Update app.py

app.py

@@ -1,38 +1,81 @@
+from huggingface_hub import InferenceClient
 import gradio as gr
 
+client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.3")
 
-# Assign GPU to the overall spaces
-def main():
+def format_prompt(message, history, system_prompt=None):
+    prompt = "<s>"
+    for user_prompt, bot_response in history:
+        prompt += f"[INST] {user_prompt} [/INST]"
+        prompt += f" {bot_response}</s> "
+    if system_prompt:
+        prompt += f"[SYS] {system_prompt} [/SYS]"
+    prompt += f"[INST] {message} [/INST]"
+    return prompt
 
-    # Define the components using the specified URLs
-    with gr.Blocks() as chatbot:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/Mistral-7B-Instruct-v0.3' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+def generate(
+    prompt, history, system_prompt=None, temperature=0.2, max_new_tokens=1024, top_p=0.95, repetition_penalty=1.0,
+):
+    temperature = float(temperature)
+    if temperature < 1e-2:
+        temperature = 1e-2
+    top_p = float(top_p)
 
-    with gr.Blocks() as imagegpt:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/ImageGPT-4XL' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+    generate_kwargs = dict(
+        temperature=temperature,
+        max_new_tokens=max_new_tokens,
+        top_p=top_p,
+        repetition_penalty=repetition_penalty,
+        do_sample=True,
+        seed=42,
+    )
 
-    with gr.Blocks() as imageengine4k:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/DALLE-4K' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+    formatted_prompt = format_prompt(prompt, history, system_prompt)
 
-    with gr.Blocks() as imageengine2k:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/DALLE-2K' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+    stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    output = ""
 
-    with gr.Blocks() as mediadownloader:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/All-In-One-Downloader' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+    for response in stream:
+        output += response.token.text
+        yield output
+    return output
 
-    with gr.Blocks() as photofakes:
-        gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/Face-Swap-Roop' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+demo = gr.ChatInterface(
+    fn=generate,
+    title="",
+    theme="xiaobaiyuan/theme_brief"
+)
 
-    # Assemble the components into a single Gradio interface
-    with gr.Blocks(theme="xiaobaiyuan/theme_brief", title="GPT 4o") as demo:
-        gr.Markdown("# GPT 4o")
-        gr.TabbedInterface(
+
+# Define the components using the specified URLs
+with gr.Blocks() as chatbot:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/Mistral-7B-Instruct-v0.3' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+with gr.Blocks() as imagegpt:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/ImageGPT-4XL' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+with gr.Blocks() as imageengine4k:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/DALLE-4K' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+with gr.Blocks() as imageengine2k:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/DALLE-2K' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+with gr.Blocks() as mediadownloader:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/All-In-One-Downloader' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+with gr.Blocks() as photofakes:
+    gr.HTML("<iframe src='https://huggingface.co/spaces/prithivMLmods/Face-Swap-Roop' width='100%' height='1000px' style='border-radius: 8px;'></iframe>")
+
+# Assemble the components into a single Gradio interface
+with gr.Blocks(theme="xiaobaiyuan/theme_brief", title="GPT 4o") as demo:
+    gr.Markdown("# GPT 4o")
+    gr.TabbedInterface(
             [chatbot, imagegpt, imageengine4k, imageengine2k, mediadownloader, photofakes], 
             ['ChatBot', 'ImageGPT', 'Image Engine 4K', 'Image Engine 2K', 'Media Downloader', 'PhotoFakes']
         )
 
-    # Launch the demo
-    demo.queue().launch(debug=True)
+# Launch the demo
+
+
 
-# Run the main function
-main()
+demo.queue().launch(show_api=False)