CapRL-Video-178K: Video Path Index for Multimodal Models

CapRL-Video-178K is a dataset providing file paths to over 97,000 video clips. The dataset is hosted by internlm on Hugging Face and was last updated on 2026-05-25. It serves as an index for video data from the LLaVA-Video-178K collection, which includes clips from sources like YouTube and ActivityNet.

Use Cases

• Training video-language models based on the referenced video path structure.
• Aligning video content with textual descriptions using the indexed video files.
• Benchmarking video understanding tasks on a collection of clips from sources like YouTube and ActivityNet.

Strengths

• Indexes over 97,000 video samples across specified splits, such as 72,970 samples in the '0_30_s_youtube_v0_1' split.
• Provides a direct link to the source LLaVA-Video-178K dataset on Hugging Face for accessing the actual video content.

Limitations

• Row count for the full dataset is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

lmms-lab/LLaVA-Video-178K on Hugging Face

Collection Method

Likely compiled as an index of video file paths from a larger multimodal collection.

Freshness

Last updated 2026-05-25 02:08:18; freshness should be verified.