README.md

---
license: apache-2.0
language:
- en
base_model:
- microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext
pipeline_tag: fill-mask
---
# Model Card for Model ID

We fine-tuned BiomedBERT using study descriptions from metagenomic projects sourced from MGnify. We applied MLM to unlabelled text data, specifically focusing on the project study descriptions. By fine-tuning the model on domain-specific text, the model now better understands the language and nuances found in metagenomics study description, which helps improve the performance of biome classification tasks.

This modelcard aims to be a base template for new models. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/modelcard_template.md?plain=1).

## Model Details

### Model Description

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



- **Developed by:** SantiagoSanchezF
- **Model type:** MLM
- **Language(s) (NLP):** English
- **License:** [More Information Needed]
- **Finetuned from model:** microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext


### Downstream Use [optional]

This model isthe base of SantiagoSanchezF/trapiche-biome-classifier


## Training Details

### Training Data

<!-- This should link to a Dataset 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. -->

[More Information Needed]

### Training Procedure

The model was domain adapted by applying masked language modeling (MLM) to a corpus of study descriptions derived from metagenomic projects in MGnify. The input text was tokenized with a maximum sequence length of 256 tokens. A data collator was configured to randomly mask 15% of the input tokens for the MLM task. Training was performed with a batch size of 8, over 3 epochs, and with a learning rate of 5e-5. 


## Citation [optional]

TBD