[AI Benchmark Alpha](http://ai-benchmark.com/alpha) is an open source python library for evaluating AI performance of various hardware platforms, including CPUs, GPUs and TPUs. The benchmark is relying on [TensorFlow](https://www.tensorflow.org) machine learning library, and is providing a lightweight and accurate solution for assessing inference and training speed for key Deep Learning models.</br></br>
In total, AI Benchmark consists of <b>42 tests</b> and <b>19 sections</b> provided below:</br>
1. MobileNet-V2 `[classification]`
2. Inception-V3 `[classification]`
3. Inception-V4 `[classification]`
4. Inception-ResNet-V2 `[classification]`
5. ResNet-V2-50 `[classification]`
6. ResNet-V2-152 `[classification]`
7. VGG-16 `[classification]`
8. SRCNN 9-5-5 `[image-to-image mapping]`
9. VGG-19 `[image-to-image mapping]`
10. ResNet-SRGAN `[image-to-image mapping]`
11. ResNet-DPED `[image-to-image mapping]`
12. U-Net `[image-to-image mapping]`
13. Nvidia-SPADE `[image-to-image mapping]`
14. ICNet `[image segmentation]`
15. PSPNet `[image segmentation]`
16. DeepLab `[image segmentation]`
17. Pixel-RNN `[inpainting]`
18. LSTM `[sentence sentiment analysis]`
19. GNMT `[text translation]`
For more information and results, please visit the project website: [http://ai-benchmark.com/alpha](http://ai-benchmark.com/alpha)</br></br>
#### Installation Instructions </br>
The benchmark requires TensorFlow machine learning library to be present in your system.
On systems that <b>do not have Nvidia GPUs</b>, run the following commands to install AI Benchmark:
```bash
pip install tensorflow
pip install ai-benchmark
```
</br>
If you want to check the <b>performance of Nvidia graphic cards</b>, run the following commands:
```bash
pip install tensorflow-gpu
pip install ai-benchmark
```
<b>`Note 1:`</b> If Tensorflow is already installed in your system, you can skip the first command.
<b>`Note 2:`</b> For running the benchmark on Nvidia GPUs, <b>`NVIDIA CUDA`</b> and <b>`cuDNN`</b> libraries should be installed first. Please find detailed instructions [here](https://www.tensorflow.org/install/gpu). </br></br>
#### Getting Started </br>
To run AI Benchmark, use the following code:
```bash
from ai_benchmark import AIBenchmark
benchmark = AIBenchmark()
results = benchmark.run()
```
Alternatively, on Linux systems you can type `ai-benchmark` in the command line to start the tests.
To run inference or training only, use `benchmark.run_inference()` or `benchmark.run_training()`. </br></br>
#### Advanced settings </br>
```bash
AIBenchmark(use_CPU=None, verbose_level=1):
```
> use_CPU=`{True, False, None}`: whether to run the tests on CPUs (if tensorflow-gpu is installed)
> verbose_level=`{0, 1, 2, 3}`: run tests silently | with short summary | with information about each run | with TF logs
```bash
benchmark.run(precision="normal"):
```
> precision=`{"normal", "high"}`: if `high` is selected, the benchmark will execute 10 times more runs for each test.
</br>
### Additional Notes and Requirements </br>
GPU with at least 2GB of RAM is required for running inference tests / 4GB of RAM for training tests.
The benchmark is compatible with both `TensorFlow 1.x` and `2.x` versions. </br></br>
### Contacts </br>
Please contact `andrey@vision.ee.ethz.ch` for any feedback or information.
Raw data
{
"_id": null,
"home_page": "",
"name": "new-ai-benchmark",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "Johnny N\u00fa\u00f1ez <jnunezca11@alumnes.ub.edu>",
"keywords": "AI,Benchmark,Tensorflow,Machine Learning,Inference,Training",
"author": "",
"author_email": "Andrey Ignatov <andrey@vision.ee.ethz.ch>",
"download_url": "",
"platform": null,
"description": "\n[AI Benchmark Alpha](http://ai-benchmark.com/alpha) is an open source python library for evaluating AI performance of various hardware platforms, including CPUs, GPUs and TPUs. The benchmark is relying on [TensorFlow](https://www.tensorflow.org) machine learning library, and is providing a lightweight and accurate solution for assessing inference and training speed for key Deep Learning models.</br></br>\n\nIn total, AI Benchmark consists of <b>42 tests</b> and <b>19 sections</b> provided below:</br>\n\n1. MobileNet-V2 `[classification]`\n2. Inception-V3 `[classification]`\n3. Inception-V4 `[classification]`\n4. Inception-ResNet-V2 `[classification]`\n5. ResNet-V2-50 `[classification]`\n6. ResNet-V2-152 `[classification]`\n7. VGG-16 `[classification]`\n8. SRCNN 9-5-5 `[image-to-image mapping]`\n9. VGG-19 `[image-to-image mapping]`\n10. ResNet-SRGAN `[image-to-image mapping]`\n11. ResNet-DPED `[image-to-image mapping]`\n12. U-Net `[image-to-image mapping]`\n13. Nvidia-SPADE `[image-to-image mapping]`\n14. ICNet `[image segmentation]`\n15. PSPNet `[image segmentation]`\n16. DeepLab `[image segmentation]`\n17. Pixel-RNN `[inpainting]`\n18. LSTM `[sentence sentiment analysis]`\n19. GNMT `[text translation]`\n\nFor more information and results, please visit the project website: [http://ai-benchmark.com/alpha](http://ai-benchmark.com/alpha)</br></br>\n\n#### Installation Instructions </br>\n\nThe benchmark requires TensorFlow machine learning library to be present in your system.\n\nOn systems that <b>do not have Nvidia GPUs</b>, run the following commands to install AI Benchmark:\n\n```bash\npip install tensorflow\npip install ai-benchmark\n```\n</br>\n\nIf you want to check the <b>performance of Nvidia graphic cards</b>, run the following commands:\n\n```bash\npip install tensorflow-gpu\npip install ai-benchmark\n```\n\n<b>`Note 1:`</b> If Tensorflow is already installed in your system, you can skip the first command.\n\n<b>`Note 2:`</b> For running the benchmark on Nvidia GPUs, <b>`NVIDIA CUDA`</b> and <b>`cuDNN`</b> libraries should be installed first. Please find detailed instructions [here](https://www.tensorflow.org/install/gpu). </br></br>\n\n#### Getting Started </br>\n\nTo run AI Benchmark, use the following code:\n\n```bash\nfrom ai_benchmark import AIBenchmark\nbenchmark = AIBenchmark()\nresults = benchmark.run()\n```\n\nAlternatively, on Linux systems you can type `ai-benchmark` in the command line to start the tests.\n\nTo run inference or training only, use `benchmark.run_inference()` or `benchmark.run_training()`. </br></br>\n\n#### Advanced settings </br>\n\n```bash\nAIBenchmark(use_CPU=None, verbose_level=1):\n```\n> use_CPU=`{True, False, None}`: whether to run the tests on CPUs (if tensorflow-gpu is installed)\n\n> verbose_level=`{0, 1, 2, 3}`: run tests silently | with short summary | with information about each run | with TF logs\n\n```bash\nbenchmark.run(precision=\"normal\"):\n```\n\n> precision=`{\"normal\", \"high\"}`: if `high` is selected, the benchmark will execute 10 times more runs for each test.\n\n</br>\n\n### Additional Notes and Requirements </br>\n\nGPU with at least 2GB of RAM is required for running inference tests / 4GB of RAM for training tests.\n\nThe benchmark is compatible with both `TensorFlow 1.x` and `2.x` versions. </br></br>\n\n### Contacts </br>\n\nPlease contact `andrey@vision.ee.ethz.ch` for any feedback or information.\n",
"bugtrack_url": null,
"license": "MIT License",
"summary": "AI Benchmark is an open source python library for evaluating AI performance of various hardware platforms, including CPUs, GPUs and TPUs.",
"version": "2.7.0",
"project_urls": null,
"split_keywords": [
"ai",
"benchmark",
"tensorflow",
"machine learning",
"inference",
"training"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "92d2dc67ec7c27085e2da5c140f663cf17efb0eb39347d9f595bd1be55a2d71e",
"md5": "cb656b27759d7f5d9511a992558b9c0f",
"sha256": "bf1dc6b9356352945bf77a910d968f87eb3da73baeebb42e2b16917b712b3148"
},
"downloads": -1,
"filename": "new_ai_benchmark-2.7.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "cb656b27759d7f5d9511a992558b9c0f",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 21548864,
"upload_time": "2024-03-10T18:53:48",
"upload_time_iso_8601": "2024-03-10T18:53:48.619344Z",
"url": "https://files.pythonhosted.org/packages/92/d2/dc67ec7c27085e2da5c140f663cf17efb0eb39347d9f595bd1be55a2d71e/new_ai_benchmark-2.7.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-03-10 18:53:48",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "new-ai-benchmark"
}
JSON