Code Trans T5 Base Source Code Summarization Sql Transfer Learning Finetune

🚀 CodeTrans Model for Source Code Summarization (SQL)

A pre - trained model for SQL source code summarization, leveraging the T5 base architecture. It offers high - quality summarization for SQL code snippets.

🚀 Quick Start

This CodeTrans model is a pre - trained model for SQL source code summarization, initially released in this repository. It is trained on tokenized SQL code functions and performs best with such inputs.

✨ Features

• Based on the t5 - base model with its own SentencePiece vocabulary model.
• Utilizes transfer - learning pre - training on 7 unsupervised datasets in the software development domain.
• Fine - tuned for the source code summarization task of SQL code snippets.

📦 Installation

No specific installation steps are provided in the original README.

💻 Usage Examples

Basic Usage

Here is how to use this model to generate SQL function documentation using Transformers SummarizationPipeline:

from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline

pipeline = SummarizationPipeline(
    model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_base_source_code_summarization_sql_transfer_learning_finetune"),
    tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_base_source_code_summarization_sql_transfer_learning_finetune", skip_special_tokens=True),
    device=0
)

tokenized_code =  "select time ( col0 ) from tab0"
pipeline([tokenized_code])

Run this example in colab notebook.

📚 Documentation

Model description

This CodeTrans model is based on the t5 - base model. It has its own SentencePiece vocabulary model. It used transfer - learning pre - training on 7 unsupervised datasets in the software development domain. It is then fine - tuned on the source code summarization task for the SQL code snippets.

Intended uses & limitations

The model could be used to generate the description for the SQL function or be fine - tuned on other SQL code tasks. It can be used on unparsed and untokenized SQL code. However, if the SQL code is tokenized, the performance should be better.

🔧 Technical Details

Training data

The supervised training tasks datasets can be downloaded on Link

Training procedure

Transfer - learning Pretraining

The model was trained on a single TPU Pod V3 - 8 for 500,000 steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder - decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre - training.

Fine - tuning

This model was then fine - tuned on a single TPU Pod V2 - 8 for 500 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing SQL code.

Evaluation results

For the source code summarization tasks, different models achieve the following results on different programming languages (in BLEU score):

Test results :

Created by Ahmed Elnaggar | LinkedIn and Wei Ding | LinkedIn