Spaces:
Runtime error
Runtime error
File size: 4,577 Bytes
fd43906 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 |
# coding=utf-8
# Copyright 2023 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
import numpy as np
import torch
from diffusers import DDIMPipeline, DDIMScheduler, UNet2DModel
from diffusers.utils.testing_utils import require_torch_gpu, slow, torch_device
from ...pipeline_params import UNCONDITIONAL_IMAGE_GENERATION_BATCH_PARAMS, UNCONDITIONAL_IMAGE_GENERATION_PARAMS
from ...test_pipelines_common import PipelineTesterMixin
torch.backends.cuda.matmul.allow_tf32 = False
class DDIMPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
pipeline_class = DDIMPipeline
params = UNCONDITIONAL_IMAGE_GENERATION_PARAMS
required_optional_params = PipelineTesterMixin.required_optional_params - {
"num_images_per_prompt",
"latents",
"callback",
"callback_steps",
}
batch_params = UNCONDITIONAL_IMAGE_GENERATION_BATCH_PARAMS
test_cpu_offload = False
def get_dummy_components(self):
torch.manual_seed(0)
unet = UNet2DModel(
block_out_channels=(32, 64),
layers_per_block=2,
sample_size=32,
in_channels=3,
out_channels=3,
down_block_types=("DownBlock2D", "AttnDownBlock2D"),
up_block_types=("AttnUpBlock2D", "UpBlock2D"),
)
scheduler = DDIMScheduler()
components = {"unet": unet, "scheduler": scheduler}
return components
def get_dummy_inputs(self, device, seed=0):
if str(device).startswith("mps"):
generator = torch.manual_seed(seed)
else:
generator = torch.Generator(device=device).manual_seed(seed)
inputs = {
"batch_size": 1,
"generator": generator,
"num_inference_steps": 2,
"output_type": "numpy",
}
return inputs
def test_inference(self):
device = "cpu"
components = self.get_dummy_components()
pipe = self.pipeline_class(**components)
pipe.to(device)
pipe.set_progress_bar_config(disable=None)
inputs = self.get_dummy_inputs(device)
image = pipe(**inputs).images
image_slice = image[0, -3:, -3:, -1]
self.assertEqual(image.shape, (1, 32, 32, 3))
expected_slice = np.array(
[1.000e00, 5.717e-01, 4.717e-01, 1.000e00, 0.000e00, 1.000e00, 3.000e-04, 0.000e00, 9.000e-04]
)
max_diff = np.abs(image_slice.flatten() - expected_slice).max()
self.assertLessEqual(max_diff, 1e-3)
@slow
@require_torch_gpu
class DDIMPipelineIntegrationTests(unittest.TestCase):
def test_inference_cifar10(self):
model_id = "google/ddpm-cifar10-32"
unet = UNet2DModel.from_pretrained(model_id)
scheduler = DDIMScheduler()
ddim = DDIMPipeline(unet=unet, scheduler=scheduler)
ddim.to(torch_device)
ddim.set_progress_bar_config(disable=None)
generator = torch.manual_seed(0)
image = ddim(generator=generator, eta=0.0, output_type="numpy").images
image_slice = image[0, -3:, -3:, -1]
assert image.shape == (1, 32, 32, 3)
expected_slice = np.array([0.1723, 0.1617, 0.1600, 0.1626, 0.1497, 0.1513, 0.1505, 0.1442, 0.1453])
assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2
def test_inference_ema_bedroom(self):
model_id = "google/ddpm-ema-bedroom-256"
unet = UNet2DModel.from_pretrained(model_id)
scheduler = DDIMScheduler.from_pretrained(model_id)
ddpm = DDIMPipeline(unet=unet, scheduler=scheduler)
ddpm.to(torch_device)
ddpm.set_progress_bar_config(disable=None)
generator = torch.manual_seed(0)
image = ddpm(generator=generator, output_type="numpy").images
image_slice = image[0, -3:, -3:, -1]
assert image.shape == (1, 256, 256, 3)
expected_slice = np.array([0.0060, 0.0201, 0.0344, 0.0024, 0.0018, 0.0002, 0.0022, 0.0000, 0.0069])
assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2
|