remove top k slider

main.py

@@ -12,7 +12,6 @@ repo_id = "yezdata/EmCoder"
 
 tokenizer = AutoTokenizer.from_pretrained(repo_id, trust_remote_code=True)
 model = AutoModel.from_pretrained(repo_id, trust_remote_code=True)
-max_labels = getattr(model.config, "num_labels", 28)
 
 model.eval()
 
@@ -113,7 +112,7 @@ def predict_api(request: PredictRequest):
 def health_check():
     return {"status": "healthy"}
 
-def gradio_predict(text, top_n, monte_carlo, n_samples):
+def gradio_predict(text, monte_carlo, n_samples):
     request_data = PredictRequest(text=text, monte_carlo=bool(monte_carlo), n_samples=int(n_samples))
     response = predict_api(request_data)
 
@@ -123,12 +122,10 @@ def gradio_predict(text, top_n, monte_carlo, n_samples):
         reverse=True
     )
 
-    top_preds = sorted_preds[:int(top_n)]
-    
     standard_rows = []
     mc_rows = []
     
-    for label_name, metrics in top_preds:
+    for label_name, metrics in sorted_preds:
         if monte_carlo:
             prob = metrics["mean_probability"]
             mc_rows.append([
@@ -168,13 +165,6 @@ with gr.Blocks(title="EmCoder - Probabilistic Emotion Recognition") as ui:
                 placeholder="Input text for classification...", 
                 lines=3
             )
-            top_n_slider = gr.Slider(
-                minimum=1, 
-                maximum=max_labels, 
-                value=min(5, max_labels), 
-                step=1, 
-                label="Top Emotions to display"
-            )
             use_mc = gr.Checkbox(label="Use Monte Carlo Dropout (Uncertainty Estimation)", value=False)
             mc_samples_slider = gr.Slider(
                 minimum=5, 
@@ -208,7 +198,7 @@ with gr.Blocks(title="EmCoder - Probabilistic Emotion Recognition") as ui:
 
     submit_btn.click(
         fn=gradio_predict,
-        inputs=[input_text, top_n_slider, use_mc, mc_samples_slider],
+        inputs=[input_text, use_mc, mc_samples_slider],
         outputs=[output_table_standard, output_table_mc]
     )