Arinkc/pydoc-llama-codesearchnet-curated

PyDoc-Llama: Curated Python Docstring Dataset

A curated subset of CodeSearchNet specifically prepared for supervised fine-tuning of Llama 3.1 8B to generate Google-style Python docstrings.

Dataset Summary

• Source: CodeSearchNet (Python subset, 412,178 raw examples)
• Final size: 24,970 examples
• Format: Llama 3.1 chat conversations (system + user + assistant messages)
• Target output: Google-style Python docstrings

Curation Pipeline

1. Multi-criteria quality filter (30.66% pass rate):

   • Code length: 100-2000 characters
   • Docstring length: 30-500 characters
   • Minimum 5-8 words depending on structure
   • <5% non-ASCII content
   • No TODO/FIXME markers

2. Style enforcement — accepts only:

   • Docstrings with Google sections (Args:, Returns:, Raises:)
   • Clean prose docstrings (≤4 lines, <400 chars)
   • Rejects reStructuredText, JavaDoc, and NumPy styles

3. Random sampling of 25,000 examples (seed=42)

4. AST-based docstring stripping to prevent input-output leakage. Filters out structurally-invalid examples where the "docstring" was placed mid-function rather than as the first statement.

5. Chat formatting for Llama 3.1 with system prompt

Splits

Split	Examples
Train	22,473
Validation	1,248
Test	1,249

Token Length Distribution

Measured with the Llama 3.1 tokenizer:

• Median: 201 tokens
• p90: 359 tokens
• p99: 514 tokens
• Maximum: 799 tokens

The short token lengths enable training with max_seq_length=1024, roughly halving memory requirements compared to standard 2048-token training.

Example

{
  "messages": [
    {"role": "system", "content": "You are an expert Python documentation writer..."},
    {"role": "user", "content": "Generate a Google-style docstring for this function:\n\n```python\ndef ...\n```"},
    {"role": "assistant", "content": "Returns the name of the port that this motor is connected to."}
  ]
}

Usage

from datasets import load_dataset

dataset = load_dataset("Arinkc/pydoc-llama-codesearchnet-curated")
print(dataset['train'][0])

Evaluation Results

Model trained on this dataset achieved:

• 0% hallucination rate (vs 11% for base model) on 200 held-out examples
• 100% format compliance (capital-letter starts, no preamble)
• Training loss: 2.3 → 0.63 over 4,212 steps

See GitHub repo for full results.

Limitations

• Source data quality varies. While the filter removes most low-quality examples, some grammatically awkward but well-formatted docstrings remain.
• Google-style bias. The dataset is filtered to one specific docstring style. Models fine-tuned on this data will produce Google-style output even when other styles might be appropriate.
• English only. Non-English content is filtered out.

Citation

If you use this dataset, please cite the original CodeSearchNet paper:

@article{husain2019codesearchnet,
  title={CodeSearchNet Challenge: Evaluating the State of Semantic Code Search},
  author={Husain, Hamel and Wu, Ho-Hsiang and Gazit, Tiferet and Allamanis, Miltiadis and Brockschmidt, Marc},
  journal={arXiv preprint arXiv:1909.09436},
  year={2019}
}