add EzAudio API

app.py

@@ -176,7 +176,24 @@ def get_stable_audio_open(prompt):
     except:
         raise gr.Error("Stable Audio Open space API is not ready, please try again in few minutes ")
     
-    
+def get_ezaudio(prompt):
+    try:
+        client = Client("OpenSound/EzAudio")
+        result = client.predict(
+        		text=prompt,
+        		length=10,
+        		guidance_scale=5,
+        		guidance_rescale=0.75,
+        		ddim_steps=50,
+        		eta=1,
+        		random_seed=0,
+        		randomize_seed=True,
+        		api_name="/generate_audio"
+        )
+        print(result)
+        return result
+    except:
+        raise gr.Error("EzAudio space API is not ready, please try again in few minutes ")
     
 def infer(image_in, chosen_model):
     caption = get_caption(image_in)
@@ -198,6 +215,9 @@ def infer(image_in, chosen_model):
     elif chosen_model == "Stable Audio Open" :
         stable_audio_open_result = get_stable_audio_open(caption)
         return stable_audio_open_result
+    elif chosen_model == "EzAudio" :
+        ezaudio_result = get_ezaudio(caption)
+        return ezaudio_result
 
 css="""
 #col-container{
@@ -220,7 +240,7 @@ with gr.Blocks(css=css) as demo:
         with gr.Column():
             image_in = gr.Image(sources=["upload"], type="filepath", label="Image input", value="oiseau.png")
             with gr.Row():
-                chosen_model = gr.Dropdown(label="Choose a model", choices=["MAGNet", "AudioLDM-2", "AudioGen", "Tango", "Tango 2", "Stable Audio Open"], value="AudioLDM-2")
+                chosen_model = gr.Dropdown(label="Choose a model", choices=["MAGNet", "AudioLDM-2", "AudioGen", "Tango", "Tango 2", "Stable Audio Open", "EzAudio"], value="AudioLDM-2")
                 submit_btn = gr.Button("Submit")
         with gr.Column():
             audio_o = gr.Audio(label="Audio output")