erratum: link proposed upstream fix (#50101)

README.md

@@ -64,11 +64,11 @@ Full commit hashes are in [PINS.json](PINS.json). The evaluators require those p
 
 ## Processor behavior and checks
 
-In the pinned Hugging Face feature extractor, fusion mode selects one clip with `np.random.randint` and sets `is_longer=True` when every clip in a batch fits within the 10-second window. At batch 1, this routes every clip through the long-audio branch. Larger batches also make the selected clip depend on the random state. The current short-clip protocol explicitly sets the flag to `False`; seeding controls repeatability separately.
+In the pinned Hugging Face feature extractor, fusion mode selects one clip with `np.random.randint` and sets `is_longer=True` when every clip in a batch fits within the 10-second window. At batch 1, this routes every clip through the long-audio branch. Larger batches also make the selected clip depend on the random state. The current short-clip protocol explicitly sets the flag to `False`; seeding controls repeatability separately. A fix that removes the assignment, so the flag follows clip length, is proposed in [PR #50101](https://github.com/huggingface/transformers/pull/50101). These results set the flag explicitly, so they describe the short-clip branch under either processor behaviour.
 
 On the AudioCaps fixture, the short-clip fused path scores **0.3490 T2A R@1**, compared with **0.3417** for the `laion_clap` control. The calculated difference is **0.73 pp**, within the specified **2 pp** validation tolerance. Sources: [validation summary](erratum_analysis.json) and [package control](results_control/laion__clap-htsat-fused__audiocaps__control.json).
 
-Two seeded processor-native batch-8 runs both score **0.321839** and have identical T2A per-query ranks: [run A](results_erratum/laion__clap-htsat-fused__audiocaps__e_seed_b8a.json), [run B](results_erratum/laion__clap-htsat-fused__audiocaps__e_seed_b8b.json). This supports repeatability for that tested configuration; the branch choice still affects the score. The processor behavior is documented in [transformers issue #50097](https://github.com/huggingface/transformers/issues/50097).
+Two seeded processor-native batch-8 runs both score **0.321839** and have identical T2A per-query ranks: [run A](results_erratum/laion__clap-htsat-fused__audiocaps__e_seed_b8a.json), [run B](results_erratum/laion__clap-htsat-fused__audiocaps__e_seed_b8b.json). This supports repeatability for that tested configuration; the branch choice still affects the score. The processor behavior is documented in [transformers issue #50097](https://github.com/huggingface/transformers/issues/50097), with a fix proposed in [PR #50101](https://github.com/huggingface/transformers/pull/50101).
 
 An exploratory untrimmed Clotho run scores **0.1233 fused** and **0.1435 unfused** on T2A R@1. Both are below their trimmed scores. This is one seeded run per checkpoint without intervals: [fused](results_erratum/laion__clap-htsat-fused__clotho5__b1_full.json), [unfused](results_erratum/laion__clap-htsat-unfused__clotho5__b1_full.json).