README.md

---
language: en
license: mit
model_id: Statement_Equivalence
developers: Matt Stammers
model_type: BERT-Base-Uncased
model_summary: This model Compares the similarity of two text objects. It is the first
  BERT model I have fine tuned so there may be bugs. The model labels should read
  equivalent/not-equivalent but despite mapping the id2label variables they are presently
  still displaying as label0/1 in the inference module. I may come back and fix this
  at a later date.
shared_by: Matt Stammers
finetuned_from: Glue
repo: https://huggingface.co/MattStammers/Statement_Equivalence?text=I+like+you.+I+love+you
paper: N/A
demo: N/A
direct_use: Test it out here
downstream_use: This is a standalone app
out_of_scope_use: The model will not work with any very complex sentences or to compare
  more than 3 statements
bias_risks_limitations: Biases inherent in Glue also apply here
bias_recommendations: Do not be surprised if unusual results are obtained
get_started_code: "\n    ``` python \n    # Use a pipeline as a high-level helper\n\
  \        from transformers import pipeline\n\n        pipe = pipeline(\"text-classification\"\
  , model=\"MattStammers/Statement_Equivalence\")\n    # Load model directly\n   \
  \     from transformers import AutoTokenizer, AutoModelForSequenceClassification\n\
  \n        tokenizer = AutoTokenizer.from_pretrained(\"MattStammers/Statement_Equivalence\"\
  )\n        model = AutoModelForSequenceClassification.from_pretrained(\"MattStammers/Statement_Equivalence\"\
  )\n    ```\n                        "
training_data: 'See Glue Dataset: https://huggingface.co/datasets/glue'
preprocessing: Sentence Pairs to analyse similarity
training_regime: User Defined
speeds_sizes_times: Not Relevant
testing_data: 'MRCP. Link: https://huggingface.co/datasets/SetFit/mrpc'
testing_factors: N/A
testing_metrics: N/A
results: See evaluation results.
results_summary: See Over
model_examination: Model should be interpreted with user discretion.
model_specs: Bert fine-tuned
compute_infrastructure: requires less than 4GB of GPU to run quickly
hardware: T600
hours_used: '0.1'
cloud_provider: N/A
cloud_region: N/A
co2_emitted: <1
software: Python, pytorch with transformers
citation_bibtex: N/A
citation_apa: N/A
glossary: N/A
more_information: Can be made available on request
model_card_authors: Matt Stammers
model_card_contact: Matt Stammers
model-index:
- name: statement
  results:
  - task:
      type: text-classification
    dataset:
      name: MRCP
      type: mrcp
    metrics:
    - type: accuracy
      value: 0.8480392156862745
    - type: F1-Score
      value: 0.8945578231292517
---

# Model Card for Statement_Equivalence

<!-- Provide a quick summary of what the model is/does. -->

This model Compares the similarity of two text objects. It is the first BERT model I have fine tuned so there may be bugs. The model labels should read equivalent/not-equivalent but despite mapping the id2label variables they are presently still displaying as label0/1 in the inference module. I may come back and fix this at a later date.

## Model Details

### Model Description

<!-- Provide a longer summary of what this model is. -->



- **Developed by:** Matt Stammers
- **Shared by [optional]:** Matt Stammers
- **Model type:** BERT-Base-Uncased
- **Language(s) (NLP):** en
- **License:** mit
- **Finetuned from model [optional]:** Glue

### Model Sources [optional]

<!-- Provide the basic links for the model. -->

- **Repository:** https://huggingface.co/MattStammers/Statement_Equivalence?text=I+like+you.+I+love+you
- **Paper [optional]:** N/A
- **Demo [optional]:** N/A

## Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

### Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

Test it out here

### Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

This is a standalone app

### Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

The model will not work with any very complex sentences or to compare more than 3 statements

## Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

Biases inherent in Glue also apply here

### Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Do not be surprised if unusual results are obtained

## How to Get Started with the Model

Use the code below to get started with the model.


    ``` python 
    # Use a pipeline as a high-level helper
        from transformers import pipeline

        pipe = pipeline("text-classification", model="MattStammers/Statement_Equivalence")
    # Load model directly
        from transformers import AutoTokenizer, AutoModelForSequenceClassification

        tokenizer = AutoTokenizer.from_pretrained("MattStammers/Statement_Equivalence")
        model = AutoModelForSequenceClassification.from_pretrained("MattStammers/Statement_Equivalence")
    ```
                        

## Training Details

### Training Data

<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

See Glue Dataset: https://huggingface.co/datasets/glue

### Training Procedure 

<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->

#### Preprocessing [optional]

Sentence Pairs to analyse similarity


#### Training Hyperparameters

- **Training regime:** User Defined <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->

#### Speeds, Sizes, Times [optional]

<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->

Not Relevant

## Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

### Testing Data, Factors & Metrics

#### Testing Data

<!-- This should link to a Data Card if possible. -->

MRCP. Link: https://huggingface.co/datasets/SetFit/mrpc

#### Factors

<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->

N/A

#### Metrics

<!-- These are the evaluation metrics being used, ideally with a description of why. -->

N/A

### Results

See evaluation results.

#### Summary

See Over

## Model Examination [optional]

<!-- Relevant interpretability work for the model goes here -->

Model should be interpreted with user discretion.

## Environmental Impact

<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->

Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).

- **Hardware Type:** T600
- **Hours used:** 0.1
- **Cloud Provider:** N/A
- **Compute Region:** N/A
- **Carbon Emitted:** <1

## Technical Specifications [optional]

### Model Architecture and Objective

Bert fine-tuned

### Compute Infrastructure

requires less than 4GB of GPU to run quickly

#### Hardware

T600

#### Software

Python, pytorch with transformers

## Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

N/A

**APA:**

N/A

## Glossary [optional]

<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->

N/A

## More Information [optional]

Can be made available on request

## Model Card Authors [optional]

Matt Stammers

## Model Card Contact

Matt Stammers