Update app.py

app.py

@@ -9,6 +9,12 @@ HF_TOKEN = os.environ.get("HF_TOKEN", None)
 USER_NAME = os.environ.get("USER_NAME", None)
 APP_PASSWORD = os.environ.get("APP_PASSWORD", None)
 
+# Define the model URL using the model_name
+model_url = 'https://api-inference.huggingface.co/models/h2oai/h2ogpt-16k-codellama-34b-instruct'
+
+# Create the InferenceClient
+client = InferenceClient(model_url, headers={"Authorization": f"Bearer {HF_TOKEN}"})
+
 STOP_SEQUENCES = ["\nUser:", "<|endoftext|>", " User:", "###"]
 
 EXAMPLES = [
@@ -67,7 +73,6 @@ def generate(
     return output
 
 additional_inputs=[
-    gr.Textbox("google/flan-t5-small", label="Input Model Name", type="text", elem_id="model-name-input"),
     gr.Textbox("", label="Optional system prompt"),
     gr.Slider(
         label="Temperature",
@@ -132,17 +137,4 @@ with gr.Blocks() as demo:
         additional_inputs=additional_inputs,
     )
 
-# Create a function to get the model name from the input
-def get_model_name():
-    return gr.get_value("model-name-input")
-
-# Get the model name from the input
-model_name = get_model_name()
-
-# Define the model URL using the model_name
-model_url = 'https://api-inference.huggingface.co/models/' + model_name
-
-# Create the InferenceClient
-client = InferenceClient(model_url, headers={"Authorization": f"Bearer {HF_TOKEN}"})
-
 demo.queue(concurrency_count=100, api_open=True).launch(show_api=True, auth=(USER_NAME, APP_PASSWORD))