tests/models/autoencoders/test_models_autoencoder_kl_temporal_decoder.py

# coding=utf-8
# Copyright 2025 HuggingFace Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import unittest

from diffusers import AutoencoderKLTemporalDecoder

from ...testing_utils import (
    enable_full_determinism,
    floats_tensor,
    torch_device,
)
from ..test_modeling_common import ModelTesterMixin
from .testing_utils import AutoencoderTesterMixin


enable_full_determinism()


class AutoencoderKLTemporalDecoderTests(ModelTesterMixin, AutoencoderTesterMixin, unittest.TestCase):
    model_class = AutoencoderKLTemporalDecoder
    main_input_name = "sample"
    base_precision = 1e-2

    @property
    def dummy_input(self):
        batch_size = 3
        num_channels = 3
        sizes = (32, 32)

        image = floats_tensor((batch_size, num_channels) + sizes).to(torch_device)
        num_frames = 3

        return {"sample": image, "num_frames": num_frames}

    @property
    def input_shape(self):
        return (3, 32, 32)

    @property
    def output_shape(self):
        return (3, 32, 32)

    def prepare_init_args_and_inputs_for_common(self):
        init_dict = {
            "block_out_channels": [32, 64],
            "in_channels": 3,
            "out_channels": 3,
            "down_block_types": ["DownEncoderBlock2D", "DownEncoderBlock2D"],
            "latent_channels": 4,
            "layers_per_block": 2,
        }
        inputs_dict = self.dummy_input
        return init_dict, inputs_dict

    def test_gradient_checkpointing_is_applied(self):
        expected_set = {"Encoder", "TemporalDecoder", "UNetMidBlock2D"}
        super().test_gradient_checkpointing_is_applied(expected_set=expected_set)