Skip to content

Commit 52d4ebf

Browse files
zRzRzRzRzRzRzRzRzRzRzRzRzRzR
authored
Merge branch 'huggingface:main' into cogview4_control
2 parents 29b0c81 + cc7b5b8 commit 52d4ebf

File tree

97 files changed

+1153
-481
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

97 files changed

+1153
-481
lines changed

‎.github/workflows/pr_tests_gpu.yml‎

Lines changed: 241 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,241 @@
1+
name: Fast GPU Tests on PR
2+
3+
on:
4+
pull_request:
5+
branches: main
6+
paths:
7+
- "src/diffusers/models/modeling_utils.py"
8+
- "src/diffusers/models/model_loading_utils.py"
9+
- "src/diffusers/pipelines/pipeline_utils.py"
10+
- "src/diffusers/pipeline_loading_utils.py"
11+
- "src/diffusers/loaders/lora_base.py"
12+
- "src/diffusers/loaders/lora_pipeline.py"
13+
- "src/diffusers/loaders/peft.py"
14+
workflow_dispatch:
15+
16+
concurrency:
17+
group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
18+
cancel-in-progress: true
19+
20+
env:
21+
DIFFUSERS_IS_CI: yes
22+
OMP_NUM_THREADS: 8
23+
MKL_NUM_THREADS: 8
24+
HF_HUB_ENABLE_HF_TRANSFER: 1
25+
PYTEST_TIMEOUT: 600
26+
PIPELINE_USAGE_CUTOFF: 1000000000 # set high cutoff so that only always-test pipelines run
27+
28+
jobs:
29+
setup_torch_cuda_pipeline_matrix:
30+
name: Setup Torch Pipelines CUDA Slow Tests Matrix
31+
runs-on:
32+
group: aws-general-8-plus
33+
container:
34+
image: diffusers/diffusers-pytorch-cpu
35+
outputs:
36+
pipeline_test_matrix: ${{ steps.fetch_pipeline_matrix.outputs.pipeline_test_matrix }}
37+
steps:
38+
- name: Checkout diffusers
39+
uses: actions/checkout@v3
40+
with:
41+
fetch-depth: 2
42+
- name: Install dependencies
43+
run: |
44+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
45+
python -m uv pip install -e [quality,test]
46+
- name: Environment
47+
run: |
48+
python utils/print_env.py
49+
- name: Fetch Pipeline Matrix
50+
id: fetch_pipeline_matrix
51+
run: |
52+
matrix=$(python utils/fetch_torch_cuda_pipeline_test_matrix.py)
53+
echo $matrix
54+
echo "pipeline_test_matrix=$matrix" >> $GITHUB_OUTPUT
55+
- name: Pipeline Tests Artifacts
56+
if: ${{ always() }}
57+
uses: actions/upload-artifact@v4
58+
with:
59+
name: test-pipelines.json
60+
path: reports
61+
62+
torch_pipelines_cuda_tests:
63+
name: Torch Pipelines CUDA Tests
64+
needs: setup_torch_cuda_pipeline_matrix
65+
strategy:
66+
fail-fast: false
67+
max-parallel: 8
68+
matrix:
69+
module: ${{ fromJson(needs.setup_torch_cuda_pipeline_matrix.outputs.pipeline_test_matrix) }}
70+
runs-on:
71+
group: aws-g4dn-2xlarge
72+
container:
73+
image: diffusers/diffusers-pytorch-cuda
74+
options: --shm-size "16gb" --ipc host --gpus 0
75+
steps:
76+
- name: Checkout diffusers
77+
uses: actions/checkout@v3
78+
with:
79+
fetch-depth: 2
80+
81+
- name: NVIDIA-SMI
82+
run: |
83+
nvidia-smi
84+
- name: Install dependencies
85+
run: |
86+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
87+
python -m uv pip install -e [quality,test]
88+
pip uninstall accelerate -y && python -m uv pip install -U accelerate@git+https://github.com/huggingface/accelerate.git
89+
pip uninstall transformers -y && python -m uv pip install -U transformers@git+https://github.com/huggingface/transformers.git --no-deps
90+
91+
- name: Environment
92+
run: |
93+
python utils/print_env.py
94+
- name: Extract tests
95+
id: extract_tests
96+
run: |
97+
pattern=$(python utils/extract_tests_from_mixin.py --type pipeline)
98+
echo "$pattern" > /tmp/test_pattern.txt
99+
echo "pattern_file=/tmp/test_pattern.txt" >> $GITHUB_OUTPUT
100+
101+
- name: PyTorch CUDA checkpoint tests on Ubuntu
102+
env:
103+
HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
104+
# https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
105+
CUBLAS_WORKSPACE_CONFIG: :16:8
106+
run: |
107+
pattern=$(cat ${{ steps.extract_tests.outputs.pattern_file }})
108+
python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile \
109+
-s -v -k "not Flax and not Onnx and $pattern" \
110+
--make-reports=tests_pipeline_${{ matrix.module }}_cuda \
111+
tests/pipelines/${{ matrix.module }}
112+
113+
- name: Failure short reports
114+
if: ${{ failure() }}
115+
run: |
116+
cat reports/tests_pipeline_${{ matrix.module }}_cuda_stats.txt
117+
cat reports/tests_pipeline_${{ matrix.module }}_cuda_failures_short.txt
118+
- name: Test suite reports artifacts
119+
if: ${{ always() }}
120+
uses: actions/upload-artifact@v4
121+
with:
122+
name: pipeline_${{ matrix.module }}_test_reports
123+
path: reports
124+
125+
torch_cuda_tests:
126+
name: Torch CUDA Tests
127+
runs-on:
128+
group: aws-g4dn-2xlarge
129+
container:
130+
image: diffusers/diffusers-pytorch-cuda
131+
options: --shm-size "16gb" --ipc host --gpus 0
132+
defaults:
133+
run:
134+
shell: bash
135+
strategy:
136+
fail-fast: false
137+
max-parallel: 2
138+
matrix:
139+
module: [models, schedulers, lora, others]
140+
steps:
141+
- name: Checkout diffusers
142+
uses: actions/checkout@v3
143+
with:
144+
fetch-depth: 2
145+
146+
- name: Install dependencies
147+
run: |
148+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
149+
python -m uv pip install -e [quality,test]
150+
python -m uv pip install peft@git+https://github.com/huggingface/peft.git
151+
pip uninstall accelerate -y && python -m uv pip install -U accelerate@git+https://github.com/huggingface/accelerate.git
152+
pip uninstall transformers -y && python -m uv pip install -U transformers@git+https://github.com/huggingface/transformers.git --no-deps
153+
154+
- name: Environment
155+
run: |
156+
python utils/print_env.py
157+
158+
- name: Extract tests
159+
id: extract_tests
160+
run: |
161+
pattern=$(python utils/extract_tests_from_mixin.py --type ${{ matrix.module }})
162+
echo "$pattern" > /tmp/test_pattern.txt
163+
echo "pattern_file=/tmp/test_pattern.txt" >> $GITHUB_OUTPUT
164+
165+
- name: Run PyTorch CUDA tests
166+
env:
167+
HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
168+
# https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
169+
CUBLAS_WORKSPACE_CONFIG: :16:8
170+
run: |
171+
pattern=$(cat ${{ steps.extract_tests.outputs.pattern_file }})
172+
if [ -z "$pattern" ]; then
173+
python -m pytest -n 1 -sv --max-worker-restart=0 --dist=loadfile -k "not Flax and not Onnx" tests/${{ matrix.module }} \
174+
--make-reports=tests_torch_cuda_${{ matrix.module }}
175+
else
176+
python -m pytest -n 1 -sv --max-worker-restart=0 --dist=loadfile -k "not Flax and not Onnx and $pattern" tests/${{ matrix.module }} \
177+
--make-reports=tests_torch_cuda_${{ matrix.module }}
178+
fi
179+
180+
- name: Failure short reports
181+
if: ${{ failure() }}
182+
run: |
183+
cat reports/tests_torch_cuda_${{ matrix.module }}_stats.txt
184+
cat reports/tests_torch_cuda_${{ matrix.module }}_failures_short.txt
185+
186+
- name: Test suite reports artifacts
187+
if: ${{ always() }}
188+
uses: actions/upload-artifact@v4
189+
with:
190+
name: torch_cuda_test_reports_${{ matrix.module }}
191+
path: reports
192+
193+
run_examples_tests:
194+
name: Examples PyTorch CUDA tests on Ubuntu
195+
pip uninstall transformers -y && python -m uv pip install -U transformers@git+https://github.com/huggingface/transformers.git --no-deps
196+
runs-on:
197+
group: aws-g4dn-2xlarge
198+
199+
container:
200+
image: diffusers/diffusers-pytorch-cuda
201+
options: --gpus 0 --shm-size "16gb" --ipc host
202+
steps:
203+
- name: Checkout diffusers
204+
uses: actions/checkout@v3
205+
with:
206+
fetch-depth: 2
207+
208+
- name: NVIDIA-SMI
209+
run: |
210+
nvidia-smi
211+
- name: Install dependencies
212+
run: |
213+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
214+
python -m uv pip install -e [quality,test,training]
215+
216+
- name: Environment
217+
run: |
218+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
219+
python utils/print_env.py
220+
221+
- name: Run example tests on GPU
222+
env:
223+
HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
224+
run: |
225+
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
226+
python -m uv pip install timm
227+
python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s -v --make-reports=examples_torch_cuda examples/
228+
229+
- name: Failure short reports
230+
if: ${{ failure() }}
231+
run: |
232+
cat reports/examples_torch_cuda_stats.txt
233+
cat reports/examples_torch_cuda_failures_short.txt
234+
235+
- name: Test suite reports artifacts
236+
if: ${{ always() }}
237+
uses: actions/upload-artifact@v4
238+
with:
239+
name: examples_test_reports
240+
path: reports
241+

‎.github/workflows/push_tests.yml‎

Lines changed: 0 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -1,13 +1,6 @@
11
name: Fast GPU Tests on main
22

33
on:
4-
pull_request:
5-
branches: main
6-
paths:
7-
- "src/diffusers/models/modeling_utils.py"
8-
- "src/diffusers/models/model_loading_utils.py"
9-
- "src/diffusers/pipelines/pipeline_utils.py"
10-
- "src/diffusers/pipeline_loading_utils.py"
114
workflow_dispatch:
125
push:
136
branches:
@@ -167,7 +160,6 @@ jobs:
167160
path: reports
168161

169162
flax_tpu_tests:
170-
if: ${{ github.event_name != 'pull_request' }}
171163
name: Flax TPU Tests
172164
runs-on:
173165
group: gcp-ct5lp-hightpu-8t
@@ -216,7 +208,6 @@ jobs:
216208
path: reports
217209

218210
onnx_cuda_tests:
219-
if: ${{ github.event_name != 'pull_request' }}
220211
name: ONNX CUDA Tests
221212
runs-on:
222213
group: aws-g4dn-2xlarge
@@ -265,7 +256,6 @@ jobs:
265256
path: reports
266257

267258
run_torch_compile_tests:
268-
if: ${{ github.event_name != 'pull_request' }}
269259
name: PyTorch Compile CUDA tests
270260

271261
runs-on:
@@ -309,7 +299,6 @@ jobs:
309299
path: reports
310300

311301
run_xformers_tests:
312-
if: ${{ github.event_name != 'pull_request' }}
313302
name: PyTorch xformers CUDA tests
314303

315304
runs-on:

‎docs/source/en/_toctree.yml‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -543,6 +543,10 @@
543543
title: Overview
544544
- local: api/schedulers/cm_stochastic_iterative
545545
title: CMStochasticIterativeScheduler
546+
- local: api/schedulers/ddim_cogvideox
547+
title: CogVideoXDDIMScheduler
548+
- local: api/schedulers/multistep_dpm_solver_cogvideox
549+
title: CogVideoXDPMScheduler
546550
- local: api/schedulers/consistency_decoder
547551
title: ConsistencyDecoderScheduler
548552
- local: api/schedulers/cosine_dpm

‎docs/source/en/api/pipelines/animatediff.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,10 @@ specific language governing permissions and limitations under the License.
1212

1313
# Text-to-Video Generation with AnimateDiff
1414

15+
<div class="flex flex-wrap space-x-1">
16+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
17+
</div>
18+
1519
## Overview
1620

1721
[AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning](https://arxiv.org/abs/2307.04725) by Yuwei Guo, Ceyuan Yang, Anyi Rao, Yaohui Wang, Yu Qiao, Dahua Lin, Bo Dai.

‎docs/source/en/api/pipelines/cogvideox.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -15,6 +15,10 @@
1515

1616
# CogVideoX
1717

18+
<div class="flex flex-wrap space-x-1">
19+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
20+
</div>
21+
1822
[CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://arxiv.org/abs/2408.06072) from Tsinghua University & ZhipuAI, by Zhuoyi Yang, Jiayan Teng, Wendi Zheng, Ming Ding, Shiyu Huang, Jiazheng Xu, Yuanming Yang, Wenyi Hong, Xiaohan Zhang, Guanyu Feng, Da Yin, Xiaotao Gu, Yuxuan Zhang, Weihan Wang, Yean Cheng, Ting Liu, Bin Xu, Yuxiao Dong, Jie Tang.
1923

2024
The abstract from the paper is:

‎docs/source/en/api/pipelines/consisid.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -15,6 +15,10 @@
1515

1616
# ConsisID
1717

18+
<div class="flex flex-wrap space-x-1">
19+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
20+
</div>
21+
1822
[Identity-Preserving Text-to-Video Generation by Frequency Decomposition](https://arxiv.org/abs/2411.17440) from Peking University & University of Rochester & etc, by Shenghai Yuan, Jinfa Huang, Xianyi He, Yunyang Ge, Yujun Shi, Liuhan Chen, Jiebo Luo, Li Yuan.
1923

2024
The abstract from the paper is:

‎docs/source/en/api/pipelines/control_flux_inpaint.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,10 @@ specific language governing permissions and limitations under the License.
1212

1313
# FluxControlInpaint
1414

15+
<div class="flex flex-wrap space-x-1">
16+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
17+
</div>
18+
1519
FluxControlInpaintPipeline is an implementation of Inpainting for Flux.1 Depth/Canny models. It is a pipeline that allows you to inpaint images using the Flux.1 Depth/Canny models. The pipeline takes an image and a mask as input and returns the inpainted image.
1620

1721
FLUX.1 Depth and Canny [dev] is a 12 billion parameter rectified flow transformer capable of generating an image based on a text description while following the structure of a given input image. **This is not a ControlNet model**.

‎docs/source/en/api/pipelines/controlnet.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,10 @@ specific language governing permissions and limitations under the License.
1212

1313
# ControlNet
1414

15+
<div class="flex flex-wrap space-x-1">
16+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
17+
</div>
18+
1519
ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.
1620

1721
With a ControlNet model, you can provide an additional control image to condition and control Stable Diffusion generation. For example, if you provide a depth map, the ControlNet model generates an image that'll preserve the spatial information from the depth map. It is a more flexible and accurate way to control the image generation process.

‎docs/source/en/api/pipelines/controlnet_flux.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,10 @@ specific language governing permissions and limitations under the License.
1212

1313
# ControlNet with Flux.1
1414

15+
<div class="flex flex-wrap space-x-1">
16+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
17+
</div>
18+
1519
FluxControlNetPipeline is an implementation of ControlNet for Flux.1.
1620

1721
ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.

‎docs/source/en/api/pipelines/controlnet_sd3.md‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -12,6 +12,10 @@ specific language governing permissions and limitations under the License.
1212

1313
# ControlNet with Stable Diffusion 3
1414

15+
<div class="flex flex-wrap space-x-1">
16+
<img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/>
17+
</div>
18+
1519
StableDiffusion3ControlNetPipeline is an implementation of ControlNet for Stable Diffusion 3.
1620

1721
ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.

0 commit comments

Comments
 (0)