4 ms·
That's a common issue with all video (and audio, but at longer durations, like a few minutes) models. The context is simply too long for current LLMs.
by lostmsu 1y ago
That's a common issue with all video (and audio, but at longer durations, like a few minutes) models. The context is simply too long for current LLMs.