# optimum-deepsparse
Accelerated inference of 🤗 models on CPUs using the [DeepSparse Inference Runtime](https://github.com/neuralmagic/deepsparse).
[](https://github.com/neuralmagic/optimum-deepsparse/actions/workflows/test_check.yaml)
[](https://github.com/neuralmagic/optimum-deepsparse/actions/workflows/test_nightly.yaml)
## Install
Optimum DeepSparse is a fast-moving project, and you may want to install from source.
`pip install git+https://github.com/neuralmagic/optimum-deepsparse.git`
### Installing in developer mode
If you are working on the `optimum-deepsparse` code then you should use an editable install by cloning and installing `optimum` and `optimum-deepsparse`:
```
git clone https://github.com/huggingface/optimum
git clone https://github.com/neuralmagic/optimum-deepsparse
pip install -e optimum -e optimum-deepsparse
```
Now whenever you change the code, you'll be able to run with those changes instantly.
## How to use it?
To load a model and run inference with DeepSparse, you can just replace your `AutoModelForXxx` class with the corresponding `DeepSparseModelForXxx` class.
```diff
import requests
from PIL import Image
- from transformers import AutoModelForImageClassification
+ from optimum.deepsparse import DeepSparseModelForImageClassification
from transformers import AutoFeatureExtractor, pipeline
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
model_id = "microsoft/resnet-50"
- model = AutoModelForImageClassification.from_pretrained(model_id)
+ model = DeepSparseModelForImageClassification.from_pretrained(model_id, export=True, input_shapes="[1,3,224,224]")
feature_extractor = AutoFeatureExtractor.from_pretrained(model_id)
cls_pipe = pipeline("image-classification", model=model, feature_extractor=feature_extractor)
outputs = cls_pipe(image)
```
| Supported Task | Model Class |
| ------------------------------------------- | ------------- |
| "image-classification" | DeepSparseModelForImageClassification |
| "text-classification"/"sentiment-analysis" | DeepSparseModelForSequenceClassification |
| "audio-classification" | DeepSparseModelForAudioClassification |
| "question-answering" | DeepSparseModelForQuestionAnswering |
| "image-segmentation" | DeepSparseModelForSemanticSegmentation |
If you find any issue while using those, please open an issue or a pull request.
Raw data
{
"_id": null,
"home_page": "https://github.com/neuralmagic/deepsparse",
"name": "optimum-deepsparse",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8, <3.11",
"maintainer_email": "",
"keywords": "inference,cpu,x86,arm,transformers,quantization,pruning,sparsity",
"author": "Neuralmagic, Inc.",
"author_email": "michael@neuralmagic.com",
"download_url": "https://files.pythonhosted.org/packages/b8/c3/dc8e590088b3e3d1f31f48da014c1cc852ba8bd64860abfdd45f34ff2b5b/optimum-deepsparse-0.1.0.dev1.tar.gz",
"platform": null,
"description": "# optimum-deepsparse\n\nAccelerated inference of \ud83e\udd17 models on CPUs using the [DeepSparse Inference Runtime](https://github.com/neuralmagic/deepsparse).\n\n[](https://github.com/neuralmagic/optimum-deepsparse/actions/workflows/test_check.yaml)\n[](https://github.com/neuralmagic/optimum-deepsparse/actions/workflows/test_nightly.yaml)\n\n## Install\nOptimum DeepSparse is a fast-moving project, and you may want to install from source.\n\n`pip install git+https://github.com/neuralmagic/optimum-deepsparse.git`\n\n### Installing in developer mode\n\nIf you are working on the `optimum-deepsparse` code then you should use an editable install by cloning and installing `optimum` and `optimum-deepsparse`:\n\n```\ngit clone https://github.com/huggingface/optimum\ngit clone https://github.com/neuralmagic/optimum-deepsparse\npip install -e optimum -e optimum-deepsparse\n```\n\nNow whenever you change the code, you'll be able to run with those changes instantly.\n\n\n## How to use it?\nTo load a model and run inference with DeepSparse, you can just replace your `AutoModelForXxx` class with the corresponding `DeepSparseModelForXxx` class. \n\n```diff\nimport requests\nfrom PIL import Image\n\n- from transformers import AutoModelForImageClassification\n+ from optimum.deepsparse import DeepSparseModelForImageClassification\nfrom transformers import AutoFeatureExtractor, pipeline\n\nurl = \"http://images.cocodataset.org/val2017/000000039769.jpg\"\nimage = Image.open(requests.get(url, stream=True).raw)\n\nmodel_id = \"microsoft/resnet-50\"\n- model = AutoModelForImageClassification.from_pretrained(model_id)\n+ model = DeepSparseModelForImageClassification.from_pretrained(model_id, export=True, input_shapes=\"[1,3,224,224]\")\nfeature_extractor = AutoFeatureExtractor.from_pretrained(model_id)\ncls_pipe = pipeline(\"image-classification\", model=model, feature_extractor=feature_extractor)\noutputs = cls_pipe(image)\n```\n\n| Supported Task | Model Class |\n| ------------------------------------------- | ------------- |\n| \"image-classification\" | DeepSparseModelForImageClassification |\n| \"text-classification\"/\"sentiment-analysis\" | DeepSparseModelForSequenceClassification |\n| \"audio-classification\" | DeepSparseModelForAudioClassification |\n| \"question-answering\" | DeepSparseModelForQuestionAnswering |\n| \"image-segmentation\" | DeepSparseModelForSemanticSegmentation |\n\nIf you find any issue while using those, please open an issue or a pull request.\n\n\n",
"bugtrack_url": null,
"license": "Neural Magic DeepSparse Community License, Apache",
"summary": "Optimum DeepSparse is an extension of the Hugging Face Transformers library that integrates the DeepSparse inference runtime. DeepSparse offers GPU-class performance on CPUs, making it possible to run Transformers and other deep learning models on commodity hardware with sparsity. Optimum DeepSparse provides a framework for developers to easily integrate DeepSparse into their applications, regardless of the hardware platform.",
"version": "0.1.0.dev1",
"project_urls": {
"Homepage": "https://github.com/neuralmagic/deepsparse"
},
"split_keywords": [
"inference",
"cpu",
"x86",
"arm",
"transformers",
"quantization",
"pruning",
"sparsity"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "df53313636c48365970d6093a4925ef6289c9a9b577b2bf48ca654b0494f6b83",
"md5": "b156bf0ad9a0044dfb7b82358984b63f",
"sha256": "7e5d50db484e518ea122e15f81503a8ff3feec93b6a552a8d13af76de7967a48"
},
"downloads": -1,
"filename": "optimum_deepsparse-0.1.0.dev1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "b156bf0ad9a0044dfb7b82358984b63f",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8, <3.11",
"size": 27406,
"upload_time": "2023-10-26T02:02:44",
"upload_time_iso_8601": "2023-10-26T02:02:44.122001Z",
"url": "https://files.pythonhosted.org/packages/df/53/313636c48365970d6093a4925ef6289c9a9b577b2bf48ca654b0494f6b83/optimum_deepsparse-0.1.0.dev1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "b8c3dc8e590088b3e3d1f31f48da014c1cc852ba8bd64860abfdd45f34ff2b5b",
"md5": "7de60a9bfe9fb033ca6c34f5af061b3a",
"sha256": "8597663583ba2b54e7ebe922a23ad3e0f746e3fcb0d2786d00f422f2a79e3a0c"
},
"downloads": -1,
"filename": "optimum-deepsparse-0.1.0.dev1.tar.gz",
"has_sig": false,
"md5_digest": "7de60a9bfe9fb033ca6c34f5af061b3a",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8, <3.11",
"size": 26922,
"upload_time": "2023-10-26T02:02:45",
"upload_time_iso_8601": "2023-10-26T02:02:45.556076Z",
"url": "https://files.pythonhosted.org/packages/b8/c3/dc8e590088b3e3d1f31f48da014c1cc852ba8bd64860abfdd45f34ff2b5b/optimum-deepsparse-0.1.0.dev1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-10-26 02:02:45",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "neuralmagic",
"github_project": "deepsparse",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "optimum-deepsparse"
}
JSON
