File size: 7,272 Bytes
06646a9 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 |
import random
import unittest
import numpy as np
import torch
from transformers import Qwen2_5_VLConfig, Qwen2_5_VLForConditionalGeneration, Qwen2Tokenizer
from diffusers import (
AutoencoderKLQwenImage,
FlowMatchEulerDiscreteScheduler,
QwenImageImg2ImgPipeline,
QwenImageTransformer2DModel,
)
from ...testing_utils import (
enable_full_determinism,
floats_tensor,
torch_device,
)
from ..test_pipelines_common import PipelineTesterMixin, to_np
enable_full_determinism()
class QwenImageImg2ImgPipelineFastTests(unittest.TestCase, PipelineTesterMixin):
pipeline_class = QwenImageImg2ImgPipeline
params = frozenset(["prompt", "image", "height", "width", "guidance_scale", "true_cfg_scale", "strength"])
batch_params = frozenset(["prompt", "image"])
image_params = frozenset(["image"])
image_latents_params = frozenset(["latents"])
required_optional_params = frozenset(
[
"num_inference_steps",
"generator",
"latents",
"return_dict",
"callback_on_step_end",
"callback_on_step_end_tensor_inputs",
]
)
supports_dduf = False
test_xformers_attention = False
test_attention_slicing = True
test_layerwise_casting = True
test_group_offloading = True
def get_dummy_components(self):
torch.manual_seed(0)
transformer = QwenImageTransformer2DModel(
patch_size=2,
in_channels=16,
out_channels=4,
num_layers=2,
attention_head_dim=16,
num_attention_heads=3,
joint_attention_dim=16,
guidance_embeds=False,
axes_dims_rope=(8, 4, 4),
)
torch.manual_seed(0)
z_dim = 4
vae = AutoencoderKLQwenImage(
base_dim=z_dim * 6,
z_dim=z_dim,
dim_mult=[1, 2, 4],
num_res_blocks=1,
temperal_downsample=[False, True],
latents_mean=[0.0] * 4,
latents_std=[1.0] * 4,
)
torch.manual_seed(0)
scheduler = FlowMatchEulerDiscreteScheduler()
torch.manual_seed(0)
config = Qwen2_5_VLConfig(
text_config={
"hidden_size": 16,
"intermediate_size": 16,
"num_hidden_layers": 2,
"num_attention_heads": 2,
"num_key_value_heads": 2,
"rope_scaling": {
"mrope_section": [1, 1, 2],
"rope_type": "default",
"type": "default",
},
"rope_theta": 1000000.0,
},
vision_config={
"depth": 2,
"hidden_size": 16,
"intermediate_size": 16,
"num_heads": 2,
"out_hidden_size": 16,
},
hidden_size=16,
vocab_size=152064,
vision_end_token_id=151653,
vision_start_token_id=151652,
vision_token_id=151654,
)
text_encoder = Qwen2_5_VLForConditionalGeneration(config)
tokenizer = Qwen2Tokenizer.from_pretrained("hf-internal-testing/tiny-random-Qwen2VLForConditionalGeneration")
return {
"transformer": transformer,
"vae": vae,
"scheduler": scheduler,
"text_encoder": text_encoder,
"tokenizer": tokenizer,
}
def get_dummy_inputs(self, device, seed=0):
image = floats_tensor((1, 3, 32, 32), rng=random.Random(seed)).to(device)
if str(device).startswith("mps"):
generator = torch.manual_seed(seed)
else:
generator = torch.Generator(device="cpu").manual_seed(seed)
inputs = {
"image": image,
"prompt": "dance monkey",
"negative_prompt": "bad quality",
"generator": generator,
"num_inference_steps": 2,
"guidance_scale": 3.0,
"true_cfg_scale": 1.0,
"height": 32,
"width": 32,
"max_sequence_length": 16,
"output_type": "pt",
}
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
generated_image = image[0]
self.assertEqual(generated_image.shape, (3, 32, 32))
def test_inference_batch_single_identical(self):
self._test_inference_batch_single_identical(batch_size=3, expected_max_diff=1e-1)
def test_attention_slicing_forward_pass(
self, test_max_difference=True, test_mean_pixel_difference=True, expected_max_diff=1e-3
):
if not self.test_attention_slicing:
return
components = self.get_dummy_components()
pipe = self.pipeline_class(**components)
for component in pipe.components.values():
if hasattr(component, "set_default_attn_processor"):
component.set_default_attn_processor()
pipe.to(torch_device)
pipe.set_progress_bar_config(disable=None)
generator_device = "cpu"
inputs = self.get_dummy_inputs(generator_device)
output_without_slicing = pipe(**inputs).images[0]
pipe.enable_attention_slicing(slice_size=1)
inputs = self.get_dummy_inputs(generator_device)
output_with_slicing1 = pipe(**inputs).images[0]
pipe.enable_attention_slicing(slice_size=2)
inputs = self.get_dummy_inputs(generator_device)
output_with_slicing2 = pipe(**inputs).images[0]
if test_max_difference:
max_diff1 = np.abs(to_np(output_with_slicing1) - to_np(output_without_slicing)).max()
max_diff2 = np.abs(to_np(output_with_slicing2) - to_np(output_without_slicing)).max()
self.assertLess(
max(max_diff1, max_diff2),
expected_max_diff,
"Attention slicing should not affect the inference results",
)
def test_vae_tiling(self, expected_diff_max: float = 0.2):
generator_device = "cpu"
components = self.get_dummy_components()
pipe = self.pipeline_class(**components)
pipe.to("cpu")
pipe.set_progress_bar_config(disable=None)
# Without tiling
inputs = self.get_dummy_inputs(generator_device)
inputs["height"] = inputs["width"] = 128
output_without_tiling = pipe(**inputs)[0]
# With tiling
pipe.vae.enable_tiling(
tile_sample_min_height=96,
tile_sample_min_width=96,
tile_sample_stride_height=64,
tile_sample_stride_width=64,
)
inputs = self.get_dummy_inputs(generator_device)
inputs["height"] = inputs["width"] = 128
output_with_tiling = pipe(**inputs)[0]
self.assertLess(
(to_np(output_without_tiling) - to_np(output_with_tiling)).max(),
expected_diff_max,
"VAE tiling should not affect the inference results",
)
|