ResAdapt - a Xnhyacinth Collection

Xnhyacinth 's Collections

Attention research checkpoints

Additional research data

RSIBench-Context

ResAdapt

updated about 7 hours ago

ResAdapt checkpoint, multimodal training mixture, and video-QA evaluation data.

Xnhyacinth/ResAdapt-Research-Checkpoint

Video-Text-to-Text • 8B • Updated about 14 hours ago
Xnhyacinth/Resadapt-Mix

Viewer • Updated Jun 24 • 374k • 130

Note Training mixture and media archives. Start with data/train_lens.parquet and data/TSPO-10K/val.parquet; materialize media before training.
Xnhyacinth/FBench

Viewer • Updated about 16 hours ago • 1.9k • 11

Note Video-QA evaluation tables: longvideobench/validation and videomme/test.