executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
legacy-workflowsGuides InternVideo1 legacy pretraining and downstream task routes, including VideoMAE, ViCLIP, retrieval, localization, VQA, open-set recognition, and submodule cautions.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this sub-skill when the user works with InternVideo1 or legacy InternVideo checkpoints/tasks from the 2022/2023 code areas.
| User signal | Area |
|---|---|
Video masked autoencoder, run_mae_pretraining.py, run_class_finetuning.py | VideoMAE pretraining/finetuning |
| ViCLIP, video CLIP, InternVid-10M/200M transfer | ViCLIP / Multi-Modalities pretraining |
| MSR-VTT, DiDeMo, LSMDC, MSVD, VATEX, ActivityNet retrieval | Downstream video-text retrieval |
| AVA, THUMOS14, ActivityNet localization | Spatial/temporal action localization |
| VQA, zero-shot action recognition, zero-shot multiple choice | Multi-modalities downstream |
| UCF/HMDB open-set | Open-set action recognition |
| VLN-CE | Visual-language navigation |