遵循以下最佳实践的项目将能够自愿的自我认证,并显示他们已经实现了核心基础设施计划(OpenSSF)徽章。 显示详细资料
[](https://www.bestpractices.dev/projects/8925)
<a href="https://www.bestpractices.dev/projects/8925"><img src="https://www.bestpractices.dev/projects/8925/badge"></a>
Intel® AI Reference models project demos performance of AI models on Intel hardware - CPUs and GPUs. Replicates and runs the model Intel HW based on latency, throughput and accuracy. The purpose of the Intel® AI Reference Models repository is to quickly replicate the complete software environment that demonstrates the best-known performance of each of these target model/dataset combinations. When executed in optimally configured hardware environments, these software environments showcase the AI capabilities of Intel platforms.
https://github.com/IntelAI/models/blob/master/CONTRIBUTING.md
https://github.com/intel-innersource/frameworks.ai.models.intel-models/commit/d6041bcae0766ad6e7c16dd5d2868a8630e50ab8
https://github.com/IntelAI/models/blob/master/CODE_of_CONDUCT.md
https://github.com/intel/models/blob/master/CODEOWNERS https://github.com/intel/models/blob/master/CONTRIBUTING.md
https://github.com/intel/models/tree/master#purpose-of-intel-ai-reference-models
It does not produce software. The project is only for replicating the performance of models on Intel hardware
https://github.com/intel/models/blob/master/SECURITY.md
https://github.com/intel/models/blob/master/README.md
This contains README.md for all the workloads in the repository which will help user to run any workload
https://github.com/intel/models.
The badge has been included
It is met because we have a simple readme to follow which does not include any images or videos.
We just support code and not the datasets
The project does not store any passwords for end user authentication
https://github.com/intel/models/releases
https://github.com/intel/models/issues
https://github.com/intel/models/issues The history includes the person who reported.
https://github.com/intel/models/blob/master/CONTRIBUTING.md
https://github.com/intel/models#purpose-of-intel-ai-reference-models
No native binaries are generated
No build or installation system
Directly the code is used, not compiled.
https://github.com/intel/models/tree/master/datasets/dataset_api#dataset-librarian
Yes, it installs only where the customer points to.
We don't have any package
https://github.com/intel/models/security/dependabot
The deprecated functions are mentioned here and the project owner maintains it time to time
https://github.com/intel/models/blob/master/CONTRIBUTING.md https://github.com/intel/models/tree/master/tests/cicd
https://github.com/intel/models/blob/master/CONTRIBUTING.md https://github.com/intel/models/tree/master/tests/cicd https://github.com/intel/models/tree/master/tests/unit
https://github.com/intel/models/tree/master/tests/cicd https://github.com/intel/models/blob/master/CONTRIBUTING.md
https://github.com/intel/models/tree/master/tests/cicd
We have linter and warning check which needs to be clear before merging pull request
The project does not produce a software. The project is just for the customers to reproduce results on Intel platform with best known configurations.
None of the scripts depend on cryptographic algorithms
Yes, the project will work on any type of crypto
The project does not provide or need any credentials
The project supports all network communications
We tag every release and can be used by everyone - https://github.com/intel/ai-reference-models/releases/tag/v3.2
We explain in notes with every release with a tag - https://github.com/intel/ai-reference-models/releases/tag/v3.2
The project includes all licenses related to everything. And datasets must be manually downloaded by the users.
We keep fixing the vulnerabilities very frequently to avoid security issues.
Yes, the badges have been updated the documentation of the project and we do a release once we do not have any security issues and all security scans are clear: https://github.com/intel/ai-reference-models/blob/main/SECURITY.md
We have depandabot to fix vulnerabilities (third party issues)
We have a security scan which fixes CVEs so fuzzing may not be required
后退