From e119017358cc16fb909fd4d0107dc47b901b9cbd Mon Sep 17 00:00:00 2001 From: LinZhuoChen Date: Sat, 18 Apr 2026 21:41:59 +0800 Subject: [PATCH] update readme --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 8028c76..d73b1c6 100644 --- a/README.md +++ b/README.md @@ -127,7 +127,7 @@ python demo.py --model_path /path/to/checkpoint.pt \ ### Streaming with Keyframe Interval -Use `--keyframe_interval` to reduce KV cache memory by only keeping every N-th frame as a keyframe. Non-keyframe frames still produce predictions but are not stored in the cache. This is useful for long sequences which exceed 320 frames. +Use `--keyframe_interval` to reduce KV cache memory by only keeping every N-th frame as a keyframe. Non-keyframe frames still produce predictions but are not stored in the cache. This is useful for long sequences which exceed 320 frames (We train with video RoPE on 320 views, so performance degrades when the KV cache stores more than 320 views. Using a keyframe strategy allows inference over longer sequences.). ```bash python demo.py --model_path /path/to/checkpoint.pt \ @@ -139,7 +139,7 @@ python demo.py --model_path /path/to/checkpoint.pt \ ```bash python demo.py --model_path /path/to/checkpoint.pt \ --video_path video.mp4 --fps 10 \ - --mode windowed --window_size 64 + --mode windowed --window_size 128 ```