mirror of
https://github.com/huggingface/diffusers.git
synced 2026-10-24 20:59:27 +03:00
357 lines
13 KiB
Python
357 lines
13 KiB
Python
# coding=utf-8
|
|
# Copyright 2026 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 gc
|
|
import random
|
|
|
|
import numpy as np
|
|
import pytest
|
|
import torch
|
|
from PIL import Image
|
|
from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer
|
|
|
|
from diffusers import AutoencoderKL, DDIMScheduler, DDPMScheduler, StableDiffusionUpscalePipeline, UNet2DConditionModel
|
|
|
|
from ...testing_utils import (
|
|
assert_tensors_close,
|
|
backend_empty_cache,
|
|
backend_max_memory_allocated,
|
|
backend_reset_max_memory_allocated,
|
|
backend_reset_peak_memory_stats,
|
|
enable_full_determinism,
|
|
floats_tensor,
|
|
load_image,
|
|
load_numpy,
|
|
require_accelerator,
|
|
require_torch_accelerator,
|
|
slow,
|
|
torch_device,
|
|
)
|
|
from ..pipeline_params import TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
|
|
from ..testing_utils import (
|
|
BasePipelineTesterConfig,
|
|
MemoryTesterMixin,
|
|
PipelineTesterMixin,
|
|
)
|
|
|
|
|
|
enable_full_determinism()
|
|
|
|
|
|
class StableDiffusionUpscalePipelineTesterConfig(BasePipelineTesterConfig):
|
|
pipeline_class = StableDiffusionUpscalePipeline
|
|
# `TEXT_GUIDED_IMAGE_VARIATION_PARAMS` without `height` / `width`: the output resolution follows the image.
|
|
required_input_params_in_call_signature = frozenset(
|
|
["prompt", "image", "guidance_scale", "negative_prompt", "prompt_embeds", "negative_prompt_embeds"]
|
|
)
|
|
batch_input_params = TEXT_GUIDED_IMAGE_VARIATION_BATCH_PARAMS
|
|
# The 64x64 low-resolution input is upscaled 4x.
|
|
output_shape = (3, 256, 256)
|
|
|
|
@property
|
|
def dummy_image(self):
|
|
batch_size = 1
|
|
num_channels = 3
|
|
sizes = (32, 32)
|
|
|
|
image = floats_tensor((batch_size, num_channels) + sizes, rng=random.Random(0))
|
|
return image
|
|
|
|
@property
|
|
def dummy_cond_unet_upscale(self):
|
|
torch.manual_seed(0)
|
|
model = UNet2DConditionModel(
|
|
block_out_channels=(32, 32, 64),
|
|
layers_per_block=2,
|
|
sample_size=32,
|
|
in_channels=7,
|
|
out_channels=4,
|
|
down_block_types=("DownBlock2D", "CrossAttnDownBlock2D", "CrossAttnDownBlock2D"),
|
|
up_block_types=("CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "UpBlock2D"),
|
|
cross_attention_dim=32,
|
|
# SD2-specific config below
|
|
attention_head_dim=8,
|
|
use_linear_projection=True,
|
|
only_cross_attention=(True, True, False),
|
|
num_class_embeds=100,
|
|
)
|
|
return model
|
|
|
|
@property
|
|
def dummy_vae(self):
|
|
torch.manual_seed(0)
|
|
model = AutoencoderKL(
|
|
block_out_channels=[32, 32, 64],
|
|
in_channels=3,
|
|
out_channels=3,
|
|
down_block_types=["DownEncoderBlock2D", "DownEncoderBlock2D", "DownEncoderBlock2D"],
|
|
up_block_types=["UpDecoderBlock2D", "UpDecoderBlock2D", "UpDecoderBlock2D"],
|
|
latent_channels=4,
|
|
)
|
|
return model
|
|
|
|
@property
|
|
def dummy_text_encoder(self):
|
|
torch.manual_seed(0)
|
|
config = CLIPTextConfig(
|
|
bos_token_id=0,
|
|
eos_token_id=2,
|
|
hidden_size=32,
|
|
intermediate_size=37,
|
|
layer_norm_eps=1e-05,
|
|
num_attention_heads=4,
|
|
num_hidden_layers=5,
|
|
pad_token_id=1,
|
|
vocab_size=1000,
|
|
# SD2-specific config below
|
|
hidden_act="gelu",
|
|
projection_dim=512,
|
|
)
|
|
return CLIPTextModel(config)
|
|
|
|
@property
|
|
def dummy_low_res_image(self):
|
|
image = self.dummy_image.permute(0, 2, 3, 1)[0]
|
|
return Image.fromarray(np.uint8(image)).convert("RGB").resize((64, 64))
|
|
|
|
def get_dummy_components(self):
|
|
return {
|
|
"unet": self.dummy_cond_unet_upscale,
|
|
"low_res_scheduler": DDPMScheduler(),
|
|
"scheduler": DDIMScheduler(prediction_type="v_prediction"),
|
|
"vae": self.dummy_vae,
|
|
"text_encoder": self.dummy_text_encoder,
|
|
"tokenizer": CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip"),
|
|
"safety_checker": None,
|
|
"feature_extractor": None,
|
|
"watermarker": None,
|
|
"max_noise_level": 350,
|
|
}
|
|
|
|
def get_dummy_inputs(self):
|
|
return {
|
|
"prompt": "A painting of a squirrel eating a burger",
|
|
"image": self.dummy_low_res_image,
|
|
"generator": self.get_generator(0),
|
|
"guidance_scale": 6.0,
|
|
"noise_level": 20,
|
|
"num_inference_steps": 2,
|
|
# Request torch outputs so tests compare torch tensors directly (see `BasePipelineTesterConfig`).
|
|
# Note `"pt"` images are `(batch, channels, height, width)`, unlike `"np"` (`(batch, h, w, c)`).
|
|
"output_type": "pt",
|
|
}
|
|
|
|
|
|
class TestStableDiffusionUpscalePipeline(StableDiffusionUpscalePipelineTesterConfig, PipelineTesterMixin):
|
|
def test_stable_diffusion_upscale(self):
|
|
# Run on CPU: the expected slice below is CPU-specific.
|
|
sd_pipe = self.get_pipeline()
|
|
|
|
image = sd_pipe(**self.get_dummy_inputs()).images
|
|
image_slice = image[0, -1, -3:, -3:]
|
|
|
|
assert image.shape == (1, *self.output_shape)
|
|
# fmt: off
|
|
expected_slice = torch.tensor([0.2631, 0.4038, 0.4338, 0.4254, 0.5002, 0.4831, 0.5073, 0.5619, 0.5597])
|
|
# fmt: on
|
|
assert_tensors_close(image_slice.flatten(), expected_slice, atol=1e-2)
|
|
|
|
def test_stable_diffusion_upscale_batch(self):
|
|
sd_pipe = self.get_pipeline()
|
|
|
|
inputs = self.get_dummy_inputs()
|
|
inputs["prompt"] = 2 * [inputs["prompt"]]
|
|
inputs["image"] = 2 * [inputs["image"]]
|
|
del inputs["generator"]
|
|
image = sd_pipe(**inputs).images
|
|
assert image.shape[0] == 2
|
|
|
|
inputs = self.get_dummy_inputs()
|
|
inputs["prompt"] = [inputs["prompt"]]
|
|
image = sd_pipe(**inputs, num_images_per_prompt=2).images
|
|
assert image.shape[0] == 2
|
|
|
|
def test_stable_diffusion_upscale_prompt_embeds(self):
|
|
# `encode_prompt` is called with `torch_device` below, so the pipeline has to live there too.
|
|
sd_pipe = self.get_pipeline().to(torch_device)
|
|
|
|
image = sd_pipe(**self.get_dummy_inputs()).images
|
|
|
|
inputs = self.get_dummy_inputs()
|
|
prompt = inputs.pop("prompt")
|
|
prompt_embeds, negative_prompt_embeds = sd_pipe.encode_prompt(prompt, torch_device, 1, False)
|
|
if negative_prompt_embeds is not None:
|
|
prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
|
|
|
|
inputs["image"] = [inputs["image"]]
|
|
image_from_prompt_embeds = sd_pipe(prompt_embeds=prompt_embeds, **inputs, return_dict=False)[0]
|
|
|
|
assert_tensors_close(
|
|
image_from_prompt_embeds[0, -1, -3:, -3:],
|
|
image[0, -1, -3:, -3:],
|
|
atol=1e-2,
|
|
msg="Passing `prompt_embeds` changed the output.",
|
|
)
|
|
|
|
def test_encode_prompt_works_in_isolation(self):
|
|
extra_required_param_value_dict = {
|
|
"device": torch.device(torch_device).type,
|
|
"do_classifier_free_guidance": self.get_dummy_inputs().get("guidance_scale", 1.0) > 1.0,
|
|
}
|
|
return super().test_encode_prompt_works_in_isolation(extra_required_param_value_dict)
|
|
|
|
@require_accelerator
|
|
def test_stable_diffusion_upscale_fp16(self):
|
|
"""Test that stable diffusion upscale works with fp16"""
|
|
components = self.get_dummy_components()
|
|
# put models in fp16, except vae as it overflows in fp16
|
|
components["unet"] = components["unet"].half()
|
|
components["text_encoder"] = components["text_encoder"].half()
|
|
sd_pipe = self.get_pipeline(**components).to(torch_device)
|
|
|
|
inputs = self.get_dummy_inputs()
|
|
del inputs["guidance_scale"]
|
|
del inputs["noise_level"]
|
|
image = sd_pipe(**inputs).images
|
|
|
|
assert image.shape == (1, *self.output_shape)
|
|
|
|
def test_stable_diffusion_upscale_from_save_pretrained(self, tmp_path):
|
|
sd_pipe = self.get_pipeline()
|
|
sd_pipe.save_pretrained(tmp_path)
|
|
sd_pipe_loaded = StableDiffusionUpscalePipeline.from_pretrained(tmp_path)
|
|
sd_pipe_loaded.set_progress_bar_config(disable=None)
|
|
|
|
image_slices = []
|
|
for pipe in [sd_pipe, sd_pipe_loaded]:
|
|
image = pipe(**self.get_dummy_inputs()).images
|
|
image_slices.append(image[0, -1, -3:, -3:].flatten())
|
|
|
|
assert_tensors_close(image_slices[1], image_slices[0], atol=1e-3)
|
|
|
|
|
|
class TestStableDiffusionUpscalePipelineMemory(StableDiffusionUpscalePipelineTesterConfig, MemoryTesterMixin):
|
|
"""Memory optimization tests (CPU offload, group offload, layerwise casting) for the SD upscale pipeline."""
|
|
|
|
|
|
@slow
|
|
@require_torch_accelerator
|
|
class TestStableDiffusionUpscalePipelineIntegration:
|
|
@pytest.fixture(autouse=True)
|
|
def cleanup(self):
|
|
# clean up the VRAM before and after each test
|
|
gc.collect()
|
|
backend_empty_cache(torch_device)
|
|
yield
|
|
gc.collect()
|
|
backend_empty_cache(torch_device)
|
|
|
|
def test_stable_diffusion_upscale_pipeline(self):
|
|
image = load_image(
|
|
"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main"
|
|
"/sd2-upscale/low_res_cat.png"
|
|
)
|
|
expected_image = load_numpy(
|
|
"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-upscale"
|
|
"/upsampled_cat.npy"
|
|
)
|
|
|
|
model_id = "stabilityai/stable-diffusion-x4-upscaler"
|
|
pipe = StableDiffusionUpscalePipeline.from_pretrained(model_id)
|
|
pipe.to(torch_device)
|
|
pipe.set_progress_bar_config(disable=None)
|
|
pipe.enable_attention_slicing()
|
|
|
|
prompt = "a cat sitting on a park bench"
|
|
|
|
generator = torch.manual_seed(0)
|
|
output = pipe(
|
|
prompt=prompt,
|
|
image=image,
|
|
generator=generator,
|
|
output_type="np",
|
|
)
|
|
image = output.images[0]
|
|
|
|
assert image.shape == (512, 512, 3)
|
|
assert np.abs(expected_image - image).max() < 1e-3
|
|
|
|
def test_stable_diffusion_upscale_pipeline_fp16(self):
|
|
image = load_image(
|
|
"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main"
|
|
"/sd2-upscale/low_res_cat.png"
|
|
)
|
|
expected_image = load_numpy(
|
|
"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd2-upscale"
|
|
"/upsampled_cat_fp16.npy"
|
|
)
|
|
|
|
model_id = "stabilityai/stable-diffusion-x4-upscaler"
|
|
pipe = StableDiffusionUpscalePipeline.from_pretrained(
|
|
model_id,
|
|
torch_dtype=torch.float16,
|
|
)
|
|
pipe.to(torch_device)
|
|
pipe.set_progress_bar_config(disable=None)
|
|
pipe.enable_attention_slicing()
|
|
|
|
prompt = "a cat sitting on a park bench"
|
|
|
|
generator = torch.manual_seed(0)
|
|
output = pipe(
|
|
prompt=prompt,
|
|
image=image,
|
|
generator=generator,
|
|
output_type="np",
|
|
)
|
|
image = output.images[0]
|
|
|
|
assert image.shape == (512, 512, 3)
|
|
assert np.abs(expected_image - image).max() < 5e-1
|
|
|
|
def test_stable_diffusion_pipeline_with_sequential_cpu_offloading(self):
|
|
backend_empty_cache(torch_device)
|
|
backend_reset_max_memory_allocated(torch_device)
|
|
backend_reset_peak_memory_stats(torch_device)
|
|
|
|
image = load_image(
|
|
"https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main"
|
|
"/sd2-upscale/low_res_cat.png"
|
|
)
|
|
|
|
model_id = "stabilityai/stable-diffusion-x4-upscaler"
|
|
pipe = StableDiffusionUpscalePipeline.from_pretrained(
|
|
model_id,
|
|
torch_dtype=torch.float16,
|
|
)
|
|
pipe.set_progress_bar_config(disable=None)
|
|
pipe.enable_attention_slicing(1)
|
|
pipe.enable_sequential_cpu_offload(device=torch_device)
|
|
|
|
prompt = "a cat sitting on a park bench"
|
|
|
|
generator = torch.manual_seed(0)
|
|
_ = pipe(
|
|
prompt=prompt,
|
|
image=image,
|
|
generator=generator,
|
|
num_inference_steps=5,
|
|
output_type="np",
|
|
)
|
|
|
|
mem_bytes = backend_max_memory_allocated(torch_device)
|
|
# make sure that less than 2.9 GB is allocated
|
|
assert mem_bytes < 2.9 * 10**9
|