app.py

import gradio as gr
from huggingface_hub import InferenceClient

client = InferenceClient("lambdaindie/lambdai")

css = """
.thinking-html {
    display: flex;
    align-items: center;
    gap: 8px;
    color: #666;
    font-style: italic;
    margin-bottom: 5px;
    animation: pulse 1.5s infinite;
}
.loader {
    width: 14px;
    height: 14px;
    border: 2px solid #ccc;
    border-top: 2px solid #666;
    border-radius: 50%;
    animation: spin 1s linear infinite;
}
@keyframes spin {
    to { transform: rotate(360deg); }
}
@keyframes pulse {
    0% { opacity: 1; transform: scale(1); }
    50% { opacity: 0.6; transform: scale(1.05); }
    100% { opacity: 1; transform: scale(1); }
}
"""

def respond(message, history):
    thinking = (
        "<div class='thinking-html'>"
        "<div class='loader'></div>"
        "Thinking..."
        "</div>"
    )
    yield history + [[message, thinking]]

    prompt = f"Think step by step and explain your reasoning before answering:\n\n{message}"
    response = client.chat_completion([{"role": "user", "content": prompt}], stream=False)
    output = response['choices'][0]['message']['content']

    if "\n\n" in output:
        reasoning, answer = output.split("\n\n", 1)
    else:
        reasoning, answer = "No reasoning provided.", output

    reasoning_md = f"> {reasoning.strip()}"
    final = f"{reasoning_md}\n\n{answer.strip()}"
    yield history + [[message, final]]

with gr.Blocks(css=css) as demo:
    gr.Markdown("## Lambdai-v1-1B")
    chatbot = gr.Chatbot()
    msg = gr.Textbox(label="Message")
    send = gr.Button("Send")

    send.click(respond, [msg, chatbot], chatbot)
    msg.submit(respond, [msg, chatbot], chatbot)

demo.launch()