Skip to content

fix(seedream4): add flag to ignore error on partial success - #9952

Merged
comfyanonymous merged 1 commit into
Comfy-Org:masterfrom
bigcat88:fix/api_nodes/seedream4-partitial
Sep 19, 2025
Merged

comfyanonymous merged 1 commit into
Comfy-Org:masterfrom
bigcat88:fix/api_nodes/seedream4-partitial

Conversation

@bigcat88

Copy link
Copy Markdown
Contributor

When you ask SeeDream4 to generate multiple images, sometimes their server returns only partial results, giving responses like this:

[{'url': 'img_url1', 'size': '4096x4096'},
{'url': 'img_url2', 'size': '4096x4096'},
{'error': {'code': 'InternalServiceError', 'message': 'The service encountered an unexpected internal error. Request id: xxxx'}}
]

This PR adds the ability to consider such execution successful and return what was successfully generated, ignoring errors.

The default behavior hasn't changed; instead of 'invalid key 'url' not found', the correct error is now returned:

image

@bigcat88

Copy link
Copy Markdown
Contributor Author

+label: Core

@Kosinkadink Kosinkadink added the Core Core team dependency label Sep 19, 2025
@comfyanonymous
comfyanonymous merged commit 852704c into Comfy-Org:master Sep 19, 2025
11 checks passed
@bigcat88
bigcat88 deleted the fix/api_nodes/seedream4-partitial branch September 20, 2025 06:44
Thor-ATX pushed a commit to dunklehohle/ComfyUI that referenced this pull request Sep 26, 2025
…e-update

* commit '2b7f9a8196304badb5fe58e5c734e4b182ad0fdf': (91 commits)
  Fix the failing unit test. (Comfy-Org#10037)
  Make LatentCompositeMasked work with basic video latents. (Comfy-Org#10023)
  Fix memory leak by properly detaching model finalizer (Comfy-Org#9979)
  Fix issue with .view() in HuMo. (Comfy-Org#10014)
  Add new audio nodes (Comfy-Org#9908)
  Rodin3D - add [Rodin3D Gen-2 generate] api-node (Comfy-Org#9994)
  ComfyUI version 0.3.60
  feat(api-nodes): add wan t2i, t2v, i2v nodes (Comfy-Org#9996)
  update template to 0.1.86 (Comfy-Org#9998)
  Fix bug with WanAnimateToVideo. (Comfy-Org#9990)
  Fix bug with WanAnimateToVideo node. (Comfy-Org#9988)
  add offset param (Comfy-Org#9977)
  Support for qwen edit plus model. Use the new TextEncodeQwenImageEditPlus. (Comfy-Org#9986)
  Set some wan nodes as no longer experimental. (Comfy-Org#9976)
  Lower wan memory estimation value a bit. (Comfy-Org#9964)
  Fix LoRA Trainer bugs with FP8 models. (Comfy-Org#9854)
  [Reviving Comfy-Org#5709] Add strength input to Differential Diffusion (Comfy-Org#9957)
  Add inputs for character replacement to the WanAnimateToVideo node. (Comfy-Org#9960)
  Update WanAnimateToVideo to more easily extend videos. (Comfy-Org#9959)
  fix(seedream4): add flag to ignore error on partial success (Comfy-Org#9952)
  ...
toxicwind pushed a commit to toxicwind/ComfyUI that referenced this pull request Oct 12, 2025
gmaOCR pushed a commit to gmaOCR/ComfyUI that referenced this pull request Oct 14, 2025
adlerfaulkner pushed a commit to LucaLabsInc/ComfyUI that referenced this pull request Oct 16, 2025
Subarasheese pushed a commit to Subarasheese/ComfyUI that referenced this pull request Jul 26, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Core Core team dependency

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants