initial commit

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,12 @@
+---
+title: Lama
+emoji: 🚀
+colorFrom: yellow
+colorTo: pink
+sdk: gradio
+sdk_version: 5.6.0
+app_file: app.py
+pinned: false
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,248 @@
+import gradio as gr
+import torch
+import numpy as np
+from PIL import Image
+
+from saicinpainting.evaluation.utils import move_to_device
+from saicinpainting.evaluation.refinement import refine_predict
+from saicinpainting.evaluation.data import pad_img_to_modulo
+from saicinpainting.training.trainers import load_checkpoint
+
+import numpy as np
+import torch
+import yaml
+from omegaconf import OmegaConf
+from torch.utils.data._utils.collate import default_collate
+import os
+#from gradio_imageslider import ImageSlider
+import requests
+import zipfile
+import os
+
+# URL of the file to download
+url = "https://huggingface.co/smartywu/big-lama/resolve/main/big-lama.zip"
+
+# Local filename to save the downloaded file
+local_filename = "big-lama.zip"
+
+# Directory to extract the files into
+extract_dir = "big-lama"
+
+# Check if the extracted directory already exists
+if os.path.exists(extract_dir):
+    print(f"The directory '{extract_dir}' already exists. Skipping download and extraction.")
+else:
+    # Check if the zip file already exists
+    if not os.path.exists(local_filename):
+        # Download the file
+        with requests.get(url, stream=True) as response:
+            response.raise_for_status()
+            with open(local_filename, 'wb') as f:
+                for chunk in response.iter_content(chunk_size=8192):
+                    f.write(chunk)
+        print(f"Downloaded '{local_filename}' successfully.")
+    else:
+        print(f"The file '{local_filename}' already exists. Skipping download.")
+
+    # Unzip the file
+    with zipfile.ZipFile(local_filename, 'r') as zip_ref:
+        zip_ref.extractall()
+    print(f"Extracted '{local_filename}' into '{extract_dir}' successfully.")
+
+    # Optionally, remove the zip file after extraction
+    os.remove(local_filename)
+    print(f"Removed '{local_filename}' after extraction.")
+
+generator = torch.Generator(device="cuda").manual_seed(42)
+
+size = (1024, 1024)
+
+
+def image_preprocess(image: Image, mode="RGB", return_orig=False):
+    img = np.array(image.convert(mode))
+    if img.ndim == 3:
+        img = np.transpose(img, (2, 0, 1))
+    out_img = img.astype("float32") / 255
+    if return_orig:
+        return out_img, img
+    else:
+        return out_img
+
+
+def infer(prompt, image):
+    source = image["background"].convert("RGB").resize(size)
+
+    mask = image["layers"][0]
+
+    mask = mask.point(lambda p: p > 0 and 255).split()[3]
+    mask.convert("RGB")
+
+    # binary_mask = mask.point(lambda p: 255 if p > 0 else 0)
+    # inverted_mask = ImageChops.invert(binary_mask)
+
+    # alpha_image = Image.new("RGB", source.size, (0, 0, 0))
+    # cnet_image = Image.composite(source, alpha_image, inverted_mask)
+
+    device = torch.device("cpu")
+
+    predict_config_path = "/home/naumov/lama_predict/configs/prediction/default.yaml"
+
+    with open(predict_config_path, "r") as f:
+        predict_config = OmegaConf.create(yaml.safe_load(f))
+
+    train_config_path = os.path.join(predict_config.model.path, "config.yaml")
+    with open(train_config_path, "r") as f:
+        train_config = OmegaConf.create(yaml.safe_load(f))
+
+    train_config.training_model.predict_only = True
+    train_config.visualizer.kind = "noop"
+
+    checkpoint_path = os.path.join(
+        predict_config.model.path, "models", predict_config.model.checkpoint
+    )
+
+    model = load_checkpoint(
+        train_config, checkpoint_path, strict=False, map_location="cpu"
+    )
+    model.freeze()
+    if not predict_config.get("refine", False):
+        model.to(device)
+
+    img = image_preprocess(source, mode="RGB")
+    mask = image_preprocess(mask, mode="L")
+
+    result = dict(image=img, mask=mask[None, ...])
+
+    if (
+        predict_config.dataset.pad_out_to_modulo is not None
+        and predict_config.dataset.pad_out_to_modulo > 1
+    ):
+        result["unpad_to_size"] = result["image"].shape[1:]
+        result["image"] = pad_img_to_modulo(
+            result["image"], predict_config.dataset.pad_out_to_modulo
+        )
+        result["mask"] = pad_img_to_modulo(
+            result["mask"], predict_config.dataset.pad_out_to_modulo
+        )
+
+    batch = default_collate([result])
+    if predict_config.get("refine", False):
+        assert "unpad_to_size" in batch, "Unpadded size is required for the refinement"
+        # image unpadding is taken care of in the refiner, so that output image
+        # is same size as the input image
+        cur_res = refine_predict(batch, model, **predict_config.refiner)
+        cur_res = cur_res[0].permute(1, 2, 0).detach().cpu().numpy()
+    else:
+        with torch.no_grad():
+            batch = move_to_device(batch, device)
+            batch["mask"] = (batch["mask"] > 0) * 1
+            batch = model(batch)
+            cur_res = (
+                batch[predict_config.out_key][0].permute(1, 2, 0).detach().cpu().numpy()
+            )
+            unpad_to_size = batch.get("unpad_to_size", None)
+            if unpad_to_size is not None:
+                orig_height, orig_width = unpad_to_size
+                cur_res = cur_res[:orig_height, :orig_width]
+
+    cur_res = np.clip(cur_res * 255, 0, 255).astype("uint8")
+
+    yield cur_res
+
+
+def clear_result():
+    return gr.update(value=None)
+
+
+css = """.main-div div{display:inline-flex;align-items:center;gap:.8rem;font-size:1.75rem}.main-div div h1{font-weight:900;margin-bottom:7px}.main-div p{margin-bottom:10px;font-size:94%}a{text-decoration:underline}.tabs{margin-top:0;margin-bottom:0}#gallery{min-height:20rem}
+"""
+prefix = ""
+
+title = f"""
+			<div class="main-div">
+			  <div>
+				<h1>Small Stable Diffusion V0</h1>
+			  </div>
+			  <p>
+			   Demo for <a href="https://huggingface.co/OFA-Sys/small-stable-diffusion-v0">Small Stable Diffusion V0</a> Stable Diffusion model.<br>
+			   {"Add the following tokens to your prompts for the model to work properly: <b>prefix</b>" if prefix else ""}
+			  </p>
+			  Running on {"<b>GPU 🔥</b>" if torch.cuda.is_available() else f"<b>CPU 🥶</b>. For faster inference it is recommended to <b>upgrade to GPU in <a href='https://huggingface.co/spaces/akhaliq/small-stable-diffusion-v0/settings'>Settings</a></b>"} after duplicating the space<br><br>
+			  <a style="display:inline-block" href="https://huggingface.co/spaces/akhaliq/small-stable-diffusion-v0?duplicate=true"><img src="https://bit.ly/3gLdBN6" alt="Duplicate Space"></a>
+			</div>
+		"""
+
+with gr.Blocks(css=css) as demo:
+    gr.HTML(title)
+    with gr.Row():
+        with gr.Row():
+            with gr.Column():
+                prompt = gr.Textbox(
+                    label="Prompt",
+                    info="Describe what to inpaint the mask with",
+                    lines=3,
+                )
+            with gr.Column():
+                with gr.Row():
+                    with gr.Column():
+                        run_button = gr.Button("Generate")
+    with gr.Row():
+        input_image = gr.ImageMask(
+            type="pil",
+            label="Input Image",
+            crop_size=(1024, 1024),
+            layers=False,
+            height=712,
+            width=712
+        )
+
+        result = gr.Image(
+            interactive=False,
+            label="Generated Image",
+        )
+    use_as_input_button = gr.Button("Use as Input Image", visible=False)
+
+    def use_output_as_input(output_image):
+        return gr.update(value=output_image)
+
+    use_as_input_button.click(
+        fn=use_output_as_input, inputs=[result], outputs=[input_image]
+    )
+
+    run_button.click(
+        fn=clear_result,
+        inputs=None,
+        outputs=result,
+    ).then(
+        fn=lambda: gr.update(visible=False),
+        inputs=None,
+        outputs=use_as_input_button,
+    ).then(
+        fn=infer,
+        inputs=[prompt, input_image],
+        outputs=result,
+    ).then(
+        fn=lambda: gr.update(visible=True),
+        inputs=None,
+        outputs=use_as_input_button,
+    )
+
+    prompt.submit(
+        fn=clear_result,
+        inputs=None,
+        outputs=result,
+    ).then(
+        fn=lambda: gr.update(visible=False),
+        inputs=None,
+        outputs=use_as_input_button,
+    ).then(
+        fn=infer,
+        inputs=[prompt, input_image],
+        outputs=result,
+    ).then(
+        fn=lambda: gr.update(visible=True),
+        inputs=None,
+        outputs=use_as_input_button,
+    )
+
+demo.launch()

configs/prediction/default.yaml

@@ -0,0 +1,24 @@
+indir: no  # to be overriden in CLI
+outdir: no  # to be overriden in CLI
+
+model:
+  path: big-lama  # to be overriden in CLI
+  checkpoint: best.ckpt
+
+dataset:
+  kind: default
+  img_suffix: .png
+  pad_out_to_modulo: 8
+
+device: cuda
+out_key: inpainted
+
+refine: False # refiner will only run if this is True
+refiner:
+  gpu_ids: 0,1 # the GPU ids of the machine to use. If only single GPU, use: "0,"
+  modulo: ${dataset.pad_out_to_modulo}
+  n_iters: 15 # number of iterations of refinement for each scale
+  lr: 0.002 # learning rate
+  min_side: 512 # all sides of image on all scales should be >= min_side / sqrt(2)
+  max_scales: 3 # max number of downscaling scales for the image-mask pyramid
+  px_budget: 1800000 # pixels budget. Any image will be resized to satisfy height*width <= px_budget

requirements.txt

@@ -0,0 +1,20 @@
+pyyaml
+tqdm
+numpy
+easydict==1.9.0
+scikit-image==0.17.2
+scikit-learn==0.24.2
+opencv-python
+tensorflow
+joblib
+matplotlib
+pandas
+albumentations==0.5.2
+hydra-core==1.1.0
+pytorch-lightning==1.2.9
+tabulate
+kornia==0.5.0
+webdataset
+packaging
+scikit-learn==0.24.2
+wldhx.yadisk-direct