File size: 2,131 Bytes
7b64e9f
e97aebb
0267d3c
 
7b64e9f
 
0267d3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7b64e9f
e97aebb
87e851b
0267d3c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
import os
import gradio as gr
from transformers import MarianMTModel, MarianTokenizer
from diffusers import StableDiffusionPipeline
import torch

# Set your Hugging Face API token here if needed (for private models)
HF_API_TOKEN = os.getenv("HF_API_TOKEN", None)

# Translation model name
translation_model_name = "Helsinki-NLP/opus-mt-tc-big-en-ta"

# Load translation tokenizer and model (make sure sentencepiece is installed)
translation_tokenizer = MarianTokenizer.from_pretrained(translation_model_name)
translation_model = MarianMTModel.from_pretrained(translation_model_name)

# Load stable diffusion pipeline for image generation
pipe = StableDiffusionPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    torch_dtype=torch.float16,
    revision="fp16",
    use_auth_token=HF_API_TOKEN,
)
pipe = pipe.to("cuda") if torch.cuda.is_available() else pipe.to("cpu")

def translate_tamil_to_english(tamil_text):
    # Tokenize and translate
    inputs = translation_tokenizer(tamil_text, return_tensors="pt", padding=True)
    outputs = translation_model.generate(**inputs)
    english_text = translation_tokenizer.decode(outputs[0], skip_special_tokens=True)
    return english_text

def generate_image_from_text(text):
    # Generate image from English text prompt
    image = pipe(text).images[0]
    return image

def translate_and_generate_image(tamil_text):
    english_text = translate_tamil_to_english(tamil_text)
    image = generate_image_from_text(english_text)
    return english_text, image

with gr.Blocks() as app:
    gr.Markdown("# Tamil to English Translation + Image Generation")
    
    tamil_input = gr.Textbox(label="Enter Tamil Text", lines=3)
    english_output = gr.Textbox(label="Translated English Text")
    generated_image = gr.Image(label="Generated Image")
    
    translate_btn = gr.Button("Translate and Generate Image")
    
    translate_btn.click(
        fn=translate_and_generate_image,
        inputs=[tamil_input],
        outputs=[english_output, generated_image]
    )

if __name__ == "__main__":
    app.launch(share=True)  # share=True creates a public link (optional)