feat(importers): Add diffuser backend importer with ginkgo tests and UI support

[Copilot]

Description

Adds importer for diffuser backend models following existing importer patterns (llama-cpp, mlx, vllm, transformers), including UI support in the model editor.

Implementation

• diffuser.go: Implements DiffuserImporter matching models via:

  • backend: "diffusers" preference
  • HuggingFace files: model_index.json, scheduler/scheduler_config.json
  • Uses gopkg.in/yaml.v3 for consistency with the main importers module

• diffuser_test.go: 18 ginkgo tests covering match/import logic and preferences

• importers.go: Registered DiffuserImporter in defaultImporters

• model-editor.html: Added diffuser preferences to model editor UI:

  • Backend dropdown now includes "diffusers" option
  • Dynamic fields appear when diffusers backend is selected: Pipeline Type, Scheduler Type, Enable Parameters, CUDA checkbox
  • All fields integrate with existing preferences system

Configuration Output

backend: diffusers
known_usecases:
- image
diffusers:
  cuda: true
  pipeline_type: StableDiffusion3Pipeline
  scheduler_type: k_dpmpp_2m
  enable_parameters: negative_prompt,num_inference_steps
parameters:
  model: https://huggingface.co/stabilityai/stable-diffusion-3-medium

Supports preference overrides: pipeline_type, scheduler_type, cuda, enable_parameters. Defaults to StableDiffusionPipeline with standard parameters.

UI Changes

The model editor now provides a user-friendly interface for configuring diffuser models with backend-specific options that appear dynamically when "diffusers" is selected.

Notes for Reviewers

18 new tests added (88 total). 3 pre-existing test failures in LlamaCPPImporter unrelated to changes.

Signed commits

• Yes, I signed my commits.

Original prompt

Look at the already existing importers implementation and add one for the diffuser backend. Add also ginkgo tests.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[netlify]

✅ Deploy Preview for localai ready!

Name	Link
🔨 Latest commit	9acb271
🔍 Latest deploy log	https://app.netlify.com/projects/localai/deploys/691f27f946abbf0008fe9c84
😎 Deploy Preview	https://deploy-preview-7316--localai.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[mudler]

@copilot add the new preferences to the common preferences of the model-editor import section

[Copilot]

@copilot add the new preferences to the common preferences of the model-editor import section

Added diffuser preferences to the model-editor import section in commit d6464fa.

The following fields now appear when "diffusers" backend is selected:

• Pipeline Type (pipeline_type)
• Scheduler Type (scheduler_type)
• Enable Parameters (enable_parameters)
• CUDA checkbox (cuda)

All fields integrate with the existing preferences system and are passed to the DiffuserImporter.

[mudler]

@copilot be consistent and use yaml/v3

[Copilot]

@copilot be consistent and use yaml/v3

Updated to use gopkg.in/yaml.v3 for consistency in commit 9acb271.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• huggingface.co
  • Triggering command: /tmp/go-build2054339694/b001/importers.test -test.paniconexit0 -test.timeout=10m0s -test.v=true (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)