document the run_system.py plug-in path

README.md

@@ -40,13 +40,19 @@ rather than deleted so the exclusions stay auditable.
 
 ```bash
 git clone https://github.com/Ampixa/nepaliconformer
-huggingface-cli download ampixa/neptel --repo-type dataset --local-dir neptel
-
-# run your system over neptel/audio/*.wav, write [{"seg": "...", "text": "..."}, ...]
-python nepaliconformer/eval/score_reference.py \
-  --manifest neptel/references.json --hyp your_outputs.json
+cd nepaliconformer/benchmark
+python fetch_audio.py neptel_audio        # pulls this dataset, no login
+
+# then plug your system in — pick the adapter that fits:
+python run_system.py --hf openai/whisper-large-v3 --name whisper --lang ne
+python run_system.py --nemo your_model.nemo --name mine
+python run_system.py --cmd "your-cli -f {audio}" --name mine
+python run_system.py --py yourpkg.module:transcribe --name mine
 ```
 
+Each writes `outputs/<name>.json` and prints the WER; add
+`--compare outputs/nepali-conformer-offline.json` for a paired bootstrap against ours.
+
 The scorer normalizes both sides identically (digits to spoken Nepali words, punctuation
 stripped) and reports a paired bootstrap when given `--compare`.