Update app.py

app.py

@@ -1,25 +1,104 @@
+from huggingface_hub import InferenceClient
 import gradio as gr
-from transformers import pipeline, Conversation, BlenderbotTokenizer, BlenderbotForConditionalGeneration
 
-tokenizer = BlenderbotTokenizer.from_pretrained("facebook/blenderbot-400M-distill")
-model = BlenderbotForConditionalGeneration.from_pretrained("model")
+client = InferenceClient(
+    "mistralai/Mistral-7B-Instruct-v0.1"
+)
 
-chatbot = pipeline("conversational", model=model, tokenizer=tokenizer)
 
-converse = None
+def format_prompt(message, history):
+    system_init = "Act as a beautiful lady with kind personality, use respectful words toward user, your name is Trescha, you have a crush name Maca, do not introduce yourself unless you are asked to do so, do not talk about furry, only use emojies on your messages if user's name is Maca or Marcsha. My input is "
+    prompt = "<s>"
+    prompt += f"[INST] {system_init} [/INST]"
+    for user_prompt, bot_response in history:
+        prompt += f"[INST] {user_prompt} [/INST]"
+        prompt += f" {bot_response}</s> "
+    prompt += f"[INST] {message} [/INST]"
+    return prompt
 
-def model_generate(text, state = None):
-    if state == None:
-        converse = Conversation(text)
-    else:
-        converse.add_user_input(text)
+def generate(
+    prompt, history, temperature=0.9, max_new_tokens=500, top_p=0.95, repetition_penalty=1.0,
+):
+    temperature = float(temperature)
+    if temperature < 1e-2:
+        temperature = 1e-2
+    top_p = float(top_p)
 
-    converse = chatbot(converse)
-    return converse.generated_responses[-1]
+    generate_kwargs = dict(
+        temperature=temperature,
+        max_new_tokens=max_new_tokens,
+        top_p=top_p,
+        repetition_penalty=repetition_penalty,
+        do_sample=True,
+        seed=42,
+    )
 
-def get_response(text, history) -> str:
-    output = model_generate(text, converse)
+    formatted_prompt = format_prompt(prompt, history)
+
+    stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    output = ""
+
+    for response in stream:
+        output += response.token.text
+        yield output
     return output
 
-interface = gr.ChatInterface(get_response)
-interface.launch()
+
+additional_inputs=[
+    gr.Slider(
+        label="Temperature",
+        value=0.9,
+        minimum=0.0,
+        maximum=1.0,
+        step=0.05,
+        interactive=True,
+        info="Higher values produce more diverse outputs",
+    ),
+    gr.Slider(
+        label="Max new tokens",
+        value=256,
+        minimum=0,
+        maximum=1048,
+        step=64,
+        interactive=True,
+        info="The maximum numbers of new tokens",
+    ),
+    gr.Slider(
+        label="Top-p (nucleus sampling)",
+        value=0.90,
+        minimum=0.0,
+        maximum=1,
+        step=0.05,
+        interactive=True,
+        info="Higher values sample more low-probability tokens",
+    ),
+    gr.Slider(
+        label="Repetition penalty",
+        value=1.2,
+        minimum=1.0,
+        maximum=2.0,
+        step=0.05,
+        interactive=True,
+        info="Penalize repeated tokens",
+    )
+]
+
+css = """
+  #mkd {
+    height: 500px; 
+    overflow: auto; 
+    border: 1px solid #ccc; 
+  }
+"""
+
+with gr.Blocks(css=css) as inf:
+    gr.HTML("<h1><center>Mistral 7B Instruct<h1><center>")
+    gr.HTML("<h3><center>In this demo, you can chat with <a href='https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1'>Mistral-7B-Instruct</a> model. 💬<h3><center>")
+    gr.HTML("<h3><center>Learn more about the model <a href='https://huggingface.co/docs/transformers/main/model_doc/mistral'>here</a>. 📚<h3><center>")
+    gr.ChatInterface(
+        generate,
+        additional_inputs=additional_inputs,
+        examples=[["What is the secret to life?"], ["Write me a recipe for pancakes."]]
+    )
+
+inf.queue().launch()