---

name: app-deployment description: "Install, configure, launch, and sanity-check InternGPT/InternChat services." metadata: disco-role: operating disable-model-invocation: true license: Apache 2.0

app-deployment

Use this sub-skill when the task is to install, configure, launch, containerize, or sanity-check an InternGPT/InternChat Gradio service. It owns app-level CLI planning, model-zoo placement, HTTPS/OpenAI setup, Docker GPU deployment, and static validation of --load/--tab choices.

Route detailed workflow questions elsewhere:

• Read ../visual-dialogue-tools/SKILL.md for detailed image, mask, OCR, SAM, Husky VQA, inpainting, and image-tool behavior after the service is running.
• Read ../cross-modal-generation/SKILL.md for detailed ImageBind audio/thermal/image-conditioned generation and DragGAN point-editing usage.
• Read ../video-understanding/SKILL.md for detailed video caption/action/dense-caption/TikTok workflows.

Deployment routing

• For CLI flags, supported --load classes, tab names, and launch-plan sanity checks, read references/cli-and-config.md.
• For model_zoo/, checkpoints, Husky/LLaMA conversion, SAM, LaMa, Tag2Text, GRiT, StyleGAN, ImageBind, and StableUnCLIP prerequisites, read references/model-zoo.md.
• For basic, full, DragGAN-only, HTTPS, OpenAI, and Docker GPU launch recipes, read references/deployment-recipes.md.
• For common failures such as missing model_zoo, bad LLaMA/Husky partial conversion, missing OpenAI key, OPENAI_API_BASE, certificate errors, CUDA/VRAM mismatch, Docker placeholder volumes, detectron2/OpenCV pitfalls, and -e expectations, read references/troubleshooting.md.
• Run scripts/validate_load_plan.py before recommending a launch command or compose command when the user asks which --load or --tab string to use.

Quick decision guide

1. Choose the smallest tab set that matches the user's task: DragGAN for StyleGAN point editing, Image for image upload/click/OCR/segmentation/dialogue, Audio for ImageBind audio input, and Video for video upload/analysis.
2. Choose the smallest direct --load list that provides those tabs' tools; template tools are auto-created only after their prerequisites load.
3. Prefer -e/--e-mode for constrained VRAM, but do not describe it as CPU-only: the app still expects CUDA for normal launches, and the speech model is initialized on cuda:0.
4. Check model_zoo/, certificate/, and OpenAI configuration before treating a launch failure as a Python bug.
5. For Docker, replace every placeholder volume with real host directories and keep GPU reservation/container-toolkit requirements explicit.

Safe default recipes

• Basic image-dialogue service: HuskyVQA_cuda:0,SegmentAnything_cuda:0,ImageOCRRecognition_cuda:0 with -e on a CUDA machine and a populated model_zoo/.
• Low-memory DragGAN-only HTTPS service: StyleGAN_cuda:0, --tab DragGAN, --https, and -e, plus certificate/cert.pem, certificate/key.pem, and StyleGAN checkpoint readiness.
• Full multimodal demo: use the long full-feature load string in references/deployment-recipes.md; warn that it is heavy and requires multiple model families, CUDA, and usually far more setup than a task-specific launch.