Edit model card

Description

CodeT5-small model, fine-tuned on the code summarization subtask of CodeXGLUE (Ruby programming language). This model can generate a docstring of a given function written in Ruby.

Notebook

The notebook that I used to fine-tune CodeT5 can be found here.

Usage

Here's how to use this model:

from transformers import RobertaTokenizer, T5ForConditionalGeneration

model_name = "nielsr/codet5-small-code-summarization-ruby"
tokenizer = RobertaTokenizer.from_pretrained(model_name)
model = T5ForConditionalGeneration.from_pretrained(model_name)

code = """
def update_with_file_contents(digest, filename)
      File.open(filename) do |io|
        while (chunk = io.read(1024 * 8))
          digest.update(chunk)
        end
      end
    end
"""

input_ids = tokenizer(code, return_tensors="pt").input_ids
outputs = model.generate(input_ids)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
# Update the digest with the contents of the given file
Downloads last month
3
Safetensors
Model size
60.5M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train jwlovetea/test_model