app.py

import gradio as gr
from transformers import AutoTokenizer
from transformers.utils import logging

# Enable logging to see debug messages
logging.set_verbosity_info()

# Import custom configuration and model classes
from transformers_modules.deepseek_ai.DeepSeek_R1.configuration_deepseek import DeepseekV3Config
from transformers_modules.deepseek_ai.DeepSeek_R1.modeling_deepseek import DeepseekV3Model

# Load model and tokenizer
model_name = "deepseek-ai/DeepSeek-R1"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
config = DeepseekV3Config.from_pretrained(model_name, trust_remote_code=True)
model = DeepseekV3Model.from_pretrained(model_name, config=config, trust_remote_code=True)

def classify_text(input_text):
    # Tokenize input
    inputs = tokenizer(input_text, return_tensors="pt")
    # Get model output
    outputs = model(**inputs)
    probabilities = outputs.logits.softmax(dim=-1).detach().numpy()
    return {f"Class {i}": prob for i, prob in enumerate(probabilities[0])}

# Create Gradio interface
interface = gr.Interface(
    fn=classify_text,
    inputs=gr.Textbox(label="Enter Text"),
    outputs=gr.Label(label="Class Probabilities"),
    title="DeepSeek-R1 Text Classification",
    description="A text classification app powered by DeepSeek-R1."
)

# Launch the app
interface.launch()