add correct usage in huggingface

README.md

@@ -89,6 +89,33 @@ This project fine‑tunes a Wav2Vec2 audio classifier (e.g., `facebook/wav2vec2-
 
 Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Evaluate on target devices/microphones; add noise augmentation and tune detection thresholds for deployment context.
 
+## How to use it in HuggingFace
+```bash
+from transformers import AutoFeatureExtractor, AutoModelForAudioClassification, pipeline
+
+model_id = "Amirhossein75/Keyword-Spotting"
+```
+# Option A — simple:
+```bash
+clf = pipeline("audio-classification", model=model_id)
+print(clf("path/to/1sec_16kHz.wav"))
+```
+
+# Option B — manual pre/post:
+```bash
+fe = AutoFeatureExtractor.from_pretrained(model_id)
+model = AutoModelForAudioClassification.from_pretrained(model_id)
+
+import soundfile as sf, torch
+wave, sr = sf.read("path/to/1sec_16kHz.wav")
+inputs = fe(wave, sampling_rate=sr, return_tensors="pt")
+with torch.no_grad():
+    logits = model(**inputs).logits
+pred_id = int(logits.argmax(-1))
+print(model.config.id2label[pred_id])
+
+```
+
 ## How to Get Started with the Model
 
 Use the code below to get started with the model.