Fix model card: use AudioLDMCAASteeringController + AudioLDM2 backbone

README.md

@@ -1,7 +1,6 @@
 ---
 library_name: audio-interv
 tags:
-  - ace-step
   - activation-steering
   - audio
   - audioldm
@@ -14,21 +13,24 @@ tags:
   - steering
 ---
 
-# Steering Vectors — `female_vocals` (ACE-Step)
+# Steering Vectors — `female_vocals` (AudioLDM2)
 
-Steering vectors for the **female_vocals** concept on ACE-Step, computed via contrastive activation addition (CAA).
+Steering vectors for the **female_vocals** concept on AudioLDM2, computed via contrastive activation addition (CAA).
 
 ## Quickstart
 
 ```python
-from src.steering import SteerableACEModel
-from src.steering.methods.caa import CAASteeringController
+from src.steering import SteerableAudioLDMModel, AudioLDMCAASteeringController
 
-model = SteerableACEModel(device="cuda")
-ctrl = CAASteeringController.from_pretrained("lukasz-staniszewski/audioldm2-caa-female-vocals", alpha=20)
+model = SteerableAudioLDMModel(device="cuda")
+ctrl = AudioLDMCAASteeringController.from_pretrained("lukasz-staniszewski/audioldm2-caa-female-vocals", alpha=1.0)
 
 with model.steer(ctrl):
-    audio = model.generate(prompt="instrumental music", seed=0)
+    out = model.generate(
+        prompt="instrumental music",
+        num_inference_steps=30, audio_length_in_s=10.0,
+        guidance_scale=3.5, seed=0,
+    )
 ```
 
 ## Generation config