- Add ComfyUI service to AI stack using nginx:alpine as reverse proxy
- Proxy to RunPod ComfyUI via Tailscale (100.121.199.88:8188)
- Configure Traefik routing for comfy.ai.pivoine.art
- Enable Authelia SSO middleware (net-authelia)
- Support WebSocket connections for real-time updates
- Set appropriate timeouts for image generation (300s)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Set drop_params: false in litellm_settings
- Set modify_params: false in litellm_settings
- Set drop_params: false in default_litellm_params
- Commented out LITELLM_DROP_PARAMS env var
- Removed --drop_params command flag
These settings were stripping critical streaming parameters causing
vLLM streaming responses to collapse into empty deltas
- Reverted direct orchestrator connection to WebUI
- Added stream: true parameter to qwen-2.5-7b model config
- Keep LiteLLM as single proxy for all models
- Configure WebUI with both LiteLLM and direct orchestrator API base URLs
- This bypasses LiteLLM's streaming issues for the qwen-2.5-7b model
- WebUI will now show models from both endpoints
- Allows testing if LiteLLM is the bottleneck for streaming
Related to streaming fix in RunPod models/vllm/server.py
- Changed back from hosted_vllm/qwen-2.5-7b to openai/qwen-2.5-7b
- Removed /v1 suffix from api_base (LiteLLM adds it automatically)
- Added supports_system_messages: false for vLLM compatibility
- Update api_base URLs from 100.100.108.13 to 100.121.199.88 (RunPod Tailscale IP)
- All self-hosted models (qwen-2.5-7b, flux-schnell, musicgen-medium) now route through Tailscale VPN
- Tested and verified connectivity between VPS and RunPod GPU orchestrator
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Removed outdated AI infrastructure README that referenced GPU services.
VPS AI services (Open WebUI, Crawl4AI, facefusion) are documented in compose.yaml comments.
GPU infrastructure docs are now in dedicated runpod repository.
Multi-modal AI stack (text/image/music generation) has been moved to:
Repository: ssh://git@dev.pivoine.art:2222/valknar/runpod.git
Updated ai/README.md to document:
- VPS AI services (Open WebUI, Crawl4AI, AI PostgreSQL)
- Reference to new runpod repository for GPU infrastructure
- Clear separation between VPS and GPU deployments
- Integration architecture via Tailscale VPN
Implemented a cost-optimized AI infrastructure running on single RTX 4090 GPU with
automatic model switching based on request type. This enables text, image, and
music generation on the same hardware with sequential loading.
## New Components
**Model Orchestrator** (ai/model-orchestrator/):
- FastAPI service managing model lifecycle
- Automatic model detection and switching based on request type
- OpenAI-compatible API proxy for all models
- Simple YAML configuration for adding new models
- Docker SDK integration for service management
- Endpoints: /v1/chat/completions, /v1/images/generations, /v1/audio/generations
**Text Generation** (ai/vllm/):
- Reorganized existing vLLM server into proper structure
- Qwen 2.5 7B Instruct (14GB VRAM, ~50 tok/sec)
- Docker containerized with CUDA 12.4 support
**Image Generation** (ai/flux/):
- Flux.1 Schnell for fast, high-quality images
- 14GB VRAM, 4-5 sec per image
- OpenAI DALL-E compatible API
- Pre-built image: ghcr.io/matatonic/openedai-images-flux
**Music Generation** (ai/musicgen/):
- Meta's MusicGen Medium (facebook/musicgen-medium)
- Text-to-music generation (11GB VRAM)
- 60-90 seconds for 30s audio clips
- Custom FastAPI wrapper with AudioCraft
## Architecture
```
VPS (LiteLLM) → Tailscale VPN → GPU Orchestrator (Port 9000)
↓
┌───────────────┼───────────────┐
vLLM (8001) Flux (8002) MusicGen (8003)
[Only ONE active at a time - sequential loading]
```
## Configuration Files
- docker-compose.gpu.yaml: Main orchestration file for RunPod deployment
- model-orchestrator/models.yaml: Model registry (easy to add new models)
- .env.example: Environment variable template
- README.md: Comprehensive deployment and usage guide
## Updated Files
- litellm-config.yaml: Updated to route through orchestrator (port 9000)
- GPU_DEPLOYMENT_LOG.md: Documented multi-modal architecture
## Features
✅ Automatic model switching (30-120s latency)
✅ Cost-optimized single GPU deployment (~$0.50/hr vs ~$0.75/hr multi-GPU)
✅ Easy model addition via YAML configuration
✅ OpenAI-compatible APIs for all model types
✅ Centralized routing through LiteLLM proxy
✅ GPU memory safety (only one model loaded at time)
## Usage
Deploy to RunPod:
```bash
scp -r ai/* gpu-pivoine:/workspace/ai/
ssh gpu-pivoine "cd /workspace/ai && docker compose -f docker-compose.gpu.yaml up -d orchestrator"
```
Test models:
```bash
# Text
curl http://100.100.108.13:9000/v1/chat/completions -d '{"model":"qwen-2.5-7b","messages":[...]}'
# Image
curl http://100.100.108.13:9000/v1/images/generations -d '{"model":"flux-schnell","prompt":"..."}'
# Music
curl http://100.100.108.13:9000/v1/audio/generations -d '{"model":"musicgen-medium","prompt":"..."}'
```
All models available via Open WebUI at https://ai.pivoine.art
## Adding New Models
1. Add entry to models.yaml
2. Define Docker service in docker-compose.gpu.yaml
3. Restart orchestrator
That's it! The orchestrator automatically detects and manages the new model.
## Performance
| Model | VRAM | Startup | Speed |
|-------|------|---------|-------|
| Qwen 2.5 7B | 14GB | 120s | ~50 tok/sec |
| Flux.1 Schnell | 14GB | 60s | 4-5s/image |
| MusicGen Medium | 11GB | 45s | 60-90s for 30s audio |
Model switching overhead: 30-120 seconds
## License Notes
- vLLM: Apache 2.0
- Flux.1: Apache 2.0
- AudioCraft: MIT (code), CC-BY-NC (pre-trained weights - non-commercial)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
This commit finalizes the GPU infrastructure deployment on RunPod:
- Added qwen-2.5-7b model to LiteLLM configuration
- Self-hosted on RunPod RTX 4090 GPU server
- Connected via Tailscale VPN (100.100.108.13:8000)
- OpenAI-compatible API endpoint
- Rate limits: 1000 RPM, 100k TPM
- Marked GPU deployment as COMPLETE in deployment log
- vLLM 0.6.4.post1 with custom AsyncLLMEngine server
- Qwen/Qwen2.5-7B-Instruct model (14.25 GB)
- 85% GPU memory utilization, 4096 context length
- Successfully integrated with Open WebUI at ai.pivoine.art
Infrastructure:
- Provider: RunPod Spot Instance (~$0.50/hr)
- GPU: NVIDIA RTX 4090 24GB
- Disk: 50GB local SSD + 922TB network volume
- VPN: Tailscale (replaces WireGuard due to RunPod UDP restrictions)
Model now visible and accessible in Open WebUI for end users.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Removed supports_prompt_caching parameter that was causing 400 errors.
Prompt caching is automatically enabled by Anthropic when the client
sends cache_control blocks in messages - no config needed.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Added supports_prompt_caching: true to all Claude models:
- claude-sonnet-4
- claude-sonnet-4.5
- claude-3-5-sonnet
- claude-3-opus
- claude-3-haiku
This enables Anthropic's prompt caching feature across all models,
significantly reducing latency and costs for repeated requests
with the same system prompts.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Configure LiteLLM to use existing Redis from core stack for caching:
- Enabled cache with Redis backend
- Set TTL to 1 hour for cached responses
- Uses core_redis container on default port
This will improve performance by caching API responses.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Disabled cache setting that requires Redis configuration.
Prompt caching at the Anthropic API level is still enabled
via supports_prompt_caching setting.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Reduce database logging overhead and enable prompt caching:
- Disabled verbose logging (set_verbose: false)
- Disabled spend tracking logs to reduce DB writes
- Disabled tag tracking and daily spend logs
- Removed success/failure callbacks
- Enabled prompt caching for claude-sonnet-4.5
- Set log level to ERROR only
- Removed --detailed_debug flag from command
This should significantly improve response times by eliminating
unnecessary database writes for every request.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Replace HTTP Basic Auth with Authelia ForwardAuth for consistent
authentication across infrastructure:
- Asciinema Admin (admin.asciinema.dev.pivoine.art): Removed Basic Auth,
added Authelia protection
- FaceFusion (facefusion.ai.pivoine.art): Removed Basic Auth, added
Authelia protection
Updated Authelia access control to include both services with one_factor
policy.
All services now use Authelia for authentication, eliminating the need
to manage separate Basic Auth credentials.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add Mailpit service to NET stack with web UI at mailpit.pivoine.art
- Configure Mailpit to relay all emails through IONOS SMTP
- Migrate all 11+ services to use Mailpit instead of direct IONOS SMTP:
* SEXY: Directus API
* UTIL: Joplin, Mattermost, Vaultwarden, Tandoor, Linkwarden
* DEV: Gitea, n8n, Asciinema
* AI: Open WebUI
* NET: Netdata (via msmtp)
- Centralize SMTP credentials in mailpit-relay.yaml
- Simplify service configs (no auth/TLS for internal SMTP)
- Enable email monitoring via Mailpit web UI with Basic Auth
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Replace DISABLE_SWAGGER_UI with NO_DOCS and NO_REDOC
- Following official LiteLLM documentation for disabling API docs
- Disables both Swagger UI and Redoc documentation interfaces
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add --disable_swagger flag to LiteLLM command
- Improves security by hiding API documentation interface
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Watchtower was trying to pull updates from Docker Hub for facefusion-patched:3.5.0-cpu
which only exists locally, causing spam errors. Disabled Watchtower monitoring for this
container since it's a custom-built image with NSFW filter patches.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
