Update app.py

app.py

@@ -1,11 +1,8 @@
 import os
-from threading import Thread
-from typing import Iterator
-
 import gradio as gr
 import spaces
 import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
 MAX_MAX_NEW_TOKENS = 2048
 DEFAULT_MAX_NEW_TOKENS = 1024
@@ -19,20 +16,16 @@ This Space demonstrates [L-MChat](https://huggingface.co/collections/Artples/l-m
 if not torch.cuda.is_available():
     DESCRIPTION += "\n<p>Running on CPU! This demo does not work on CPU.</p>"
 
-# Dictionary to manage model details
 model_details = {
     "Fast-Model": "Artples/L-MChat-Small",
     "Quality-Model": "Artples/L-MChat-7b"
 }
 
-# Initialize models and tokenizers based on availability
 models = {name: AutoModelForCausalLM.from_pretrained(model_id, device_map="auto") for name, model_id in model_details.items()}
 tokenizers = {name: AutoTokenizer.from_pretrained(model_id) for name, model_id in model_details.items()}
-for tokenizer in tokenizers.values():
-    tokenizer.use_default_system_prompt = False
 
 @spaces.GPU(enable_queue=True, duration=90)
-def generate(
+async def generate(
     model_choice: str,
     message: str,
     chat_history: list[tuple[str, str]],
@@ -42,86 +35,26 @@ def generate(
     top_p: float = 0.9,
     top_k: int = 50,
     repetition_penalty: float = 1.2,
-) -> Iterator[str]:
+) -> str:
     model = models[model_choice]
     tokenizer = tokenizers[model_choice]
-
-    conversation = []
-    if system_prompt:
-        conversation.append({"role": "system", "content": system_prompt})
-    for user, assistant in chat_history:
-        conversation.extend([{"role": "user", "content": user}, {"role": "assistant", "content": assistant}])
+    
+    conversation = [{"role": "system", "content": system_prompt}] if system_prompt else []
+    conversation += [{"role": "user", "content": user}, {"role": "assistant", "content": assistant} for user, assistant in chat_history]
     conversation.append({"role": "user", "content": message})
 
-    input_ids = tokenizer.apply_chat_template(conversation, return_tensors="pt", add_generation_prompt=True)
-    if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
-        input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
-        gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
+    input_ids = tokenizer(conversation, return_tensors="pt", truncation=True, max_length=MAX_INPUT_TOKEN_LENGTH).input_ids
     input_ids = input_ids.to(model.device)
 
-    streamer = TextIteratorStreamer(tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True)
-    generate_kwargs = dict(
-        {"input_ids": input_ids},
-        streamer=streamer,
-        max_new_tokens=max_new_tokens,
-        do_sample=True,
-        top_p=top_p,
-        top_k=top_k,
-        temperature=temperature,
-        num_beams=1,
-        repetition_penalty=repetition_penalty,
-    )
-    t = Thread(target=model.generate, kwargs=generate_kwargs)
-    t.start()
+    output_ids = model.generate(input_ids, max_length=MAX_INPUT_TOKEN_LENGTH + max_new_tokens, temperature=temperature, top_p=top_p, top_k=top_k, repetition_penalty=repetition_penalty)
+    output_text = tokenizer.decode(output_ids[0], skip_special_tokens=True)
 
-    outputs = []
-    for text in streamer:
-        outputs.append(text)
-        yield "".join(outputs)
+    return output_text
 
 chat_interface = gr.ChatInterface(
     theme='ehristoforu/RE_Theme',
     fn=generate,
-    additional_inputs=[
-        gr.Textbox(label="System prompt", lines=6),
-        gr.Dropdown(label="Model Choice", choices=list(model_details.keys()), value="Quality-Model"),
-        gr.Slider(
-            label="Max new tokens",
-            minimum=1,
-            maximum=MAX_MAX_NEW_TOKENS,
-            step=1,
-            value=DEFAULT_MAX_NEW_TOKENS,
-        ),
-        gr.Slider(
-            label="Temperature",
-            minimum=0.1,
-            maximum=4.0,
-            step=0.1,
-            value=0.6,
-        ),
-        gr.Slider(
-            label="Top-p (nucleus sampling)",
-            minimum=0.05,
-            maximum=1.0,
-            step=0.05,
-            value=0.9,
-        ),
-        gr.Slider(
-            label="Top-k",
-            minimum=1,
-            maximum=1000,
-            step=1,
-            value=50,
-        ),
-   gr.Slider(
-            label="Repetition penalty",
-            minimum=1.0,
-            maximum=2.0,
-            step=0.05,
-            value=1.2,
-        ),
-    ],
-    stop_btn=None,
+    additional_inputs=[gr.Textbox(label="System prompt", lines=6), gr.Dropdown(label="Model Choice", choices=list(model_details.keys()), value="Quality-Model")],
     examples=[
         ["Hello there! How are you doing?"],
         ["Can you explain briefly to me what is the Python programming language?"],
@@ -136,4 +69,4 @@ with gr.Blocks(css="style.css") as demo:
     chat_interface.render()
 
 if __name__ == "__main__":
-    demo.queue(max_size=20).launch()
+    demo.launch()