File size: 3,622 Bytes
00dac32 2752b73 00dac32 2752b73 00dac32 a2a5285 00dac32 8566ed8 00dac32 4106448 00dac32 |
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 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 |
import os
import gradio as gr
from huggingface_hub import InferenceClient
class HaikuGenerator:
def __init__(self):
HUGGINGFACE_API_TOKEN = os.getenv("HUGGINGFACE_API_TOKEN")
self.text_client = InferenceClient(token=HUGGINGFACE_API_TOKEN, model="HuggingFaceH4/zephyr-7b-beta")
self.image_client = InferenceClient(token=HUGGINGFACE_API_TOKEN, model="stabilityai/stable-diffusion-xl-base-1.0")
def generate_haiku(self, prompt):
system_message = "You are a Haiku generator."
response = ""
messages = [{"role": "system", "content": system_message},
{"role": "user", "content": prompt}]
# Get the chat completion response
message = self.text_client.chat_completion(
messages,
max_tokens=30,
stream=False,
temperature=0.7,
top_p=0.95,
)
# Extract the generated content
response = message['choices'][0]['message']['content']
return response.strip()
def text_to_image(self, prompt, style):
# Modify the prompt based on the selected style
if style == "Japanese":
prompt += ", in Japanese art style"
elif style == "oil painting":
prompt += ", in the style of an oil painting"
image = self.image_client.text_to_image(prompt)
return image
def gradio_interface(self):
# Custom CSS to apply a Japanese-style font
custom_css = """
body {
font-family: 'Sawarabi Mincho', serif !important;
}
h1, h6 {
font-family: 'Sawarabi Mincho', serif !important;
}
.button {
font-family: 'Sawarabi Mincho', serif !important;
}
"""
with gr.Blocks(theme='earneleh/paris', css=custom_css) as demo:
gr.HTML("""
<center><h1 style="color:#C73E3A">HaiKool - Haiku Poem and Image Generator</h1></center>""")
gr.HTML("""
<center><h6 style="color:#C73E3A">Generate a Haiku poem and an image based on it - Please note that loading time can be up to 1 minute.</h6></center>""")
with gr.Column(elem_id="col-container"):
haiku_output = gr.Textbox(label="Generated Haiku", interactive=False)
image_output = gr.Image()
with gr.Row(elem_id="col-container"):
with gr.Column():
prompt = gr.Textbox(show_label=False, placeholder="Enter a prompt for the Haiku")
with gr.Column():
style = gr.Dropdown(label="Select Image Style", choices=["default", "Japanese", "oil painting"], value="default")
with gr.Column():
generate_button = gr.Button("Generate Haiku and Image", elem_classes="button")
# Define the function that integrates both steps: Haiku generation and image creation
def generate_haiku_and_image(prompt, style):
haiku = self.generate_haiku(prompt)
image = self.text_to_image(haiku, style)
return haiku, image
generate_button.click(generate_haiku_and_image, inputs=[prompt, style], outputs=[haiku_output, image_output])
prompt.submit(generate_haiku_and_image, inputs=[prompt, style], outputs=[haiku_output, image_output])
demo.launch(debug=True)
if __name__ == "__main__":
haiku_generator = HaikuGenerator()
haiku_generator.gradio_interface()
|