| Name | deepsynth JSON |
| Version |
0.1.3
JSON |
| download |
| home_page | https://github.com/venombolteop/deepsynth |
| Summary | A Python package for synthesizing deep learning models using advanced techniques like NAS, Hyperparameter Optimization, and more. |
| upload_time | 2024-09-17 14:38:53 |
| maintainer | None |
| docs_url | None |
| author | venombolteop |
| requires_python | <4.0,>=3.10 |
| license | MIT |
| keywords |
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# DeepSynth
**DeepSynth** is a cutting-edge Python package for synthesizing deep learning models using advanced techniques such as Neural Architecture Search (NAS), Hyperparameter Optimization, Transfer Learning, and more. Designed to streamline the model-building process, DeepSynth enables users to create high-performance models with minimal manual effort.
## Key Features
- **Neural Architecture Search (NAS)**: Explore various model architectures automatically to identify the best configuration for your task.
- **Hyperparameter Optimization**: Fine-tune model hyperparameters using state-of-the-art optimization algorithms.
- **Transfer Learning**: Utilize pre-trained models and adapt them to your specific use case.
- **Model Generation**: Create new model architectures based on predefined constraints and datasets.
- **Visualization Tools**: Visualize model performance, architecture, and optimization progress with built-in tools.
## Installation
To get started with DeepSynth, you need to install the package and its dependencies. You can do this by using the `requirements.txt` file or directly installing via pip.
### Using `requirements.txt`
1. Clone the repository or download the `requirements.txt` file.
2. Install the dependencies:
```bash
pip install -r requirements.txt
```
### Using pip
Install DeepSynth directly from PyPI:
```bash
pip install deepsynth
```
## Usage
Here are some examples of how to use various features of DeepSynth:
### Neural Architecture Search (NAS)
```python
from deepsynth.nas import NeuralArchitectureSearch
import numpy as np
# Prepare sample data
X_train = np.random.rand(1000, 20) # 1000 samples, 20 features
y_train = np.random.randint(2, size=1000) # Binary target
# Define search space
search_space = [
{'layers': 2, 'units': 64},
{'layers': 3, 'units': 128},
{'layers': 4, 'units': 256}
]
# Perform NAS
nas = NeuralArchitectureSearch(search_space)
best_model = nas.search_best_model(X_train, y_train)
print("Best model found by NAS:")
best_model.summary()
```
### Hyperparameter Optimization
```python
from deepsynth.hyperparameter_optimization import HyperparameterOptimization
from hyperopt import hp
def objective_function(params):
from deepsynth.model_generation import ModelGeneration
model = ModelGeneration(params['num_layers'], params['units_per_layer']).generate_model()
model.fit(X_train, y_train, epochs=5, verbose=0)
loss, accuracy = model.evaluate(X_train, y_train, verbose=0)
return {'loss': loss, 'status': 'ok'}
# Define hyperparameter search space
space = {
'num_layers': hp.choice('num_layers', [2, 3, 4]),
'units_per_layer': hp.choice('units_per_layer', [32, 64, 128])
}
# Optimize hyperparameters
optimizer = HyperparameterOptimization(objective_function)
best_params = optimizer.optimize(space)
print("Best hyperparameters found:")
print(best_params)
```
### Transfer Learning
```python
from deepsynth.transfer_learning import TransferLearning
# Load and adapt a pre-trained model
transfer_learning = TransferLearning()
model = transfer_learning.load_model()
print("Transfer learning model:")
model.summary()
```
### Model Generation
```python
from deepsynth.model_generation import ModelGeneration
# Generate a new model
num_layers = 3
units_per_layer = 128
model_generator = ModelGeneration(num_layers, units_per_layer)
model = model_generator.generate_model()
print("Generated model:")
model.summary()
```
### Visualization
```python
from deepsynth.visualization import Visualization
# Example training history
history = {
'accuracy': [0.1, 0.3, 0.5, 0.7, 0.9],
'val_accuracy': [0.15, 0.35, 0.55, 0.75, 0.85]
}
Visualization.plot_model_performance(history)
```
## Examples
For more detailed examples and advanced use cases, check out the `examples/` directory in the repository. The `example_usage.py` file includes various functionalities and applications of DeepSynth.
## Contributing
We welcome contributions to DeepSynth! If you would like to contribute, please follow these steps:
1. **Fork the repository**: Create a personal copy of the repository on GitHub.
2. **Create a branch**: Develop your changes on a new branch.
3. **Commit changes**: Make and commit your changes with clear messages.
4. **Push changes**: Push your branch to your forked repository.
5. **Create a pull request**: Submit a pull request describing your changes and improvements.
## License
DeepSynth is licensed under the MIT License. See the [LICENSE](LICENSE) file for more details.
---
## Contact & Social Media
- **Telegram Channel**: [Channel](https://t.me/VenomOwners)
- **Telegram Group**: [Group](https://t.me/Venom_Chatz)
- **Owner's Telegram**: [Telegram](https://t.me/K_4ip)
- **Instagram**: [Instagram](https://instagram.com/venom_owners)
- **GitHub**: [Github](https://github.com/venombolteop/deepsynth)
- **Email**: ayush20912@gmail.com
For further questions, support, or to report issues, please open an issue on the [GitHub repository](https://github.com/venombolteop/deepsynth) or contact the maintainers.
Raw data
{
"_id": null,
"home_page": "https://github.com/venombolteop/deepsynth",
"name": "deepsynth",
"maintainer": null,
"docs_url": null,
"requires_python": "<4.0,>=3.10",
"maintainer_email": null,
"keywords": null,
"author": "venombolteop",
"author_email": "venombolteop@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/ac/8b/594c4b7144459980026cacc21e43eeeeaa136f11d49cee635fbc9cf6337b/deepsynth-0.1.3.tar.gz",
"platform": null,
"description": "# DeepSynth\n\n**DeepSynth** is a cutting-edge Python package for synthesizing deep learning models using advanced techniques such as Neural Architecture Search (NAS), Hyperparameter Optimization, Transfer Learning, and more. Designed to streamline the model-building process, DeepSynth enables users to create high-performance models with minimal manual effort.\n\n## Key Features\n\n- **Neural Architecture Search (NAS)**: Explore various model architectures automatically to identify the best configuration for your task.\n- **Hyperparameter Optimization**: Fine-tune model hyperparameters using state-of-the-art optimization algorithms.\n- **Transfer Learning**: Utilize pre-trained models and adapt them to your specific use case.\n- **Model Generation**: Create new model architectures based on predefined constraints and datasets.\n- **Visualization Tools**: Visualize model performance, architecture, and optimization progress with built-in tools.\n\n## Installation\n\nTo get started with DeepSynth, you need to install the package and its dependencies. You can do this by using the `requirements.txt` file or directly installing via pip.\n\n### Using `requirements.txt`\n\n1. Clone the repository or download the `requirements.txt` file.\n2. Install the dependencies:\n\n ```bash\n pip install -r requirements.txt\n ```\n\n### Using pip\n\nInstall DeepSynth directly from PyPI:\n\n```bash\npip install deepsynth\n```\n\n## Usage\n\nHere are some examples of how to use various features of DeepSynth:\n\n### Neural Architecture Search (NAS)\n\n```python\nfrom deepsynth.nas import NeuralArchitectureSearch\nimport numpy as np\n\n# Prepare sample data\nX_train = np.random.rand(1000, 20) # 1000 samples, 20 features\ny_train = np.random.randint(2, size=1000) # Binary target\n\n# Define search space\nsearch_space = [\n {'layers': 2, 'units': 64},\n {'layers': 3, 'units': 128},\n {'layers': 4, 'units': 256}\n]\n\n# Perform NAS\nnas = NeuralArchitectureSearch(search_space)\nbest_model = nas.search_best_model(X_train, y_train)\nprint(\"Best model found by NAS:\")\nbest_model.summary()\n```\n\n### Hyperparameter Optimization\n\n```python\nfrom deepsynth.hyperparameter_optimization import HyperparameterOptimization\nfrom hyperopt import hp\n\ndef objective_function(params):\n from deepsynth.model_generation import ModelGeneration\n model = ModelGeneration(params['num_layers'], params['units_per_layer']).generate_model()\n model.fit(X_train, y_train, epochs=5, verbose=0)\n loss, accuracy = model.evaluate(X_train, y_train, verbose=0)\n return {'loss': loss, 'status': 'ok'}\n\n# Define hyperparameter search space\nspace = {\n 'num_layers': hp.choice('num_layers', [2, 3, 4]),\n 'units_per_layer': hp.choice('units_per_layer', [32, 64, 128])\n}\n\n# Optimize hyperparameters\noptimizer = HyperparameterOptimization(objective_function)\nbest_params = optimizer.optimize(space)\nprint(\"Best hyperparameters found:\")\nprint(best_params)\n```\n\n### Transfer Learning\n\n```python\nfrom deepsynth.transfer_learning import TransferLearning\n\n# Load and adapt a pre-trained model\ntransfer_learning = TransferLearning()\nmodel = transfer_learning.load_model()\nprint(\"Transfer learning model:\")\nmodel.summary()\n```\n\n### Model Generation\n\n```python\nfrom deepsynth.model_generation import ModelGeneration\n\n# Generate a new model\nnum_layers = 3\nunits_per_layer = 128\nmodel_generator = ModelGeneration(num_layers, units_per_layer)\nmodel = model_generator.generate_model()\nprint(\"Generated model:\")\nmodel.summary()\n```\n\n### Visualization\n\n```python\nfrom deepsynth.visualization import Visualization\n\n# Example training history\nhistory = {\n 'accuracy': [0.1, 0.3, 0.5, 0.7, 0.9],\n 'val_accuracy': [0.15, 0.35, 0.55, 0.75, 0.85]\n}\nVisualization.plot_model_performance(history)\n```\n\n## Examples\n\nFor more detailed examples and advanced use cases, check out the `examples/` directory in the repository. The `example_usage.py` file includes various functionalities and applications of DeepSynth.\n\n## Contributing\n\nWe welcome contributions to DeepSynth! If you would like to contribute, please follow these steps:\n\n1. **Fork the repository**: Create a personal copy of the repository on GitHub.\n2. **Create a branch**: Develop your changes on a new branch.\n3. **Commit changes**: Make and commit your changes with clear messages.\n4. **Push changes**: Push your branch to your forked repository.\n5. **Create a pull request**: Submit a pull request describing your changes and improvements.\n\n## License\n\nDeepSynth is licensed under the MIT License. See the [LICENSE](LICENSE) file for more details.\n\n---\n\n## Contact & Social Media\n\n- **Telegram Channel**: [Channel](https://t.me/VenomOwners)\n- **Telegram Group**: [Group](https://t.me/Venom_Chatz)\n- **Owner's Telegram**: [Telegram](https://t.me/K_4ip)\n- **Instagram**: [Instagram](https://instagram.com/venom_owners)\n- **GitHub**: [Github](https://github.com/venombolteop/deepsynth)\n- **Email**: ayush20912@gmail.com\n\nFor further questions, support, or to report issues, please open an issue on the [GitHub repository](https://github.com/venombolteop/deepsynth) or contact the maintainers.\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A Python package for synthesizing deep learning models using advanced techniques like NAS, Hyperparameter Optimization, and more.",
"version": "0.1.3",
"project_urls": {
"Documentation": "https://github.com/venombolteop/deepsynth/wiki",
"Homepage": "https://github.com/venombolteop/deepsynth",
"Repository": "https://github.com/venombolteop/deepsynth"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "e2a2a454370a7e419a130bbd03315837deab50680497b71b3a913c569ea971f8",
"md5": "79f42de96b5afb9c557d529e5c853524",
"sha256": "48fa0cdfabde1b8d901d6bf0a7334d9dcca7f1647271be3172c99a99aeec48e4"
},
"downloads": -1,
"filename": "deepsynth-0.1.3-py3-none-any.whl",
"has_sig": false,
"md5_digest": "79f42de96b5afb9c557d529e5c853524",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": "<4.0,>=3.10",
"size": 6770,
"upload_time": "2024-09-17T14:38:52",
"upload_time_iso_8601": "2024-09-17T14:38:52.721885Z",
"url": "https://files.pythonhosted.org/packages/e2/a2/a454370a7e419a130bbd03315837deab50680497b71b3a913c569ea971f8/deepsynth-0.1.3-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "ac8b594c4b7144459980026cacc21e43eeeeaa136f11d49cee635fbc9cf6337b",
"md5": "db6d518b27001a0569cbcade2fc51018",
"sha256": "e04f11e695ef3e4fb559da70b9c1c20eaf6ccfc37ab1a57ca25a3c3eb4878593"
},
"downloads": -1,
"filename": "deepsynth-0.1.3.tar.gz",
"has_sig": false,
"md5_digest": "db6d518b27001a0569cbcade2fc51018",
"packagetype": "sdist",
"python_version": "source",
"requires_python": "<4.0,>=3.10",
"size": 4788,
"upload_time": "2024-09-17T14:38:53",
"upload_time_iso_8601": "2024-09-17T14:38:53.732147Z",
"url": "https://files.pythonhosted.org/packages/ac/8b/594c4b7144459980026cacc21e43eeeeaa136f11d49cee635fbc9cf6337b/deepsynth-0.1.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-09-17 14:38:53",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "venombolteop",
"github_project": "deepsynth",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [],
"lcname": "deepsynth"
}
