anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

[Model] Add Cache-DiT support for FLUX.1-Kontext-dev (#4205)

Browse Source 
Signed-off-by: yifan <Yifan.Tian@colorado.edu>
Signed-off-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
Co-authored-by: Didan Deng <33117903+wtomin@users.noreply.github.com>
This commit is contained in:
Yifan TianandDidan Deng authored and GitHub committed 2026-07-22 14:06:12 +08:00
1 parent e39f4dddfb
commit 24eb4ab89c
3 files changed
+19 -1
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

docs/user_guide/diffusion_features.md
+1 -1
View File
@@ -116,7 +116,7 @@ The following tables show which models support each feature:
| **FLUX.1-dev** | ✅ | ✅ | ❌ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ |
| **FLUX.1-schnell** | ❌ | ✅ | ❌ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ |
| **FLUX.2-klein** | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ |
| **FLUX.1-Kontext-dev** | ❌ | ❌ | ❌ | ❌ | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| **FLUX.1-Kontext-dev** | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| **FLUX.2-dev** | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| **GLM-Image** | ❌ | ❌ | ❌ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| **Hidream-I1-Full** | ❌ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
tests/e2e/online_serving/test_flux_kontext_expansion.py
+13
View File
@@ -45,6 +45,19 @@ def _get_diffusion_feature_cases(model: str):
id="parallel_cfg_2",
marks=PARALLEL_FEATURE_MARKS,
),
pytest.param(
OmniServerParams(
model=model,
server_args=[
"--tensor-parallel-size",
"2",
"--enable-cpu-offload",
"--cache-backend",
"cache_dit",
],
),
id="parallel_002",
),
]
vllm_omni/diffusion/models/flux/pipeline_flux_kontext.py
+5
View File
@@ -641,6 +641,11 @@ class FluxKontextPipeline(
num_inference_steps = req.sampling_params.num_inference_steps or num_inference_steps
sigmas = req.sampling_params.sigmas or sigmas
guidance_scale = req.sampling_params.guidance_scale or guidance_scale
num_images_per_prompt = (
req.sampling_params.num_outputs_per_prompt
if req.sampling_params.num_outputs_per_prompt > 0
else num_images_per_prompt
)
true_cfg_scale = (
req.sampling_params.true_cfg_scale if req.sampling_params.true_cfg_scale is not None else true_cfg_scale
)