Stsb M Mt Es Distiluse Base Multilingual Cased V1

🚀 Test Model for Semantic Textual Similarity

This is a test model fine - tuned with Spanish datasets to understand and benchmark STS models.

🚀 Quick Start

This is a test model that was fine - tuned using the Spanish datasets from stsb_multi_mt in order to understand and benchmark STS models. You may use it as any other STS trained model to extract sentence embeddings. Check Sentence Transformers documentation.

✨ Features

• Built based on distiluse - base - multilingual - cased - v1 and fine - tuned for Semantic Textual Similarity task.
• Trained with Spanish datasets from stsb_multi_mt.

📦 Installation

No specific installation steps are provided in the original README.

💻 Usage Examples

You may use it as any other STS trained model to extract sentence embeddings. Check Sentence Transformers documentation.

📚 Documentation

Model and training data description

This model was built taking distiluse - base - multilingual - cased - v1 and training it on a Semantic Textual Similarity task using a modified version of the training script for STS from Sentece Transformers (the modified script is included in the repo). It was trained using the Spanish datasets from stsb_multi_mt which are the STSBenchmark datasets automatically translated to other languages using deepl.com. Refer to the dataset repository for more details.

Intended uses & limitations

This model was built just as a proof - of - concept on STS fine - tuning using Spanish data and no specific use other than getting a sense on how this training works.

Training procedure

This model was trained using this Colab Notebook

Evaluation results

Evaluating distiluse - base - multilingual - cased - v1 on the Spanish test dataset before training results in:

2021-07-06 17:44:46 - EmbeddingSimilarityEvaluator: Evaluating the model on  dataset:
2021-07-06 17:45:00 - Cosine-Similarity :	Pearson: 0.7662	Spearman: 0.7583
2021-07-06 17:45:00 - Manhattan-Distance:	Pearson: 0.7805	Spearman: 0.7772
2021-07-06 17:45:00 - Euclidean-Distance:	Pearson: 0.7816	Spearman: 0.7778
2021-07-06 17:45:00 - Dot-Product-Similarity:	Pearson: 0.6610	Spearman: 0.6536

While the fine - tuned version with the defaults of the training script and the Spanish training dataset results in:

2021-07-06 17:49:22 - EmbeddingSimilarityEvaluator: Evaluating the model on stsb-multi-mt-test dataset:
2021-07-06 17:49:24 - Cosine-Similarity :	Pearson: 0.8265	Spearman: 0.8207
2021-07-06 17:49:24 - Manhattan-Distance:	Pearson: 0.8131	Spearman: 0.8190
2021-07-06 17:49:24 - Euclidean-Distance:	Pearson: 0.8129	Spearman: 0.8190
2021-07-06 17:49:24 - Dot-Product-Similarity:	Pearson: 0.7773	Spearman: 0.7692

In our STS Evaluation repository we compare the performance of this model with other models from Sentence Transformers and Tensorflow Hub using the standard STSBenchmark and the 2017 STSBenchmark Task 3 for Spanish.

Resources

• Training dataset stsb_multi_mt
• Sentence Transformers Semantic Textual Similarity
• Check sts_eval for a comparison with Tensorflow and Sentence - Transformers models
• Check the development environment to run the scripts and evaluation

📄 License

No license information is provided in the original README.