app.py

import gradio as gr
from huggingface_hub import InferenceClient

MODELS = {
    "tiny-LLM": "omurberaisik/tiny-LLM",
    "HoloCom 13B HQ": "omurberaisik/holocomn13b_high_quality",
    "HoloCom Lite v8": "omurberaisik/holocom-lite-v8_test_ai",
    "HoloCom NB7 (merged)": "omurberaisik/holocomnb7-merged",
    "HoloCom NB7 (base)": "omurberaisik/holocomnb7",
    "DNA 7B Genomic": "omurberaisik/dna-7b-genomic",
    "BioTransformer 7B v2": "omurberaisik/biotransformer-7b-v2",
    "BioTransformer 1B": "omurberaisik/biotransformer-1b-v1",
}

def parse_history(history):
    messages = []
    for item in history:
        if isinstance(item, dict):
            messages.append({"role": item["role"], "content": item["content"]})
        else:
            user_msg, bot_msg = item
            messages.append({"role": "user", "content": user_msg})
            if bot_msg:
                messages.append({"role": "assistant", "content": bot_msg})
    return messages

def respond(message, history, model_name, max_tokens, temperature, top_p):
    model_id = MODELS[model_name]
    client = InferenceClient(model_id)

    messages = parse_history(history)
    messages.append({"role": "user", "content": message})

    response = ""
    try:
        for chunk in client.chat_completion(
            messages=messages,
            max_tokens=max_tokens,
            temperature=temperature,
            top_p=top_p,
            stream=True,
        ):
            token = chunk.choices[0].delta.content
            if token:
                response += token
                yield response
    except Exception:
        prompt = ""
        for item in messages[:-1]:
            role = "User" if item["role"] == "user" else "Assistant"
            prompt += f"{role}: {item['content']}\n"
        prompt += f"User: {message}\nAssistant:"
        try:
            for chunk in client.text_generation(
                prompt,
                max_new_tokens=max_tokens,
                temperature=temperature,
                top_p=top_p,
                stream=True,
            ):
                response += chunk
                yield response
        except Exception as e:
            yield f"❌ Hata: {str(e)}\n\nModel yüklenmemiş olabilir, biraz bekleyip tekrar deneyin."


with gr.Blocks(title="Ömür Bera Işık — Model Hub") as demo:

    gr.Markdown("# ∞ Apeiron Model Hub\n**omurberaisik** tarafından geliştirilen tüm modeller.")

    with gr.Row():
        with gr.Column(scale=1):
            model_selector = gr.Dropdown(
                choices=list(MODELS.keys()),
                value="tiny-LLM",
                label="Model Seç"
            )
            max_tokens = gr.Slider(64, 1024, value=256, step=64, label="Max Tokens")
            temperature = gr.Slider(0.1, 2.0, value=0.7, step=0.1, label="Temperature")
            top_p = gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Top-p")

        with gr.Column(scale=3):
            gr.ChatInterface(
                fn=respond,
                additional_inputs=[model_selector, max_tokens, temperature, top_p],
            )

    gr.Markdown("🤗 [HuggingFace Profile](https://huggingface.co/omurberaisik) · Apeiron LM")

demo.launch()