Update app.py

app.py

@@ -1,11 +1,71 @@
 import gradio as gr
-from diffusers import DiffusionPipeline
+import jax
+import numpy as np
+import jax.numpy as jnp
+from flax.jax_utils import replicate
+from flax.training.common_utils import shard
+from diffusers.utils import load_image
+from PIL import Image
+from diffusers import FlaxStableDiffusionControlNetPipeline, FlaxControlNetModel
 
-pipeline = DiffusionPipeline.from_pretrained("jax-diffusers-event/canny-coyo1m")
 
-def infer(prompt):
-    image = pipeline(prompt).images[0]
-    return image
+def image_grid(imgs, rows, cols):
+    w, h = imgs[0].size
+    grid = Image.new("RGB", size=(cols * w, rows * h))
+    for i, img in enumerate(imgs):
+        grid.paste(img, box=(i % cols * w, i // cols * h))
+    return grid
 
 
-gr.Interface(pipeline, inputs="text", outputs="image").launch()
+def create_key(seed=0):
+    return jax.random.PRNGKey(seed)
+
+
+rng = create_key(0)
+
+
+def canny_filter(image):
+
+    ## TODO: Implement canny filter here
+    return canny_image
+
+def infer(prompts, negative_prompts, image):
+    
+    
+    # load control net and stable diffusion v1-5
+    controlnet, controlnet_params = FlaxControlNetModel.from_pretrained(
+        "jax-diffusers-event/canny-coyo1m", from_pt=True, dtype=jnp.float32
+    )
+    pipe, params = FlaxStableDiffusionControlNetPipeline.from_pretrained(
+        "runwayml/stable-diffusion-v1-5", controlnet=controlnet, from_pt=True, dtype=jnp.float32
+    )
+    params["controlnet"] = controlnet_params
+    
+    num_samples = jax.device_count()
+    rng = jax.random.split(rng, jax.device_count())
+    canny_image = canny_filter(image)
+    
+    prompt_ids = pipe.prepare_text_inputs([prompts] * num_samples)
+    negative_prompt_ids = pipe.prepare_text_inputs([negative_prompts] * num_samples)
+    processed_image = pipe.prepare_image_inputs([canny_image] * num_samples)
+    
+    p_params = replicate(params)
+    prompt_ids = shard(prompt_ids)
+    negative_prompt_ids = shard(negative_prompt_ids)
+    processed_image = shard(processed_image)
+    
+    output = pipe(
+        prompt_ids=prompt_ids,
+        image=processed_image,
+        params=p_params,
+        prng_seed=rng,
+        num_inference_steps=50,
+        neg_prompt_ids=negative_prompt_ids,
+        jit=True,
+    ).images
+    
+    output_images = pipe.numpy_to_pil(np.asarray(output.reshape((num_samples,) + output.shape[-3:])))
+    output_images = image_grid(output_images, num_samples // 4, 4)
+    return output_images
+
+gr.Interface(pipeline, inputs=["text", "text", "image"], outputs="gallery").launch()