app.py

 # import gradio as gr

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer, pipeline
from threading import Thread

model_id = "rasyosef/Llama-3.2-180M-Amharic-Instruct"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

llama_am = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    pad_token_id=tokenizer.pad_token_id,
    eos_token_id=tokenizer.eos_token_id
  )

# Function that accepts a prompt and generates text using the phi2 pipeline
def generate(message, chat_history, max_new_tokens=64):

  history = []

  for sent, received in chat_history:
    history.append({"role": "user", "content": sent})
    history.append({"role": "assistant", "content": received})

  history.append({"role": "user", "content": message})
  #print(history)

  if len(tokenizer.apply_chat_template(history)) > 512:
    yield "chat history is too long"
  else:
    # Streamer
    streamer = TextIteratorStreamer(tokenizer=tokenizer, skip_prompt=True, skip_special_tokens=True, timeout=300.0)
    thread = Thread(target=llama_am,
                    kwargs={
                        "text_inputs":history,
                        "max_new_tokens":max_new_tokens,
                        "repetition_penalty":1.15,
                        "streamer":streamer
                        }
                    )
    thread.start()

    generated_text = ""
    for word in streamer:
      generated_text += word
      response = generated_text.strip()

      yield response

# Chat interface with gradio
with gr.Blocks() as demo:
  gr.Markdown("""
  # Llama 3.2 180M Amharic Chatbot Demo

  This chatbot was created using [Llama-3.2-180M-Amharic-Instruct](https://huggingface.co/rasyosef/Llama-3.2-180M-Amharic-Instruct), a finetuned version of my 180 million parameter [Llama 3.2 180M Amharic](https://huggingface.co/rasyosef/Llama-3.2-180M-Amharic) transformer model.
  """)

  tokens_slider = gr.Slider(8, 256, value=64, label="Maximum new tokens", info="A larger `max_new_tokens` parameter value gives you longer text responses but at the cost of a slower response time.")

  chatbot = gr.ChatInterface(
    chatbot=gr.Chatbot(height=400),
    fn=generate,
    additional_inputs=[tokens_slider],
    stop_btn=None,
    cache_examples=False,
    examples=[
        ["ሰላም፣ እንዴት ነህ?"],
        ["የኢትዮጵያ ዋና ከተማ ስም ምንድን ነው?"],
        ["የኢትዮጵያ የመጨረሻው ንጉስ ማን ነበሩ?"],
        ["የአማርኛ ግጥም ፃፍልኝ"],
        ["ተረት ንገረኝ\n\nጅብና አንበሳ"],
        ["አንድ አስቂኝ ቀልድ ንገረኝ"],
        ["የተሰጠው ጽሑፍ አስተያየት ምን አይነት ነው? 'አዎንታዊ'፣ 'አሉታዊ' ወይም 'ገለልተኛ' የሚል ምላሽ ስጥ። 'አሪፍ ፊልም ነበር'"],
        ["የፈረንሳይ ዋና ከተማ ስም ምንድን ነው?"],
        ["አሁን የአሜሪካ ፕሬዚዳንት ማን ነው?"],
        ["ሶስት የአፍሪካ ሀገራት ጥቀስልኝ"],
        ["3 የአሜሪካ መሪዎችን ስም ጥቀስ"],
        ["5 የአሜሪካ ከተማዎችን ጥቀስ"],
        ["አምስት የአውሮፓ ሀገሮችን ጥቀስልኝ"],
        ["በ ዓለም ላይ ያሉትን 7 አህጉራት ንገረኝ"]
      ]
  )

demo.queue().launch(debug=True,share=True)
# from huggingface_hub import InferenceClient

# """
# For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
# """
# client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")


# def respond(
#     message,
#     history: list[tuple[str, str]],
#     system_message,
#     max_tokens,
#     temperature,
#     top_p,
# ):
#     messages = [{"role": "system", "content": system_message}]

#     for val in history:
#         if val[0]:
#             messages.append({"role": "user", "content": val[0]})
#         if val[1]:
#             messages.append({"role": "assistant", "content": val[1]})

#     messages.append({"role": "user", "content": message})

#     response = ""

#     for message in client.chat_completion(
#         messages,
#         max_tokens=max_tokens,
#         stream=True,
#         temperature=temperature,
#         top_p=top_p,
#     ):
#         token = message.choices[0].delta.content

#         response += token
#         yield response


# """
# For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
# """
# demo = gr.ChatInterface(
#     respond,
#     additional_inputs=[
#         gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
#         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
#         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
#         gr.Slider(
#             minimum=0.1,
#             maximum=1.0,
#             value=0.95,
#             step=0.05,
#             label="Top-p (nucleus sampling)",
#         ),
#     ],
# )


# if __name__ == "__main__":
#     demo.launch()