Pretrained model on programming language python using the t5 large model architecture. It was first released in this repository. This model is trained on tokenized python code functions: it works best with tokenized python functions.
This CodeTrans model is based on the
t5-large model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
The model could be used to generate the description for the Python function or be fine-tuned on other Python code tasks. It can be used on unparsed and untokenized Python code. However, if the Python code is tokenized, the performance should be better.
Here is how to use this model to generate Python function documentation using Transformers SummarizationPipeline:
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline pipeline = SummarizationPipeline( model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_large_source_code_summarization_python_multitask"), tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_large_source_code_summarization_python_multitask", skip_special_tokens=True), device=0 ) tokenized_code = '''with open ( CODE_STRING , CODE_STRING ) as in_file : buf = in_file . readlines ( ) with open ( CODE_STRING , CODE_STRING ) as out_file : for line in buf : if line == " ; Include this text " : line = line + " Include below " out_file . write ( line ) ''' pipeline([tokenized_code])
Run this example in colab notebook.
The supervised training tasks datasets can be downloaded on Link
The model was trained on a single TPU Pod V3-8 for 80,000 steps, 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. (We have trained in total 260,000 steps.)
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
|Language / Model||Python||SQL||C#|
|State of the art||--||18.40||20.50|
Select AutoNLP in the “Train” menu to fine-tune this model automatically.
- Downloads last month