Create app.py

app.py

@@ -0,0 +1,210 @@
+"""
+UPDATED JANUARY 21 - 2025
+To fix the HF API TOKEN STUFF___DONE
+"""
+
+import datetime, random, string
+import gradio as gr
+#from openai import OpenAI
+from gradio_client import Client
+from PIL import Image
+from rich.console import Console
+import os
+from huggingface_hub import InferenceClient
+
+
+console = Console(width=80)
+theme=gr.themes.Default(primary_hue="blue", secondary_hue="pink",
+                        font=[gr.themes.GoogleFont("Oxanium"), "Arial", "sans-serif"]) 
+
+def checkHFT(hf_token):
+    if 'hf_' in hf_token:
+        return gr.Row(visible=True),gr.Row(visible=True),gr.Row(visible=True),gr.Row(visible=True),"✅HF TOKEN detected"
+   
+    else:
+        gr.Warning("⚠️ You don't have a Hugging Face Token set")
+        return gr.Row(visible=False),gr.Row(visible=False),gr.Row(visible=False),gr.Row(visible=False), "⚠️ You don't have a Hugging Face Token set"  
+    
+
+def writehistory(filename,text):
+    """
+    save a string into a logfile with python file operations
+    filename -> str pathfile/filename
+    text -> str, the text to be written in the file
+    """
+    with open(f'{filename}', 'a', encoding='utf-8') as f:
+        f.write(text)
+        f.write('\n')
+    f.close()
+
+def genRANstring(n):
+    """
+    n = int number of char to randomize
+    Return -> str, the filename with n random alphanumeric charachters
+    """
+    N = n
+    res = ''.join(random.choices(string.ascii_uppercase +
+                                string.digits, k=N))
+    return f'Logfile_{res}.txt'
+
+LOGFILENAME = genRANstring(5)
+
+################## STABLE DIFFUSION PROMPT ##############################
+def createSDPrompt(token,headers):
+    #bruteText = bruteText.replace('\n\n','\n')
+    SD_prompt = f'''Create a prompt for Stable Diffusion based on the information below. Return only the prompt.\n---\n{headers}\n\nPROMPT:'''
+    client = InferenceClient(token=token)
+    messages = [{"role": "user", "content": SD_prompt}]
+    completion = client.chat.completions.create(
+        model="Qwen/Qwen2.5-72B-Instruct",
+        messages=messages,
+        max_tokens=500
+    )
+    print(completion.choices[0].message.content)
+    ImageGEN_prompt = completion.choices[0].message.content
+    return ImageGEN_prompt
+
+############### CREATE IMAGE ##########################
+def CreateImage(token,ImageGEN_prompt):
+    from gradio_client import Client
+    from gradio_client import handle_file
+    from PIL import Image
+    client = Client("stabilityai/stable-diffusion-3.5-large",hf_token=token)
+    result = client.predict(
+            prompt=ImageGEN_prompt,
+        negative_prompt='blur',
+            seed=0,
+            randomize_seed=True,
+            width=1360,
+            height=768,
+            guidance_scale=4.5,
+            num_inference_steps=30,
+            api_name="/infer"
+    )
+    ############ SAVE IMAGE ##########################
+    from gradio_client import handle_file
+    temp = result[0]
+    from PIL import Image
+    image = Image.open(temp)
+    imagename = datetime.datetime.strftime(datetime.datetime.now(),'IMage_%Y-%m-%d_%H-%M-%S.png')
+    image.save(imagename)
+    print(f'Image saved as {imagename}...')
+    return image, imagename
+
+def openDIR():
+    import os
+    current_directory = os.getcwd()
+    print("Current Directory:", current_directory)
+    os.system(f'start explorer "{current_directory}"')
+
+############# TWEET GENERATION #########################
+def createTweets(token,bruteText):   
+    Tweet_prompt = f"Read the following newsletter. rewrite it into 3 twitter posts in English, in progression.\n---\n{bruteText}"
+    from rich.console import Console
+    console = Console(width=80)
+    # using https://huggingface.co/spaces/eswardivi/phi-4
+    client = Client("eswardivi/phi-4",hf_token=token)
+    result = client.predict(
+            message=Tweet_prompt,
+            param_2=0.7,
+            param_3=True,
+            param_4=512,
+            api_name="/chat"
+    )
+    print(result)
+    from rich.console import Console
+    console = Console(width=80)
+    tweet1 = result.split('1:**')[1].split('\n\n')[0]
+    tweet2 = result.split('2:**')[1].split('\n\n')[0]
+    tweet3 = result.split('3:**')[1]
+    console.print(tweet1)
+    console.rule()
+    console.print(tweet2)
+    console.rule()
+    console.print(tweet3)
+    console.rule()
+    return tweet1,tweet2, tweet3
+
+#OR
+def createTweets2(token,bruteText):
+    # Using https://huggingface.co/spaces/Qwen/Qwen2.5-72B-Instruct
+    Tweet_prompt = f"Read the following newsletter. rewrite it into 3 twitter posts in English, in progression.\n---\n{bruteText}"
+    client = Client("Qwen/Qwen2.5-72B-Instruct",hf_token=token)
+    result = client.predict(
+            query=Tweet_prompt,
+            history=[],
+            system="You are Qwen, created by Alibaba Cloud. You are a helpful assistant.",
+            api_name="/model_chat"
+    )
+    twitposts = result[1][0][1]
+    console.print(twitposts)
+
+    tweet1 = twitposts.split('Post 1:')[1].split('\n\n')[0]
+    tweet2 = twitposts.split('Post 2:')[1].split('\n\n')[0]
+    tweet3 = twitposts.split('Post 3:')[1]
+    console.print(tweet1)
+    console.rule()
+    console.print(tweet2)
+    console.rule()
+    console.print(tweet3)
+    console.rule()
+    return twitposts
+
+
+with gr.Blocks(fill_width=True,theme=theme) as demo:
+    # INTERFACE
+    with gr.Row(variant='panel'):
+        with gr.Column(scale=2):
+            gr.Image('gradioLOGO.png',width=260)
+        with gr.Column(scale=4):
+            gr.HTML(
+        f"""<h1 style="text-align:center">Advanced POST creation with GRADIO and HF API</h1>""")
+            alertTEXT = gr.Text("⚠️✅You don't have a Hugging Face Token set",container=False,show_label=False,)         
+        with gr.Column(scale=2):
+            TOKEN = gr.Textbox(lines=1,label='Your HF token',scale=1)
+            btn_token = gr.Button("Validate HF token", variant='secondary',size='lg',scale=1)
+            
+             
+    with gr.Row(visible=False) as row1:
+        #HYPERPARAMETERS
+        with gr.Column(scale=1):
+            CREATE_SDP = gr.Button(variant='huggingface',value='Generate Prompt')
+            GEN_IMAGE = gr.Button(value='Generate Image',variant='primary')
+            gr.Markdown('---')
+            OPEN_FOLDER = gr.Button(variant='secondary',value='Open Image Folder')
+            clear = gr.ClearButton()
+        #CHATBOT AREA    
+        with gr.Column(scale=3):    
+            headers = gr.Textbox(lines=8,label='Header of the Article')
+         
+    with gr.Row(visible=False) as row2:
+        with gr.Column(scale=2):
+            SDPrompt = gr.Textbox(lines=8,label='Generated prompt Stable Diffusion')
+            ImageFilename = gr.Textbox(lines=2,label='Generated Image Filename',show_copy_button=True)
+        with gr.Column(scale=3):
+            SDImage = gr.Image(type='pil',label='Generated Image',show_download_button=True, show_fullscreen_button=True,)
+
+    with gr.Row(visible=False) as row3:
+        gr.Markdown('---')
+
+    with gr.Row(visible=False) as row4:
+        #TWITTERPOSTS CREATION SECTION
+        with gr.Column(scale=2):
+            body = gr.Textbox(lines=12,label='Body of the Article')
+            CREATE_TWEET = gr.Button(variant='huggingface',value='Generate Tweets')
+        #TWEET RESULTS AREA    
+        with gr.Column(scale=1):    
+            tweets1 = gr.Textbox(lines=5,label='🐦 TWEET #1 - 1️⃣',show_copy_button=True)
+            tweets2 = gr.Textbox(lines=5,label='🐦 TWEET #2 - 2️⃣',show_copy_button=True)
+            tweets3 = gr.Textbox(lines=5,label='🐦 TWEET #3 - 3️⃣',show_copy_button=True)                  
+
+    CREATE_SDP.click(createSDPrompt, [TOKEN,headers], [SDPrompt])
+    GEN_IMAGE.click(CreateImage, [TOKEN,SDPrompt], [SDImage,ImageFilename])    #CreateImage
+    OPEN_FOLDER.click(openDIR, [], [])    #Open Current directory
+    CREATE_TWEET.click(createTweets,[TOKEN,body],[tweets1,tweets2,tweets3])
+    btn_token.click(checkHFT,[TOKEN],[row1,row2,row3,row4,alertTEXT])
+
+
+
+if __name__ == "__main__":
+    demo.launch()