app.py

from huggingface_hub import InferenceClient
import gradio as gr

client = InferenceClient(
    "mistralai/Mixtral-8x7B-Instruct-v0.1"
)

def format_prompt(message, history):
    prompt = "<s>"
    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 generate_from_srt(file_path, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0):
    # Assuming the file content is directly usable as a prompt
    with open(file_path, "r", encoding="utf-8") as file:
        file_content = file.load()
    
    # Process the SRT file content as needed before using it as a prompt
    # For example, extracting text and removing timestamps if necessary
    # Here, directly using the file content for simplicity
    
    temperature = float(temperature)
    if temperature < 1e-2:
        temperature = 1e-2
    top_p = float(top_p)

    generate_kwargs = dict(
        temperature=temperature,
        max_new_tokens=max_new_tokens,
        top_p=top_p,
        repetition_penalty=repetition_penalty,
        do_sample=True,
        seed=42,
    )

    formatted_prompt = format_prompt(file_content, [])
    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
    return output

def handle_file(file, **kwargs):
    # Save the uploaded file temporarily to read it
    file_path = file.name
    with open(file_path, "wb") as f:
        f.write(file.read())
    
    return generate_from_srt(file_path, **kwargs)

iface = gr.Interface(
    fn=handle_file,
    inputs=gr.File(label="Upload SRT File"),
    outputs="text",
    title="SRT File Translation",
    concurrency_limit=20,
)

iface.launch()