explicit Exponential Derivative-free Solvers (VE Data Prediction) stage 3
-    Arxiv: https://arxiv.org/abs/2305.14267
-    '''
-    extra_args = {} if extra_args is None else extra_args
-    seed = extra_args.get("seed", None)
-    noise_sampler = default_noise_sampler(x, seed=seed) if noise_sampler is None else noise_sampler
-    s_in = x.new_ones([x.shape[0]])
-
-    inject_noise = eta > 0 and s_noise > 0
-
-    for i in trange(len(sigmas) - 1, disable=disable):
-        denoised = model(x, sigmas[i] * s_in, **extra_args)
-        if callback is not None:
-            callback({'x': x, 'i': i, 'sigma': sigmas[i], 'sigma_hat': sigmas[i], 'denoised': denoised})
-        if sigmas[i + 1] == 0:
-            x = denoised
-        else:
-            t, t_next = -sigmas[i].log(), -sigmas[i + 1].log()
-            h = t_next - t
-            h_eta = h * (eta + 1)
-            s_1 = t + r_1 * h
-            s_2 = t + r_2 * h
-            sigma_s_1, sigma_s_2 = s_1.neg().exp(), s_2.neg().exp()
-
-            coeff_1, coeff_2, coeff_3 = (-r_1 * h_eta).expm1(), (-r_2 * h_eta).expm1(), (-h_eta).expm1()
-            if inject_noise:
-                noise_coeff_1 = (-2 * r_1 * h * eta).expm1().neg().sqrt()
-                noise_coeff_2 = ((-2 * r_1 * h * eta).expm1() - (-2 * r_2 * h * eta).expm1()).sqrt()
-                noise_coeff_3 = ((-2 * r_2 * h * eta).expm1() - (-2 * h * eta).expm1()).sqrt()
-                noise_1, noise_2, noise_3 = noise_sampler(sigmas[i], sigma_s_1), noise_sampler(sigma_s_1, sigma_s_2), noise_sampler(sigma_s_2, sigmas[i + 1])
-
-            # Step 1
-            x_2 = (coeff_1 + 1) * x - coeff_1 * denoised
+            x_2 = sigma_s_1 / sigmas[i] * (-r * h * eta).exp() * x - alpha_s_1 * coeff_1 * denoised
             if inject_noise:
                 x_2 = x_2 + sigma_s_1 * (noise_coeff_1 * noise_1) * s_noise
             denoised_2 = model(x_2, sigma_s_1 * s_in, **extra_args)
 
             # Step 2
-            x_3 = (coeff_2 + 1) * x - coeff_2 * denoised + (r_2 / r_1) * (coeff_2 / (r_2 * h_eta) + 1) * (denoised_2 - denoised)
+            denoised_d = (1 - fac) * denoised + fac * denoised_2
+            x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x - alpha_t * coeff_2 * denoised_d
+            if inject_noise:
+                x = x + sigmas[i + 1] * (noise_coeff_2 * noise_1 + noise_coeff_1 * noise_2) * s_noise
+    return x
+
+
+@torch.no_grad()
+def sample_seeds_3(model, x, sigmas, extra_args=None, callback=None, disable=None, eta=1., s_noise=1., noise_sampler=None, r_1=1./3, r_2=2./3):
+    """SEEDS-3 - Stochastic Explicit Exponential Derivative-free Solvers (VP Data Prediction) stage 3.
+    arXiv: https://arxiv.org/abs/2305.14267
+    """
+    extra_args = {} if extra_args is None else extra_args
+    seed = extra_args.get("seed", None)
+    noise_sampler = default_noise_sampler(x, seed=seed) if noise_sampler is None else noise_sampler
+    s_in = x.new_ones([x.shape[0]])
+
+    inject_noise = eta > 0 and s_noise > 0
+
+    model_sampling = model.inner_model.model_patcher.get_model_object('model_sampling')
+    sigma_fn = partial(half_log_snr_to_sigma, model_sampling=model_sampling)
+    lambda_fn = partial(sigma_to_half_log_snr, model_sampling=model_sampling)
+    sigmas = offset_first_sigma_for_snr(sigmas, model_sampling)
+
+    for i in trange(len(sigmas) - 1, disable=disable):
+        denoised = model(x, sigmas[i] * s_in, **extra_args)
+        if callback is not None:
+            callback({'x': x, 'i': i, 'sigma': sigmas[i], 'sigma_hat': sigmas[i], 'denoised': denoised})
+        if sigmas[i + 1] == 0:
+            x = denoised
+        else:
+            lambda_s, lambda_t = lambda_fn(sigmas[i]), lambda_fn(sigmas[i + 1])
+            h = lambda_t - lambda_s
+            h_eta = h * (eta + 1)
+            lambda_s_1 = lambda_s + r_1 * h
+            lambda_s_2 = lambda_s + r_2 * h
+            sigma_s_1, sigma_s_2 = sigma_fn(lambda_s_1), sigma_fn(lambda_s_2)
+
+            # alpha_t = sigma_t * exp(log(alpha_t / sigma_t)) = sigma_t * exp(lambda_t)
+            alpha_s_1 = sigma_s_1 * lambda_s_1.exp()
+            alpha_s_2 = sigma_s_2 * lambda_s_2.exp()
+            alpha_t = sigmas[i + 1] * lambda_t.exp()
+
+            coeff_1, coeff_2, coeff_3 = (-r_1 * h_eta).expm1(), (-r_2 * h_eta).expm1(), (-h_eta).expm1()
+            if inject_noise:
+                # 0 < r_1 < r_2 < 1
+                noise_coeff_1 = (-2 * r_1 * h * eta).expm1().neg().sqrt()
+                noise_coeff_2 = (-r_1 * h * eta).exp() * (-2 * (r_2 - r_1) * h * eta).expm1().neg().sqrt()
+                noise_coeff_3 = (-r_2 * h * eta).exp() * (-2 * (1 - r_2) * h * eta).expm1().neg().sqrt()
+                noise_1, noise_2, noise_3 = noise_sampler(sigmas[i], sigma_s_1), noise_sampler(sigma_s_1, sigma_s_2), noise_sampler(sigma_s_2, sigmas[i + 1])
+
+            # Step 1
+            x_2 = sigma_s_1 / sigmas[i] * (-r_1 * h * eta).exp() * x - alpha_s_1 * coeff_1 * denoised
+            if inject_noise:
+                x_2 = x_2 + sigma_s_1 * (noise_coeff_1 * noise_1) * s_noise
+            denoised_2 = model(x_2, sigma_s_1 * s_in, **extra_args)
+
+            # Step 2
+            x_3 = sigma_s_2 / sigmas[i] * (-r_2 * h * eta).exp() * x - alpha_s_2 * coeff_2 * denoised + (r_2 / r_1) * alpha_s_2 * (coeff_2 / (r_2 * h_eta) + 1) * (denoised_2 - denoised)
             if inject_noise:
                 x_3 = x_3 + sigma_s_2 * (noise_coeff_2 * noise_1 + noise_coeff_1 * noise_2) * s_noise
             denoised_3 = model(x_3, sigma_s_2 * s_in, **extra_args)
 
             # Step 3
-            x = (coeff_3 + 1) * x - coeff_3 * denoised + (1. / r_2) * (coeff_3 / h_eta + 1) * (denoised_3 - denoised)
+            x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x - alpha_t * coeff_3 * denoised + (1. / r_2) * alpha_t * (coeff_3 / h_eta + 1) * (denoised_3 - denoised)
             if inject_noise:
                 x = x + sigmas[i + 1] * (noise_coeff_3 * noise_1 + noise_coeff_2 * noise_2 + noise_coeff_1 * noise_3) * s_noise
     return x

comfy/ldm/cosmos/blocks.py

@@ -26,16 +26,6 @@ from torch import nn
 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={}, operations=None):
     if name == "I":
         return nn.Identity()

comfy/ldm/cosmos/position_embedding.py

@@ -66,15 +66,16 @@ class VideoRopePosition3DEmb(VideoPositionEmb):
         h_extrapolation_ratio: float = 1.0,
         w_extrapolation_ratio: float = 1.0,
         t_extrapolation_ratio: float = 1.0,
+        enable_fps_modulation: bool = True,
         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
+        self.enable_fps_modulation = enable_fps_modulation
 
         dim = head_dim
         dim_h = dim // 6 * 2
@@ -132,21 +133,19 @@ class VideoRopePosition3DEmb(VideoPositionEmb):
         temporal_freqs = 1.0 / (t_theta**self.dim_temporal_range.to(device=device))
 
         B, T, H, W, _ = B_T_H_W_C
+        seq = torch.arange(max(H, W, T), dtype=torch.float, device=device)
         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)
+        half_emb_h = torch.outer(seq[:H].to(device=device), h_spatial_freqs)
+        half_emb_w = torch.outer(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)
+        if fps is None or self.enable_fps_modulation is False:  # image case
+            half_emb_t = torch.outer(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_t = torch.outer(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)

comfy/ldm/cosmos/predict2.py

@@ -0,0 +1,864 @@
+# original code from: https://github.com/nvidia-cosmos/cosmos-predict2
+
+import torch
+from torch import nn
+from einops import rearrange
+from einops.layers.torch import Rearrange
+import logging
+from typing import Callable, Optional, Tuple
+import math
+
+from .position_embedding import VideoRopePosition3DEmb, LearnablePosEmbAxis
+from torchvision import transforms
+
+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
+
+
+# ---------------------- Feed Forward Network -----------------------
+class GPT2FeedForward(nn.Module):
+    def __init__(self, d_model: int, d_ff: int, device=None, dtype=None, operations=None) -> None:
+        super().__init__()
+        self.activation = nn.GELU()
+        self.layer1 = operations.Linear(d_model, d_ff, bias=False, device=device, dtype=dtype)
+        self.layer2 = operations.Linear(d_ff, d_model, bias=False, device=device, dtype=dtype)
+
+        self._layer_id = None
+        self._dim = d_model
+        self._hidden_dim = d_ff
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        x = self.layer1(x)
+
+        x = self.activation(x)
+        x = self.layer2(x)
+        return x
+
+
+def torch_attention_op(q_B_S_H_D: torch.Tensor, k_B_S_H_D: torch.Tensor, v_B_S_H_D: torch.Tensor) -> torch.Tensor:
+    """Computes multi-head attention using PyTorch's native implementation.
+
+    This function provides a PyTorch backend alternative to Transformer Engine's attention operation.
+    It rearranges the input tensors to match PyTorch's expected format, computes scaled dot-product
+    attention, and rearranges the output back to the original format.
+
+    The input tensor names use the following dimension conventions:
+
+    - B: batch size
+    - S: sequence length
+    - H: number of attention heads
+    - D: head dimension
+
+    Args:
+        q_B_S_H_D: Query tensor with shape (batch, seq_len, n_heads, head_dim)
+        k_B_S_H_D: Key tensor with shape (batch, seq_len, n_heads, head_dim)
+        v_B_S_H_D: Value tensor with shape (batch, seq_len, n_heads, head_dim)
+
+    Returns:
+        Attention output tensor with shape (batch, seq_len, n_heads * head_dim)
+    """
+    in_q_shape = q_B_S_H_D.shape
+    in_k_shape = k_B_S_H_D.shape
+    q_B_H_S_D = rearrange(q_B_S_H_D, "b ... h k -> b h ... k").view(in_q_shape[0], in_q_shape[-2], -1, in_q_shape[-1])
+    k_B_H_S_D = rearrange(k_B_S_H_D, "b ... h v -> b h ... v").view(in_k_shape[0], in_k_shape[-2], -1, in_k_shape[-1])
+    v_B_H_S_D = rearrange(v_B_S_H_D, "b ... h v -> b h ... v").view(in_k_shape[0], in_k_shape[-2], -1, in_k_shape[-1])
+    return optimized_attention(q_B_H_S_D, k_B_H_S_D, v_B_H_S_D, in_q_shape[-2], skip_reshape=True)
+
+
+class Attention(nn.Module):
+    """
+    A flexible attention module supporting both self-attention and cross-attention mechanisms.
+
+    This module implements a multi-head attention layer that can operate in either self-attention
+    or cross-attention mode. The mode is determined by whether a context dimension is provided.
+    The implementation uses scaled dot-product attention and supports optional bias terms and
+    dropout regularization.
+
+    Args:
+        query_dim (int): The dimensionality of the query vectors.
+        context_dim (int, optional): The dimensionality of the context (key/value) vectors.
+            If None, the module operates in self-attention mode using query_dim. Default: None
+        n_heads (int, optional): Number of attention heads for multi-head attention. Default: 8
+        head_dim (int, optional): The dimension of each attention head. Default: 64
+        dropout (float, optional): Dropout probability applied to the output. Default: 0.0
+        qkv_format (str, optional): Format specification for QKV tensors. Default: "bshd"
+        backend (str, optional): Backend to use for the attention operation. Default: "transformer_engine"
+
+    Examples:
+        >>> # Self-attention with 512 dimensions and 8 heads
+        >>> self_attn = Attention(query_dim=512)
+        >>> x = torch.randn(32, 16, 512)  # (batch_size, seq_len, dim)
+        >>> out = self_attn(x)  # (32, 16, 512)
+
+        >>> # Cross-attention
+        >>> cross_attn = Attention(query_dim=512, context_dim=256)
+        >>> query = torch.randn(32, 16, 512)
+        >>> context = torch.randn(32, 8, 256)
+        >>> out = cross_attn(query, context)  # (32, 16, 512)
+    """
+
+    def __init__(
+        self,
+        query_dim: int,
+        context_dim: Optional[int] = None,
+        n_heads: int = 8,
+        head_dim: int = 64,
+        dropout: float = 0.0,
+        device=None,
+        dtype=None,
+        operations=None,
+    ) -> None:
+        super().__init__()
+        logging.debug(
+            f"Setting up {self.__class__.__name__}. Query dim is {query_dim}, context_dim is {context_dim} and using "
+            f"{n_heads} heads with a dimension of {head_dim}."
+        )
+        self.is_selfattn = context_dim is None  # self attention
+
+        context_dim = query_dim if context_dim is None else context_dim
+        inner_dim = head_dim * n_heads
+
+        self.n_heads = n_heads
+        self.head_dim = head_dim
+        self.query_dim = query_dim
+        self.context_dim = context_dim
+
+        self.q_proj = operations.Linear(query_dim, inner_dim, bias=False, device=device, dtype=dtype)
+        self.q_norm = operations.RMSNorm(self.head_dim, eps=1e-6, device=device, dtype=dtype)
+
+        self.k_proj = operations.Linear(context_dim, inner_dim, bias=False, device=device, dtype=dtype)
+        self.k_norm = operations.RMSNorm(self.head_dim, eps=1e-6, device=device, dtype=dtype)
+
+        self.v_proj = operations.Linear(context_dim, inner_dim, bias=False, device=device, dtype=dtype)
+        self.v_norm = nn.Identity()
+
+        self.output_proj = operations.Linear(inner_dim, query_dim, bias=False, device=device, dtype=dtype)
+        self.output_dropout = nn.Dropout(dropout) if dropout > 1e-4 else nn.Identity()
+
+        self.attn_op = torch_attention_op
+
+        self._query_dim = query_dim
+        self._context_dim = context_dim
+        self._inner_dim = inner_dim
+
+    def compute_qkv(
+        self,
+        x: torch.Tensor,
+        context: Optional[torch.Tensor] = None,
+        rope_emb: Optional[torch.Tensor] = None,
+    ) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        q = self.q_proj(x)
+        context = x if context is None else context
+        k = self.k_proj(context)
+        v = self.v_proj(context)
+        q, k, v = map(
+            lambda t: rearrange(t, "b ... (h d) -> b ... h d", h=self.n_heads, d=self.head_dim),
+            (q, k, v),
+        )
+
+        def apply_norm_and_rotary_pos_emb(
+            q: torch.Tensor, k: torch.Tensor, v: torch.Tensor, rope_emb: Optional[torch.Tensor]
+        ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+            q = self.q_norm(q)
+            k = self.k_norm(k)
+            v = self.v_norm(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
+
+        q, k, v = apply_norm_and_rotary_pos_emb(q, k, v, rope_emb)
+
+        return q, k, v
+
+    def compute_attention(self, q: torch.Tensor, k: torch.Tensor, v: torch.Tensor) -> torch.Tensor:
+        result = self.attn_op(q, k, v)  # [B, S, H, D]
+        return self.output_dropout(self.output_proj(result))
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        context: Optional[torch.Tensor] = None,
+        rope_emb: Optional[torch.Tensor] = None,
+    ) -> torch.Tensor:
+        """
+        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.compute_qkv(x, context, rope_emb=rope_emb)
+        return self.compute_attention(q, k, v)
+
+
+class Timesteps(nn.Module):
+    def __init__(self, num_channels: int):
+        super().__init__()
+        self.num_channels = num_channels
+
+    def forward(self, timesteps_B_T: torch.Tensor) -> torch.Tensor:
+        assert timesteps_B_T.ndim == 2, f"Expected 2D input, got {timesteps_B_T.ndim}"
+        timesteps = timesteps_B_T.flatten().float()
+        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 rearrange(emb, "(b t) d -> b t d", b=timesteps_B_T.shape[0], t=timesteps_B_T.shape[1])
+
+
+class TimestepEmbedding(nn.Module):
+    def __init__(self, in_features: int, out_features: int, use_adaln_lora: bool = False, device=None, dtype=None, 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.in_dim = in_features
+        self.out_dim = out_features
+        self.linear_1 = operations.Linear(in_features, out_features, bias=not use_adaln_lora, device=device, dtype=dtype)
+        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, device=device, dtype=dtype)
+        else:
+            self.linear_2 = operations.Linear(out_features, out_features, bias=False, device=device, dtype=dtype)
+
+    def forward(self, sample: torch.Tensor) -> Tuple[torch.Tensor, Optional[torch.Tensor]]:
+        emb = self.linear_1(sample)
+        emb = self.activation(emb)
+        emb = self.linear_2(emb)
+
+        if self.use_adaln_lora:
+            adaln_lora_B_T_3D = emb
+            emb_B_T_D = sample
+        else:
+            adaln_lora_B_T_3D = None
+            emb_B_T_D = emb
+
+        return emb_B_T_D, adaln_lora_B_T_3D
+
+
+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: int,
+        temporal_patch_size: int,
+        in_channels: int = 3,
+        out_channels: int = 768,
+        device=None, dtype=None, 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=False, device=device, dtype=dtype
+            ),
+        )
+        self.dim = in_channels * spatial_patch_size * spatial_patch_size * temporal_patch_size
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        """
+        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
+        ), f"H,W {(H, W)} should be divisible by spatial_patch_size {self.spatial_patch_size}"
+        assert T % self.temporal_patch_size == 0
+        x = self.proj(x)
+        return x
+
+
+class FinalLayer(nn.Module):
+    """
+    The final layer of video DiT.
+    """
+
+    def __init__(
+        self,
+        hidden_size: int,
+        spatial_patch_size: int,
+        temporal_patch_size: int,
+        out_channels: int,
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        device=None, dtype=None, operations=None
+    ):
+        super().__init__()
+        self.layer_norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.linear = operations.Linear(
+            hidden_size, spatial_patch_size * spatial_patch_size * temporal_patch_size * out_channels, bias=False, device=device, dtype=dtype
+        )
+        self.hidden_size = hidden_size
+        self.n_adaln_chunks = 2
+        self.use_adaln_lora = use_adaln_lora
+        self.adaln_lora_dim = adaln_lora_dim
+        if use_adaln_lora:
+            self.adaln_modulation = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(hidden_size, adaln_lora_dim, bias=False, device=device, dtype=dtype),
+                operations.Linear(adaln_lora_dim, self.n_adaln_chunks * hidden_size, bias=False, device=device, dtype=dtype),
+            )
+        else:
+            self.adaln_modulation = nn.Sequential(
+                nn.SiLU(), operations.Linear(hidden_size, self.n_adaln_chunks * hidden_size, bias=False, device=device, dtype=dtype)
+            )
+
+    def forward(
+        self,
+        x_B_T_H_W_D: torch.Tensor,
+        emb_B_T_D: torch.Tensor,
+        adaln_lora_B_T_3D: Optional[torch.Tensor] = None,
+    ):
+        if self.use_adaln_lora:
+            assert adaln_lora_B_T_3D is not None
+            shift_B_T_D, scale_B_T_D = (
+                self.adaln_modulation(emb_B_T_D) + adaln_lora_B_T_3D[:, :, : 2 * self.hidden_size]
+            ).chunk(2, dim=-1)
+        else:
+            shift_B_T_D, scale_B_T_D = self.adaln_modulation(emb_B_T_D).chunk(2, dim=-1)
+
+        shift_B_T_1_1_D, scale_B_T_1_1_D = rearrange(shift_B_T_D, "b t d -> b t 1 1 d"), rearrange(
+            scale_B_T_D, "b t d -> b t 1 1 d"
+        )
+
+        def _fn(
+            _x_B_T_H_W_D: torch.Tensor,
+            _norm_layer: nn.Module,
+            _scale_B_T_1_1_D: torch.Tensor,
+            _shift_B_T_1_1_D: torch.Tensor,
+        ) -> torch.Tensor:
+            return _norm_layer(_x_B_T_H_W_D) * (1 + _scale_B_T_1_1_D) + _shift_B_T_1_1_D
+
+        x_B_T_H_W_D = _fn(x_B_T_H_W_D, self.layer_norm, scale_B_T_1_1_D, shift_B_T_1_1_D)
+        x_B_T_H_W_O = self.linear(x_B_T_H_W_D)
+        return x_B_T_H_W_O
+
+
+class Block(nn.Module):
+    """
+    A transformer block that combines self-attention, cross-attention and MLP layers with AdaLN modulation.
+    Each component (self-attention, cross-attention, MLP) has its own layer normalization and AdaLN modulation.
+
+    Parameters:
+        x_dim (int): Dimension of input features
+        context_dim (int): Dimension of context features for cross-attention
+        num_heads (int): Number of attention heads
+        mlp_ratio (float): Multiplier for MLP hidden dimension. Default: 4.0
+        use_adaln_lora (bool): Whether to use AdaLN-LoRA modulation. Default: False
+        adaln_lora_dim (int): Hidden dimension for AdaLN-LoRA layers. Default: 256
+
+    The block applies the following sequence:
+    1. Self-attention with AdaLN modulation
+    2. Cross-attention with AdaLN modulation
+    3. MLP with AdaLN modulation
+
+    Each component uses skip connections and layer normalization.
+    """
+
+    def __init__(
+        self,
+        x_dim: int,
+        context_dim: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        use_adaln_lora: bool = False,
+        adaln_lora_dim: int = 256,
+        device=None,
+        dtype=None,
+        operations=None,
+    ):
+        super().__init__()
+        self.x_dim = x_dim
+        self.layer_norm_self_attn = operations.LayerNorm(x_dim, elementwise_affine=False, eps=1e-6, device=device, dtype=dtype)
+        self.self_attn = Attention(x_dim, None, num_heads, x_dim // num_heads, device=device, dtype=dtype, operations=operations)
+
+        self.layer_norm_cross_attn = operations.LayerNorm(x_dim, elementwise_affine=False, eps=1e-6, device=device, dtype=dtype)
+        self.cross_attn = Attention(
+            x_dim, context_dim, num_heads, x_dim // num_heads, device=device, dtype=dtype, operations=operations
+        )
+
+        self.layer_norm_mlp = operations.LayerNorm(x_dim, elementwise_affine=False, eps=1e-6, device=device, dtype=dtype)
+        self.mlp = GPT2FeedForward(x_dim, int(x_dim * mlp_ratio), device=device, dtype=dtype, operations=operations)
+
+        self.use_adaln_lora = use_adaln_lora
+        if self.use_adaln_lora:
+            self.adaln_modulation_self_attn = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(x_dim, adaln_lora_dim, bias=False, device=device, dtype=dtype),
+                operations.Linear(adaln_lora_dim, 3 * x_dim, bias=False, device=device, dtype=dtype),
+            )
+            self.adaln_modulation_cross_attn = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(x_dim, adaln_lora_dim, bias=False, device=device, dtype=dtype),
+                operations.Linear(adaln_lora_dim, 3 * x_dim, bias=False, device=device, dtype=dtype),
+            )
+            self.adaln_modulation_mlp = nn.Sequential(
+                nn.SiLU(),
+                operations.Linear(x_dim, adaln_lora_dim, bias=False, device=device, dtype=dtype),
+                operations.Linear(adaln_lora_dim, 3 * x_dim, bias=False, device=device, dtype=dtype),
+            )
+        else:
+            self.adaln_modulation_self_attn = nn.Sequential(nn.SiLU(), operations.Linear(x_dim, 3 * x_dim, bias=False, device=device, dtype=dtype))
+            self.adaln_modulation_cross_attn = nn.Sequential(nn.SiLU(), operations.Linear(x_dim, 3 * x_dim, bias=False, device=device, dtype=dtype))
+            self.adaln_modulation_mlp = nn.Sequential(nn.SiLU(), operations.Linear(x_dim, 3 * x_dim, bias=False, device=device, dtype=dtype))
+
+    def forward(
+        self,
+        x_B_T_H_W_D: torch.Tensor,
+        emb_B_T_D: torch.Tensor,
+        crossattn_emb: torch.Tensor,
+        rope_emb_L_1_1_D: Optional[torch.Tensor] = None,
+        adaln_lora_B_T_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_B_T_H_W_D = x_B_T_H_W_D + extra_per_block_pos_emb
+
+        if self.use_adaln_lora:
+            shift_self_attn_B_T_D, scale_self_attn_B_T_D, gate_self_attn_B_T_D = (
+                self.adaln_modulation_self_attn(emb_B_T_D) + adaln_lora_B_T_3D
+            ).chunk(3, dim=-1)
+            shift_cross_attn_B_T_D, scale_cross_attn_B_T_D, gate_cross_attn_B_T_D = (
+                self.adaln_modulation_cross_attn(emb_B_T_D) + adaln_lora_B_T_3D
+            ).chunk(3, dim=-1)
+            shift_mlp_B_T_D, scale_mlp_B_T_D, gate_mlp_B_T_D = (
+                self.adaln_modulation_mlp(emb_B_T_D) + adaln_lora_B_T_3D
+            ).chunk(3, dim=-1)
+        else:
+            shift_self_attn_B_T_D, scale_self_attn_B_T_D, gate_self_attn_B_T_D = self.adaln_modulation_self_attn(
+                emb_B_T_D
+            ).chunk(3, dim=-1)
+            shift_cross_attn_B_T_D, scale_cross_attn_B_T_D, gate_cross_attn_B_T_D = self.adaln_modulation_cross_attn(
+                emb_B_T_D
+            ).chunk(3, dim=-1)
+            shift_mlp_B_T_D, scale_mlp_B_T_D, gate_mlp_B_T_D = self.adaln_modulation_mlp(emb_B_T_D).chunk(3, dim=-1)
+
+        # Reshape tensors from (B, T, D) to (B, T, 1, 1, D) for broadcasting
+        shift_self_attn_B_T_1_1_D = rearrange(shift_self_attn_B_T_D, "b t d -> b t 1 1 d")
+        scale_self_attn_B_T_1_1_D = rearrange(scale_self_attn_B_T_D, "b t d -> b t 1 1 d")
+        gate_self_attn_B_T_1_1_D = rearrange(gate_self_attn_B_T_D, "b t d -> b t 1 1 d")
+
+        shift_cross_attn_B_T_1_1_D = rearrange(shift_cross_attn_B_T_D, "b t d -> b t 1 1 d")
+        scale_cross_attn_B_T_1_1_D = rearrange(scale_cross_attn_B_T_D, "b t d -> b t 1 1 d")
+        gate_cross_attn_B_T_1_1_D = rearrange(gate_cross_attn_B_T_D, "b t d -> b t 1 1 d")
+
+        shift_mlp_B_T_1_1_D = rearrange(shift_mlp_B_T_D, "b t d -> b t 1 1 d")
+        scale_mlp_B_T_1_1_D = rearrange(scale_mlp_B_T_D, "b t d -> b t 1 1 d")
+        gate_mlp_B_T_1_1_D = rearrange(gate_mlp_B_T_D, "b t d -> b t 1 1 d")
+
+        B, T, H, W, D = x_B_T_H_W_D.shape
+
+        def _fn(_x_B_T_H_W_D, _norm_layer, _scale_B_T_1_1_D, _shift_B_T_1_1_D):
+            return _norm_layer(_x_B_T_H_W_D) * (1 + _scale_B_T_1_1_D) + _shift_B_T_1_1_D
+
+        normalized_x_B_T_H_W_D = _fn(
+            x_B_T_H_W_D,
+            self.layer_norm_self_attn,
+            scale_self_attn_B_T_1_1_D,
+            shift_self_attn_B_T_1_1_D,
+        )
+        result_B_T_H_W_D = rearrange(
+            self.self_attn(
+                # normalized_x_B_T_HW_D,
+                rearrange(normalized_x_B_T_H_W_D, "b t h w d -> b (t h w) d"),
+                None,
+                rope_emb=rope_emb_L_1_1_D,
+            ),
+            "b (t h w) d -> b t h w d",
+            t=T,
+            h=H,
+            w=W,
+        )
+        x_B_T_H_W_D = x_B_T_H_W_D + gate_self_attn_B_T_1_1_D * result_B_T_H_W_D
+
+        def _x_fn(
+            _x_B_T_H_W_D: torch.Tensor,
+            layer_norm_cross_attn: Callable,
+            _scale_cross_attn_B_T_1_1_D: torch.Tensor,
+            _shift_cross_attn_B_T_1_1_D: torch.Tensor,
+        ) -> torch.Tensor:
+            _normalized_x_B_T_H_W_D = _fn(
+                _x_B_T_H_W_D, layer_norm_cross_attn, _scale_cross_attn_B_T_1_1_D, _shift_cross_attn_B_T_1_1_D
+            )
+            _result_B_T_H_W_D = rearrange(
+                self.cross_attn(
+                    rearrange(_normalized_x_B_T_H_W_D, "b t h w d -> b (t h w) d"),
+                    crossattn_emb,
+                    rope_emb=rope_emb_L_1_1_D,
+                ),
+                "b (t h w) d -> b t h w d",
+                t=T,
+                h=H,
+                w=W,
+            )
+            return _result_B_T_H_W_D
+
+        result_B_T_H_W_D = _x_fn(
+            x_B_T_H_W_D,
+            self.layer_norm_cross_attn,
+            scale_cross_attn_B_T_1_1_D,
+            shift_cross_attn_B_T_1_1_D,
+        )
+        x_B_T_H_W_D = result_B_T_H_W_D * gate_cross_attn_B_T_1_1_D + x_B_T_H_W_D
+
+        normalized_x_B_T_H_W_D = _fn(
+            x_B_T_H_W_D,
+            self.layer_norm_mlp,
+            scale_mlp_B_T_1_1_D,
+            shift_mlp_B_T_1_1_D,
+        )
+        result_B_T_H_W_D = self.mlp(normalized_x_B_T_H_W_D)
+        x_B_T_H_W_D = x_B_T_H_W_D + gate_mlp_B_T_1_1_D * result_B_T_H_W_D
+        return x_B_T_H_W_D
+
+
+class MiniTrainDIT(nn.Module):
+    """
+    A clean impl of DIT that can load and  reproduce the training results of the original DIT model in~(cosmos 1)
+    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.
+        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.
+        crossattn_emb_channels (int): Number of embedding channels for 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.
+        min_fps (int): Minimum frames per second.
+        max_fps (int): Maximum frames per second.
+        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_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.
+    """
+
+    def __init__(
+        self,
+        max_img_h: int,
+        max_img_w: int,
+        max_frames: int,
+        in_channels: int,
+        out_channels: int,
+        patch_spatial: int,  # tuple,
+        patch_temporal: int,
+        concat_padding_mask: bool = True,
+        # attention settings
+        model_channels: int = 768,
+        num_blocks: int = 10,
+        num_heads: int = 16,
+        mlp_ratio: float = 4.0,
+        # cross attention settings
+        crossattn_emb_channels: int = 1024,
+        # positional embedding settings
+        pos_emb_cls: str = "sincos",
+        pos_emb_learnable: bool = False,
+        pos_emb_interpolation: str = "crop",
+        min_fps: int = 1,
+        max_fps: int = 30,
+        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_h_extrapolation_ratio: float = 1.0,
+        extra_w_extrapolation_ratio: float = 1.0,
+        extra_t_extrapolation_ratio: float = 1.0,
+        rope_enable_fps_modulation: bool = True,
+        image_model=None,
+        device=None,
+        dtype=None,
+        operations=None,
+    ) -> None:
+        super().__init__()
+        self.dtype = dtype
+        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.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.min_fps = min_fps
+        self.max_fps = max_fps
+        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_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.rope_enable_fps_modulation = rope_enable_fps_modulation
+
+        self.build_pos_embed(device=device, dtype=dtype)
+        self.use_adaln_lora = use_adaln_lora
+        self.adaln_lora_dim = adaln_lora_dim
+        self.t_embedder = nn.Sequential(
+            Timesteps(model_channels),
+            TimestepEmbedding(model_channels, model_channels, use_adaln_lora=use_adaln_lora, device=device, dtype=dtype, operations=operations,),
+        )
+
+        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,
+            device=device, dtype=dtype, operations=operations,
+        )
+
+        self.blocks = nn.ModuleList(
+            [
+                Block(
+                    x_dim=model_channels,
+                    context_dim=crossattn_emb_channels,
+                    num_heads=num_heads,
+                    mlp_ratio=mlp_ratio,
+                    use_adaln_lora=use_adaln_lora,
+                    adaln_lora_dim=adaln_lora_dim,
+                    device=device, dtype=dtype, operations=operations,
+                )
+                for _ in range(num_blocks)
+            ]
+        )
+
+        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,
+            device=device, dtype=dtype, operations=operations,
+        )
+
+        self.t_embedding_norm = operations.RMSNorm(model_channels, eps=1e-6, device=device, dtype=dtype)
+
+    def build_pos_embed(self, device=None, dtype=None) -> 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,
+            max_fps=self.max_fps,
+            min_fps=self.min_fps,
+            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,
+            enable_fps_modulation=self.rope_enable_fps_modulation,
+            device=device,
+        )
+        self.pos_embedder = cls_type(
+            **kwargs,  # type: ignore
+        )
+
+        if self.extra_per_block_abs_pos_emb:
+            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
+            kwargs["dtype"] = dtype
+            self.extra_pos_embedder = LearnablePosEmbAxis(
+                **kwargs,  # type: ignore
+            )
+
+    def prepare_embedded_sequence(
+        self,
+        x_B_C_T_H_W: torch.Tensor,
+        fps: Optional[torch.Tensor] = None,
+        padding_mask: Optional[torch.Tensor] = None,
+    ) -> Tuple[torch.Tensor, Optional[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 None:
+                padding_mask = torch.zeros(x_B_C_T_H_W.shape[0], 1, x_B_C_T_H_W.shape[3], x_B_C_T_H_W.shape[4], dtype=x_B_C_T_H_W.dtype, device=x_B_C_T_H_W.device)
+            else:
+                padding_mask = transforms.functional.resize(
+                    padding_mask, list(x_B_C_T_H_W.shape[-2:]), interpolation=transforms.InterpolationMode.NEAREST
+                )
+            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, dtype=x_B_C_T_H_W.dtype)
+        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
+        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 unpatchify(self, x_B_T_H_W_M: torch.Tensor) -> torch.Tensor:
+        x_B_C_Tt_Hp_Wp = rearrange(
+            x_B_T_H_W_M,
+            "B T H W (p1 p2 t C) -> B C (T t) (H p1) (W p2)",
+            p1=self.patch_spatial,
+            p2=self.patch_spatial,
+            t=self.patch_temporal,
+        )
+        return x_B_C_Tt_Hp_Wp
+
+    def forward(
+        self,
+        x: torch.Tensor,
+        timesteps: torch.Tensor,
+        context: torch.Tensor,
+        fps: Optional[torch.Tensor] = None,
+        padding_mask: Optional[torch.Tensor] = None,
+        **kwargs,
+    ):
+        x_B_C_T_H_W = x
+        timesteps_B_T = timesteps
+        crossattn_emb = context
+        """
+        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
+        """
+        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_B_C_T_H_W,
+            fps=fps,
+            padding_mask=padding_mask,
+        )
+
+        if timesteps_B_T.ndim == 1:
+            timesteps_B_T = timesteps_B_T.unsqueeze(1)
+        t_embedding_B_T_D, adaln_lora_B_T_3D = self.t_embedder[1](self.t_embedder[0](timesteps_B_T).to(x_B_T_H_W_D.dtype))
+        t_embedding_B_T_D = self.t_embedding_norm(t_embedding_B_T_D)
+
+        # for logging purpose
+        affline_scale_log_info = {}
+        affline_scale_log_info["t_embedding_B_T_D"] = t_embedding_B_T_D.detach()
+        self.affline_scale_log_info = affline_scale_log_info
+        self.affline_emb = t_embedding_B_T_D
+        self.crossattn_emb = crossattn_emb
+
+        if extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D is not None:
+            assert (
+                x_B_T_H_W_D.shape == extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D.shape
+            ), f"{x_B_T_H_W_D.shape} != {extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D.shape}"
+
+        block_kwargs = {
+            "rope_emb_L_1_1_D": rope_emb_L_1_1_D.unsqueeze(1).unsqueeze(0),
+            "adaln_lora_B_T_3D": adaln_lora_B_T_3D,
+            "extra_per_block_pos_emb": extra_pos_emb_B_T_H_W_D_or_T_H_W_B_D,
+        }
+        for block in self.blocks:
+            x_B_T_H_W_D = block(
+                x_B_T_H_W_D,
+                t_embedding_B_T_D,
+                crossattn_emb,
+                **block_kwargs,
+            )
+
+        x_B_T_H_W_O = self.final_layer(x_B_T_H_W_D, t_embedding_B_T_D, adaln_lora_B_T_3D=adaln_lora_B_T_3D)
+        x_B_C_Tt_Hp_Wp = self.unpatchify(x_B_T_H_W_O)
+        return x_B_C_Tt_Hp_Wp

comfy/ldm/flux/model.py

@@ -161,6 +161,9 @@ class Flux(nn.Module):
                     if add is not None:
                         img += add
 
+        if img.dtype == torch.float16:
+            img = torch.nan_to_num(img, nan=0.0, posinf=65504, neginf=-65504)
+
         img = torch.cat((txt, img), 1)
 
         for i, block in enumerate(self.single_blocks):

comfy/ldm/lightricks/model.py

@@ -262,8 +262,8 @@ class CrossAttention(nn.Module):
         self.heads = heads
         self.dim_head = dim_head
 
-        self.q_norm = operations.RMSNorm(inner_dim, dtype=dtype, device=device)
-        self.k_norm = operations.RMSNorm(inner_dim, dtype=dtype, device=device)
+        self.q_norm = operations.RMSNorm(inner_dim, eps=1e-5, dtype=dtype, device=device)
+        self.k_norm = operations.RMSNorm(inner_dim, eps=1e-5, dtype=dtype, device=device)
 
         self.to_q = operations.Linear(query_dim, inner_dim, bias=True, dtype=dtype, device=device)
         self.to_k = operations.Linear(context_dim, inner_dim, bias=True, dtype=dtype, device=device)

comfy/ldm/modules/attention.py

@@ -786,7 +786,7 @@ class BasicTransformerBlock(nn.Module):
             for p in patch:
                 n = p(n, extra_options)
 
-        x += n
+        x = n + x
         if "middle_patch" in transformer_patches:
             patch = transformer_patches["middle_patch"]
             for p in patch:
@@ -826,12 +826,12 @@ class BasicTransformerBlock(nn.Module):
             for p in patch:
                 n = p(n, extra_options)
 
-        x += n
+        x = n + x
         if self.is_res:
             x_skip = x
         x = self.ff(self.norm3(x))
         if self.is_res:
-            x += x_skip
+            x = x_skip + x
 
         return x
 

comfy/model_base.py

@@ -16,25 +16,28 @@
     along with this program.  If not, see <https://www.gnu.org/licenses/>.
 """
 
-import logging
 import math
+
+import logging
+import torch
 from enum import Enum
 from typing import TypeVar, Type, Protocol, Any, Optional
 
-import torch
-
 from . import conds
 from . import latent_formats
 from . import model_management
 from . import ops
 from . import utils
+from .conds import CONDRegular, CONDConstant
 from .ldm.ace.model import ACEStepTransformer2DModel
 from .ldm.audio.dit import AudioDiffusionTransformer
 from .ldm.audio.embedders import NumberConditioner
 from .ldm.aura.mmdit import MMDiT as AuraMMDiT
 from .ldm.cascade.stage_b import StageB
 from .ldm.cascade.stage_c import StageC
+from .ldm.chroma import model as chroma_model
 from .ldm.cosmos.model import GeneralDIT
+from .ldm.cosmos.predict2 import MiniTrainDIT
 from .ldm.flux import model as flux_model
 from .ldm.genmo.joint_model.asymm_models_joint import AsymmDiTJoint
 from .ldm.hidream.model import HiDreamImageTransformer2DModel
@@ -49,8 +52,10 @@ from .ldm.modules.diffusionmodules.upscaling import ImageConcatWithNoiseAugmenta
 from .ldm.modules.encoders.noise_aug_modules import CLIPEmbeddingNoiseAugmentation
 from .ldm.pixart.pixartms import PixArtMS
 from .ldm.wan.model import WanModel, VaceWanModel, CameraWanModel
-from .ldm.chroma import model as chroma_model
 from .model_management_types import ModelManageable
+from .model_sampling import CONST, ModelSamplingDiscreteFlow, ModelSamplingFlux, IMG_TO_IMG
+from .model_sampling import StableCascadeSampling, COSMOS_RFLOW, ModelSamplingCosmosRFlow, V_PREDICTION, \
+    ModelSamplingContinuousEDM, ModelSamplingDiscrete, EPS, EDM, ModelSamplingContinuousV
 from .ops import Operations
 from .patcher_extension import WrapperExecutor, WrappersMP, get_all_wrappers
 
@@ -65,9 +70,7 @@ class ModelType(Enum):
     V_PREDICTION_CONTINUOUS = 7
     FLUX = 8
     IMG_TO_IMG = 9
-
-
-from .model_sampling import EPS, V_PREDICTION, EDM, ModelSamplingDiscrete, ModelSamplingContinuousEDM, StableCascadeSampling, CONST, ModelSamplingDiscreteFlow, ModelSamplingContinuousV, ModelSamplingFlux, IMG_TO_IMG
+    FLOW_COSMOS = 10
 
 
 def model_sampling(model_config, model_type):
@@ -98,6 +101,9 @@ def model_sampling(model_config, model_type):
         s = ModelSamplingFlux
     elif model_type == ModelType.IMG_TO_IMG:
         c = IMG_TO_IMG
+    elif model_type == ModelType.FLOW_COSMOS:
+        c = COSMOS_RFLOW
+        s = ModelSamplingCosmosRFlow
 
     class ModelSampling(s, c):
         pass
@@ -1038,6 +1044,44 @@ class CosmosVideo(BaseModel):
         return latent_image * ((sigma ** 2 + self.model_sampling.sigma_data ** 2) ** 0.5)
 
 
+class CosmosPredict2(BaseModel):
+    def __init__(self, model_config, model_type=ModelType.FLOW_COSMOS, image_to_video=False, device=None):
+        super().__init__(model_config, model_type, device=device, unet_model=MiniTrainDIT)
+        self.image_to_video = image_to_video
+        if self.image_to_video:
+            self.concat_keys = ("mask_inverted",)
+
+    def extra_conds(self, **kwargs):
+        out = super().extra_conds(**kwargs)
+        cross_attn = kwargs.get("cross_attn", None)
+        if cross_attn is not None:
+            out['c_crossattn'] = CONDRegular(cross_attn)
+
+        denoise_mask = kwargs.get("concat_mask", kwargs.get("denoise_mask", None))
+        if denoise_mask is not None:
+            out["denoise_mask"] = CONDRegular(denoise_mask)
+
+        out['fps'] = CONDConstant(kwargs.get("frame_rate", None))
+        return out
+
+    def process_timestep(self, timestep, x, denoise_mask=None, **kwargs):
+        if denoise_mask is None:
+            return timestep
+        condition_video_mask_B_1_T_1_1 = denoise_mask.mean(dim=[1, 3, 4], keepdim=True)
+        c_noise_B_1_T_1_1 = 0.0 * (1.0 - condition_video_mask_B_1_T_1_1) + timestep.reshape(timestep.shape[0], 1, 1, 1, 1) * condition_video_mask_B_1_T_1_1
+        out = c_noise_B_1_T_1_1.squeeze(dim=[1, 3, 4])
+        return out
+
+    def scale_latent_inpaint(self, sigma, noise, latent_image, **kwargs):
+        sigma = sigma.reshape([sigma.shape[0]] + [1] * (len(noise.shape) - 1))
+        sigma_noise_augmentation = 0  # TODO
+        if sigma_noise_augmentation != 0:
+            latent_image = latent_image + noise
+        latent_image = self.model_sampling.calculate_input(torch.tensor([sigma_noise_augmentation], device=latent_image.device, dtype=latent_image.dtype), latent_image)
+        sigma = (sigma / (sigma + 1))
+        return latent_image / (1.0 - sigma)
+
+
 class Lumina2(BaseModel):
     def __init__(self, model_config, model_type=ModelType.FLOW, device=None):
         super().__init__(model_config, model_type, device=device, unet_model=NextDiT)
@@ -1151,6 +1195,7 @@ class WAN21_Vace(WAN21):
         out['vace_strength'] = conds.CONDConstant(vace_strength)
         return out
 
+
 class WAN21_Camera(WAN21):
     def __init__(self, model_config, model_type=ModelType.FLOW, image_to_video=False, device=None):
         super(WAN21, self).__init__(model_config, model_type, device=device, unet_model=CameraWanModel)
@@ -1163,6 +1208,7 @@ class WAN21_Camera(WAN21):
             out['camera_conditions'] = conds.CONDRegular(camera_conditions)
         return out
 
+
 class Hunyuan3Dv2(BaseModel):
     def __init__(self, model_config, model_type=ModelType.FLOW, device=None):
         super().__init__(model_config, model_type, device=device, unet_model=Hunyuan3Dv2Model)
@@ -1199,6 +1245,7 @@ class HiDream(BaseModel):
             out['image_cond'] = conds.CONDNoiseShape(self.process_latent_in(image_cond))
         return out
 
+
 class Chroma(Flux):
     def __init__(self, model_config, model_type=ModelType.FLUX, device=None):
         super().__init__(model_config, model_type, device=device, unet_model=chroma_model.Chroma)
@@ -1211,6 +1258,7 @@ class Chroma(Flux):
             out['guidance'] = conds.CONDRegular(torch.FloatTensor([guidance]))
         return out
 
+
 class ACEStep(BaseModel):
     def __init__(self, model_config, model_type=ModelType.FLOW, device=None):
         super().__init__(model_config, model_type, device=device, unet_model=ACEStepTransformer2DModel)

comfy/model_detection.py

@@ -411,6 +411,58 @@ def detect_unet_config(state_dict, key_prefix, metadata=None):
         dit_config["text_emb_dim"] = 2048
         return dit_config
 
+    if '{}blocks.0.mlp.layer1.weight'.format(key_prefix) in state_dict_keys:  # Cosmos predict2
+        dit_config = {}
+        dit_config["image_model"] = "cosmos_predict2"
+        dit_config["max_img_h"] = 240
+        dit_config["max_img_w"] = 240
+        dit_config["max_frames"] = 128
+        concat_padding_mask = True
+        dit_config["in_channels"] = (state_dict['{}x_embedder.proj.1.weight'.format(key_prefix)].shape[1] // 4) - int(concat_padding_mask)
+        dit_config["out_channels"] = 16
+        dit_config["patch_spatial"] = 2
+        dit_config["patch_temporal"] = 1
+        dit_config["model_channels"] = state_dict['{}x_embedder.proj.1.weight'.format(key_prefix)].shape[0]
+        dit_config["concat_padding_mask"] = concat_padding_mask
+        dit_config["crossattn_emb_channels"] = 1024
+        dit_config["pos_emb_cls"] = "rope3d"
+        dit_config["pos_emb_learnable"] = True
+        dit_config["pos_emb_interpolation"] = "crop"
+        dit_config["min_fps"] = 1
+        dit_config["max_fps"] = 30
+
+        dit_config["use_adaln_lora"] = True
+        dit_config["adaln_lora_dim"] = 256
+        if dit_config["model_channels"] == 2048:
+            dit_config["num_blocks"] = 28
+            dit_config["num_heads"] = 16
+        elif dit_config["model_channels"] == 5120:
+            dit_config["num_blocks"] = 36
+            dit_config["num_heads"] = 40
+
+        if dit_config["in_channels"] == 16:
+            dit_config["extra_per_block_abs_pos_emb"] = False
+            dit_config["rope_h_extrapolation_ratio"] = 4.0
+            dit_config["rope_w_extrapolation_ratio"] = 4.0
+            dit_config["rope_t_extrapolation_ratio"] = 1.0
+        elif dit_config["in_channels"] == 17: # img to video
+            if dit_config["model_channels"] == 2048:
+                dit_config["extra_per_block_abs_pos_emb"] = False
+                dit_config["rope_h_extrapolation_ratio"] = 3.0
+                dit_config["rope_w_extrapolation_ratio"] = 3.0
+                dit_config["rope_t_extrapolation_ratio"] = 1.0
+            elif dit_config["model_channels"] == 5120:
+                dit_config["rope_h_extrapolation_ratio"] = 2.0
+                dit_config["rope_w_extrapolation_ratio"] = 2.0
+                dit_config["rope_t_extrapolation_ratio"] = 0.8333333333333334
+
+        dit_config["extra_h_extrapolation_ratio"] = 1.0
+        dit_config["extra_w_extrapolation_ratio"] = 1.0
+        dit_config["extra_t_extrapolation_ratio"] = 1.0
+        dit_config["rope_enable_fps_modulation"] = False
+
+        return dit_config
+
     if '{}input_blocks.0.0.weight'.format(key_prefix) not in state_dict_keys:
         return None
 

comfy/model_management.py

@@ -1189,7 +1189,7 @@ def pytorch_attention_flash_attention():
     global ENABLE_PYTORCH_ATTENTION
     if ENABLE_PYTORCH_ATTENTION:
         # TODO: more reliable way of checking for flash attention?
-        if is_nvidia():  # pytorch flash attention only works on Nvidia
+        if is_nvidia():
             return True
         if is_intel_xpu():
             return True
@@ -1206,7 +1206,7 @@ def force_upcast_attention_dtype():
     upcast = args.force_upcast_attention
 
     macos_version = mac_version()
-    if macos_version is not None and ((14, 5) <= macos_version < (16,)):  # black image bug on recent versions of macOS
+    if macos_version is not None and ((14, 5) <= macos_version):  # black image bug on recent versions of macOS, I don't think it's ever getting fixed
         upcast = True
 
     if upcast:

comfy/model_sampling.py

@@ -97,6 +97,25 @@ class IMG_TO_IMG(X0):
     def calculate_input(self, sigma, noise):
         return noise
 
+class COSMOS_RFLOW:
+    def calculate_input(self, sigma, noise):
+        sigma = (sigma / (sigma + 1))
+        sigma = sigma.view(sigma.shape[:1] + (1,) * (noise.ndim - 1))
+        return noise * (1.0 - sigma)
+
+    def calculate_denoised(self, sigma, model_output, model_input):
+        sigma = (sigma / (sigma + 1))
+        sigma = sigma.view(sigma.shape[:1] + (1,) * (model_output.ndim - 1))
+        return model_input * (1.0 - sigma) - model_output * sigma
+
+    def noise_scaling(self, sigma, noise, latent_image, max_denoise=False):
+        sigma = sigma.view(sigma.shape[:1] + (1,) * (noise.ndim - 1))
+        noise = noise * sigma
+        noise += latent_image
+        return noise
+
+    def inverse_noise_scaling(self, sigma, latent):
+        return latent
 
 class ModelSamplingDiscrete(torch.nn.Module):
     def __init__(self, model_config=None, zsnr=None):
@@ -375,3 +394,15 @@ class ModelSamplingFlux(torch.nn.Module):
         if percent >= 1.0:
             return 0.0
         return flux_time_shift(self.shift, 1.0, 1.0 - percent)
+
+
+class ModelSamplingCosmosRFlow(ModelSamplingContinuousEDM):
+    def timestep(self, sigma):
+        return sigma / (sigma + 1)
+
+    def sigma(self, timestep):
+        sigma_max = self.sigma_max
+        if timestep >= (sigma_max / (sigma_max + 1)):
+            return sigma_max
+
+        return timestep / (1 - timestep)

comfy/sd.py

@@ -1128,6 +1128,27 @@ def load_state_dict_guess_config(sd, output_vae=True, output_clip=True, output_c
 
 
 def load_diffusion_model_state_dict(sd, model_options: dict = None, ckpt_path: Optional[str] = ""):  # load unet in diffusers or regular format
+    """
+    Loads a UNet diffusion model from a state dictionary, supporting both diffusers and regular formats.
+
+    Args:
+        sd (dict): State dictionary containing model weights and configuration
+        model_options (dict, optional): Additional options for model loading. Supports:
+            - dtype: Override model data type
+            - custom_operations: Custom model operations
+            - fp8_optimizations: Enable FP8 optimizations
+
+    Returns:
+        ModelPatcher: A wrapped model instance that handles device management and weight loading.
+        Returns None if the model configuration cannot be detected.
+
+    The function:
+    1. Detects and handles different model formats (regular, diffusers, mmdit)
+    2. Configures model dtype based on parameters and device capabilities
+    3. Handles weight conversion and device placement
+    4. Manages model optimization settings
+    5. Loads weights and returns a device-managed model instance
+    """
     if model_options is None:
         model_options = {}
     dtype = model_options.get("dtype", None)

comfy/sd1_clip.py

@@ -556,7 +556,7 @@ class SDTokenizer:
         self.tokenizer_path = tokenizer_path
         self.tokenizer: PreTrainedTokenizerBase | SPieceTokenizer = tokenizer_class.from_pretrained(tokenizer_path, **tokenizer_args)
         self.max_length = tokenizer_data.get("{}_max_length".format(embedding_key), max_length)
-        self.min_length = min_length
+        self.min_length = tokenizer_data.get("{}_min_length".format(embedding_key), min_length)
         self.end_token = None
         self.min_padding = min_padding
 

comfy/supported_models.py

@@ -976,6 +976,52 @@ class CosmosI2V(CosmosT2V):
         return out
 
 
+class CosmosT2IPredict2(supported_models_base.BASE):
+    unet_config = {
+        "image_model": "cosmos_predict2",
+        "in_channels": 16,
+    }
+
+    sampling_settings = {
+        "sigma_data": 1.0,
+        "sigma_max": 80.0,
+        "sigma_min": 0.002,
+    }
+
+    unet_extra_config = {}
+    latent_format = latent_formats.Wan21
+
+    memory_usage_factor = 1.0
+
+    supported_inference_dtypes = [torch.bfloat16, torch.float32]
+
+    def __init__(self, unet_config):
+        super().__init__(unet_config)
+        self.memory_usage_factor = (unet_config.get("model_channels", 2048) / 2048) * 0.9
+
+    def get_model(self, state_dict, prefix="", device=None):
+        out = model_base.CosmosPredict2(self, device=device)
+        return out
+
+    def clip_target(self, state_dict=None):
+        if state_dict is None:
+            state_dict = {}
+        pref = self.text_encoder_key_prefix[0]
+        t5_detect = sd3_clip.t5_xxl_detect(state_dict, "{}t5xxl.transformer.".format(pref))
+        return supported_models_base.ClipTarget(cosmos.CosmosT5Tokenizer, cosmos.te(**t5_detect))
+
+
+class CosmosI2VPredict2(CosmosT2IPredict2):
+    unet_config = {
+        "image_model": "cosmos_predict2",
+        "in_channels": 17,
+    }
+
+    def get_model(self, state_dict, prefix="", device=None):
+        out = model_base.CosmosPredict2(self, image_to_video=True, device=device)
+        return out
+
+
 class Lumina2(supported_models_base.BASE):
     unet_config = {
         "image_model": "lumina2",
@@ -1230,6 +1276,6 @@ class ACEStep(supported_models_base.BASE):
         return supported_models_base.ClipTarget(ace.AceT5Tokenizer, ace.AceT5Model)
 
 
-models = [LotusD, 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, HunyuanVideoSkyreelsI2V, HunyuanVideoI2V, HunyuanVideo, CosmosT2V, CosmosI2V, Lumina2, WAN21_T2V, WAN21_I2V, WAN21_FunControl2V, WAN21_Vace, WAN21_Camera, Hunyuan3Dv2mini, Hunyuan3Dv2, HiDream, Chroma, ACEStep]
+models = [LotusD, 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, HunyuanVideoSkyreelsI2V, HunyuanVideoI2V, HunyuanVideo, CosmosT2V, CosmosI2V, CosmosT2IPredict2, CosmosI2VPredict2, Lumina2, WAN21_T2V, WAN21_I2V, WAN21_FunControl2V, WAN21_Vace, WAN21_Camera, Hunyuan3Dv2mini, Hunyuan3Dv2, HiDream, Chroma, ACEStep]
 
 models += [SVD_img2vid]

comfy/weight_adapter/__init__.py

@@ -1,4 +1,4 @@
-from .base import WeightAdapterBase
+from .base import WeightAdapterBase, WeightAdapterTrainBase
 from .lora import LoRAAdapter
 from .loha import LoHaAdapter
 from .lokr import LoKrAdapter
@@ -15,3 +15,9 @@ adapters: list[type[WeightAdapterBase]] = [
     OFTAdapter,
     BOFTAdapter,
 ]
+
+__all__ = [
+    "WeightAdapterBase",
+    "WeightAdapterTrainBase",
+    "adapters"
+] + [a.__name__ for a in adapters]

comfy/weight_adapter/base.py

@@ -12,12 +12,20 @@ class WeightAdapterBase:
     weights: list[torch.Tensor]
 
     @classmethod
-    def load(cls, x: str, lora: dict[str, torch.Tensor]) -> Optional["WeightAdapterBase"]:
+    def load(cls, x: str, lora: dict[str, torch.Tensor], alpha: float, dora_scale: torch.Tensor) -> Optional["WeightAdapterBase"]:
         raise NotImplementedError
 
     def to_train(self) -> "WeightAdapterTrainBase":
         raise NotImplementedError
 
+    @classmethod
+    def create_train(cls, weight, *args) -> "WeightAdapterTrainBase":
+        """
+        weight: The original weight tensor to be modified.
+        *args: Additional arguments for configuration, such as rank, alpha etc.
+        """
+        raise NotImplementedError
+
     def calculate_weight(
         self,
         weight,
@@ -33,10 +41,22 @@ class WeightAdapterBase:
 
 
 class WeightAdapterTrainBase(nn.Module):
+    # We follow the scheme of PR #7032
     def __init__(self):
         super().__init__()
 
-    # [TODO] Collaborate with LoRA training PR #7032
+    def __call__(self, w):
+        """
+        w: The original weight tensor to be modified.
+        """
+        raise NotImplementedError
+
+    def passive_memory_usage(self):
+        raise NotImplementedError("passive_memory_usage is not implemented")
+
+    def move_to(self, device):
+        self.to(device)
+        return self.passive_memory_usage()
 
 
 def weight_decompose(dora_scale, weight, lora_diff, alpha, strength, intermediate_dtype, function):
@@ -102,3 +122,14 @@ def pad_tensor_to_shape(tensor: torch.Tensor, new_shape: list[int]) -> torch.Ten
     padded_tensor[new_slices] = tensor[orig_slices]
 
     return padded_tensor
+
+
+def tucker_weight_from_conv(up, down, mid):
+    up = up.reshape(up.size(0), up.size(1))
+    down = down.reshape(down.size(0), down.size(1))
+    return torch.einsum("m n ..., i m, n j -> i j ...", mid, up, down)
+
+
+def tucker_weight(wa, wb, t):
+    temp = torch.einsum("i j ..., j r -> i r ...", t, wb)
+    return torch.einsum("i j ..., i r -> r j ...", temp, wa)

comfy/weight_adapter/lora.py

@@ -3,7 +3,56 @@ from typing import Optional
 
 import torch
 from ..model_management import cast_to_device
-from .base import WeightAdapterBase, weight_decompose, pad_tensor_to_shape
+from .base import (
+    WeightAdapterBase,
+    WeightAdapterTrainBase,
+    weight_decompose,
+    pad_tensor_to_shape,
+    tucker_weight_from_conv,
+)
+
+
+class LoraDiff(WeightAdapterTrainBase):
+    def __init__(self, weights):
+        super().__init__()
+        mat1, mat2, alpha, mid, dora_scale, reshape = weights
+        out_dim, rank = mat1.shape[0], mat1.shape[1]
+        rank, in_dim = mat2.shape[0], mat2.shape[1]
+        if mid is not None:
+            convdim = mid.ndim - 2
+            layer = (
+                torch.nn.Conv1d,
+                torch.nn.Conv2d,
+                torch.nn.Conv3d
+            )[convdim]
+        else:
+            layer = torch.nn.Linear
+        self.lora_up = layer(rank, out_dim, bias=False)
+        self.lora_down = layer(in_dim, rank, bias=False)
+        self.lora_up.weight.data.copy_(mat1)
+        self.lora_down.weight.data.copy_(mat2)
+        if mid is not None:
+            self.lora_mid = layer(mid, rank, bias=False)
+            self.lora_mid.weight.data.copy_(mid)
+        else:
+            self.lora_mid = None
+        self.rank = rank
+        self.alpha = torch.nn.Parameter(torch.tensor(alpha), requires_grad=False)
+
+    def __call__(self, w):
+        org_dtype = w.dtype
+        if self.lora_mid is None:
+            diff = self.lora_up.weight @ self.lora_down.weight
+        else:
+            diff = tucker_weight_from_conv(
+                self.lora_up.weight, self.lora_down.weight, self.lora_mid.weight
+            )
+        scale = self.alpha / self.rank
+        weight = w + scale * diff.reshape(w.shape)
+        return weight.to(org_dtype)
+
+    def passive_memory_usage(self):
+        return sum(param.numel() * param.element_size() for param in self.parameters())
 
 logger = logging.getLogger(__name__)
 
@@ -14,6 +63,21 @@ class LoRAAdapter(WeightAdapterBase):
         self.loaded_keys = loaded_keys
         self.weights = weights
 
+    @classmethod
+    def create_train(cls, weight, rank=1, alpha=1.0):
+        out_dim = weight.shape[0]
+        in_dim = weight.shape[1:].numel()
+        mat1 = torch.empty(out_dim, rank, device=weight.device, dtype=weight.dtype)
+        mat2 = torch.empty(rank, in_dim, device=weight.device, dtype=weight.dtype)
+        torch.nn.init.kaiming_uniform_(mat1, a=5**0.5)
+        torch.nn.init.constant_(mat2, 0.0)
+        return LoraDiff(
+            (mat1, mat2, alpha, None, None, None)
+        )
+
+    def to_train(self):
+        return LoraDiff(self.weights)
+
     @classmethod
     def load(
         cls,

comfy_api_nodes/nodes_bfl.py

@@ -346,20 +346,6 @@ class FluxKontextProImageNode(ComfyNodeABC):
             },
         }
 
-    @classmethod
-    def VALIDATE_INPUTS(cls, aspect_ratio: str):
-        try:
-            validate_aspect_ratio(
-                aspect_ratio,
-                minimum_ratio=cls.MINIMUM_RATIO,
-                maximum_ratio=cls.MAXIMUM_RATIO,
-                minimum_ratio_str=cls.MINIMUM_RATIO_STR,
-                maximum_ratio_str=cls.MAXIMUM_RATIO_STR,
-            )
-        except Exception as e:
-            return str(e)
-        return True
-
     RETURN_TYPES = (IO.IMAGE,)
     DESCRIPTION = cleandoc(__doc__ or "")  # Handle potential None value
     FUNCTION = "api_call"
@@ -380,6 +366,13 @@ class FluxKontextProImageNode(ComfyNodeABC):
         unique_id: Union[str, None] = None,
         **kwargs,
     ):
+        aspect_ratio = validate_aspect_ratio(
+            aspect_ratio,
+            minimum_ratio=self.MINIMUM_RATIO,
+            maximum_ratio=self.MAXIMUM_RATIO,
+            minimum_ratio_str=self.MINIMUM_RATIO_STR,
+            maximum_ratio_str=self.MAXIMUM_RATIO_STR,
+        )
         if input_image is None:
             validate_string(prompt, strip_whitespace=False)
         operation = SynchronousOperation(
@@ -395,13 +388,7 @@ class FluxKontextProImageNode(ComfyNodeABC):
                 guidance=round(guidance, 1),
                 steps=steps,
                 seed=seed,
-                aspect_ratio=validate_aspect_ratio(
-                    aspect_ratio,
-                    minimum_ratio=self.MINIMUM_RATIO,
-                    maximum_ratio=self.MAXIMUM_RATIO,
-                    minimum_ratio_str=self.MINIMUM_RATIO_STR,
-                    maximum_ratio_str=self.MAXIMUM_RATIO_STR,
-                ),
+                aspect_ratio=aspect_ratio,
                 input_image=(
                     input_image
                     if input_image is None

comfy_config/types.py

@@ -1,4 +1,4 @@
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, field_validator
 from pydantic_settings import BaseSettings, SettingsConfigDict
 from typing import List, Optional
 
@@ -50,6 +50,7 @@ class ComfyConfig(BaseModel):
     icon: str = Field(default="", alias="Icon")
     models: List[Model] = Field(default_factory=list, alias="Models")
     includes: List[str] = Field(default_factory=list)
+    web: Optional[str] = None
 
 
 class License(BaseModel):
@@ -66,6 +67,18 @@ class ProjectConfig(BaseModel):
     license: License = Field(default_factory=License)
     urls: URLs = Field(default_factory=URLs)
 
+    @field_validator('license', mode='before')
+    @classmethod
+    def validate_license(cls, v):
+        if isinstance(v, str):
+            return License(text=v)
+        elif isinstance(v, dict):
+            return License(**v)
+        elif isinstance(v, License):
+            return v
+        else:
+            return License()
+
 
 class PyProjectConfig(BaseModel):
     project: ProjectConfig = Field(default_factory=ProjectConfig)
@@ -77,4 +90,4 @@ class PyProjectSettings(BaseSettings):
 
     tool: dict = Field(default_factory=dict)
 
-    model_config = SettingsConfigDict()
+    model_config = SettingsConfigDict(extra='allow')

comfy_extras/nodes/nodes_cosmos.py

@@ -1,5 +1,6 @@
 import torch
 
+import comfy.latent_formats
 import comfy.model_management
 import comfy.utils
 from comfy.node_helpers import export_custom_nodes
@@ -79,4 +80,49 @@ class CosmosImageToVideoLatent(CustomNode):
         return (out_latent,)
 
 
+class CosmosPredict2ImageToVideoLatent(CustomNode):
+    @classmethod
+    def INPUT_TYPES(s):
+        return {"required": {"vae": ("VAE",),
+                             "width": ("INT", {"default": 848, "min": 16, "max": MAX_RESOLUTION, "step": 16}),
+                             "height": ("INT", {"default": 480, "min": 16, "max": MAX_RESOLUTION, "step": 16}),
+                             "length": ("INT", {"default": 93, "min": 1, "max": MAX_RESOLUTION, "step": 4}),
+                             "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096}),
+                             },
+                "optional": {"start_image": ("IMAGE",),
+                             "end_image": ("IMAGE",),
+                             }}
+
+    RETURN_TYPES = ("LATENT",)
+    FUNCTION = "encode"
+
+    CATEGORY = "conditioning/inpaint"
+
+    def encode(self, vae, width, height, length, batch_size, start_image=None, end_image=None):
+        latent = torch.zeros([1, 16, ((length - 1) // 4) + 1, height // 8, width // 8], device=comfy.model_management.intermediate_device())
+        if start_image is None and end_image is None:
+            out_latent = {}
+            out_latent["samples"] = latent
+            return (out_latent,)
+
+        mask = torch.ones([latent.shape[0], 1, ((length - 1) // 4) + 1, latent.shape[-2], latent.shape[-1]], device=comfy.model_management.intermediate_device())
+
+        if start_image is not None:
+            latent_temp = vae_encode_with_padding(vae, start_image, width, height, length, padding=1)
+            latent[:, :, :latent_temp.shape[-3]] = latent_temp
+            mask[:, :, :latent_temp.shape[-3]] *= 0.0
+
+        if end_image is not None:
+            latent_temp = vae_encode_with_padding(vae, end_image, width, height, length, padding=0)
+            latent[:, :, -latent_temp.shape[-3]:] = latent_temp
+            mask[:, :, -latent_temp.shape[-3]:] *= 0.0
+
+        out_latent = {}
+        latent_format = comfy.latent_formats.Wan21()
+        latent = latent_format.process_out(latent) * mask + latent * (1.0 - mask)
+        out_latent["samples"] = latent.repeat((batch_size,) + (1,) * (latent.ndim - 1))
+        out_latent["noise_mask"] = mask.repeat((batch_size,) + (1,) * (mask.ndim - 1))
+        return (out_latent,)
+
+
 export_custom_nodes()

comfy_extras/nodes/nodes_model_advanced.py

@@ -198,7 +198,7 @@ class ModelSamplingContinuousEDM:
     @classmethod
     def INPUT_TYPES(s):
         return {"required": {"model": ("MODEL",),
-                             "sampling": (["v_prediction", "edm", "edm_playground_v2.5", "eps"],),
+                             "sampling": (["v_prediction", "edm", "edm_playground_v2.5", "eps", "cosmos_rflow"],),
                              "sigma_max": ("FLOAT", {"default": 120.0, "min": 0.0, "max": 1000.0, "step": 0.001, "round": False}),
                              "sigma_min": ("FLOAT", {"default": 0.002, "min": 0.0, "max": 1000.0, "step": 0.001, "round": False}),
                              }}
@@ -211,6 +211,7 @@ class ModelSamplingContinuousEDM:
     def patch(self, model, sampling, sigma_max, sigma_min):
         m = model.clone()
 
+        sampling_base = comfy.model_sampling.ModelSamplingContinuousEDM
         latent_format = None
         sigma_data = 1.0
         sampling_type = comfy.model_sampling.EPS
@@ -225,8 +226,11 @@ class ModelSamplingContinuousEDM:
             sampling_type = comfy.model_sampling.EDM
             sigma_data = 0.5
             latent_format = comfy.latent_formats.SDXL_Playground_2_5()
+        elif sampling == "cosmos_rflow":
+            sampling_type = comfy.model_sampling.COSMOS_RFLOW
+            sampling_base = comfy.model_sampling.ModelSamplingCosmosRFlow
 
-        class ModelSamplingAdvanced(comfy.model_sampling.ModelSamplingContinuousEDM, sampling_type):
+        class ModelSamplingAdvanced(sampling_base, sampling_type):
             pass
 
         model_sampling = ModelSamplingAdvanced(model.model.model_config)

comfy_extras/nodes_train.py

@@ -0,0 +1,709 @@
+import datetime
+import json
+import logging
+import os
+
+import numpy as np
+import safetensors
+import torch
+from PIL import Image, ImageDraw, ImageFont
+from PIL.PngImagePlugin import PngInfo
+import torch.utils.checkpoint
+import tqdm
+
+import comfy.samplers
+import comfy.sd
+import comfy.utils
+import comfy.model_management
+import comfy_extras.nodes_custom_sampler
+import folder_paths
+import node_helpers
+from comfy.cli_args import args
+from comfy.comfy_types.node_typing import IO
+from comfy.weight_adapter import adapters
+
+
+class TrainSampler(comfy.samplers.Sampler):
+
+    def __init__(self, loss_fn, optimizer, loss_callback=None):
+        self.loss_fn = loss_fn
+        self.optimizer = optimizer
+        self.loss_callback = loss_callback
+
+    def sample(self, model_wrap, sigmas, extra_args, callback, noise, latent_image=None, denoise_mask=None, disable_pbar=False):
+        self.optimizer.zero_grad()
+        noise = model_wrap.inner_model.model_sampling.noise_scaling(sigmas, noise, latent_image, False)
+        latent = model_wrap.inner_model.model_sampling.noise_scaling(
+            torch.zeros_like(sigmas),
+            torch.zeros_like(noise, requires_grad=True),
+            latent_image,
+            False
+        )
+
+        # Ensure model is in training mode and computing gradients
+        # x0 pred
+        denoised = model_wrap(noise, sigmas, **extra_args)
+        try:
+            loss = self.loss_fn(denoised, latent.clone())
+        except RuntimeError as e:
+            if "does not require grad and does not have a grad_fn" in str(e):
+                logging.info("WARNING: This is likely due to the model is loaded in inference mode.")
+        loss.backward()
+        if self.loss_callback:
+            self.loss_callback(loss.item())
+
+        self.optimizer.step()
+        # torch.cuda.memory._dump_snapshot("trainn.pickle")
+        # torch.cuda.memory._record_memory_history(enabled=None)
+        return torch.zeros_like(latent_image)
+
+
+class BiasDiff(torch.nn.Module):
+    def __init__(self, bias):
+        super().__init__()
+        self.bias = bias
+
+    def __call__(self, b):
+        org_dtype = b.dtype
+        return (b.to(self.bias) + self.bias).to(org_dtype)
+
+    def passive_memory_usage(self):
+        return self.bias.nelement() * self.bias.element_size()
+
+    def move_to(self, device):
+        self.to(device=device)
+        return self.passive_memory_usage()
+
+
+def load_and_process_images(image_files, input_dir, resize_method="None"):
+    """Utility function to load and process a list of images.
+
+    Args:
+        image_files: List of image filenames
+        input_dir: Base directory containing the images
+        resize_method: How to handle images of different sizes ("None", "Stretch", "Crop", "Pad")
+
+    Returns:
+        torch.Tensor: Batch of processed images
+    """
+    if not image_files:
+        raise ValueError("No valid images found in input")
+
+    output_images = []
+    w, h = None, None
+
+    for file in image_files:
+        image_path = os.path.join(input_dir, file)
+        img = node_helpers.pillow(Image.open, image_path)
+
+        if img.mode == "I":
+            img = img.point(lambda i: i * (1 / 255))
+        img = img.convert("RGB")
+
+        if w is None and h is None:
+            w, h = img.size[0], img.size[1]
+
+        # Resize image to first image
+        if img.size[0] != w or img.size[1] != h:
+            if resize_method == "Stretch":
+                img = img.resize((w, h), Image.Resampling.LANCZOS)
+            elif resize_method == "Crop":
+                img = img.crop((0, 0, w, h))
+            elif resize_method == "Pad":
+                img = img.resize((w, h), Image.Resampling.LANCZOS)
+            elif resize_method == "None":
+                raise ValueError(
+                    "Your input image size does not match the first image in the dataset. Either select a valid resize method or use the same size for all images."
+                )
+
+        img_array = np.array(img).astype(np.float32) / 255.0
+        img_tensor = torch.from_numpy(img_array)[None,]
+        output_images.append(img_tensor)
+
+    return torch.cat(output_images, dim=0)
+
+
+class LoadImageSetNode:
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "images": (
+                    [
+                        f
+                        for f in os.listdir(folder_paths.get_input_directory())
+                        if f.endswith((".png", ".jpg", ".jpeg", ".webp", ".bmp", ".gif", ".jpe", ".apng", ".tif", ".tiff"))
+                    ],
+                    {"image_upload": True, "allow_batch": True},
+                )
+            },
+            "optional": {
+                "resize_method": (
+                    ["None", "Stretch", "Crop", "Pad"],
+                    {"default": "None"},
+                ),
+            },
+        }
+
+    INPUT_IS_LIST = True
+    RETURN_TYPES = ("IMAGE",)
+    FUNCTION = "load_images"
+    CATEGORY = "loaders"
+    EXPERIMENTAL = True
+    DESCRIPTION = "Loads a batch of images from a directory for training."
+
+    @classmethod
+    def VALIDATE_INPUTS(s, images, resize_method):
+        filenames = images[0] if isinstance(images[0], list) else images
+
+        for image in filenames:
+            if not folder_paths.exists_annotated_filepath(image):
+                return "Invalid image file: {}".format(image)
+        return True
+
+    def load_images(self, input_files, resize_method):
+        input_dir = folder_paths.get_input_directory()
+        valid_extensions = [".png", ".jpg", ".jpeg", ".webp", ".bmp", ".gif", ".jpe", ".apng", ".tif", ".tiff"]
+        image_files = [
+            f
+            for f in input_files
+            if any(f.lower().endswith(ext) for ext in valid_extensions)
+        ]
+        output_tensor = load_and_process_images(image_files, input_dir, resize_method)
+        return (output_tensor,)
+
+
+class LoadImageSetFromFolderNode:
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "folder": (folder_paths.get_input_subfolders(), {"tooltip": "The folder to load images from."})
+            },
+            "optional": {
+                "resize_method": (
+                    ["None", "Stretch", "Crop", "Pad"],
+                    {"default": "None"},
+                ),
+            },
+        }
+
+    RETURN_TYPES = ("IMAGE",)
+    FUNCTION = "load_images"
+    CATEGORY = "loaders"
+    EXPERIMENTAL = True
+    DESCRIPTION = "Loads a batch of images from a directory for training."
+
+    def load_images(self, folder, resize_method):
+        sub_input_dir = os.path.join(folder_paths.get_input_directory(), folder)
+        valid_extensions = [".png", ".jpg", ".jpeg", ".webp"]
+        image_files = [
+            f
+            for f in os.listdir(sub_input_dir)
+            if any(f.lower().endswith(ext) for ext in valid_extensions)
+        ]
+        output_tensor = load_and_process_images(image_files, sub_input_dir, resize_method)
+        return (output_tensor,)
+
+
+def draw_loss_graph(loss_map, steps):
+    width, height = 500, 300
+    img = Image.new("RGB", (width, height), "white")
+    draw = ImageDraw.Draw(img)
+
+    min_loss, max_loss = min(loss_map.values()), max(loss_map.values())
+    scaled_loss = [(l - min_loss) / (max_loss - min_loss) for l in loss_map.values()]
+
+    prev_point = (0, height - int(scaled_loss[0] * height))
+    for i, l in enumerate(scaled_loss[1:], start=1):
+        x = int(i / (steps - 1) * width)
+        y = height - int(l * height)
+        draw.line([prev_point, (x, y)], fill="blue", width=2)
+        prev_point = (x, y)
+
+    return img
+
+
+def find_all_highest_child_module_with_forward(model: torch.nn.Module, result = None, name = None):
+    if result is None:
+        result = []
+    elif hasattr(model, "forward") and not isinstance(model, (torch.nn.ModuleList, torch.nn.Sequential, torch.nn.ModuleDict)):
+        result.append(model)
+        logging.debug(f"Found module with forward: {name} ({model.__class__.__name__})")
+        return result
+    name = name or "root"
+    for next_name, child in model.named_children():
+        find_all_highest_child_module_with_forward(child, result, f"{name}.{next_name}")
+    return result
+
+
+def patch(m):
+    if not hasattr(m, "forward"):
+        return
+    org_forward = m.forward
+    def fwd(args, kwargs):
+        return org_forward(*args, **kwargs)
+    def checkpointing_fwd(*args, **kwargs):
+        return torch.utils.checkpoint.checkpoint(
+            fwd, args, kwargs, use_reentrant=False
+        )
+    m.org_forward = org_forward
+    m.forward = checkpointing_fwd
+
+
+def unpatch(m):
+    if hasattr(m, "org_forward"):
+        m.forward = m.org_forward
+        del m.org_forward
+
+
+class TrainLoraNode:
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "model": (IO.MODEL, {"tooltip": "The model to train the LoRA on."}),
+                "latents": (
+                    "LATENT",
+                    {
+                        "tooltip": "The Latents to use for training, serve as dataset/input of the model."
+                    },
+                ),
+                "positive": (
+                    IO.CONDITIONING,
+                    {"tooltip": "The positive conditioning to use for training."},
+                ),
+                "batch_size": (
+                    IO.INT,
+                    {
+                        "default": 1,
+                        "min": 1,
+                        "max": 10000,
+                        "step": 1,
+                        "tooltip": "The batch size to use for training.",
+                    },
+                ),
+                "steps": (
+                    IO.INT,
+                    {
+                        "default": 16,
+                        "min": 1,
+                        "max": 100000,
+                        "tooltip": "The number of steps to train the LoRA for.",
+                    },
+                ),
+                "learning_rate": (
+                    IO.FLOAT,
+                    {
+                        "default": 0.0005,
+                        "min": 0.0000001,
+                        "max": 1.0,
+                        "step": 0.000001,
+                        "tooltip": "The learning rate to use for training.",
+                    },
+                ),
+                "rank": (
+                    IO.INT,
+                    {
+                        "default": 8,
+                        "min": 1,
+                        "max": 128,
+                        "tooltip": "The rank of the LoRA layers.",
+                    },
+                ),
+                "optimizer": (
+                    ["AdamW", "Adam", "SGD", "RMSprop"],
+                    {
+                        "default": "AdamW",
+                        "tooltip": "The optimizer to use for training.",
+                    },
+                ),
+                "loss_function": (
+                    ["MSE", "L1", "Huber", "SmoothL1"],
+                    {
+                        "default": "MSE",
+                        "tooltip": "The loss function to use for training.",
+                    },
+                ),
+                "seed": (
+                    IO.INT,
+                    {
+                        "default": 0,
+                        "min": 0,
+                        "max": 0xFFFFFFFFFFFFFFFF,
+                        "tooltip": "The seed to use for training (used in generator for LoRA weight initialization and noise sampling)",
+                    },
+                ),
+                "training_dtype": (
+                    ["bf16",  "fp32"],
+                    {"default": "bf16", "tooltip": "The dtype to use for training."},
+                ),
+                "lora_dtype": (
+                    ["bf16", "fp32"],
+                    {"default": "bf16", "tooltip": "The dtype to use for lora."},
+                ),
+                "existing_lora": (
+                    folder_paths.get_filename_list("loras") + ["[None]"],
+                    {
+                        "default": "[None]",
+                        "tooltip": "The existing LoRA to append to. Set to None for new LoRA.",
+                    },
+                ),
+            },
+        }
+
+    RETURN_TYPES = (IO.MODEL, IO.LORA_MODEL, IO.LOSS_MAP, IO.INT)
+    RETURN_NAMES = ("model_with_lora", "lora", "loss", "steps")
+    FUNCTION = "train"
+    CATEGORY = "training"
+    EXPERIMENTAL = True
+
+    def train(
+        self,
+        model,
+        latents,
+        positive,
+        batch_size,
+        steps,
+        learning_rate,
+        rank,
+        optimizer,
+        loss_function,
+        seed,
+        training_dtype,
+        lora_dtype,
+        existing_lora,
+    ):
+        mp = model.clone()
+        dtype = node_helpers.string_to_torch_dtype(training_dtype)
+        lora_dtype = node_helpers.string_to_torch_dtype(lora_dtype)
+        mp.set_model_compute_dtype(dtype)
+
+        latents = latents["samples"].to(dtype)
+        num_images = latents.shape[0]
+
+        with torch.inference_mode(False):
+            lora_sd = {}
+            generator = torch.Generator()
+            generator.manual_seed(seed)
+
+            # Load existing LoRA weights if provided
+            existing_weights = {}
+            existing_steps = 0
+            if existing_lora != "[None]":
+                lora_path = folder_paths.get_full_path_or_raise("loras", existing_lora)
+                # Extract steps from filename like "trained_lora_10_steps_20250225_203716"
+                existing_steps = int(existing_lora.split("_steps_")[0].split("_")[-1])
+                if lora_path:
+                    existing_weights = comfy.utils.load_torch_file(lora_path)
+
+            all_weight_adapters = []
+            for n, m in mp.model.named_modules():
+                if hasattr(m, "weight_function"):
+                    if m.weight is not None:
+                        key = "{}.weight".format(n)
+                        shape = m.weight.shape
+                        if len(shape) >= 2:
+                            alpha = float(existing_weights.get(f"{key}.alpha", 1.0))
+                            dora_scale = existing_weights.get(
+                                f"{key}.dora_scale", None
+                            )
+                            for adapter_cls in adapters:
+                                existing_adapter = adapter_cls.load(
+                                    n, existing_weights, alpha, dora_scale
+                                )
+                                if existing_adapter is not None:
+                                    break
+                            else:
+                                # If no existing adapter found, use LoRA
+                                # We will add algo option in the future
+                                existing_adapter = None
+                                adapter_cls = adapters[0]
+
+                            if existing_adapter is not None:
+                                train_adapter = existing_adapter.to_train().to(lora_dtype)
+                            else:
+                                # Use LoRA with alpha=1.0 by default
+                                train_adapter = adapter_cls.create_train(
+                                    m.weight, rank=rank, alpha=1.0
+                                ).to(lora_dtype)
+                            for name, parameter in train_adapter.named_parameters():
+                                lora_sd[f"{n}.{name}"] = parameter
+
+                            mp.add_weight_wrapper(key, train_adapter)
+                            all_weight_adapters.append(train_adapter)
+                        else:
+                            diff = torch.nn.Parameter(
+                                torch.zeros(
+                                    m.weight.shape, dtype=lora_dtype, requires_grad=True
+                                )
+                            )
+                            diff_module = BiasDiff(diff)
+                            mp.add_weight_wrapper(key, BiasDiff(diff))
+                            all_weight_adapters.append(diff_module)
+                            lora_sd["{}.diff".format(n)] = diff
+                    if hasattr(m, "bias") and m.bias is not None:
+                        key = "{}.bias".format(n)
+                        bias = torch.nn.Parameter(
+                            torch.zeros(m.bias.shape, dtype=lora_dtype, requires_grad=True)
+                        )
+                        bias_module = BiasDiff(bias)
+                        lora_sd["{}.diff_b".format(n)] = bias
+                        mp.add_weight_wrapper(key, BiasDiff(bias))
+                        all_weight_adapters.append(bias_module)
+
+            if optimizer == "Adam":
+                optimizer = torch.optim.Adam(lora_sd.values(), lr=learning_rate)
+            elif optimizer == "AdamW":
+                optimizer = torch.optim.AdamW(lora_sd.values(), lr=learning_rate)
+            elif optimizer == "SGD":
+                optimizer = torch.optim.SGD(lora_sd.values(), lr=learning_rate)
+            elif optimizer == "RMSprop":
+                optimizer = torch.optim.RMSprop(lora_sd.values(), lr=learning_rate)
+
+            # Setup loss function based on selection
+            if loss_function == "MSE":
+                criterion = torch.nn.MSELoss()
+            elif loss_function == "L1":
+                criterion = torch.nn.L1Loss()
+            elif loss_function == "Huber":
+                criterion = torch.nn.HuberLoss()
+            elif loss_function == "SmoothL1":
+                criterion = torch.nn.SmoothL1Loss()
+
+            # setup models
+            for m in find_all_highest_child_module_with_forward(mp.model.diffusion_model):
+                patch(m)
+            comfy.model_management.load_models_gpu([mp], memory_required=1e20, force_full_load=True)
+
+            # Setup sampler and guider like in test script
+            loss_map = {"loss": []}
+            def loss_callback(loss):
+                loss_map["loss"].append(loss)
+                pbar.set_postfix({"loss": f"{loss:.4f}"})
+            train_sampler = TrainSampler(
+                criterion, optimizer, loss_callback=loss_callback
+            )
+            guider = comfy_extras.nodes_custom_sampler.Guider_Basic(mp)
+            guider.set_conds(positive)  # Set conditioning from input
+            ss = comfy_extras.nodes_custom_sampler.SamplerCustomAdvanced()
+
+            # yoland: this currently resize to the first image in the dataset
+
+            # Training loop
+            torch.cuda.empty_cache()
+            try:
+                for step in (pbar:=tqdm.trange(steps, desc="Training LoRA", smoothing=0.01, disable=not comfy.utils.PROGRESS_BAR_ENABLED)):
+                    # Generate random sigma
+                    sigma = mp.model.model_sampling.percent_to_sigma(
+                        torch.rand((1,)).item()
+                    )
+                    sigma = torch.tensor([sigma])
+
+                    noise = comfy_extras.nodes_custom_sampler.Noise_RandomNoise(step * 1000 + seed)
+
+                    indices = torch.randperm(num_images)[:batch_size]
+                    ss.sample(
+                        noise, guider, train_sampler, sigma, {"samples": latents[indices].clone()}
+                    )
+            finally:
+                for m in mp.model.modules():
+                    unpatch(m)
+            del ss, train_sampler, optimizer
+            torch.cuda.empty_cache()
+
+            for adapter in all_weight_adapters:
+                adapter.requires_grad_(False)
+
+            for param in lora_sd:
+                lora_sd[param] = lora_sd[param].to(lora_dtype)
+
+            return (mp, lora_sd, loss_map, steps + existing_steps)
+
+
+class LoraModelLoader:
+    def __init__(self):
+        self.loaded_lora = None
+
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "model": ("MODEL", {"tooltip": "The diffusion model the LoRA will be applied to."}),
+                "lora": (IO.LORA_MODEL, {"tooltip": "The LoRA model to apply to the diffusion model."}),
+                "strength_model": ("FLOAT", {"default": 1.0, "min": -100.0, "max": 100.0, "step": 0.01, "tooltip": "How strongly to modify the diffusion model. This value can be negative."}),
+            }
+        }
+
+    RETURN_TYPES = ("MODEL",)
+    OUTPUT_TOOLTIPS = ("The modified diffusion model.",)
+    FUNCTION = "load_lora_model"
+
+    CATEGORY = "loaders"
+    DESCRIPTION = "Load Trained LoRA weights from Train LoRA node."
+    EXPERIMENTAL = True
+
+    def load_lora_model(self, model, lora, strength_model):
+        if strength_model == 0:
+            return (model, )
+
+        model_lora, _ = comfy.sd.load_lora_for_models(model, None, lora, strength_model, 0)
+        return (model_lora, )
+
+
+class SaveLoRA:
+    def __init__(self):
+        self.output_dir = folder_paths.get_output_directory()
+
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "lora": (
+                    IO.LORA_MODEL,
+                    {
+                        "tooltip": "The LoRA model to save. Do not use the model with LoRA layers."
+                    },
+                ),
+                "prefix": (
+                    "STRING",
+                    {
+                        "default": "loras/ComfyUI_trained_lora",
+                        "tooltip": "The prefix to use for the saved LoRA file.",
+                    },
+                ),
+            },
+            "optional": {
+                "steps": (
+                    IO.INT,
+                    {
+                        "forceInput": True,
+                        "tooltip": "Optional: The number of steps to LoRA has been trained for, used to name the saved file.",
+                    },
+                ),
+            },
+        }
+
+    RETURN_TYPES = ()
+    FUNCTION = "save"
+    CATEGORY = "loaders"
+    EXPERIMENTAL = True
+    OUTPUT_NODE = True
+
+    def save(self, lora, prefix, steps=None):
+        full_output_folder, filename, counter, subfolder, filename_prefix = folder_paths.get_save_image_path(prefix, self.output_dir)
+        if steps is None:
+            output_checkpoint = f"{filename}_{counter:05}_.safetensors"
+        else:
+            output_checkpoint = f"{filename}_{steps}_steps_{counter:05}_.safetensors"
+        output_checkpoint = os.path.join(full_output_folder, output_checkpoint)
+        safetensors.torch.save_file(lora, output_checkpoint)
+        return {}
+
+
+class LossGraphNode:
+    def __init__(self):
+        self.output_dir = folder_paths.get_temp_directory()
+
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "loss": (IO.LOSS_MAP, {"default": {}}),
+                "filename_prefix": (IO.STRING, {"default": "loss_graph"}),
+            },
+            "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"},
+        }
+
+    RETURN_TYPES = ()
+    FUNCTION = "plot_loss"
+    OUTPUT_NODE = True
+    CATEGORY = "training"
+    EXPERIMENTAL = True
+    DESCRIPTION = "Plots the loss graph and saves it to the output directory."
+
+    def plot_loss(self, loss, filename_prefix, prompt=None, extra_pnginfo=None):
+        loss_values = loss["loss"]
+        width, height = 800, 480
+        margin = 40
+
+        img = Image.new(
+            "RGB", (width + margin, height + margin), "white"
+        )  # Extend canvas
+        draw = ImageDraw.Draw(img)
+
+        min_loss, max_loss = min(loss_values), max(loss_values)
+        scaled_loss = [(l - min_loss) / (max_loss - min_loss) for l in loss_values]
+
+        steps = len(loss_values)
+
+        prev_point = (margin, height - int(scaled_loss[0] * height))
+        for i, l in enumerate(scaled_loss[1:], start=1):
+            x = margin + int(i / steps * width)  # Scale X properly
+            y = height - int(l * height)
+            draw.line([prev_point, (x, y)], fill="blue", width=2)
+            prev_point = (x, y)
+
+        draw.line([(margin, 0), (margin, height)], fill="black", width=2)  # Y-axis
+        draw.line(
+            [(margin, height), (width + margin, height)], fill="black", width=2
+        )  # X-axis
+
+        font = None
+        try:
+            font = ImageFont.truetype("arial.ttf", 12)
+        except IOError:
+            font = ImageFont.load_default()
+
+        # Add axis labels
+        draw.text((5, height // 2), "Loss", font=font, fill="black")
+        draw.text((width // 2, height + 10), "Steps", font=font, fill="black")
+
+        # Add min/max loss values
+        draw.text((margin - 30, 0), f"{max_loss:.2f}", font=font, fill="black")
+        draw.text(
+            (margin - 30, height - 10), f"{min_loss:.2f}", font=font, fill="black"
+        )
+
+        metadata = None
+        if not args.disable_metadata:
+            metadata = PngInfo()
+            if prompt is not None:
+                metadata.add_text("prompt", json.dumps(prompt))
+            if extra_pnginfo is not None:
+                for x in extra_pnginfo:
+                    metadata.add_text(x, json.dumps(extra_pnginfo[x]))
+
+        date = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
+        img.save(
+            os.path.join(self.output_dir, f"{filename_prefix}_{date}.png"),
+            pnginfo=metadata,
+        )
+        return {
+            "ui": {
+                "images": [
+                    {
+                        "filename": f"{filename_prefix}_{date}.png",
+                        "subfolder": "",
+                        "type": "temp",
+                    }
+                ]
+            }
+        }
+
+
+NODE_CLASS_MAPPINGS = {
+    "TrainLoraNode": TrainLoraNode,
+    "SaveLoRANode": SaveLoRA,
+    "LoraModelLoader": LoraModelLoader,
+    "LoadImageSetFromFolderNode": LoadImageSetFromFolderNode,
+    "LossGraphNode": LossGraphNode,
+}
+
+NODE_DISPLAY_NAME_MAPPINGS = {
+    "TrainLoraNode": "Train LoRA",
+    "SaveLoRANode": "Save LoRA Weights",
+    "LoraModelLoader": "Load LoRA Model",
+    "LoadImageSetFromFolderNode": "Load Image Dataset from Folder",
+    "LossGraphNode": "Plot Loss Graph",
+}

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "comfyui"
-version = "0.3.40"
+version = "0.3.41"
 description = "An installable version of ComfyUI"
 readme = "README.md"
 authors = [
@@ -104,6 +104,8 @@ dependencies = [
     "scikit-learn>=1.4.1",
     # everything that is a torch extension will need setuptools, so just include it
     "setuptools",
+    "alembic",
+    "SQLAlchemy",
 ]
 
 [build-system]

tests/unit/folder_paths_test/misc_test.py

@@ -0,0 +1,51 @@
+import pytest
+import os
+import tempfile
+from folder_paths import get_input_subfolders, set_input_directory
+
+@pytest.fixture(scope="module")
+def mock_folder_structure():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        # Create a nested folder structure
+        folders = [
+            "folder1",
+            "folder1/subfolder1",
+            "folder1/subfolder2",
+            "folder2",
+            "folder2/deep",
+            "folder2/deep/nested",
+            "empty_folder"
+        ]
+
+        # Create the folders
+        for folder in folders:
+            os.makedirs(os.path.join(temp_dir, folder))
+
+        # Add some files to test they're not included
+        with open(os.path.join(temp_dir, "root_file.txt"), "w") as f:
+            f.write("test")
+        with open(os.path.join(temp_dir, "folder1", "test.txt"), "w") as f:
+            f.write("test")
+
+        set_input_directory(temp_dir)
+        yield temp_dir
+
+
+def test_gets_all_folders(mock_folder_structure):
+    folders = get_input_subfolders()
+    expected = ["folder1", "folder1/subfolder1", "folder1/subfolder2",
+                "folder2", "folder2/deep", "folder2/deep/nested", "empty_folder"]
+    assert sorted(folders) == sorted(expected)
+
+
+def test_handles_nonexistent_input_directory():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        nonexistent = os.path.join(temp_dir, "nonexistent")
+        set_input_directory(nonexistent)
+        assert get_input_subfolders() == []
+
+
+def test_empty_input_directory():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        set_input_directory(temp_dir)
+        assert get_input_subfolders() == []  # Empty since we don't include root