4bit

app.py

@@ -1,103 +1,145 @@
-from huggingface_hub import InferenceClient
+import os
+from threading import Thread
+from typing import Iterator
+
 import gradio as gr
+import spaces
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
 
-client = InferenceClient(
-    "tenyx/TenyxChat-7B-v1"
-)
+MAX_MAX_NEW_TOKENS = 2048
+DEFAULT_MAX_NEW_TOKENS = 1024
+MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
+
+DESCRIPTION = """
+TenyxChat-8x7B-v1 is part of the TenyxChat series, models trained to function as useful assistants. 
+The model is obtained via direct preference tuning using Tenyx's advanced fine-tuning technology. Model details available at [Hugging Face](https://huggingface.co/tenyx/TenyxChat-7B-v1). **It is currently loaded in 4-bit**.
+"""
+
+
+LICENSE = """
+<p/>
+---
+This demo is governed by the license available at https://huggingface.co/spaces/tenyx/TenyxChat-8x7B-v1/blob/main/LICENSE.txt."""
 
 
-def format_prompt(message, history):
-  prompt = "<s>"
-  for user_prompt, bot_response in history:
-    prompt += f"[INST] {user_prompt} [/INST]"
-    prompt += f" {bot_response}"
-  prompt += f"[INST] {message} [/INST]"
-  return prompt
+if not torch.cuda.is_available():
+    DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
 
+
+if torch.cuda.is_available():
+    model_id = "tenyx/TenyxChat-8x7B-v1"
+    #model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", torch_dtype=torch.bfloat16)
+    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
+    tokenizer = AutoTokenizer.from_pretrained(model_id)
+    tokenizer.use_default_system_prompt = False
+
+
+@spaces.GPU
 def generate(
-    prompt, history, system_prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
-):
-    temperature = float(temperature)
-    if temperature < 1e-2:
-        temperature = 1e-2
-    top_p = float(top_p)
+    message: str,
+    chat_history: list[tuple[str, str]],
+    system_prompt: str,
+    max_new_tokens: int = 1024,
+    temperature: float = 0.6,
+    top_p: float = 0.9,
+    top_k: int = 50,
+    repetition_penalty: float = 1.2,
+) -> Iterator[str]:
+    conversation = [{"role": "system", "content": "You are a helpful assistant developed by Tenyx."}]
+    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.append({"role": "user", "content": message})
+    
+    # if not message.strip():
+    #     return "It looks like your message is empty. How can I assist you today?"
 
+    input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt = True, return_tensors="pt")
+    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 = input_ids.to(model.device)
+
+    streamer = TextIteratorStreamer(tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True)
     generate_kwargs = dict(
-        temperature=temperature,
+        {"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,
-        do_sample=True,
-        seed=42,
     )
+    t = Thread(target=model.generate, kwargs=generate_kwargs)
+    t.start()
 
-    formatted_prompt = format_prompt(f"{system_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
-
-
-additional_inputs=[
-    gr.Textbox(
-        label="System Prompt",
-        max_lines=1,
-        interactive=True,
-    ),
-    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",
-    )
-]
+    outputs = []
+    for text in streamer:
+        outputs.append(text)
+        yield "".join(outputs)
 
-examples=[["I'm planning a vacation to Japan. Can you suggest a one-week itinerary including must-visit places and local cuisines to try?", None, None, None, None, None, ],
-          ["Can you write a short story about a time-traveling detective who solves historical mysteries?", None, None, None, None, None,],
-          ["I'm trying to learn French. Can you provide some common phrases that would be useful for a beginner, along with their pronunciations?", None, None, None, None, None,],
-          ["I have chicken, rice, and bell peppers in my kitchen. Can you suggest an easy recipe I can make with these ingredients?", None, None, None, None, None,],
-          ["Can you explain how the QuickSort algorithm works and provide a Python implementation?", None, None, None, None, None,],
-          ["What are some unique features of Rust that make it stand out compared to other systems programming languages like C++?", None, None, None, None, None,],
-         ]
 
-gr.ChatInterface(
+chat_interface = gr.ChatInterface(
     fn=generate,
-    chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
-    additional_inputs=additional_inputs,
-    title="Mixtral 46.7B",
-    examples=examples,
-    concurrency_limit=20,
-).launch(show_api=False)
+    additional_inputs=[
+        gr.Textbox(label="System prompt", lines=6),
+        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=0.1,
+        ),
+        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,
+    examples=[
+        ["Hello there! How are you doing?"],
+        ["Can you explain briefly to me what is the Python programming language?"],
+        ["Explain the plot of Cinderella in a sentence."],
+        ["How many hours does it take a man to eat a Helicopter?"],
+        ["Write a 100-word article on 'Benefits of Open-Source in AI research'"],
+    ],
+)
+
+with gr.Blocks(css="style.css") as demo:
+    gr.Markdown(DESCRIPTION)
+    gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button")
+    chat_interface.render()
+    gr.Markdown(LICENSE)
+
+if __name__ == "__main__":
+    demo.queue(max_size=20).launch()
+
+