Update

app.py

@@ -6,7 +6,7 @@ import gradio as gr
 # iface = gr.Interface(fn=greet, inputs="text", outputs="text")
 # iface.launch()
 
-import spaces
+# import spaces
 import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM, Trainer, TrainingArguments, pipeline, set_seed
 
@@ -42,29 +42,18 @@ examples=[
 #     else:
 #         print("You downvoted this response: " + data.value)
         
-@spaces.GPU
+# @spaces.GPU
 def generate_response(message, chatbot, system_prompt="",):
     set_seed(SEED)
 
-    if system_prompt != "":
-        input_prompt = f"<s>[INST] <<SYS>>\n{system_prompt}\n<</SYS>>\n\n "
-    else:
-        input_prompt = f"<s>[INST] "
-
-    for interaction in chatbot:
-        input_prompt = input_prompt + str(interaction[0]) + " [/INST] " + str(interaction[1]) + " </s><s>[INST] "
-
-    input_prompt = input_prompt + str(message) + " [/INST] "
-    print(input_prompt)
-
-    tokenized_prompt = tokenizer(input_prompt, return_tensors="pt", padding=True, truncation=True, max_length=128)
-    # print(tokenized_prompt)
+    tokenized_prompt = tokenizer(message, return_tensors="pt", padding=True, truncation=True, max_length=128)
+    print(tokenized_prompt)
 
     output_sequences = model.generate(**tokenized_prompt, max_length=1024, num_return_sequences=1)
     decoded_output = tokenizer.decode(output_sequences[0], skip_special_tokens=True)
-    # print(decoded_output)
+    print(decoded_output)
 
-    return decoded_output
+    yield decoded_output
 
 chatbot_stream = gr.Chatbot()
 chat_interface_stream = gr.ChatInterface(generate_response, 
@@ -84,4 +73,5 @@ with gr.Blocks() as demo:
     # chatbot_stream.like(vote, None, None)
     chat_interface_stream.render()
         
-demo.queue(max_size=100).launch()
+if __name__ == "__main__":
+    demo.queue().launch(share=True)