update

app.py

@@ -39,7 +39,7 @@ def run_inference(model_name, input_text, num_tokens=6):
         
         model_name = model_name.split("/")[1]
         start_time = time.time()
-        if input_text is None : 
+        if input_text is None or input_text == "": 
             return "Please provide an input text for the model"
         result = subprocess.run(
             f"python run_inference.py -m models/{model_name}/ggml-model-i2_s.gguf -p \"{input_text}\" -n {num_tokens} -temp 0",
@@ -65,7 +65,7 @@ def run_transformers(model_name, input_text, num_tokens):
     # Load the model and tokenizer dynamically if needed (commented out for performance)
     # if model_name=="TinyLlama/TinyLlama-1.1B-Chat-v1.0" : 
     print(input_text)
-    if input_text is None : 
+    if input_text is None or input_text == "": 
         return "Please provide an input text for the model", None
     tokenizer = AutoTokenizer.from_pretrained(model_name)
     model = AutoModelForCausalLM.from_pretrained(model_name)
@@ -125,8 +125,8 @@ def interface():
         with gr.Row():
             transformer_model_dropdown = gr.Dropdown(
                 label="Select Transformers Model",
-                choices=["TinyLlama/TinyLlama-1.1B-Chat-v1.0"],  # Replace with actual models
-                value="TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+                choices=["TinyLlama/TinyLlama_v1.1"],  # Replace with actual models
+                value="TinyLlama/TinyLlama_v1.1",
                 interactive=True
             )
             compare_button = gr.Button("Run Transformers Inference", elem_id="compare-button")