Upload folder using huggingface_hub

README.md

@@ -0,0 +1,137 @@
+# MakeItColor: Image Colorization Model
+
+## Model Description
+**MakeItColor** is a deep learning model designed for automatic image colorization. It accepts grayscale images as input and generates vivid, realistic colorized outputs. Built with a PyTorch-based Convolutional Neural Network (CNN) architecture, it is seamlessly integrated with the **ModelScope** framework for easy deployment across various applications.
+
+This model is inspired by and builds upon the work of [DDColor](https://github.com/piddnad/DDColor), utilizing a dual-encoder approach and trained on the **ImageNet-Val5k** dataset.
+
+---
+
+## Task
+- **Image Colorization**
+
+## Framework
+- **PyTorch**, **ModelScope**
+
+## Model Type
+- **Convolutional Neural Network (CNN)**
+
+## Input
+- **Grayscale images** (single-channel)
+
+## Output
+- **Colorized images** (RGB format)
+
+---
+
+## Installation
+Make sure you have **Python 3.7+** installed. Then, install the required dependencies:
+
+```bash
+!pip install gradio
+!pip install opencv-python
+!pip install modelscope==1.12.0 
+!pip install datasets==2.14.7
+!pip install pillow
+!pip install numpy
+!pip install gradio-imageslider
+```
+
+---
+
+## Usage
+
+You can easily use **MakeItColor** through the ModelScope pipeline:
+
+```python
+import cv2
+from modelscope.pipelines import pipeline
+from modelscope.utils.constant import Tasks
+
+# Initialize the colorization pipeline
+img_colorization = pipeline(Tasks.image_colorization, model='your-username/makeitcolor')
+
+# Load a grayscale image
+img_path = 'input.jpg'
+
+# Run colorization
+result = img_colorization(img_path)
+
+# Save the colorized image
+cv2.imwrite('result.png', result['output_img'])
+```
+
+> **Note**:  
+> - Ensure that the input image (`input.jpg`) is a proper grayscale (single-channel) image.  
+> - The output (`result.png`) will be a standard RGB image.
+
+---
+
+## Model Files
+
+The repository contains the following files:
+
+- `pytorch_model.pt`: Pre-trained model weights.
+- `configuration.json`: Model configuration file for ModelScope integration.
+- `README.md`: This documentation file.
+
+---
+
+## Inference Requirements
+
+- **Hardware**:  
+  - CPU (supported)  
+  - GPU (recommended for faster inference)
+
+- **Software Dependencies**:  
+  - `modelscope`  
+  - `opencv-python`  
+  - `torch`
+
+---
+
+## Input Format
+
+- Grayscale images (`.png`, `.jpg`, etc.)
+
+### Example
+
+1. Prepare a grayscale image (e.g., `input.jpg`).
+2. Run the provided example code.
+3. Check the output file (`result.png`) for the colorized result.
+
+---
+
+## Limitations
+
+- The model may struggle with highly complex, ambiguous, or abstract grayscale images.
+- Performance and output quality depend on the clarity and details of the input.
+- Primarily optimized for **natural images**; results may vary for synthetic or artistic inputs.
+
+---
+
+## Credits
+
+This work builds upon and was inspired by the [DDColor project](https://github.com/piddnad/DDColor).  
+**MakeItColor** leverages a dual-encoder strategy from DDColor and is trained on the **ImageNet-Val5k** dataset.
+
+Special thanks to the creators of DDColor for their foundational contributions.
+
+---
+
+## License
+
+This project is licensed under the **Apache License 2.0**.
+
+---
+
+## Contact
+
+For issues, questions, or feedback, feel free to:
+
+- Open an issue on the [Hugging Face repository](#).
+- Contact the maintainer directly at: **[[email protected]](mailto:[email protected])**
+
+---
+
+**Developed by Muhammad Noman**

configuration.json

@@ -0,0 +1,67 @@
+{
+    "framework": "pytorch",
+
+    "task": "image-colorization",
+
+    "pipeline": {
+        "type": "ddcolor-image-colorization"
+    },
+
+    "model": {
+        "type": "ddcolor"
+    },
+
+    "dataset": {
+        "name": "imagenet-val5k-image",
+        "dataroot_gt": "val5k/",
+        "filename_tmpl": "{}",
+        "scale": 1,
+        "gt_size": 256
+    },
+
+    "train": {
+        "dataloader": {
+            "batch_size_per_gpu": 4,
+            "workers_per_gpu": 4,
+            "shuffle": true
+        },
+        "optimizer": {
+            "type": "AdamW",
+            "lr": 1e-6,
+            "weight_decay": 0.01,
+            "betas": [0.9, 0.99]
+        },
+        "lr_scheduler": {
+            "type": "CosineAnnealingLR",
+            "T_max": 200000,
+            "eta_min": 1e-7
+        },
+        "max_epochs": 2,
+        "hooks": [{
+                "type": "CheckpointHook",
+                "interval": 1
+            },
+            {
+                "type": "TextLoggerHook",
+                "interval": 1
+            },
+            {
+                "type": "IterTimerHook"
+            },
+            {
+                "type": "EvaluationHook",
+                "interval": 1
+            }
+        ]
+    },
+
+    "evaluation": {
+        "dataloader": {
+            "batch_size_per_gpu": 8,
+            "workers_per_gpu": 1,
+            "shuffle": false
+        },
+        "metrics": "image-colorization-metric"
+    }
+
+}

pytorch_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17c460d7e55b32a598370621d77173be59e03c24b0823f06821db23a50c263ce
+size 911950059