Update README.md

README.md

@@ -13,7 +13,7 @@ contents of dataset:
 cyber_pipette # dataset root path
   └── data/
       ├── metadata_ID1_240808.json
-      ├── segment_ids_ID1_240808.bin # for each frame segment_ids[i] uniquely points to the segment index that frame i came from. You may want to use this to separate non-contiguous frames from different videos (transitions).
+      ├── segment_ids_ID1_240808.bin # for each frame segment_ids uniquely points to the segment index that frame i came from. You may want to use this to separate non-contiguous frames from different videos (transitions).
       ├── videos_ID1_240808.bin # 16x16 image patches at 30hz, each patch is vector-quantized into 2^18 possible integer values. These can be decoded into 256x256 RGB images using the provided magvit2.ckpt weights.
       ├── ...
   └── ...