valknar/runpod
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
219 Commits 1 Branch 0 Tags
main
Commit Graph

7 Commits

Author SHA1 Message Date
Sebastian Krüger
2189697734 refactor: remove type field from models_huggingface.yaml and include type in dest paths
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 14s
Details 
- Prepended ComfyUI model type folder (checkpoints/, clip/, vae/, etc.) to all dest paths
- Removed separate 'type' field from all model entries
- Consolidated SD3.5 duplicate entries (5 → 1)
- Simplified model configuration by embedding directory structure directly in destination paths

This change eliminates the need to parse the 'type' field separately in artifact_huggingface_download.sh,
making the configuration more explicit and easier to understand.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-25 19:19:42 +01:00
Sebastian Krüger
ff6c1369ae feat: add three new SDXL/SD1.5 image generation models
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 14s
Details 
Add the following models to models_huggingface.yaml:
- John6666/diving-illustrious-real-asian-v50-sdxl: SDXL fine-tune for photorealistic Asian subjects
- playgroundai/playground-v2.5-1024px-aesthetic: High-aesthetic 1024px SDXL-based model
- Lykon/dreamshaper-8: Versatile SD1.5 fine-tune with multi-style support

All models marked as non-essential and will download .safetensors files.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-25 17:29:04 +01:00
Sebastian Krüger
aa2cc5973b fix: update CogVideoX models to link sharded transformer files
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 14s
Details 
- CogVideoX-5b: Link 2 shards + index.json instead of single non-existent file
- CogVideoX-5b-I2V: Link 3 shards + index.json instead of single non-existent file  - Fixes link command failures for these video generation models
- Shards are now properly symlinked to ComfyUI diffusion_models directory

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-25 15:47:09 +01:00
Sebastian Krüger
6efb55c59f feat: add complete HunyuanVideo and Wan2.2 video generation integration
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 15s
Details 
Integrated 35+ video generation models and 13 production workflows from ComfyUI docs tutorials for state-of-the-art text-to-video and image-to-video generation.

Models Added (models_huggingface.yaml):
- HunyuanVideo (5 models): Original T2V/I2V (720p), v1.5 (720p/1080p) with Qwen 2.5 VL
- Wan2.2 diffusion models (18 models):
  - 5B TI2V hybrid (8GB VRAM, efficient)
  - 14B variants: T2V, I2V (high/low noise), Animate, S2V (FP8/BF16), Fun Camera/Control (high/low noise)
- Support models (12): VAEs, UMT5-XXL, CLIP Vision H, Wav2Vec2, LLaVA encoders
- LoRA accelerators (4): Lightx2v 4-step distillation for 5x speedup

Workflows Added (comfyui/workflows/image-to-video/):
- HunyuanVideo (5 workflows): T2V original, I2V v1/v2 (webp embedded), v1.5 T2V/I2V (JSON)
- Wan2.2 (8 workflows): 5B TI2V, 14B T2V/I2V/FLF2V/Animate/S2V/Fun Camera/Fun Control
- Asset files (10): Reference images, videos, audio for workflow testing

Custom Nodes Added (arty.yml):
- ComfyUI-KJNodes: Kijai optimizations for HunyuanVideo/Wan2.2 (FP8 scaling, video helpers)
- comfyui_controlnet_aux: ControlNet preprocessors (Canny, Depth, OpenPose, MLSD) for Fun Control
- ComfyUI-GGUF: GGUF quantization support for memory optimization

VRAM Requirements:
- HunyuanVideo original: 24GB (720p T2V/I2V, 129 frames, 5s generation)
- HunyuanVideo 1.5: 30-60GB (720p/1080p, improved quality with Qwen 2.5 VL)
- Wan2.2 5B: 8GB (efficient dual-expert architecture with native offloading)
- Wan2.2 14B: 24GB (high-quality video generation, all modes)

Note: Wan2.2 Fun Inpaint workflow not available in official templates repository (404).

Tutorial Sources:
- https://docs.comfy.org/tutorials/video/hunyuan/hunyuan-video
- https://docs.comfy.org/tutorials/video/hunyuan/hunyuan-video-1-5
- https://docs.comfy.org/tutorials/video/wan/wan2_2
- https://docs.comfy.org/tutorials/video/wan/wan2-2-animate
- https://docs.comfy.org/tutorials/video/wan/wan2-2-s2v
- https://docs.comfy.org/tutorials/video/wan/wan2-2-fun-camera
- https://docs.comfy.org/tutorials/video/wan/wan2-2-fun-control

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-25 10:43:39 +01:00
Sebastian Krüger
513062623c feat: integrate ACE Step music generation with 19-language support
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 14s
Details 
Added ACE Step v1 3.5B model for state-of-the-art music generation:
- 15x faster than LLM baselines with superior structural coherence
- Supports 19 languages (en, zh, ja, ko, fr, es, de, it, pt, ru + 9 more)
- Voice cloning, lyric alignment, and multi-genre capabilities

Changes:
- Added ACE Step models to models_huggingface.yaml (checkpoint + Chinese RAP LoRA)
- Added ComfyUI_ACE-Step custom node to arty.yml with installation script
- Created 4 comprehensive workflows in comfyui/workflows/text-to-music/:
  * acestep-simple-t2m-v1.json - Basic 60s text-to-music generation
  * acestep-multilang-t2m-v1.json - 19-language music generation
  * acestep-remix-m2m-v1.json - Music-to-music remixing with style transfer
  * acestep-chinese-rap-v1.json - Chinese hip-hop with specialized LoRA

🤖 Generated with Claude Code (https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-25 08:40:17 +01:00
Sebastian Krüger
f2186db78e feat: integrate ComfyUI_DiffRhythm extension with 7 models and 4 workflows
All checks were successful
Build and Push RunPod Docker Image / build-and-push (push) Successful in 15s
Details 
- Add DiffRhythm to arty.yml references and setup/comfyui-nodes
- Install espeak-ng system dependency for phoneme processing
- Add 7 DiffRhythm models to models_huggingface.yaml with file mappings:
  * ASLP-lab/DiffRhythm-1_2 (95s generation)
  * ASLP-lab/DiffRhythm-full (4m45s generation)
  * ASLP-lab/DiffRhythm-base
  * ASLP-lab/DiffRhythm-vae
  * OpenMuQ/MuQ-MuLan-large
  * OpenMuQ/MuQ-large-msd-iter
  * FacebookAI/xlm-roberta-base
- Create 4 comprehensive workflows:
  * diffrhythm-simple-t2m-v1.json (basic 95s text-to-music)
  * diffrhythm-full-length-t2m-v1.json (4m45s full-length)
  * diffrhythm-reference-based-v1.json (style transfer with reference audio)
  * diffrhythm-random-generation-v1.json (no-prompt random generation)
- Update storage requirements: 90GB essential, 149GB total

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-24 09:50:45 +01:00
Sebastian Krüger
e29f77c90b feat: add dedicated CivitAI NSFW model downloader 
- Add models_civitai.yaml with 6 NSFW SDXL checkpoints
- Create artifact_civitai_download.sh with beautiful purple/magenta CLI
- Update .env.example with CIVITAI_API_KEY documentation
- Update CLAUDE.md with CivitAI usage instructions
- Rename comfyui_models.yaml to models_huggingface.yaml for clarity

Features:
- Dedicated config and downloader for CivitAI models
- Same elegant architecture as HuggingFace downloader
- Retry logic, rate limiting, progress bars
- Models: LUSTIFY, Pony Diffusion V6, RealVisXL, etc.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-23 17:58:25 +01:00