Dataset · nampdn-ai
1.6M short code snippets for reasoning training
nampdn-ai/tiny-codes
Synthetic short code examples across 13 programming languages and databases to help LLMs learn reasoning.
What's inside
A synthetic corpus of approximately 1.6 million short and clear English code snippets. Covers Python, TypeScript, JavaScript, Ruby, Julia, Rust, C++, Bash, Java, C#, Go, as well as query languages Cypher (graph databases) and SQL (relational databases). Collected specifically to train LLMs to bridge natural language and programming.
What you can do with it
- SFT small code models on short examples
- Train LoRA for code reasoning tasks
- Verify model code generation based on natural language descriptions
- Tasks
- Text generation
- Languages
- English
- Size
- 1M–10M rows
- Downloads this month
- 1.2K
- Likes
- 310
- Access
- on request on Hugging Face
Models trained on it
- IvmeLabs/Ivme-Conversate-N-v1-BaseText generation 687
- daekeun-ml/phi-2-ko-v0.1Text generation 149
- amusktweewt/tiny-model-700M-chatText generation 147
- SykoSLM/SykoLLM-V6.1Text generation 123
- SykoSLM/SykoLLM-Phi3-TestText generation 81
- zhhan/adapter-Phi-3-mini-4k-instruct_code_writingText generation 43
- SykoSLM/SykoLLM-V5.8-MiniText generation 31
- daekeun-ml/Phi-3-medium-4k-instruct-ko-poc-gguf-v0.1Text generation 30
How to load it
from datasets import load_dataset
ds = load_dataset("nampdn-ai/tiny-codes")