change gradio demo

demo/app_januspro.py

@@ -11,27 +11,50 @@ import time
 # import spaces  # Import spaces for ZeroGPU compatibility
 
 
-# Load model and processor
+# Global variables to store model and processor (initially for 7B)
-model_path = "deepseek-ai/Janus-Pro-7B"
+vl_gpt = None
-config = AutoConfig.from_pretrained(model_path)
+vl_chat_processor = None
-language_config = config.language_config
+tokenizer = None
-language_config._attn_implementation = 'eager'
-vl_gpt = AutoModelForCausalLM.from_pretrained(model_path,
-                                             language_config=language_config,
-                                             trust_remote_code=True)
-if torch.cuda.is_available():
-    vl_gpt = vl_gpt.to(torch.bfloat16).cuda()
-else:
-    vl_gpt = vl_gpt.to(torch.float16)
-
-vl_chat_processor = VLChatProcessor.from_pretrained(model_path)
-tokenizer = vl_chat_processor.tokenizer
 cuda_device = 'cuda' if torch.cuda.is_available() else 'cpu'
+current_model_path = "deepseek-ai/Janus-Pro-7B" # Default model
+
+def load_model_components(model_path):
+    global vl_gpt, vl_chat_processor, tokenizer, current_model_path # Declare current_model_path as global here
+
+    if vl_gpt is not None and current_model_path == model_path:
+        print(f"Using cached model: {model_path}")
+        return vl_gpt, vl_chat_processor, tokenizer
+
+    print(f"Loading model: {model_path}")
+    config = AutoConfig.from_pretrained(model_path)
+    language_config = config.language_config
+    language_config._attn_implementation = 'eager'
+    vl_gpt_local = AutoModelForCausalLM.from_pretrained(model_path,
+                                                 language_config=language_config,
+                                                 trust_remote_code=True)
+    if torch.cuda.is_available():
+        vl_gpt_local = vl_gpt_local.to(torch.bfloat16).cuda()
+    else:
+        vl_gpt_local = vl_gpt_local.to(torch.float16)
+
+    vl_chat_processor_local = VLChatProcessor.from_pretrained(model_path)
+    tokenizer_local = vl_chat_processor_local.tokenizer
+
+    vl_gpt = vl_gpt_local
+    vl_chat_processor = vl_chat_processor_local
+    tokenizer = tokenizer_local
+    current_model_path = model_path
+    print(f"Model loaded: {model_path}")
+    return vl_gpt, vl_chat_processor, tokenizer
+
 
 @torch.inference_mode()
 # @spaces.GPU(duration=120)
 # Multimodal Understanding function
-def multimodal_understanding(image, question, seed, top_p, temperature):
+def multimodal_understanding(model_name, image, question, seed, top_p, temperature):
+    # Load model based on selection
+    load_model_components(model_name)
+
     # Clear CUDA cache before generating
     torch.cuda.empty_cache()
 
@@ -114,7 +137,6 @@ def generate(input_ids,
             inputs_embeds = img_embeds.unsqueeze(dim=1)
 
 
-
     patches = vl_gpt.gen_vision_model.decode_code(generated_tokens.to(dtype=torch.int),
                                                  shape=[parallel_size, 8, width // patch_size, height // patch_size])
 
@@ -133,10 +155,10 @@ def unpack(dec, width, height, parallel_size=5):
 
 @torch.inference_mode()
 # @spaces.GPU(duration=120)  # Specify a duration to avoid timeout
-def generate_image(prompt,
+def generate_image(model_name, prompt, seed, guidance, t2i_temperature, parallel_size_slider):
-                   seed=None,
+    # Load model based on selection
-                   guidance=5,
+    load_model_components(model_name)
-                   t2i_temperature=1.0):
+
     # Clear CUDA cache and avoid tracking gradients
     torch.cuda.empty_cache()
     # Set the seed for reproducible results
@@ -146,7 +168,7 @@ def generate_image(prompt,
         np.random.seed(seed)
     width = 384
     height = 384
-    parallel_size = 5
+    parallel_size = int(parallel_size_slider) # Use slider value for parallel_size
 
     with torch.no_grad():
         messages = [{'role': '<|User|>', 'content': prompt},
@@ -173,71 +195,75 @@ def generate_image(prompt,
 
 # Gradio interface
 with gr.Blocks() as demo:
-    gr.Markdown(value="# Multimodal Understanding")
+    gr.Markdown(value="# Multimodal Model Demo: Janus-Pro-7B & 1B")
-    with gr.Row():
-        image_input = gr.Image()
-        with gr.Column():
-            question_input = gr.Textbox(label="Question")
-            und_seed_input = gr.Number(label="Seed", precision=0, value=42)
-            top_p = gr.Slider(minimum=0, maximum=1, value=0.95, step=0.05, label="top_p")
-            temperature = gr.Slider(minimum=0, maximum=1, value=0.1, step=0.05, label="temperature")
 
-    understanding_button = gr.Button("Chat")
+    model_selector = gr.Dropdown(
-    understanding_output = gr.Textbox(label="Response")
+        ["deepseek-ai/Janus-Pro-7B", "deepseek-ai/Janus-Pro-1B"],
-
+        value="deepseek-ai/Janus-Pro-7B", label="Select Model"
-    examples_inpainting = gr.Examples(
-        label="Multimodal Understanding examples",
-        examples=[
-            [
-                "explain this meme",
-                "images/doge.png",
-            ],
-            [
-                "Convert the formula into latex code.",
-                "images/equation.png",
-            ],
-        ],
-        inputs=[question_input, image_input],
     )
 
+    with gr.Tab("Multimodal Understanding"):
+        with gr.Row():
+            image_input = gr.Image()
+            with gr.Column():
+                question_input = gr.Textbox(label="Question")
+                und_seed_input = gr.Number(label="Seed", precision=0, value=42)
+                top_p = gr.Slider(minimum=0, maximum=1, value=0.95, step=0.05, label="top_p")
+                temperature = gr.Slider(minimum=0, maximum=1, value=0.1, step=0.05, label="temperature")
 
-    gr.Markdown(value="# Text-to-Image Generation")
+        understanding_button = gr.Button("Chat")
+        understanding_output = gr.Textbox(label="Response")
 
+        examples_inpainting = gr.Examples(
+            label="Multimodal Understanding examples",
+            examples=[
+                [
+                    "explain this meme",
+                    "images/doge.png",
+                ],
+                [
+                    "Convert the formula into latex code.",
+                    "images/equation.png",
+                ],
+            ],
+            inputs=[question_input, image_input],
+        )
 
+    with gr.Tab("Text-to-Image Generation"):
+        with gr.Row():
+            cfg_weight_input = gr.Slider(minimum=1, maximum=10, value=5, step=0.5, label="CFG Weight")
+            t2i_temperature = gr.Slider(minimum=0, maximum=1, value=1.0, step=0.05, label="temperature")
+            parallel_size_slider = gr.Slider(minimum=1, maximum=5, value=5, step=1, label="Parallel Size") # New slider
 
-    with gr.Row():
+        prompt_input = gr.Textbox(label="Prompt. (Prompt in more detail can help produce better images!)")
-        cfg_weight_input = gr.Slider(minimum=1, maximum=10, value=5, step=0.5, label="CFG Weight")
+        seed_input = gr.Number(label="Seed (Optional)", precision=0, value=12345)
-        t2i_temperature = gr.Slider(minimum=0, maximum=1, value=1.0, step=0.05, label="temperature")
 
-    prompt_input = gr.Textbox(label="Prompt. (Prompt in more detail can help produce better images!)")
+        generation_button = gr.Button("Generate Images")
-    seed_input = gr.Number(label="Seed (Optional)", precision=0, value=12345)
 
-    generation_button = gr.Button("Generate Images")
+        image_output = gr.Gallery(label="Generated Images", columns=2, rows=2, height=300)
 
-    image_output = gr.Gallery(label="Generated Images", columns=2, rows=2, height=300)
+        examples_t2i = gr.Examples(
-
+            label="Text to image generation examples.",
-    examples_t2i = gr.Examples(
+            examples=[
-        label="Text to image generation examples.",
+                "Master shifu racoon wearing drip attire as a street gangster.",
-        examples=[
+                "The face of a beautiful girl",
-            "Master shifu racoon wearing drip attire as a street gangster.",
+                "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
-            "The face of a beautiful girl",
+                "A glass of red wine on a reflective surface.",
-            "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
+                "A cute and adorable baby fox with big brown eyes, autumn leaves in the background enchanting,immortal,fluffy, shiny mane,Petals,fairyism,unreal engine 5 and Octane Render,highly detailed, photorealistic, cinematic, natural colors.",
-            "A glass of red wine on a reflective surface.",
+                "The image features an intricately designed eye set against a circular backdrop adorned with ornate swirl patterns that evoke both realism and surrealism. At the center of attention is a strikingly vivid blue iris surrounded by delicate veins radiating outward from the pupil to create depth and intensity. The eyelashes are long and dark, casting subtle shadows on the skin around them which appears smooth yet slightly textured as if aged or weathered over time.\n\nAbove the eye, there's a stone-like structure resembling part of classical architecture, adding layers of mystery and timeless elegance to the composition. This architectural element contrasts sharply but harmoniously with the organic curves surrounding it. Below the eye lies another decorative motif reminiscent of baroque artistry, further enhancing the overall sense of eternity encapsulated within each meticulously crafted detail. \n\nOverall, the atmosphere exudes a mysterious aura intertwined seamlessly with elements suggesting timelessness, achieved through the juxtaposition of realistic textures and surreal artistic flourishes. Each component\u2014from the intricate designs framing the eye to the ancient-looking stone piece above\u2014contributes uniquely towards creating a visually captivating tableau imbued with enigmatic allure.",
-            "A cute and adorable baby fox with big brown eyes, autumn leaves in the background enchanting,immortal,fluffy, shiny mane,Petals,fairyism,unreal engine 5 and Octane Render,highly detailed, photorealistic, cinematic, natural colors.",
+            ],
-            "The image features an intricately designed eye set against a circular backdrop adorned with ornate swirl patterns that evoke both realism and surrealism. At the center of attention is a strikingly vivid blue iris surrounded by delicate veins radiating outward from the pupil to create depth and intensity. The eyelashes are long and dark, casting subtle shadows on the skin around them which appears smooth yet slightly textured as if aged or weathered over time.\n\nAbove the eye, there's a stone-like structure resembling part of classical architecture, adding layers of mystery and timeless elegance to the composition. This architectural element contrasts sharply but harmoniously with the organic curves surrounding it. Below the eye lies another decorative motif reminiscent of baroque artistry, further enhancing the overall sense of eternity encapsulated within each meticulously crafted detail. \n\nOverall, the atmosphere exudes a mysterious aura intertwined seamlessly with elements suggesting timelessness, achieved through the juxtaposition of realistic textures and surreal artistic flourishes. Each component\u2014from the intricate designs framing the eye to the ancient-looking stone piece above\u2014contributes uniquely towards creating a visually captivating tableau imbued with enigmatic allure.",
+            inputs=prompt_input,
-        ],
+        )
-        inputs=prompt_input,
-    )
 
     understanding_button.click(
         multimodal_understanding,
-        inputs=[image_input, question_input, und_seed_input, top_p, temperature],
+        inputs=[model_selector, image_input, question_input, und_seed_input, top_p, temperature], # Added model_selector
         outputs=understanding_output
     )
 
     generation_button.click(
         fn=generate_image,
-        inputs=[prompt_input, seed_input, cfg_weight_input, t2i_temperature],
+        inputs=[model_selector, prompt_input, seed_input, cfg_weight_input, t2i_temperature, parallel_size_slider], # Added model_selector and parallel_size_slider
         outputs=image_output
     )