Skip to content

Latest commit

 

History

History
177 lines (109 loc) · 4.06 KB

model_card.md

File metadata and controls

177 lines (109 loc) · 4.06 KB
Raw

Model Card For [Model Name]

==============================

This documentation is aimed to help communicate key information about a machine learning model. This model card has been borrowed and adapted from Annotated Hugging Face Model Card. A model card can also be generated through huggingface_hub library which is a python interface.

Last updated : August 2023.

Provide a short summary of model.

Table of Contents

This is intended to make it easier to navigate the different sections.

Model Details

This section describes important background information of the developers and the model.

Provide a longer summary of what this model is. Include details such as the input and output of the model.

Model Description

  • Developed by:
  • Model type:
  • Language(s) (NLP):
  • License:

Some other optional details that can be added :

  • Shared by :
  • Finetuned from model:

Model Source

This section is optional, and is used to provide direct sources to the users.

  • Repository:
  • Paper:
  • Demo:

Model Type

This section is optional.

Model version

This section is optional.

Intended Uses

This section aims to describe how the model was intended to be used in various applied contexts, the intended users of the model and those that will be affected by it.

Primary uses

Provide the direct uses of the model, when it is not fine-tuned or any post-processing is done.

Secondary uses

Provide a description of uses of the model after fine-tuning is done, or it is added to a larger ecosytem.

Out-Of-Scope uses

Provide a description of misuse of the model.

Bias, Risks, and Limitations

Provide any forseeable harm, technical and sociotechnical limitations.

Technical Specifications

This heading is optional.

Model Architecture

Provide a description of the model specifications

Compute Infrasture

  • Hardware:
  • Software:

Training Details

This section is intended to provide information that describes the training of the model, and can allow users to replicate the training.

Training data

Training Procedure/Process

Training Hyperparameters

Evaluation

This section is intended to provide information that describes the evaluation protocol that was followed, what was measured and the final results obtained.

Testing Data

Factors

Metrics

Results

Environmental Impact

This provides a summary of the factors that contribute to the environmental impact. An estimation can be generated on ML CO2 Impact.

  • Hardware Type:
  • Hours used:
  • Cloud Provider:
  • Compute Region:
  • Carbon Emitted:

How to Get Started with the Model

Provide the code needed to use the model.

 get started code

Citations

Provide the preferred citation (usually a paper).

BibTeX


@misc{name_year_modeltype,
   title={Model card title},
   author={Lastname, Firstname (and Lastname, Firstname and...)},
   year={year},
   url={this URL}
}

APA

Author {Lastname, Initial}.(Date). Title. url

More Information

This is optional and should address any additional information that the users might require.

Author Details

This is optional and provides the necessary information about the authors of the model card.

  • Written by :
  • Contact details :