README.md

---
language:
- code
- en
license:
- mit
- apache-2.0
- bsd
tags:
- code
- pretraining
- code-generation
- instruct
- luck-spark
- moe
- Github
size_categories:
- 1K<n<10K
task_categories:
- text-generation
pretty_name: Luck Spark 1B - High Quality Code Dataset
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: text
    dtype: string
  - name: repo
    dtype: string
  - name: path
    dtype: string
  - name: language
    dtype: string
  - name: hash
    dtype: string
  - name: score
    dtype: float64
  - name: stars
    dtype: int64
  splits:
  - name: train
    num_bytes: 282833845
    num_examples: 47982
  download_size: 112741739
  dataset_size: 282833845
---

# Luck Spark 1B - High Quality Code Dataset

**The first quality-scored, star-agnostic code dataset for training 1B MoE code models.**

Unlike The Stack / CodeParrot that filter by stars, this dataset scores every file **by its content (0-10)**. A 2-star well-documented library scores higher than a 10k-star minified file. Continuously updated by an autonomous bot.

**Repo:** `ahmetggg/luck-spark-1b-code-dataset` | **Bot:** `github_to_hf_bot.py` | **License:** Permissive only (MIT / Apache-2.0 / BSD / Unlicense)

## Why This Dataset is Different?

| Feature | This Dataset | Others (Stack, etc.) |
|---------|--------------|----------------------|
| **Filter** | Content Quality Score 0-10 | Stars > 100 |
| **Low-star gems** | ✅ Kept if quality 7+ | ❌ Discarded |
| **Quality transparency** | `score` column for every file | No score |
| **Dedup** | SHA256 + diversity check | Basic |
| **Execution check** | AST parse + structure | None |
| **Live** | Bot updates daily | Static dump |

**Quality Score (0-10) breakdown:**
- `+3` AST parse + has function/class + docstring
- `+2` Comment ratio 5-40% (documented, not spam)
- `+1` Ideal size 500-20k chars
- `+1` Diversity (unique lines >60%)
- `+1` Weak star bonus `log10(stars+1)*0.5` (max 1 point)
- `- fail` minified, auto-generated, binary, 0/50 diversity

- `score <5` → discarded (trash)
- `score 5-7` → kept locally, not pushed (medium)
- `score 7+` → **pushed to HF** (high quality only)

You can see the exact scorer: `quality_score()` in `github_to_hf_bot.py:26`

## Dataset Structure

```python
{
  "text": "import math\nclass Calculator:\n    ...",  # raw code
  "repo": "ahmetggg/example-repo",                  # source repo
  "path": "src/calc.py",                            # file path
  "language": ".py",                                # .py/.js/.rs/.go/.java/.cpp/.ts
  "hash": "a1b2c3d4e5f6g7h8",                        # SHA256 dedup
  "score": 7.4,                                     # 0-10 quality
  "stars": 12                                       # repo stars at scrape time
}
```

**Languages:** Python, JavaScript, Rust, Go, Java, C++, TypeScript (balanced, no star bias)

## Usage

```python
from datasets import load_dataset

# Load high-quality only (7+ already filtered)
ds = load_dataset("ahmetggg/luck-spark-1b-code-dataset")
print(ds)
# DatasetDict({ train: Dataset({ num_rows: 1000+, features: [...] }) })

# Filter even stricter (e.g., 8+)
high = ds["train"].filter(lambda x: x["score"] >= 8)
print(f"Elite: {len(high)} files")

# Language split
py = ds["train"].filter(lambda x: x["language"] == ".py")

# For pretraining (raw text)
from transformers import AutoTokenizer
tok = AutoTokenizer.from_pretrained("ahmetggg/luck-spark-1b")
texts = ds["train"]["text"]
```

**For Luck Spark 1B training:**
```bash
# Pretrain: use raw text
# Instruct: use text + auto-generated instruction (coming soon)
# RL: execution-verified subset (score 8+)
```

## Stats (Live)

- **Total repos scanned:** 616+ (7 languages × 3 pages, growing)
- **Files kept:** ~60% (0/50 for trash repos, 34/50 for gems)
- **Avg score:** 6.2 - 7.6 (pushed avg >7.0)
- **Dedup:** SHA256, ~5% duplicates removed
- **Licenses:** MIT / Apache-2.0 / BSD / Unlicense only (commercial safe)

*Updated continuously. Last bot run: see commit history.*

## Collection Method

1. GitHub Search API: `language:python license:mit` (no star filter, `sort:updated`)
2. Tree API: max 50 files / repo, `<500KB`, allowed extensions
3. Raw download + `quality_score()` -> keep 5+, push 7+
4. Arrow/Parquet -> `push_to_hub` every 1000 files

No manual curation. Fully autonomous, reproducible.

## Limitations & Ethics

- Only permissive licenses. No GPL/copyleft. Check `repo` field before commercial use.
- Code may contain biases from GitHub. Filter `score` for your use-case.
- No PII scrubbing beyond GitHub public data. Report issues via Discussions.

## Citation

```bibtex
@dataset{luck_spark_1b_2026,
  title={Luck Spark 1B High Quality Code Dataset},
  author={ahmetggg},
  year={2026},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/ahmetggg/luck-spark-1b-code-dataset}
}
```

## Roadmap

- [x] Quality-scored v1 (7+ push)
- [ ] Execution-verified subset (`python -m py_compile` + tests)
- [ ] Instruction pairs (`explain this code` / `complete this function`)
- [ ] 100B tokens target for 1B MoE pretraining

Built for **Luck Spark 1B (Mamba + MoE, Executor + Architect)** - open source, HF first.

*Questions? Open a Discussion on HF or check `github_to_hf_bot.py` for the exact logic.*