Some checks are pending
Publish to Comfy registry / Publish Custom Node to registry (push) Waiting to run
82 lines
2.3 KiB
Python
82 lines
2.3 KiB
Python
import ast
|
|
from pathlib import Path
|
|
|
|
import torch
|
|
|
|
|
|
def _load_make_ref_video_class():
|
|
path = Path(__file__).parents[1] / "nodes" / "ltxv.py"
|
|
tree = ast.parse(path.read_text(encoding="utf-8"))
|
|
class_node = next(
|
|
node for node in tree.body
|
|
if isinstance(node, ast.ClassDef) and node.name == "LTXVMakeRefVideo"
|
|
)
|
|
module = ast.Module(body=[class_node], type_ignores=[])
|
|
|
|
class _ComfyNode:
|
|
pass
|
|
|
|
class _NodeOutput:
|
|
def __init__(self, *values):
|
|
self.values = values
|
|
|
|
namespace = {
|
|
"io": type("IO", (), {"ComfyNode": _ComfyNode, "NodeOutput": _NodeOutput}),
|
|
"torch": torch,
|
|
}
|
|
exec(compile(module, str(path), "exec"), namespace)
|
|
return namespace["LTXVMakeRefVideo"]
|
|
|
|
|
|
def _labeled_images(count: int) -> torch.Tensor:
|
|
return torch.arange(count, dtype=torch.float32).reshape(count, 1, 1, 1)
|
|
|
|
|
|
def test_single_image_is_used_as_background_for_all_frames():
|
|
node = _load_make_ref_video_class()
|
|
|
|
frames = node._expand_frames(_labeled_images(1), 41)
|
|
|
|
assert frames[:, 0, 0, 0].tolist() == [0.0] * 41
|
|
|
|
|
|
def test_last_image_occupies_final_latent_section():
|
|
node = _load_make_ref_video_class()
|
|
|
|
frames = node._expand_frames(_labeled_images(3), 41)
|
|
|
|
assert frames[:, 0, 0, 0].tolist() == [0.0] * 17 + [1.0] * 16 + [2.0] * 8
|
|
|
|
|
|
def test_subjects_are_distributed_by_latent_sections_before_background():
|
|
node = _load_make_ref_video_class()
|
|
|
|
frames = node._expand_frames(_labeled_images(5), 41)
|
|
|
|
assert frames[:, 0, 0, 0].tolist() == (
|
|
[0.0] * 9 + [1.0] * 8 + [2.0] * 8 + [3.0] * 8 + [4.0] * 8
|
|
)
|
|
|
|
|
|
def test_single_list_value_preserves_existing_image_batch_behavior():
|
|
node = _load_make_ref_video_class()
|
|
batch = _labeled_images(3)
|
|
|
|
output = node.execute([batch], [41]).values[0]
|
|
|
|
assert output[:, 0, 0, 0].tolist() == [0.0] * 17 + [1.0] * 16 + [2.0] * 8
|
|
|
|
|
|
def test_multiple_list_values_fit_subjects_to_the_background_canvas():
|
|
node = _load_make_ref_video_class()
|
|
subject = torch.zeros(1, 2, 4, 3)
|
|
background = torch.full((1, 4, 4, 3), 0.25)
|
|
|
|
output = node.execute([subject, background], [17]).values[0]
|
|
|
|
assert output.shape == (17, 4, 4, 3)
|
|
assert torch.all(output[0, 0] == 1)
|
|
assert torch.all(output[0, 1:3] == 0)
|
|
assert torch.all(output[0, 3] == 1)
|
|
assert torch.all(output[-1] == 0.25)
|