Add profiling script for GPU bottleneck analysis in EasyCom extraction. Introduce command-line options for GPU ID, log directory, and sampling period. Implement stage profiling to measure extraction performance and synchronize CUDA operations. Enhance CausalSourceAV class with new parameters for decoder threads and open readers, optimizing video decoding process.

README.md

@@ -680,6 +680,18 @@ conversion can differ slightly; validate a short Decord/PyNvVideoCodec pair
 before extracting every session, and do not mix decoder backends across
 train/test splits.
 
+For CPU Decord, the direct reader intentionally indexes each clip and releases
+that temporary reader, then keeps an LRU of at most two live readers during
+extraction. Each reader defaults to one decode thread. These limits avoid one
+worker retaining a thread pool for every one-minute EasyCom clip and can be
+overridden with `--decoder-threads` and `--max-open-readers`. The recommended
+full-run settings are:
+
+```text
+--decoder decord --decoder-threads 1 --max-open-readers 2
+--frame-cache-size 64
+```
+
 Matching `states.shape[0]` to the target is necessary but not sufficient.
 The direct reader also validates EasyCom's 20-Hz source inventory, source PTS
 monotonicity, dense duration, crop bounds, audio sample count, and causal